摘要:

Russian Chemical Reviews 70 (6) 449 ± 469 (2001) Femtochemistry OMSarkisov, S Ya Umanskii Contents I. Introduction II. Characteristic features of femtosecond pulses III. The main objectives of femtochemistry IV. Experimental methods of femtochemistry V. Dynamics of intramolecular processes and of transition states in a chemical transformation VI. The kinetics of ultrafast chemical reactions VII. Control of intramolecular dynamics and an elementary chemical event VIII. The prospects of femtochemistry Abstract. � chemistry of field new a to devoted is review The The review is devoted to a new field of chemistry � femtochemistry. of possibilities principle, in New, femtochemistry. New, in principle, possibilities of investigation investigation and ultrashort of means by reactions chemical of control and control of chemical reactions by means of ultrashort coherent coherent light illustrated are concepts general The considered. are pulses light pulses are considered.The general concepts are illustrated by by numerous dynamics the of studies femtochemical of examples numerous examples of femtochemical studies of the dynamics and and kinetics and gas the in reactions chemical elementary of kinetics of elementary chemical reactions in the gas and condensed condensed phases. are femtochemistry of prospects The phases. The prospects of femtochemistry are discussed. discussed. The bibliography includes 172 references. The bibliography includes 172 references. I. Introduction How does an elementary chemical reaction occur? What are intermediate configurations of a reacting system? How can the route and the rate of a chemical reaction be controlled? Chemists have asked themselves these questions throughout the history of the development of chemistry.However, the profundity of both the questions and the answers depends on the level of under- standing of elementary processes and on the experimental capa- bilities. Until quite recently, in all experiments dealing with elemen- tary reactions the time evolution of either reactants or reaction products has been recorded. However, on the path from reactants to products, the molecular system passes through various struc- tures (states) which can no longer be considered as reactants but cannot yet be regarded as products. From the experimenter's viewpoint, the range of interatomic distances corresponding to these structures is a `black box' because the experiment does not provide any information about these structures or their change.Chemists call this range of interatomic distances the transition state,1 and the time evolution of atom configurations in this `black box' is called transition state dynamics. The passage of a reacting molecular system through the transition state represents an elementary step of a chemical reaction. The reaction path is OMSarkisov, S Ya Umanskii N N Semenov Institute of Chemical Physics, Russian Academy of Sciences, ul. Kosygina 4, 117334 Moscow, Russian Federation. Fax (7- 095) 938 24 84. Tel. (7-095) 939 74 97. E-mail: sarkisov@femto.chph.ras.ru (OMSarkisov) Tel. (7-095) 939 74 36.E-mail: unan@center.chph.ras.ru (S Ya Umanskii) Received 27 February 2001 Uspekhi Khimii 70 (6) 515 ± 538 (2001); translated by Z P Bobkova #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n06ABEH000664 449 449 451 451 452 457 462 466 usually characterised by the dependence of the potential energy on interatomic distances. Theorists try to reconstruct the transition state dynamics from the results of investigations of the kinetics of reactants and products. However, it turned out that this often cannot be done unambiguously; experimental data are needed. The problem lies in the fact that the residence time of a molecular system in the `black box' is very short, *100 fs (1 fs=10715 s). Therefore, special laser equipment is required to detect the events taking place over this 100-fs period.The appearance of a new tool�light pulses with durations of 10 ± 100 fs � provided the possibility of detecting tiny changes, about 0.1A, in the distances between atoms in a molecular system; this triggered a new line of research in chemical physics, namely, femtochemistry. This field of investigations makes it possible to answer the questions posed above at a new, in principle, level; to verify some fundamental assumptions which we use; to distinguish the regions where they are no longer valid; and to discover new ways of controlling a chemical transformation. In 1999, an American scientist A Zewail was awarded the Nobel Prize for studies in the field of femtochemistry. Although femtochemistry is `young' (its age can be estimated as 12 years), several reviews have been published (see, for example, publications 2 ±6).In this review, we attempt to examine, from different standpoints, the rapidly developing approaches and concepts as well as new problems arising in this field. II. Characteristic features of femtosecond pulses Before discussing the problems of femtochemistry, let us consider characteristic features of ultrashort light pulses and new exper- imental capacities brought about by these features. In a system of coordinates the z axis of which coincides with the direction of propagation of a laser pulse, the components of the electric field strength E(t)=[Ex(t),Ey(t),Ez(t)] of a femto- second light pulse for a specified point in space can be written in the form Ex(t)=Ex0f(t)cos[o0t+dx+a(t)], (1) Ey(t) = Ey0f(t)cos[o0t+dy+a(t)], Ez(t)=0.In these equations, Ex0 and Ey0 are the maximum values of the x and y components of the field strength, respectively, f(t) is the450 time dependence of the envelope of the pulse, t is the current time, o0 is the carrier light frequency related to the carrier wavelength l0 in the following way: l0=2pc , o0 dx and dy are the phase constants and a(t) is the function describing the frequency modulation over the pulse duration. The polarisation of light in the pulse is characterised by the ratio Ex0/Ey0 and by the dx and dy values. The pulse envelope f(t)!pAAAAAAAA, where I(t) is the time dependence of light intensity IOtU in the pulse, can have different forms. However, irrespective of this, it is determined by pulse duration (t).Most often, Gaussian pulses with the envelope f(t)=exp (2) ¢§t 22 ln2 , t2 and the frequency modulation (3) a(t)=12 gt2, where g is the temporal chirp (the parameter determining the rate of phase modulation) are considered. When t=t/2, the light intensity in the pulse is equal to half its maximum. Thus, parameters characterising a femtosecond pulse include pulse duration t, field amplitude, carrier frequency and the function a(t) (expressed in most cases in terms of the temporal chirp g). It is known that a field can also be specified by spectral characteristics (4) ej (o)= EjOtU expOiotUdt, O a? ¢§? where j=x, y. Usually, the functions f(t) and exp[ia(t)] vary much more slowly than exp(io0t); therefore, it can be written that (5) ej (o)=1 exp (7idj)Fj (o7o0), 2 where (6) f(t)exp {i [(o7o0)t7a(t)]}dt.Fj (o7o0)=Ej 0¢§O a?? (7) The function Fj (o7o0) is complex-valued and can be repre- sented in the form Fj (o7o0)=|Fj (o7o0)|exp[ij(o7o0)], where |Fj (o7o0)|!pAAAAAAAAAA SOoU , S(o) is the the pulse power spectrum measured experimentally. The spectral phase j(o7o0) is related to the functions f(t) and a(t) in a complex manner. For Gaussian pulses, these functions are determined by relations (2) and (3), respectively. Therefore, j(o7o0)=12 b(o7o0)2, (8) b= 4gt4 16Oln 2U2 a g2t4 . The b parameter is called spectral chirp. An important spectral characteristic of a femtosecond pulse is the width of its spectrum at the half-height of the amplitude (Do). The spectral width is determined by the uncertainty relation (9) tDo=const.The value of const depends on the shape of the pulse and the chirp envelope. For a Gaussian pulse with g=0 (this pulse is usually referred to as transform-limited), const=2.773. In the case of OM Sarkisov, S Ya Umanskii chirped pulses (i.e., those with g=0), the constant can acquire arbitrarily large epending on the chirp value. Now we shall list the main experimental capacities resulting from the properties of femtosecond pulses. First, due to the short pulse duration, evolution of processes can be studied with a temporal resolution even higher than the duration of the pulse itself. When the velocity of an atom is 105 cm s71 and t=10714 s, changes in the internuclear distances of about 0.1 A can be detected. This means that not only the formation of products but also the time evolution of the nuclear configuration can be monitored with high accuracy in real time within the time scale of vibrational motion.Second, since the spectral width of femtosecond pulses is rather large, several quantum states with different energies can be excited simultaneously. For example, a transform-limited pulse with t=10714 s has the spectral width Do=1100 cm71. This light pulse can excite simultaneously several vibrational quantum states. Third, a coherent light pulse can be obtained. In combination with the aforesaid, this means that coherent excitation of several vibrational or rotational states is possible. This new type of an excited state is called a non-stationary quantum state or a coherent nuclear wave packet.{ A specific feature of these states is that several eigen energy states with definite relative phases of atomic motion in a molecule can be subjected simultaneously to coherent excitation.The scheme of non-stationary and stationary quantum states of a molecule is shown in Fig. 1. b a E hDn 12 Ground electronic state Figure 1. Non-stationary (a) and stationary (b) vibrational quantum states of a molecule. The horizontal lines mark unoccupied (1) and occupied (2) stationary vibrational states. Fourth, due to high intensity (peak power) of femtosecond pulses with the pulse energy being relatively low, multiphoton absorption processes giving rise to highly excited molecular systems can be easily accomplished and numerous methods of non-linear spectroscopy can be used without heating the sample.Under the action of such pulses on a gas or a solid, light pulses with a broad spectral range (supercontinuum) and X-ray and electron pulses are generated. It should also be noted that light field intensities (1010 V cm71) exceeding even the intensities of intramolecular fields can be attained rather easily. Thus, the use of femtosecond light pulses makes it possible to attain high temporal resolution, to create coherent non-stationary quantum states, to produce highly excited molecules, to act upon the potential energy surface (PES) and to generate ultrashort light, electron and X-ray pulses. { The notion of wave packet was introduced in 1926 by E SchroE dinger; however, it has virtually not been used in chemistry.Femtochemistry III.The main objectives of femtochemistry How does the study of reactions within the femtosecond time scale pertain to practical chemistry? Indeed, the duration of an elemen- tary reaction in practical chemistry is much longer than 100 fs. This is due to the fact that an elementary reaction includes both a fast chemical event and the preceding slower intra- and intermo- lecular energy exchange processes. For an unimolecular chemical event to occur, the following processes should take place: Activation of molecules Energy localisa- tion on chemical bonds Energy redistribution over degrees of freedom Products Chemical event Activated molecules should possess a sufficiently high energy, which can be acquired either upon collisions, or due to a chemical reaction, or upon absorption of photons.Usually, this energy is redistributed very rapidly (in about 0.1 ± 1 ps) over different vibrational degrees of freedom of the molecule. Then a fluctuation should occur, after which a certain energy is concentrated in the required site of the molecule, for example, in the form of vibrations of atoms of the bond to be cleaved. Only after this, does the chemical event take place. The time scale of the chemical event itself is 10 ± 100 fs. The specific features of new instrumentation noted above determine the following lines of research in femtochemistry: study of the dynamics of intramolecular processes and the transition state in a chemical transformation; study of the kinetics of ultra- fast chemical reactions; control of the intramolecular dynamics and of the path of an elementary chemical reaction.Now we shall briefly explain what is the purpose of each of these lines of investigation. The goal of the first line is to obtain more detailed and substantiated views on the nuclear dynamics in a molecular system (either reacting or not). This refers both to the dynamics of intramolecular processes (coherent motion of nuclei, energy redistribution or localisation, etc.) and time evolution of the configurations of the molecular system on passing from reactants to products. In this connection, the possibility of creating a coherent wave packet (rotational, vibrational or electronic) in the molecular system is of fundamental importance. The most significant for chemistry is the vibrational wave packet in which several vibrational states undergo simultaneously coherent exci- tation. In this case, it can be said that intramolecular dynamics of nuclei is the time evolution of the vibration-rotational wave packets.A significant feature of this description is that phase characteristics of the nuclear motion are taken into account. However, it is known that after some time, dephasing takes place; as a consequence, the nuclear dynamics becomes incoherent and the regular nuclear motion switches to stochastic motion. The purpose of the second line is to gain information on the kinetics of ultrafast reactions. Within the framework of this line of research, investigators are interested in the time evolution of the reactant or product concentration rather than in the phase characteristics of the nuclear motion.The features of femtosecond pulses most significant for this line of research include short duration, which allows implementation of techniques with high temporal resolution (using these techniques, one can study ultra- fast reactions), and high intensity (this allows excitation of molecules with visible light to high energies). The third line of research in femtochemistry is the most `ambitious'. In this case, all the above-mentioned features of femtosecond pulses are significant. The possibility of creating femtosecond pulses opens new, in principle, opportunities for controlling intramolecular processes and the path of elementary 451 chemical reactions. The idea of this control is based on the fact that, by using femtosecond pulses, one can switch the reacting system to a transition state with a definite initial wave function and then control the time evolution of this state.Due to the coherence of the light pulse, the initial state of the system can be created as a well-defined coherent wave packet. By changing the carrier frequency and the amplitude and phase characteristics of a femtosecond light pulse, it is possible to create in the initial (zero) instant of time different nuclear wave packets, the dynamics of which can be substantially different. The high temporal resolution allows researchers to `monitor' the evolution of the transition state in order to act upon the system PES by one more femtosecond light pulse at the required point in time.IV. Experimental methods of femtochemistry Typical experimental methods of femtochemistry are based on the use of two femtosecond pulses according to the `pump ± probe' pattern (Fig. 2). The first pulse (l1) creates a wave packet in the electronically excited state. After some period of time, a second pulse (l2) probes the changes that have taken place. After the probe pulse, the molecular system passes into a different electronic state [either upper (Fig. 2 a) or lower (Fig. 2 b) in energy] and, owing to this transition, the response of the system to the action of the pump and probe pulses as a function of the time delay between them is recorded. The `pump ± probe' pattern makes it possible to detect both the dynamics of the motion of atoms (wavelength l2) and the kinetics of product formation (wavelength l02).b a U U l02 l2 l2 l1 l1 R Figure 2. Simplest patterns of the `pump ± probe' technique. U is poten- tial energy, R is the distance between the nuclei, l1 is the wavelength at the centre of the pump pulse spectrum, l2 is the wavelength at the centre of the spectrum of the pulse probing the transition state dynamics; l02 is the wavelength at the centre of the spectrum of the pulse probing the reaction products. A typical scheme of the setup is shown in Fig. 3. A laser (as a rule, a dye or titanium-sapphire laser) usually pumped by a continuous-wave argon or solid-state laser generates consecutive femtosecond light pulses. The energy of these pulses is increased in a laser amplifier. The amplified pulse passes through a compensa- tor of the envelope velocity dispersion (not shown in the Figure) and is then split into two components.One component is used as the pump pulse and the other, as the probe pulse. Both pulses pass through facilities which allow one to change the pulse parameters (carrier frequency, chirp, duration). The variable time delay between the pump and probe pulses is ensured by an optical delay line in which the optical path length of one of the pulses can vary. The pump and probe pulses are focussed on the specimen. The response to the two pulses as a function of the delay between them is recorded using a detector. The characteristics recorded as the response can include photoinduced absorption,7±9 fluores- cence,10 ± 12 stimulated radaiation,13 miltiphoton ionisation 14452 12 3 4 l1 5 6 l2 7 Figure 3.Experimental setup used for the `pump ¡¾ probe' experiments. (1) generator of femtosecond pulses; (2) optical amplifier of the pulse energy; (3, 4) devices for changing the parameters of femtosecond light pulses, (5) optical line of the time delay between the pump and probe pulses; (6) reaction zone, (7) device for recording the response of the molecular system to the action of two pulses. (often recorded using time-of-flight mass spectrometry techni- ques 15, 16), rotation of the polarisation plane, 17 etc. It is worth noting that the high intensity of femtosecond pulses permits utilisation of any non-linear spectroscopy technique developed to date in which several pulses are used.18, 19 A special group of methods includes those in which the second pulse is focussed onto a target in order to generate ultrashort super- continuum,20, 21 electron 22 ¡¾ 25 or X-ray 26, 27 pulses, which then serve as the probe pulses.Supercontinuum pulses are used most effectively to study solutions. Primary processes in solutions result in broad absorp- tion spectra of reactants, intermediate states and products. These spectra overlap, which markedly hampers interpretation of exper- imental data obtained at a particular wavelength. Regarding the time evolution of the intermediate states, more information is provided by an approach according to which time evolution of a broad spectrum of photoinduced absorption is recorded. This requires an ultrashort pulse which emits light over a broad spectral range.A supercontinuum the spectrum of which covers the whole visible region can serve as the source of such a pulse in femto- chemistry. The methods that make use of this approach are documented.20, 21 Ultrashort pulses of electrons and X-rays are also used as probe pulses. In this case, diffraction of electrons or X-rays after reflection from the specimen under study is detected. By recording the diffraction pattern of electrons or X-rays at various points in time after the action of the pump light pulse, one can gain information on the structure of intermediates and the dynamics of structural changes. The methods based on electron diffraction are applicable only to gas media, while those based on X-ray diffraction are suitable also for condensed media.For example, electron diffraction has been used to determine the structure of an intermediate species.22 The diffraction of X-rays has been used as yet only to detect the dynamics of structural changes in the absence of a chemical reaction. Thus time evolution of the arrangement of cadmium atoms in films of organic compounds heated by femtosecond light pulses has been studied.26 There are publications in which absorption of the probe X-ray pulse is used to investigate photochemical reactions. For example, a study by Raksi et al.27 deals with photodissociation of SF6 molecules. OM Sarkisov, S Ya Umanskii V. Dynamics of intramolecular processes and of transition states in a chemical transformation It has been noted above that, due to the great spectral width, a coherent femtosecond light pulse excites a non-stationary quan- tum state, viz., a coherent wave packet (rotational, vibrational or electronic).The dynamics of nuclear motion shows itself in the time evolution of this packet. High temporal resolution permits real-time detection of the dynamics of wave packets. A great number of theoretical studies devoted to the dynamics of coherent nuclear wave packets both in di- and polyatomic systems have been published.28 ¡¾ 36 Let us consider some properties of vibrational wave packets in relation to diatomic molecules in the absence of a chemical transformation. Usually wave packets of this type are formed in some electronically excited state j1i, to which dipole radiation transi- tions from the ground electronic state j0i are allowed.For simplicity, we assume that the pump pulse has a Gaussian shape, carrier frequency o0 , duration t and temporal chirp g. Let us also assume that the crucial contribution to the wave packet is made by the ground vibrational state (the state with the vibrational quantum number u=0) of the ground electronic state. Then at the instant of termination of the action of the light pulse (tf&t), the vibrational wave packet C1(R, tf) (R is the distance between the nuclei) in the electronically excited state j1i has the form (10) C1(R, tf)= j1uORUCuOtf; t; gU. u X C ¢§i e1utf ¢§ i bOt; gUae1u ¢§ Oo0 ¢§ Te1Ua2 6 h2h2 Here j1u(R) are the vibrational wave functions of the molecule in the electronic state j1i corresponding to the vibrational quantum number u and the vibration energy e1u measured in relation to the minimum of the electronic potential curve of the state j1i, while the complex coefficients are

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC

摘要:

Russian Chemical Reviews 70 (6) 471 ± 490 (2001) Ultrafast nonradiative transitions between higher excited states in organic molecules V L Ermolaev Contents I. Introduction II. Single-photon excited fluorescence from higher excited Sn levels to the ground-state level III. Fluorescence due to transitions from higher excited Sn levels under conditions of stepwise excitation of molecules IV. Lifetimes of higher excited Sn states estimated from fluorescence quantum yields and absorption integrals V. Delayed fluorescence from higher excited Sn states due to triplet ± triplet annihilation VI. Lifetimes of higher excited Sn states of molecules estimated from homogeneous broadening of absorption spectra VII. Lifetimes of higher excited Sn states estimated from picosecond and femtosecond spectroscopy data VIII.Ultrafast nonradiative transitions involving triplet states IX. Conclusion Abstract. electronic between transitions nonradiative Ultrafast Ultrafast nonradiative transitions between electronic levels fast with competing molecules, dye and aromatic of levels of aromatic and dye molecules, competing with fast photo- photo- chemical photoisomer- transfer, proton or electron reactions chemical reactions (e.g., ., electron or proton transfer, photoisomer- isation, the of determination of results The considered. are etc.), ), are considered. The results of determination of the rate higher between transitions nonradiative for constants rate constants for nonradiative transitions between higher excited excited singlet are methods different using states triplet and singlet and triplet states using different methods are summarised.summarised. The the of terms in interpreted are data experimental The experimental data are interpreted in terms of the exchange- exchange- resonance The transitions. nonradiative of theory resonance theory of nonradiative transitions. The bibliography bibliography includes references 189 includes 189 references. I. Introduction In his lecture at the XIX Mendeleev Congress academician A L Buchachenko noted that the development of experimental techniques that use short laser pulses has widened the horizons of chemistry. Fresh ideas thus introduced into chemistry have been embodied into femtochemistry.1 The subject of femtochemical studies is ultrafast reactions, such as proton or electron transfer in acid ± base (donor ± acceptor) pairs, cis ± trans-isomerisation with detection of the twist-conformation, photogeneration and relaxa- tion of electron-hole pairs in semiconductors, elementary stages in photosynthesis, molecular decomposition, etc.2, 3 Topicality of research into ultrafast processes is emphasised by the fact that the 1999 Nobel Prize in chemistry was awarded toA H Zewail, one of the originators of femtochemistry.Single-photon excitation of a complex organic molecule to a higher excited singlet state (HESS) Sn (n>1) is followed by transient relaxation processes. These processes, as well as trans- formations of excess electronic energy thus supplied to the molecule, are of considerable interest in studying the mechanisms of photochemical reactions in the vapour phase and in solutions.Most of the photochemical reactions of organic compounds in solutions are known to proceed involving the first excited singlet V L Ermolaev The State Unitary Enterprise `All-Russian Scientific Centre, S I Vavilov State Optical Institute', Birzhevaya Liniya 12, 199034 St. Petersburg, Russian Federation. Fax (7-812) 328 37 20. Tel. (7-812) 328 46 08 (ext. 426). E-mail: ermol@soi.spb.su Received 2 February 2001 Uspekhi Khimii 70 (6) 539 ± 561 (2001); translated by AMRaevsky #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n06ABEH000657 471 473 476 481 481 484 486 487 488 (S1) and triplet (T1) states.Both of them usually have rather long, on the molecular scale, lifetimes (from several nanoseconds for the S1 states to hundreds of microseconds in liquids and tens of seconds in solid solutions for the T1 states). However, the characteristic times of many fast photochemical reactions (elec- tron and proton transfer, isomerisation, etc.) are much shorter. This suggests the involvement of other, upper excited states of the molecules (whose lifetimes are much shorter compared to those of the S1 and T1 states) in these reactions. Some of the fast reactions are considered in this review as examples. It should be noted that the possibility of estimating the HESS lifetimes has appeared only in the last decades.Nevertheless, a large body of experimental data obtained by direct and indirect methods has been accumu- lated to date. Therefore, it is of interest to analyse them, to assess their reliability and to consider possible mechanisms of excitation relaxation. S I Vavilov was the first to obtain experimental data which allowed estimation of the rates of nonradiative transitions between the HESS of complex organic molecules (dyes) in solutions. He had shown experimentally 4, 5 that the quantum yields of dye fluorescence in solutions are independent of the wavelength of exciting radiation. In other words, light emission (fluorescence) always occurs from the first excited (fluorescent) S1 level irrespective of the level to which the molecule was excited after absorption of a light quantum (the Vavilov law).Attempts to observe emission from the HESS of complex molecules under standard conditions have failed for long. This means that emission from these levels (with rate constants of the order of 107 ± 108 s71) is accompanied by a competing fast non- radiative process (with a rate constant of at least 1012 s71). This interpretation is also supported by the fact that the fluorescence spectrum of a complex molecule in solution is independent of the wavelength of exciting light. After establishment of the triplet nature of the metastable phosphorescent state of aromatic molecules 6, 7 the Vavilov law was supplemented by Kasha's rule,8 according to which emission of light from complex molecules always occurs from the lowest excited level of a given spin multiplicity.The Vavilov law and Kasha's rule can be illustrated by a diagram of singlet (Sn) and triplet (Tn) vibronic energy levels of complex organic molecules (Fig. 1). Fulfilment of the Vavilov law and Kasha's rule means that the rate constants for nonradiative transitions (knr) from the Sn and Tn (n52) levels to the S1 and T1 levels, respectively, are472 S4 T3 kr 40 knr 43 S3 kr 30 knr 32 S2 kr 20 T2 knr 21 S1 kisc T1 hnexc kfl kph kabs knr 12 S0 Figure 1. An energy level diagram illustrating transitions between vibronic levels of organic molecules. Here and in Figs 2, 5 ¡¾ 7 and 9 thick (1) and thin (2) lines denote electronic and vibronic energy levels, respectively.Notations of singlet energy levels: S0 is the ground-state level and S1, S2, S3 and S4 are the first (fluorescent), second, third and fourth excited energy levels, respectively. Notations of triplet energy levels: T1, T2 and T3 are the lowest (phosphorescent), second and third excited levels, respectively. Solid and dashed vertical lines respectively denote radiative and nonradiative transitions. Notations of rate constants: kr30 is the rate constant for the S3?S0 radiative transition (the radiative rate constant), kr10:kfl is the usual fluorescence rate constant, knr32 is the rate constant for the S3?S2 nonradiative transition (internal conversion), kisc is the rate constant for intersystem crossing (spin-forbidden S1?T1 nonradiative transition), kph is the phosphores- cence rate constant, knr is the rate constant for nonradiative transition from the lowest triplet to the ground singlet state and kabs is the absorption rate constant.three for more orders of magnitude higher than the corresponding rate constants for radiative transitions (radiative rate constants, kr) and intersystem crossing rate constants. However, this esti- mate says nothing about the rate constants for particular non- radiative transitions, e.g., Sn?Sn71, Sn71?Sn72 , etc. However, knowledge of these rate constants is of interest for both research into fundamental problems of luminescence, photochemistry and photobiology and estimation of limiting energy characteristics of, e.g., dye lasers and Q switches.Recently, some experimental methods have been developed that allowed the obtaining of quantitative information on the rates of nonradiative transitions between higher excited energy levels of molecules. These are as follows: 1. Measurements of fluorescence spectra and yields from upper excited levels by direct single-quantum excitation of corre- sponding energy levels. 2. Measurements of Sn ? S0 luminescence spectra and yields using a stepwise two-photon excitation technique, S0+hn1 ? S1+hn2 ? Sn (n>1). 3. Measurements of the fluorescence spectra from higher excited singlet levels populated via triplet ¡¾ triplet annihilation. 4. HESS lifetime estimation from the results of (i) experiments on `hole burning' in absorption spectra at low temperatures, (ii) linewidth measurements in fluorescence excitation spectra in supersonic jets or (iii) bandwidth measurements in structured emission spectra from higher excited levels. 5.Direct HESS lifetime measurements using picosecond or femtosecond spectroscopy. 6. Estimation of the rates of ultrafast nonradiative transitions involving triplet states. Before presenting the results of particular experimental stud- ies on the rates of nonradiative transitions we will briefly outline elements of modern quantum-mechanical theories of nonradiative transitions in complex organic molecules. According to formal theory of nonradiative transitions,9 ¡¾11 excitation of a molecule V L Ermolaev leads to its transition to a non-stationary excited state.Owing to short duration of the excitation pulse and high density of final vibronic states inherent in polyatomic molecules, nonradiative de- activation of the initial non-stationary state occurs exponentially. The rate constant for such a process, knr, is calculated using first- order, time-dependent perturbation theory. This process corre- sponds to the `statistical limit' case.12 If the adiabatic approximation holds, the wave function of the system, Cma(r, Q), can be written as the product of electron wave function, jma(r, Q), and vibrational wave function, wa(Q), Cma(r,Q)=jm(r,Q)wa(Q), where r(r1, r2, r3,...) are the electron coordinates, Q(Q1, Q2, Q3,...) are the coordinates of atomic nuclei and the indicesmand a denote the energy level and vibrational state, respectively. Usually, the operator of nonadiabaticity, Vma,nb (m and n denote the energy levels and a and b denote the corresponding vibrational states), is considered as a perturbation inducing the nonradiative transition.Approximate expression for the matrix element of transition between the excited electronic state n and the state m with a lower energy has the form: Vma, nb& q qQ1 i O Xh2 &¢§ Mi jnOr;QU q wnbOQUdQ+ qQ1 maOQUjmOr;QU w a 12 i O Xh2 Mi q2 jnOr;QUwnbOQUdQ, qQ21 maOQUjmOr;QU w where Mi is the mass of the nucleus of ith atom. Neglecting the term containing the second derivatives of electron wave functions and simplifying this expression,13 we get Vma; nb&jmOr;QU q jnOr;QUa hjmOr;QUjqU=qQijjnOr;QUi , qQi EmOr;QU ¢§ EnOr;QU where U(r,Q) is the Coulomb energy of the interaction between electrons and nuclei and E(r,Q) is the adiabatic potential for the excited state.The total rate of a nonradiative transition between the states n and m can be written as the sum of partial transition probabilities k ma; nb ma ¢§ EnbU. r nm a 2p h a; b 2pOnbUdOE X V Here, p(nb) is the population of the initial vibronic state, d is the Dirac delta function and Ema and Enb are the energies of the vibronic states ma and nb, respectively. The aforesaid represents a framework of the general theory of nonradiative transitions.In the text below, we will consider the exchange-resonance theory of nonradiative transitions in ions and complex molecules developed by Sveshnikova, the author of this review and Bodunov.14 ¡¾ 16 According to this theory, all vibrations of a molecule under study are divided into two types, namely, promoting-accepting vibrations and accepting ones. The first type of vibrations comprises high-frequency, strongly anharmonic vibrations responsible for the vibrational absorption spectrum of the molecule in both fundamental and overtone regions. These vibrations are usually almost inactive in vibronic molecular spectra since corresponding Franck ¡¾ Condon integrals are close to zero. The second type of vibrations comprises those responsible for the vibronic spectrum of the molecule due to nonzero Franck ¡¾ Condon integrals.Often, these vibrations appear to be totally symmetric and are therefore inactive in the vibrational absorption spectrum of the molecule. For instance, the totally symmetric vibrations of aromatic molecules associated with changes in the C7C distances belong to accepting vibrations. Bodunov 15 used the operator of nonadiabaticity as a pertur- bation inducing the nonradiative transition. He divided molecularUltrafast nonradiative transitions between higher excited states in organic molecules knr& fOE UsOEUE¢§Osa1UdE. vibrations into two above-mentioned types and showed that the expression for the rate constant for the nonradiative transition between two electronic levels, knr, can be written using the overlap integral of the normalised vibronic emission spectrum of a molecule due to transition between the corresponding levels, f (E), and the vibrational absorption spectrum of the same molecule, s (E) 2 O qU hjm qQk jni Here, the horizontal line denotes averaging over the first type of vibrations; s=3 for dipole and s=5 for quadrupole radiation.This formula is similar to the FoE rster ¡¾ Dexter formulae for the nonradiative energy transfer between energy donor and acceptor, except that the transfer occurs intramolecularly to decrease the electron energy and to increase the vibrational energy. Validity of the exchange-resonance theory of nonradiative transitions was experimentally proved taking lanthanide ions in solutions,14, 17 transition-metal complexes 17, 18 and (with some limitations) com- plex aromatic molecules 17 ¡¾ 19 as examples.It should be noted that this theory was developed assuming a weak coupling between the energy levels involved in the transition, though this is not necessarily true in the case of HESS. In this review, we will analyse the results of experimental studies from the standpoint of the exchange-resonance theory of nonradiative transitions, since this theory predicts that knr is proportional to kr. Unfortunately, the rate constants for radiative transitions between higher excited singlet states are unknown for most of the molecules studied; nevertheless, in some instances such correlations can be found.Recently, a number of reviews concerning fast photophysical and photochemical processes in complex organic molecules have been published (see, e.g., excellent reviews 20, 21). However, the authors of these studies placed the emphasis on the vibrational energy redistribution after excitation of the molecules, whereas the problem of electronic relaxation was only briefly outlined. On the other hand, a large body of information on electronic relaxation reported in numerous studies allows some general- isation of the results obtained. II. Single-photon excited fluorescence from higher excited Sn levels to the ground-state level Beer and Longuet-Higgins 22 were the first who reported a strongly supported example of violation of the Vaviov law and Kasha's rule.They studied fluorescence of an azulene solution under conditions of single-quantum excitation and provided convincing proofs of the fact that the fluorescence observed after excitation of the solution into the second or shorter-wavelength absorption bands is due to transition from the second excited to the ground-state level (S2 ? S0). Azulene fluorescence was observed in the near UV spectral region (at about 28 000 cm71), though the long-wavelength absorption band of this compound lies near 14 000 cm71. These experimental observations were confirmed in more recent studies of azulene and a large number of its derivatives.23 ¡¾ 25 The quantum yields of S2 ? S0 fluores- cence of a number of azulene derivatives were found to be as high as several percent while the S2 state lifetimes were as long as 1.4 ns.The results obtained in these studies were briefly outlined in a monograph.26 fl 20=450100. fl 21 Much weaker was the S2 ? S1 fluorescence of azulene observed in the region from 14 000 to 9500 cm71 (710 ¡¾ 1050 nm).27 ¡¾ 30 Thorough measurements allowed 31 the determination of the quantum yield ratio of the S2 ? S0 and S2 ? S1 fluorescence (qfl 20 and qfl 21, respectively) qqThis value characterises the ratio of the rate constants for radiative transitions from the S2 level to the ground-state and first 473 excited levels. From the standpoint of the exchange-resonance theory of nonradiative transitions 14 the low probability of S2 ? S1 internal conversion is due to not only the wide energy gap between these levels and minor changes in the molecular configuration in the S1 state compared to those for the molecular configuration in the S2 state, but also forbiddenness of the radiative transition between these levels.Hirata and Lim 32 studied fluorescence of azulene and perdeu- terated azulene vapours and, in particular, its dependence on the excess vibrational energy in the S2 state. They suggested that a direct internal conversion from the second excited singlet state to the ground state can occur in the azulene molecule bypassing the S1 state. Unfortunately, this assumption was not strengthened by convincing proofs, so that the arguments adduced by the authors had little force.S2 / S0-Fluorescence excitation spectra of azulene cooled in a supersonic jet have been studied.33 It was found that the fluorescence excitation spectrum of the S2 / S0 transition exhib- its a fine vibrational structure. The fluorescence excitation spectra for transitions from the ground-state level to the S3 and S4 levels were found to be substantially broadened due to fast conversion from these levels to the S2 state. Direct measurements of fluorescence decay times from the S2 state of 2-haloazulene and 1,3-dihaloazulene molecules in solution have been carried out.34 It was shown that intersystem crossing to the triplet manifold with the rate constant kisc(S2?Tn) can occur in the molecules of compounds studied in addition to the S2?S1 internal conversion and S2 ? S0 and S2 ? S1 transitions.Table 1 lists the fluorescence quantum yields, energy level positions [E(S2)], energy gaps (DE ), rate constants for radiative and nonradiative transitions between different energy levels as well as the second excited singlet state lifetimes, t(S2), for azulene and a number of its derivatives.31 ¡¾ 51 For comparison, mention may be made that the S1 ? S0 fluorescence of azulene in solution (*14 000 cm71) is characterised by very low quantum yield (*1074 ¡¾1076) (see Refs 24 and 35) and a decay time, t(S1), estimated at 1.9 (see Ref. 36) or 1 ps (see Ref. 37). The dynamics of isolated azulene molecules in the S1 state has been studied in detail by femtosecond-resolved mass spectrometry in molecular beams.38 The dephasing time of azulene was found to be shorter than 100 fs.With an excess vibrational energy of *2000 cm71, the S1 ? S0 internal conversion time in azulene vapours was estimated at 900100 fs. It was concluded that, despite the wide energy gap (*14 000 cm71), such a high internal conversion rate can be rationalised by considering the conical intersection of the potential energy surfaces (PES) of the S1 and S0 states near the PES minimum of the S1 state, as was suggested earlier.22, 39 Luminescence studies 40 of azulene derivatives allowed the determination of the order in which light emission appears. At low temperatures, an increase in the energy of the S1 level and narrowing of the DE(S2¡¾S1) gap lead to the appearance of the S1 ? S0 fluorescence.At room temperature, the S2 ? S0 fluo- rescence remains the dominant process. At higher temperatures, the usual S1?S0 fluorescence prevails. A feature of azulene derivatives is the low-temperature T1 ? S0 phosphorescence. Interesting results were obtained in studies of the fluorescence of 1,3,5,7-tetra-tert-butylindacene (TTBI), a representative of nonalternant antiaromatic hydrocarbons.41 Similarly to azulene, TTBI has a low intensity transition in the near IR region [E(S1)=9300 cm71] and has an intense, allowed transition in the visible spectral region [E(S2)=18 300 cm71]. The authors of this study 41 succeeded in observing the S2 ? S0 fluorescence of a TTBI solution in cyclohexane (see Table 1).A salient feature of this compound is violation of the cascade character of internal conversion, so that the S2?S0 nonradiative transition appears to be more probable than the S2?S1 transition. In accordance with the proposed exchange-resonance mechanism of nonradiative transitions,15 this violation can be suggested to be due to sub- stantial increase in the probability of the S2 ? S0 radiative474 Table 1. Characteristics of transitions from the second singlet level for a number of organic molecules. Solvent Compound E(S2) /cm71 DE(S27S1) /cm71 cyclohexane Azulene 14 000 14 000 14 300 12 400 9100 28 300 isopentane (at 190 K) 28 300 27 000 718 300 cyclohexane ethanol cyclohexane 1,3-Dichloroazulene 1,3-Dibromoazulene 1,3,5,7-Tetra-tert- butylindacene Xanthione 7970 24 000 7500 11 000 23 400 26 800 fluorinated hydrocarbons trimethylpentane fluorinated hydrocarbons 2,2,3,3-Tetramethyl- indanethione Note. Measurements were mostly carried out by direct single-quantum excitation at room temperature.a The qfl 21 value, b the kr 21 value, c the knr 20 value. transition compared to that of the S2 ? S1 transition and to the larger value of the Franck ± Condon factor for the former tran- sition. According to quantum-mechanical calculations,41 the geo- metric parameters of the TTBI molecule in the S2 and S1 states are close, but they differ appreciably from those for the S0 state. Fluorescent properties of solutions of another representative of antiaromatic compounds, 1,3,5-tri-tert-butylpentalene, have been studied theoretically and experimentally.42 Emission observed in the region near 27 000 cm71 with a quantum yield of 261073 after excitation by light with l=313 nm was assigned to the S3 ? S0 transition.The interpretation of the results obtained in this study appeared to be the subject of a discussion.4, 44 The role of conical intersection of the PES of the S1 and S0 states in quenching of usual fluorescence was pointed out.44 Astudy of fluorescence from the second excited singlet level of [18]-annulene (18A), a representative of alternant hydrocarbons, and its monofluoro derivative (F18A) in 3-methylpentane matrix at 4 Khas been reported.45 Symmetric (D6h) molecular skeleton of 18A comprises 18 carbon atoms while the S1 and S2 states of the molecule are in the regions of 12 500 and 21 400 cm71, respec- tively.Both 18A and F18A were found to be S2 ? S0 fluorescent with rather high yields, the latter being also S1 ? S0 fluorescent (in this case, the fluorescence yield was 100 times lower). Nickel and Hertzberg 46 studied fluorescence of biphenylene (an antiar- omatic hydrocarbon) in 3-methylpentane solution at 130 K. They found that the fluorescence yield from the second excited level is more than two orders of magnitude lower than that of the S1 ? S0 fluorescence, namely, qfl(S1 ? S0)=2.361074. Thioketones represent another class of compounds charac- terised by rather high fluorescence yields for transitions of their molecules from the second excited singlet state.47 ± 50 Huber and Mahaney were the first who reported the emission from the S2 level of aromatic thioketone molecules taking xanthione as an example.47 Shortly, their paper was followed by a series of related studies carried out using a large number of alkyarylthioke- tones.48 ± 50 For thioketone molecules, the orbital forbidden S1(n,p*) / S0 transition is near 15 000 cm71 (the molar extinc- tion coefficient at the absorption band maximum, emax, is rescence.(The latter statement is incorrect since it is known 55, 57 20 mol71 litre cm71), while the allowed S2(p,p*) / S0 transition that for BPT the S1 fluorescence intensity is *0.01 of the is near 23 500 cm71 (emax=16 000 mol71 litre cm71).49, 51 phosphorescence intensity at room temperature.) The relative quantum yields of nonradiative transitions between different excited singlet levels can also be estimated by comparing the absorption and fluorescence excitation spectra from different lower-lying energy levels.The internal conversion and emission quantum yields (qnr and qr, respectively) for the three thioketones are listed in Table 2. The S2 ? S0 fluorescence bands are observed in the violet spectral region (*22 500 cm71) while those of the long-wave- length S1 / S0 transition are observed in the near IR spectral region (*15 000 cm71). It should also be noted that no S1 ? S0 fluorescence was found. On the other hand, T1 ? S0 phosphor- escence is observed even at room temperature, with a quantum yield (in frozen solution) can be as high as 10%.Fluorescence from the second excited state of xanthione and related molecules is strongly quenched upon addition of saturated V L Ermolaev Ref. 1078 knr 21 /s71 kr 20 /s71 qfl 20 t(S2) /ps 2.26107 3.161072 6.9 5.9 6.934 7 35 31 34 34 46103 (see c) 41 1400 6.861075 (see a) 4.06104 (see b) 1400 830 3102.5 5.861072 4.461072 961075 3.66107 5.56107 108 51 175 1.461073 5.7610 86106 49 51 45 880 2.16105 9.7 16107 1.66108 561074 1.461071 and aromatic hydrocarbons, which is indicative of a photochem- ical reaction between the thioketones and hydrocarbons.50 ± 54 The highest S2 ? S0 fluorescence yields were obtained in solutions of perfluoroalkanes (e.g., qfl 20=0.014 for xanthione at room tem- perature, see Table 1).The dipole S2 ? S1 radiative transition in the xanthione molecule is forbidden.49 This substantially decreases the probabilities of nonradiative transitions in accord- ance with the proposed exchange-resonance mechanism of these transitions.15 A solution of 4-H-1-benzopyran-4-thione (BPT) in perfluoro- 1,3-dimethylcyclohexane has been studied in detail.55 ± 57 It was shown that the phosphorescence quantum yield is independent of the frequency of exciting light up to 36 300 cm71 (the S0 ? S4 transition). This fact contradicts the conclusion drawn previ- ously 58, 59 that the phosphorescence yield decreases by 30% upon excitation into the S0?S1 band as compared to direct excitation to the T1 state.On the other hand, a 25% decrease in the S2 ? S0 fluorescence yield under excitation into the S0 ? S4 band was found. This was associated with changes in the rate of internal conversion from the S2 state;57 is it also implied that excitation to upper states leads to an increase in the vibrational energy of the molecule. Intersystem crossing from the HESS to the triplet manifold (namely, to higher excited triplet levels) due to the strong spin-orbit coupling, followed by fast internal conversion to the T1 state also seems not to be improbable in thioketones. Maciejewski et al.60 proposed a method for determination of quantum yields of intramolecular nonradiative transitions, qnr(Sm ? Sm7i) (i=1, 2, ..., m), between different energy levels of complex organic molecules. The method was evaluated taking three aromatic thioketones dissolved in a perfluorinated hydro- carbon and directly excited to the S3 state as examples. To determine the qnr(Sm ? Sm7i) values, the relative fluorescence quantum yields qr(Sm71), qr(Sm72), ... qr(S1) from all lower lying energy levels should be known. The authors of this study used the phosphorescence yield of thioketones, qr(T1), instead of the usual fluorescence yield, qr(S1), assuming that the T1 state is populated exclusively via the S1 state and thioketones exhibit no S1 fluo- The starting point of the study 60 was to rationalise marked violations of the cascade mechanism of internal conversion observed even for molecules with rather wide energy gaps (thisUltrafast nonradiative transitions between higher excited states in organic molecules Table 2.Characteristics of nonradiative and radiative transitions in thioketones dissolved in perfluoro-1,3-dimethylcyclohexane, obtained at room temperature.60 Xanthione-h8 Transition characteristics 4H-1-Benzo- pyran-4-thione Xanthione- 1,3,4,5,6,8-d6 12 900 7 7 (0.74) 0.90 0.72 0.74 (0.82) 0.0 0.11 (*0.0) 0.04 (?) 0.10 (?) 0.15 (*0.0) 0.24 (?) 0.96 0.85 0.88 (1.0) 0.0 0.10 (*0.0) 0.14 (?) 0.042 0.014 0.023 0.078 0.038 0.076 0.0 0.0 0.0 (0.01) 606 178 210 16 000 16 000 16 800 8000 7900 8600 4900 5000 4800 12 900 13 400 qnr(S4?S2) qnr(S3?S2) qnr(S3?S1) qnr(S3?S0) qnr(S2?S1) qnr(S2?S0) qr(S2?S0) qr(T1?S0) qr(S1?S0) t(S2) /ps E(S1) /cm71 DE(S27S1) /cm71 DE(S37S2) /cm71 DE(S37S1) /cm71 Note.The data taken from Ref. 57 are given in parentheses. particularly refers to quantum yields of the S3?S0 and S2?S0 nonradiative transitions, see Table 2). The authors cited suggested that these violations can be accounted for either by large differ- ences between the electronic matrix elements appearing in the expression for the rate constant of nonradiative transition or by conical intersection of the PES when moving over the hypersur- face of the S3 state towards a distorted local minimum. In interpretation proposed in the study,16 the above-mentioned matrix elements represent the rate constants for radiative tran- sitions between corresponding energy levels.Conical intersection of the PES provides a direct transition to the ground state. Difficulties in measuring qnr should be pointed out. First, this is a strong overlap of the bands corresponding to population of different HESS in the absorption spectrum, especially if the intensities of neighbouring bands differ substantially, as is the case of the S3?S0 and S2?S0 transitions in thioketones. Because of this, information on violation of cascade mechanism of non- radiative transitions between excited levels in particular molecules including thioketones should be considered very carefully. For instance, Nickel 61 thoroughly studied nonradiative transitions in BPT and came to conclusion that there is no good reason to assume direct internal conversion from the S2 and S3 states of thioketone molecules to the ground state.{ Fast appearance of an impurity (a photochemical reaction product) in the thioketone solution has also been reported.56 The method proposed by Maciejewski et al.60 allows the obtaining of complete information on the nonradiative transitions between all energy levels only if direct single-quantum excitation of a HESS will give rise to fluorescence from all lower-lying levels.However, this is possible (and only for the S2 state!) for a limited number of molecules (thioketones, several porphyrins, etc.). If direct single-quantum excitation gives rise only to S1 fluorescence, one can only point to deviations from the Vavilov law and Kasha's rule which were repeatedly checked using many molecules.Bright S2(p,p*) ? S0 fluorescence of 1,8-(10,80)-naphthalyl- naphthalene solutions, characterised by qfl=0.18 and t=3 ns was observed,62 despite the narrow gap between S2 and S1 levels [DE(S27S1)=5000 cm71]. However, such a high quantum yield of fluorescence due to transition from the S2 level in combination with the narrow energy gap seems to be unrealistic. Presumably, in this case there was a contribution of impurity or product of a photochemical reaction. Similarly to thioketones and azulene, no S1(n,p*) ? S0 usual fluorescence was observed, only phosphor- escence was found in this case.{ Private communication. 475 Direct single-quantum excitation of porphin derivatives also gives rise to the S2 ? S0 (or S3 ? S0) fluorescence. This fluo- rescence (*23 000 cm71) was first studied by Gouterman et al.63 The fluorescence spectrum was found to be a mirror image of intense Soret absorption band. More detailed studies of this fluorescence have shown 64, 65 that the highest quantum yield (qfl 20=1.261073) is observed for Zn-tetrabenzoporphyrin sol- utions. For metal-free tetrabenzoporphyrin, the absolute quan- tum yield of `blue' fluorescence is nearly an order of magnitude lower (1.861074). It was also found that tetrabenzoporphyrin complexes with metals with partly filled d-shell exhibit no `blue' fluorescence detectable under conditions of single-quantum exci- tation. In a study 66 of Co(II) and Ni(II) complexes with crown ethers and phthalocyanine derivatives it was shown that the S2 ? S1 internal conversion time is shorter than 500 fs.Effect of complex- ation of meso-tetraphenylporphyrins (TPP) with trivalent Y, Lu and Th on the `blue' fluorescence has been studied.67 All the complexes under study, viz., YTPP(acac), LuTPP(acac) and ThTPP(acac), exhibit fluorescence from the S2 level with nearly equal quantum yields (0.001). In contrast to this, the quantum yield of usual fluorescence of these complexes decreases substan- tially in the above-mentioned order, which indicates that inter- system crossing has no effect on quenching of the S2 level of porphyrins.No fluorescence from the second singlet level was found for free base TPP.67, 68 In spite of nearly equal energy gaps DE(S27S1) for metal complexes with octaethylporphyrins (OEP) and TPP, such OEP as H2OEP, AlClOEP and ZnOEP exhibit no S2 ? S0 fluorescence.68 S2 ? S0-Fluorescence of tetra-para-tol- ylporphyrin complexes with Sm(III), Eu(III), Gd(III), Tb(III), Yb(III) and Lu(III) in ethanol has been studied.69 Weak quenching (by factors of *1.5 to 3.0) of fluorescence of paramagnetic complexes with Gd(III), Sm(III) and Tb(III) as well as strong quenching of fluorescence for the complexes with Eu(III) and Yb(III) was explained by the appearance of charge-transfer states lying between the S1 and S2 levels on the energy scale.[Among lanthanide ions, Eu(III) and Yb(III) are known to possess the highest electron affinities.] Due to the short lifetime of the S2 state, `blue' fluorescence of porphyrins is strongly polarised even in solutions of low viscosity (in contrast to usual S1?S0 fluorescence of organic com- pounds).60 This feature was used 70 to separate out the S2?S0 fluorescence. Frequency modulation was performed using polar- isers while signal measurements were carried out with an amplifier with synchronous detection. This experimental procedure allows one to separate out the desired fluorescence against the back- ground of the impurity usual fluorescence, since the impurities emit natural light and their signal is not modulated. The quantum yield of S2 fluorescence of Zn-ethioporphyrin under conditions of single-quantum excitation was 7.561075.An up-conversion study of the S2 fluorescence decay and S1 fluorescence rise of Zn(TPP) complexes in ethanol has been reported.71 It was shown that the decay of the S2 fluorescence at 430 nm and the rise of the S1 fluorescence near 600 nm are characterised by the same time constant of 2.3 ps. A more complex picture is observed for the spectral region between these bands, which indicates a contribu- tion of `hot' luminescence. S2?S0-Fluorescence studies of dye molecules under condi- tions of single-quantum excitation have been carried out taking Brilliant Green72 and Malachite Green 73, 74 dyes as examples. Lipsky et al.75 ± 77 studied single-quantum induced fluores- cence of aromatic molecules from the second and upper levels. Alkylbenzenes were excited with a Hg resonance line at l=184.9 nm (the vacuum UV spectral region).Very weak S2?S0 (or S3?S0) fluorescence bands appeared on the short- wavelength wing of the usual fluorescence spectra. To single them out, quenchers of usual fluorescence (CCl4, CCl3H, cyclo-C7F14) were added to the solutions. In contrast to usual fluorescence,476 Table 3. Characteristics of nonradiative transitions from the second (or upper) singlet levels in a number of organic molecules. Solvent Compound E(S2) /cm71 DE(S27S1) /cm71 7400 23 300 Zn-Tetrabenzoporphyrin 7 7200 Zn-Ethioporphyrin Tetrabenzoporphyrin Brilliant Green p-Xylene DMF "ethanol, 77 K 21 800 41 000 748 800 (see a) isooctane "" Naphthalene 7 65 7 7 (1.661073)7 7 7 63 24 450 7 7 1.861074 6300 6600 (34 400) 7800 (see b) 7 7 (561076) 44 050 (see a) 7 7 (2610± 5) 7000 Pyrene """ 36 000 (see a) Note.Measurements were carried out using direct single-quantum excitation technique at room temperature. The values obtained from measurements in the vapour phase are given in parentheses. a The E(S3) value, b the E(S3±S2) value, c the kr 30 value. almost no quenching of the S2 and S3 fluorescence occurs due to the short lifetimes of these states. Presumably, impurity emission that precludes the observation of fluorescence from HESS of the molecules under study is also quenched simultaneously.Selected characteristics of nonradiative transitions in these molecules are listed in Table 3. High rates of internal conversion and small differences between them in solutions and as vapours, indicative of intramolecular process, are noteworthy. Carotenoids play an important role in many processes occur- ring in living organisms. In particular, they serve as pigment protectors in cells of green plants and purple bacteria against their destruction by singlet oxygen. A systematic study of fluorescence of b-carotenes and their analogues, linear polyenes, has been carried out.78 It was found that the energy gap DE(S17S0) becomes narrower and the yield of usual fluorescence decreases as the polyene chain length increases. The fluorescence yield decreases from 761073 for the polyene with five conjugated bonds down to 461076 for b-carotene (solutions in hexane at room temperature).Among the compounds under study, the longest molecular chains were those of decapreno-b-carotene and dodecapreno-b-carotene with 15 and 19 double bonds, respectively. These compounds exhibited only S2 fluorescence with the quantum yield qfl 20&561075. The S2 state lifetime, t(S2), estimated from qfl20 was *100 fs. Therefore, despite the decrease in the usual fluorescence yield with lengthening of the molecular chain, fluorescence from the second excited level of these molecules remains nearly constant. Weak S2?S0 fluorescence was observed in an anthracene crystal directly excited by a resonance Hg line at l=254 nm.79 The ratio of fluorescence yields from the S2 and S1 states was estimated at *561076 and t(S2) was roughly estimated at &10713 s.However, measurements of fluorescence spectra and quantum yields of fluorescence due to transitions from the second and upper excited singlet states under conditions of direct single- quantum excitation are impossible for most of fluorescent com- plex organic compounds. This is due to the low fluorescence quantum yields, superimposition of impurity luminescence in solvents and in insufficiently purified substances under study and to the Rayleigh and Raman scattering. In conclusion, mention may be made of an interesting example of fluorescence from a triplet rather than singlet higher excited state.80 Irradiation of 9,10-dibromoanthracene and 9-bromoan- thracene solutions in undegassed hexane by a high-power Xe lamp resulted in weak fluorescence near 840 nm for the former and near 865 nm for the latter compound.This phenomenon was suggested to be due to the T2?T1 radiative transition. Fluorescence was excited in the regions of conventional absorption spectra of these V L Ermolaev Ref. knr 21 /s71 kr 20 /s71 qfl 20 t(S2) /ps 3 3.361011 5.26108 1.261073 7.561075 2.561012 0.42 561011 0.1 756107 2.96108 1.061074 2.961075 1.061013 2.561014 3.06109 (see c) 0.004 1.261075 2.261075 7 7 7 70 70 72 75 (1.361075)7 7 (1.561014) 7775 7 7 7 77 7 7 (861013) 76 7 7 7 77 7 7 7 76 561076 compounds.The T2 levels were populated via S1?T2 intersystem crossing. The T2 levels of the molecules under study are near the corresponding S1 levels on the energy scale and the energy gaps, D(T27T1), are ^12 000 cm71. The T2?T1 fluorescence quan- tum yield was estimated at 861077 and the T2 state lifetime, t(T2), was estimated at ^10711 s.80 III. Fluorescence due to transitions from higher excited Sn levels under conditions of stepwise excitation of molecules The appearance of pulsed laser excitation sources allowed studies of ultrafast relaxation processes from HESS using the stepwise excitation technique. Essentially, the method is as follows. Exci- tation of a molecule by the first quantum leads to population of the first (fluorescent) excited singlet state S1.Being in the thermalised S1 state, the molecule absorbs the second quantum, thus populating an upper excited singlet state (Fig. 2) S0+hn1?S1+hn2?Sn (n52). This excitation technique allows easy observation of fluores- cence spectra corresponding to Sn?S0 transitions (n52) with quantum yields of the order of 1076± 1077. It appears to be S4 kr 40 knr 43 T3 S3 kr 30 knr 32 hnII S2 kr 20 T2 knr 21 S1 kisc T1 hnI kfl kph kabs knr S0 Figure 2. An energy level diagram illustrating nonradiative transitions between higher excited singlet states under conditions of two-quantum stepwise excitation and fluorescence due to transitions from the HESS to the ground state of the molecule; hnI and hnII are the first and second excitation quanta, respectively.Ultrafast nonradiative transitions between higher excited states in organic molecules particularly convenient when using tunable lasers generating nanosecond pulses and operating at repetition rates from 5 to 50 Hz.The method has the advantage that stepwise excitation has no effect on the fluorescent impurity molecules in the compound under study and in the solvent, whose absorption spectra are observed in the region of Sn/S0 (n52) transitions. In addition, the spectrum under study is not overlapped with the Raman spectrum since the latter is shifted up to 3000 cm71 with respect to the former. To simplify the interpretation of the results obtained, nanosecond pulses were used with moderate radiation flux levels.(In this case, the use of picosecond and femtosecond laser pulses is less appropriate since it is accompanied by an increase in the contribution of two-photon excitation which affects also the impurity molecules). The best combination of molecular proper- ties for performing stepwise excitation studies includes sufficiently high fluorescence yields, S1 state lifetimes of the order of several nanoseconds and rather large absorption cross-sections for the Sn/S1 transitions.81 The first studies on fluorescence spectra corresponding to transitions from higher excited states to the ground state obtained by stepwise excitation technique date back to the late 1960s and early 1970s.82 ± 86 In particular, Terenin et al.82 observed blue fluorescence of a chlorophyll solution excited by a high power ruby laser.Violet fluorescence of an aluminum phthalocyanine chloride solution was studied under similar conditions.83 It was also found that the action of high-power radiation of a ruby laser on a solution of cyanine dye in methanol leads to the appearance of blue fluorescence with its intensity proportional to the squared exciting radiation intensity.84 These results were subjected to criticism.87, 88 Blue fluorescence was suggested to be due to photochemical decomposition products of the dye. However, blue fluorescence was also observed in a CW excitation study of cryptocyanine.89 The authors of this study found that the blue and red fluorescence excitation spectra match each other and the UV absorption spectrum in the UV spectral region. The quantum yield of the blue fluorescence was 1074.89 The potentialities of the stepwise excitation technique for studying fluorescence from higher excited states of complex molecules have been demonstrated in a series of studies by Lin, Topp et al.90 ± 95 They used a dye laser with an output of 0.1 to 0.5 mJ per*5 ns pulse, pumped by a nitrogen laser operated at a repetition rate of up to 50 Hz.Fluorescence due to transitions from higher excited singlet states was detected using a monochro- mator and a photomultiplier with synchronous detection (the signal was averaged over 100 ± 300 pulses). To eliminate scattered visible fluorescence, a light filter was placed in front of the monochromator.The advantages of consecutive stepwise excita- tion over simultaneous excitation by picosecond pulses with the same energy at the same wavelength were demonstrated taking rubrene as an example. Stepwise excitation experiments can be performed even using a high-power CW laser. For instance, the S3?S0 fluorescence spectrum of tetracene (a solution in methylcyclohexane) was measured using an Ar laser with an output of 0.8 W at 233 K by focusing the laser light on a flow-through cell.96 This is the first observation of fluorescence accompanying the transition from HESS of aromatic hydrocarbon molecules under conditions of stepwise excitation. In studies of Sn?S0 fluorescence spectra of rhodamine B and rhodamine 6G (solutions in ethanol) at 300 K Lin and Topp 91 observed three prominent bands in the region 250 ± 400 nm for each compound, at 275, 320 and 382 nm for rhodamine B and at 270, 310 and 377 nm for rhodamine 6G.The shapes of these band systems were found to be nearly mirror images of the three-band systems corresponding to electronic transitions in the absorption spectra of the dyes. The fluorescence band of rhodamine 6G at 377 nm matches that measured in the earlier 85, 86 and more recent 97 studies. Polarisation spectrum measurements confirmed the presence of three electronic transitions in the observed fluorescence spectrum. It should be noted that positions of the 477 fluorescence bands of the rhodamines at 382 and 377 nm remain virtually unchanged upon variation of the second quantum frequency from 22 000 to 18 000 cm71. Stepwise excitation technique was used to study a large number of aromatic hydrocarbon molecules characterised by several electronic transitions in this spectral region.92 ± 95 It was shown that electronic relaxation of higher excited singlet states of organic molecules occurs until vibrational equilibration is estab- lished after the excitation event.Studies of Sn?S0 transitions in the 3,4,9,10-dibenzopyrene molecule revealed the appearance of a fine vibrational structure in the spectral region adjacent to the effective excitation frequency [nef=n070(S1)+n2] after excitation by the second quantum (540 ± 610 nm). It was found that this fine structure depends on the frequency of the second excitation quantum.The spacing between the observed bands (146060 cm71) is close to that of the principal progression in the S1?S0 fluorescence spectrum (144050 cm71). Topp and Lin 93 suggested that the observed vibrational structure of the spectrum is caused by the anti-Stokes Raman scattering; however, they did not rule out that a phenomenon similar to the Personov effect 98, 99 is observed in this case. Distinctions between the structures of the spectra measured near the excitation level have been found for centrosymmetric molecules and those without a centre of symmetry.94 Under conditions of two-step excitation, selection rules forbid transitions of centrosymmetric molecules in those states from which radiative transitions to the S0 ground state are allowed.Indeed, judging from the sum of the S1/S0 transition frequency and the second absorbed quantum frequency, in this case the fluorescence spec- trum begins `some distance' from its conventional position. Stepwise excitation studies of a large number of fluorescent dyes and aromatic molecules in solutions at room temperature and at 77 K have been carried out.100 ± 106 It was found that intramolecular nonradiative transitions from higher excited sin- glet levels in dyes and most complex aromatic compounds in solutions occur in a cascade manner, i.e., consecutively involving all the lower-lying electronic levels. This conclusion was drawn based on comparison of the shape of fluorescence spectra of organic molecules excited by the second quantum with varied frequency in states with different energies.The Sn?S0 (n52) fluorescence spectra of Magdale Red dye (a solution in ethanol) at 77 K are shown in Fig. 3 while those of 1,2-benzanthracene (a solution in ethanol) at 293 K are presented in Fig. 4. The spectra shown in Figs 3 and 4 exhibit prominent bands that are mirror images of corresponding Sn/S0 absorption bands. In addition, the intensity ratio of the Sn?S0 and Sn7k?S0 bands (k<n) remains constant as the energy of the second quantum varies. The ratio of the areas under the bands corresponding to different Sn?S0 transitions characterises the relative quantum yield of a given fluorescence transition.Given the absolute quantum yield of a particular intramolecular transition, it is possible to determine the absolute quantum yields for other transitions. The corresponding radiative rate constant can be found from the area under that absorption band which is the mirror image of the fluorescence band of the given transition. Then, one can estimate the Sn state lifetimes t(Sn)=tr(Sn/S0) qfl(Sn?S0). Here, tr (Sn/S0) is the inverse of the rate constant for the Sn?S0 radiative transition and qfl(Sn?S0) is the fluorescence quantum yield. The fluorescence quantum yields and decay times thus obtained are listed in Table 4. The results of analysis of the spectra presented in Figs 3 and 4 and those reported elsewhere 97, 100 ± 106 are sufficient to allow the following conclusions.First, stepwise excitation of a molecule to an Sn (n52) state leads to the appearance of the spectral bands corresponding to the Sn?S0, Sn71?S0, Sn72?S0 , etc. transitions in the fluores- cence spectrum of the molecule.478 1074 e /mol71 litre cm71 4 P 1 0.2 2 2 0 70.2 0 40 Figure 3. Magdale Red spectra (a solution in ethanol). The Sn/S0 absorption spectrum at 77 K (1); polarisation spectrum of S1?S0 fluorescence excitation as function of excitation frequency at 293 K (a solution in glycerol) (2), see also a separate P scale; S1?S0 fluorescence spectrum at 77 K (3) and Sn?S0 (n52) fluorescence spectra (4 ± 6) obtained under conditions of stepwise excitation from the S1 state at lexc=1064, 550 and 525 nm, respectively.Double vertical arrows denote the energy levels of Sn electronic states on the frequency scale. Single arrows denote the energy levels achieved under conditions of stepwise excitation. N is the number of absorbed quanta. Second, for most of molecules (except for those characterised by strongly different equilibrium nuclear configurations in the S1 and S0 states) each band in the Sn/S0 absorption spectrum has a nearly mirror image in the Sn?S0 fluorescence spectrum. Third, if excitation of a molecule by the second quantum leads to population of an upper Sk (k>n) state and if the Sn levels are populated by Sk?Sk71?Sn electronic relaxation, the shape of fluorescence spectra and, in particular, the shape of particular bands corresponding to Sn?S0 transitions and the band intensity ratios are in most cases independent of the energy of the second quantum. Fourth, distortion of the fluorescence spectra from higher excited states is observed only if the total energy of the S1?S0 transition and the second excitation quantum is larger than the 1074 e /mol71 litre cm71 86 4 4 S6/S0 2 6 3 510 520 44 Figure 4.Spectra of 1,2-benzanthracene (a solution in ethanol) at 293 K.72 The Sn/S0 absorption (1) and S1?S0 fluorescence (2) spectra. The Sn?S0 (n52) fluorescence spectra obtained under conditions of stepwise excitation from the S1 state at lexc=510, 520, 586 and 340 nm, respectively (3 ± 6); enlarged representation of the Sn/S0 absorption spectrum(7) (8 times along the ordinate axis).The spectral width of the instrumental response function was 200 cm71. qN/qn 3 S3 S2 S1 525 30 40 1073 n /cm71 ionization energy of the molecule under study in solution (see Figs 3 and 4). To obtain additional experimental verification of the assump- tion of cascade character of intramolecular electronic relaxation, the absolute quantum yields (qfl n0) of Sn?S0 fluorescence were measured at different energies of the second excitation quantum. By the quantum yield of fluorescence due to intramolecular transitions from higher excited electronic states under conditions of stepwise excitation is meant the ratio of the number of the quanta emitted in the form of Sn?S0 fluorescence to the number of molecules excited to the Sn state.Determination of qfl n0 requires knowledge of the number of absorbed (the Sn/S1 transition) and emitted (Sn?S0 transition) quanta. The former parameter is complicated to calculate. Previously,97, 100 ± 106 it was estimated 1 5 S4?S0 S4/S0 586 32 36 40 V L Ermolaev qN/qn 6 5 S3 S2 1064 4 550 30 1073 n /cm71 qN/qn 2 7 S2/S0 28 1073 n /cm71Ultrafast nonradiative transitions between higher excited states in organic molecules Table 4. Characteristics of transitions from the HESS of organic molecules (fluorescence spectra and yields were measured using stepwise excitation technique). Compound Solvent�ethanol, 77 K Brilliant Green Rhodamine B Magdale Red Solvent�ethanol, 293 K Rhodamine 6G Pyrene 1,2-Benzanthracene 9,10-Di-n-propylanthracene a-Naphthylphenyloxazole 3,4.9,10-Dibenzopyrene Note.The values estimated from the linewidths in structured fluorescence spectra and unreliable results due to difficulties of the determination of the emission rate constant owing to changes in the molecular configuration in the excited state are given in parentheses. from changes in the intensity of transmitted laser radiation with the frequency n2 upon turning on a laser with the frequency n1 and from the amplitude of additional opto-acoustical signal which appeared upon simultaneous illumination of the object under study by two laser pulses.107 The fluorescence quantum yield was also estimated 108 from the following relationship Ifl n0 qfl n0 à Ifl 10s1nIex 2tr1 where Ifl n0 and Ifl 10 are the integrated intensities of the Sn and S1 fluorescence, respectively; s1n is the absorption cross-section for the Sn/S1 transition, Iex 2 is the power density of exciting radiation with the frequency n2 and tr1 is the radiative lifetime of the S1 state.Table 5 lists the results obtained for a-naphthylphenyloxa- zole, pyrene and 1,2-benzanthracene. As can be seen, the qfl n0 values are independent of energy of the second quantum within the experimental error. This is a strong argument in favour of the cascade mechanism of electronic relaxation of HESS. Let us analyse the spectra of fluorescence accompanying transitions from the HESS of most of the molecules studied, obtained under conditions of stepwise excitation.The following picture is observed in the region adjacent to the frequency corresponding to the sum of the energies of the S1 state and the second excitation quantum. The Sn? S0 fluorescence spectra (Sn is the state directly populated by absorption of the second quantum) exhibit a more pronounced vibrational structure com- pared to the Sn/S0 absorption spectra of the same molecules obtained under the same conditions and to the fluorescence n DEn0 /cm71 DEn,n71 /cm71 6300 4500 5100 1800 3100 4100 6000 2200 21 800 30 600 22 800 24 600 27 700 31 800 23 300 25 500 24234523 56007 7 7 7 2750 6000 4700 *3500 1100 4000 4500 <3000 7800 5700 2300 4600 4000 3500 3900 24 300 25 650 28 400 36 300 41 000 *44 500 34 500 38 500 43 000 37 700 35 300 41 000 25 500 30 100 34 100 37 600 41 500 23445645622323456 , 479 Ref.kr n0 /s71 qfl n0 tn /ps knr(Sn?Sn71) /s71 97 97 2 97 0.024 0.3 0.28 0.25 0.03 0.18 0.33 561011 4.261013 3.361012 3.661012 4.061012 3.361013 5.661012 3.061012 56107 8.36107 16107 1.46107 46107 1.46108 8.36106 1.36107 161074 261076 361076 461076 161075 461076 1.561076 4.561076 0.2 261076 5.061012 16107 97 97 56107 46108 56108 *36108 6.76108 56108 76108 1.46108 107 104 161075 761076 1.861075 *661076 461075 1.361075 1.361075 361075 461076 5.661076 95 97 70.2 0.02 0.035 *0.022 0.06 0.026 0.018 0.023 (0.024) (0.010) 0.2 >0.1 0.0870.04 5.061012 5.061013 2.961013 *4.561013 1.761013 3.861013 5.661013 4.361013 (1.66108) 7 (6.76108) 7 7 7 7 7 7 7 7 7 7 7 7 7 2.26109 7 561075 Table 5.Absolute Sn?S0 fluorescence quantum yields (qfl n0) measured using stepwise excitation technique.97, 104 Compound 106qfl n0 n2 /cm71 n1 /cm71 Energy level to be excited (see a) (see b) a-Naphthyl- 27500 phenyl- oxazole Pyrene 26000 5.61.5 4.91.5 5.81.5 3.51.0 4.00.8 3.50.8 4.20.8 3.01.0 25900 1,2-Benz- anthracene 16 950 18 200 19 250 16 950 18 200 19 250 16 950 18 200 19 250 4.01.5 5.0358 358 4.51.5 6.02.0 777 S3 S3 S4 S4 S5 S6 S5 S6 S6 358 Note.Measurements were carried out in ethanol solutions at 293 K. a Given are the qfl 206106 values (nmax&33 200 cm71) for a-naphthyloxa- zole and qfl 406106 values (nmax&34 500 cm71) for pyrene and 1,2- benzanthracene; b given are the qfl 306106 values (nmax&39 300 cm71) for a-naphthyloxazole and qfl 506106 values (nmax&41 500 cm71) for pyrene. spectra from the HESS populated via relaxation from higher excited levels (see Figs 3 and 4). Positions of all the fluorescence bands observed depend on the frequency of the second quantum. On the other hand, neither the structure nor the relative intensities of the fluorescence bands corresponding to transitions from the lower-lying HESS popu- lated via relaxation remain unchanged (see curves 4 ± 6 in Fig.3 and curves 4 and 5 in Fig. 4).480 If the frequency of the second excitation quantum varies within Dn2<1500 cm71, the entire spectral region in question (it will be remembered that this is a region of the fluorescence spectrum) is shifted simultaneously with the position of the level to be populated. An increase in Dn2 leads to substantial redistrib- ution of vibronic band intensities. Further increase in the fre- quency n2 (by several thousands of cm71) causes the appearance of spectral bands corresponding to emission of nearly the same intensity; however, the new bands exhibit no pronounced vibra- tional structure. The appearance of structured Sn?S0 fluorescence spectra can probably be rationalised by the effect of selective excitation of molecules with a particular configuration into a broad Sn/S0 absorption band by spectrally narrow (*0.5 cm71) radiation with the frequency n2.Due to ultrashort lifetimes of the Sn states (10 ± 100 fs) of those molecules which exhibit structured spectra, inhomogeneous line broadening in the Sn/S1 absorption spec- trum can be considered as manifestation of specific molecular conformations due to low-frequency, out-of-plane vibrations of the aromatic rings. This explanation is also supported by the fact that Brilliant Green and Zn-tetrabenzoporphyrin exhibit no structured spectra upon excitation of the relatively long-lived (*1 ± 3 ps) S2 state.On the other hand, stepwise excitation of Brilliant Green by the second quantum (l2=500 ± 620 nm) and population of the S5 level (t5424 fs) leads to the appearance of a structured spectrum. This is consistent with the assumption that the structured spectra are observed upon transitions from electronically and vibration- ally unrelaxed states, while the excited states with the picosecond lifetimes undergo at least partial vibrational relaxation. It was assumed 97, 103 that the effect in question (the appear- ance of structured spectra) is to some extent similar to the Personov effect, i.e., the appearance of the S1?S0 fluorescence spectra with well-resolved fine structure under the action of selective laser excitation of organic molecules in solid solutions at 4 K (see Refs 98 and 109).There are some distinctions between these effects. First, a structured spectrum appears in the case where one electronic transition (Sn/S1) is selectively excited while another transition (Sn?S0) is used for the detection of structured emission. Conven- tional Personov effect is not observed in this situation since the two transitions are uncorrelated. Second, structured spectra are observed in liquid solutions without cooling them down to the liquid helium temperature (though liquid solutions can be consid- ered as solid ones during a time of the order of tens of femto- seconds). Third, corresponding vibronic levels are noticeably broadened (up to hundreds of cm71) due to the short lifetimes of the excited states, whereas in the case of the Personov effect the linewidths are equal to hundredth and thousandth fractions of cm71. Finally, structured spectra appeared under conditions of stepwise excitation are observed only for those molecules for which the 070 band of the S1 ? S0 transition is only slightly inhomogeneously broadened.The appearance of a structured fluorescence spectrum under conditions of stepwise excitation was explained in another man- ner.110 The authors of this study assumed that the observed fluorescence is due to anti-Stokes resonant vibronic Raman scattering (ARVRS). An energy level diagram illustrating the proposed 110 formation mechanism of structured emission spec- trum is shown in Fig.5. ARVRS occurs if excitation of a molecule to the S1 state after absorption of the first quantum is followed by absorption of the second excitation quantum with the frequency n2. As a result, transition of the molecule to a vibrational electronic ground-state level is accompanied by scattering of a quantum corresponding to the frequency n2+n0± 0 (S17S0)7 Snini. Later, the authors of other studies 108, 111, 112 supported this explanation and pointed to the similarity of the observed spectrum and the usual fluorescence spectrum shifted towards the high- frequency region. V L Ermolaev S3 hnII hnII hnII S2 S1 hn hn hn hnI S0 Figure 5. An energy level diagram illustrating the ARVRS mechanism.Dashed lines denote virtual energy levels involved in the Raman scatter- ing. Vertical lines corresponding to ARVRS are denoted as hn. Nevertheless, we dissent from the proposed 110 explanation for the experimental observations. Let us adduce some arguments in favour of the interpretation of structured emission appearing under conditions of stepwise excitation of organic molecules as fluorescence. If an excited level is populated via relaxation from upper levels rather than by direct excitation, the integrated intensity of the structured spectrum is nearly the same as the emission intensity in the same spectral region.97 The bands in the structured spectrum are much broader than those in the S1?S0 fluorescence spectrum. On the other hand, the ARVRS mechanism requires that the width of the short-wave- length band in the structured spectrum obtained under conditions of stepwise excitation and the width of the 0 ± 0 band in the S1?S0 fluorescence spectrum be equal.If a molecule in the S1 state is excited by a second quantum and if the energy of this quantum is sufficiently high for the molecule to undergo photoionisation, the fluorescence yield from higher excited states should decrease dramatically. It is just the situation that is observed in the experiments. For instance, in the case of rhodamine B the fluorescence yield from HESS decreases by more than an order of magnitude as the frequency n2 is varied from 19 000 cm71 to n2>32 000 cm71.Assuming that structured emission is due to ARVRS, there is no grounds for such a large decrease in the emission intensity, since the process involves only virtual high-lying states. Nonradiative energy transfer from a higher excited state of 1,2-benzanthracene, which was stepwise excited to the region of structured spectrum, to benzene and xylene molecules has been observed.112 This transition was interpreted as ARVRS. However, the authors of this study noted that nonradiative energy transfer from the virtual levels involved in ARVRS is rather strange. The problem of `hot' fluorescence bands in emission spectra due to transitions from HESS has been considered.113 If a molecule excited to an upper singlet state by the second quantum relaxes to a lower-lying HESS, a substantial excess vibrational energy is stored in the latter state. Hence, it is logical to expect the appearance of `hot' fluorescence bands in the emission spectrum from this state.If the HESS lifetimes are shorter than the time of vibrational energy redistribution within a complex molecule, the intensities of `hot' fluorescence bands are determined by the population probability of the states in question via internal conversion from upper electronic states and by the rate of the initial stage of energy exchange between the optically active and inactive vibrations. At first glance, the intensities of `hot' bands inUltrafast nonradiative transitions between higher excited states in organic molecules the emission spectrum must be strongly dependent on the excess vibrational energy in the emitting Sn (n>1) state and, hence, on the energy of the second quantum.However, as can be seen in Fig. 4, the S4?S0 fluorescence bandshape of 1,2-benzanthracene depends only slightly on the second-quantum energy when it is varied from 17 100 to 28 200 cm71 (for the S4 state, this corre- sponds to an excess vibrational energy lying between 8400 and 19 500 cm71). Considering a `hot' fluorescence band lying 1400 cm71 above the 0 ¡¾ 0 band of the S4?S0 transition, Lyu- bimtsev 113 has come to the conclusion that cascade relaxation passes through an optically active vibrational state (1400 cm71) with a unit probability. IV. Lifetimes of higher excited Sn states estimated from fluorescence quantum yields and absorption integrals Determination of a HESS lifetime requires knowledge of the ratio of absolute quantum yield of Sn?S0 fluorescence to the number of absorbed second quanta and the radiative rate constant for the same transition.The latter parameter can be estimated using the absorption integral for the band corresponding to the Sn/S0 transition.114, 115 tn=qfl n0 . kr n0 A number of studies concerning this subject have been reported.97, 100, 106 It was found that the HESS lifetimes of organic molecules in solutions vary from several picoseconds for the S2 states of certain molecules to 10 ¡¾ 20 fs for the Sn states, n=3¡¾6. The tn values of the order of tens of femtoseconds mean that the electronic relaxation rates of these states are higher than the intramolecular vibrational relaxation rates which were estimated at hundreds of femtoseconds.19, 116, 117 Elsaesser and Kaiser 20 proposed to distinguish between two types of vibrational excita- tion relaxation, viz., fast relaxation for the energy redistribution between isoenergetic vibrational states and slow relaxation corre- sponding to the relaxation to lower-lying energy levels.Transfer of excess vibrational energy from the molecule under study to the environment (solvent molecules) proceeds at lower rates and is characterised by the times of the order of several tens of pico- seconds.19, 116 Therefore, the cascade relaxation of the HESS of organic molecules (except for the S2 state of a certain molecules) can be considered as occurring with conservation of the energy supplied to the molecules. Variation of the excess vibrational energy was found to have a slight effect on the band position and shape in the Sn?S0 fluorescence spectra, though in some cases this energy was varied by 10 000 cm71 and more.Electronic relaxation times of the HESS of aromatic and dye molecules are nearly equal to the period of optically active vibrations (n=1200 ¡¾ 1600 cm71) or even shorter. To verify such interesting results, attempts have been under- taken at estimating the HESS lifetimes using other methods. For instance, the tn values were found from the line broadening in the structured spectra appearing upon selective excitation by the second quantum of the HESS lying near the initially excited state.97, 104 The spectral widths of vibronic bands (the first vibronic band) in such spectra were compared with the spectral width of the 0 ¡¾ 0 band in the usual fluorescence spectrum.In those cases where the 0 ¡¾ 0 bands were only slightly inhomogeneously broadened (pyrene, 1,2-benzanthracene, phenanthrene, T=77 K), the structured spectra were additionally broadened. Assuming that the broadening of the upper (Sn) and lower (S1, S0) levels is uncorrelated, the broadening Dn 0n0 due to the short lifetime of the Sn state (i.e., the homogeneous width) was estimated using the formula Dn0n0=qAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA aDnn0OSn ¢§ S0Ua2 ¢§ aDn10OS1 ¢§ S0Ua2, where Dnn0 and Dn10 are the halfwidths of the corresponding bands.The broadening calculated using this formula (300 cm71 for the S6 states of pyrene and 1,2-benzanthracene) corresponds to the lifetime of these states, t6=18 fs. These t6 values are in reasonable agreement with the estimates obtained for pyrene (t6=15 ¡¾ 30 fs) and 1,2-benzanthracene (t6=18 fs) from the absolute quantum yields of fluorescence. The rate constants for nonradiative transitions between the HESS are calculated using the absolute quantum yields of the Sn?S0 fluorescence and the radiative rate constants estimated from the band integral in the Sn/S0 absorption spectrum. It should be noted that the results of calculations of the absorption band integral refer to the molecule which is in the S0 state and has its equilibrium configuration.However, the molecule excited from the S1 state to a HESS after absorption of the second quantum retains the configuration of the former state. If the equilibrium molecular configurations in the S1 and S0 states differ only slightly, these estimates of tn are justified. However, in some cases (e.g., p-terphenyl or 2,5-diphenyloxazole) the molecular configuration in the excited (S1) state is substantially different from the ground-state (S0) configuration. This manifests itself in different patterns of their absorption and usual fluorescence spectra. According to X-ray diffraction data, the ground-state structure of the p-terphenyl molecule is nonplanar.118 On the other hand, quantum-chemical calculations 119 showed that in the case of fluorescence emission this molecule has a nearly planar configuration.Clearly, the oscillator strengths of electronic tran- sitions and the corresponding energy levels can vary over a wide range upon changes in the molecule geometry (Fig. 6). S3 S2 hnII S1 kr 30 hnI S0 Figure 6. An energy level diagram illustrating two-quantum stepwise excitation of a molecule in the case of variations of the molecular configuration or solvate environment during the lifetime of the fluorescent (S1) state. The energy levels belonging to the triplet manifold are not shown. V. Delayed fluorescence from higher excited Sn states due to triplet ¡¾ triplet annihilation As has been shown experimentally,120 annihilation of two mole- cules being in the lowest triplet state (the so-called triplet ¡¾ triplet annihilation) can lead to generation of both the first and upper excited singlet (and triplet) states of the molecules: T1+T1?Sn+S0 , n51. S03 knr 32 S02 knr 21 S01 kr 10 hnfl S00 481482 This process leads to the appearance of not only the usual S1?S0 fluorescence, but also the Sn?S0 (n>2) delayed fluo- rescence due to triplet ± triplet annihilation. This feature was used to detect the phenomenon in question.The first successful experi- ments were carried out taking 1,2-benzanthracene and fluoran- thene dissolved in paraffin oil as examples.A schematical energy level diagram of the interacting molecules is shown in Fig. 7. Triplet ± triplet annihilation studies of a number of aromatic compounds, e.g., naphthalene (S2), phenanthrene (S3), tripheny- lene (S3), 1,2-benzanthracene (S3), pyrene (S2), chrysene (S3), fluoranthene (S4) andN-methylcarbazole (S4, S2?), anthraquinone (S3, S4), xanthone (S3) have been carried out.121 ± 124 (The singlet state generated after triplet ± triplet annihilation is given in paren- theses.) Triplet ± triplet annihilation induced fluorescence from the HESS is observed using a phosphoroscope. This allows effective suppression of the stray exciting light, though the usual fluores- cence light [it is nearly four orders of magnitude more intense than the Sn?S0 (n51) fluorescence] remains.In contrast to the stepwise excitation method, triplet ± triplet annihilation technique does not allow variation of the energy of the second quantum over a wide range. However, in the former case it is difficult to observe the region near the short-wavelength limit of the usual fluores- cence in the fluorescence spectra from HESS, since this spectral region is, as a rule, used for excitation of the molecule under study to the S1 state by the first quantum. As a result, the stray light precludes measurements of the weak Sn?S0 (n51) fluorescence spectrum. Triplet ± triplet annihilation techniques provides much better conditions for the observation of this spectral region (Fig. 8). Delayed fluorescence spectra of 1,2-benzanthracene, fluoran- thene, pyrene and chrysene solutions in methylcyclohexane at 193 Kwere obtained by triplet ± triplet annihilation technique and corrected for the quantum spectral sensitivity of the experimental setup.121 All the above-mentioned compounds exhibit clearly seen S4 S3 S2 S1 kisc T1 kr 40 hnabs kfl kr 30 kph S0 kt Figure 7.Schematic representation of energy level diagram for non- radiative transitions between higher excited singlet states of a molecule (fluorescence corresponding to transitions from the HESS to the ground state due to triplet ± triplet annihilation). Vertical solid and dashed lines denote radiative and intramolecular nonradiative transitions, respectively. The system of dashed lines linked by upward-looking brackets (see the bottom of the diagram) corresponds to the interaction between two triplet molecules accompanied by energy transfer from one to the other molecule and excitation of the latter to a higher excited singlet state (kt). The energy level diagram for the system of higher excited triplet states of the molecules and the scheme illustrating their triplet ± triplet annihilation induced excitation are not shown.S4 knr 43 S3 knr 32 S2 knr 21 S1 T1 hnabs S0 e /mol71 litre cm71 T1 107 105 103 101log II0 0.8 0.4 0.0 17 Figure 8. Spectrum of delayed fluorescence due to triplet ± triplet anni- hilation (1) and absorption (2) spectrum of 1,2-benzanthracene (a 561075 mol litre71 solution in methylcyclohexane) measured at lexc=365 ± 366 nm (T=193 K) (I0 is the initial intensity).The `as-measured' (I ) and expanded (9150 times along the ordinate axis, II ) representation of the spectra.121 Sn?S0 fluorescence spectra. In some instances, the Sn?S0 fluorescence band appears to be a mirror image of the Sn/S0 absorption band, as in the stepwise excitation studies. Sn?S0-Fluorescence spectra of N-methylcarbazole (NMC) dissolved in heptane and toluene (Tol) have been studied.123 Triplet ± triplet annihilation was found to induce both fluores- cence from HESS of the NMC molecules and sensitised fluores- cence of toluene, which is due to the nonradiative energy transfer from the HESS of N-methylcarbazole to the toluene molecule.This process can be schematically described as follows: T1(NMC)+T1(NMC )?Sn(NMC)+S0(NMC), Sn(NMC)?S0( NMC)+hnfl , Sn(NMC)+S0(Tol)?S0(NMC)+S1(Tol), S1(Tol)?S0(Tol)+hnsens fl . Triplet ± triplet annihilation of Cd-tetrabenzoporphyrin mol- ecules also induces delayed fluorescence from the S2 state.125 In this case, the triplet state was sensitised by triplet ± triplet energy transfer from Pd-tetraphenylporphyrin molecules which exhibit only weak S2 fluorescence [qfl(S2?S0)<1075]. Hetero-triplet ± triplet annihilation of azulene and fluoran- thene solution in isopentane leads to the appearance of fluores- cence from the second excited singlet level of azulene.126, 127 These studies allowed the determination of the lifetime of the lowest triplet state for azulene (48 ms) and the quantum yields of intersystem crossing for this molecule, q(S1?T1)=461076 and q(S2?T1)=461072.The rate constants for intersystem crossing, corrected for the lifetimes of azulene molecule in the S1 and S2 states (see Table 1), were found to be k(S1?T1)= 26106 s71 and k(S2?T1)=2.86107 s71. Delayed S2?S0 flu- orescence of 4,6,8-trimethylazulene due to homo-triplet ± triplet annihilation was observed 128 at 27 680 cm71 using sensitisation of triplet states by adding phenazine to the solution (the T1 energy is 15 630 cm71). Hetero-triplet ± triplet annihilation of anthracene and xanthone molecules was used 129 to study fluorescence from the S6 (1Bá3u) level of anthracene, perdeuterated anthracene and a 12 S1 S2 b I 25 21 V L Ermolaev qN=qn S4 S3 2T1 II 1073 n /cm71 29Ultrafast nonradiative transitions between higher excited states in organic molecules 9,10-dimethylanthracene in their solutions in 1,1,2-trichlorotri- fluoroethane.The intensity ratio of the S6?S0 fluorescence to the S1?S0 usual fluorescence was found to be*1076. Triplet ± trip- let annihilation of biphenylene (an antiaromatic compound) sensitised with N-methylcarbazole in an isopentane ± cyclopen- tane mixture at 178 K allowed the detection 130 of very weak S2?S0 fluorescence of this compound, which was not previously observed using direct excitation techniques (see Section II). This indicates a rather high sensitivity of the triplet ± triplet annihila- tion method.A method was proposed for increasing the intensity ratio of the fluorescence from the HESS and usual fluorescence.123 It involves the addition of quenchers of the S1 state to the solution under study. The potentialities of this method were demonstrated taking delayed fluorescence due to triplet ± triplet annihilation of pyrene in the presence of efficient quenchers (diethylaniline and triethylamine) as an example. (The additives should be efficient quenchers of the lowest excited singlet state rather than the triplet state and be transparent to delayed Sn?S0 fluorescence due to triplet ± triplet annihilation. They should also not enter into photochemical reactions with the compound under study.) This experimental technique was used in precise measurements of the S2?S0 fluorescence spectrum of pyrene.123 Interesting results were obtained 56 in studies of annihilation delayed fluorescence of 4H-1-benzopyran-4-thione (BPT) in per- fluoro-1,3-dimethylcyclohexane.The experimentally determined probability of generation of the S2 state by triplet ± triplet annihi- lation was found to be 0.03. This is much lower than the statistical coefficient for the creation of a singlet state by triplet ± triplet annihilation (0.25), obtained assuming that the quintet state is inaccessible from `energy' considerations. The authors of this study mentioned that the probability obtained (0.03) is of the same order of magnitude as the known parameters for anthracene (0.08) and phenanthrene (0.02).The observed generation of the S2 state of BPT by triplet ± triplet annihilation was interpreted 56 as a special case of FoÈ rster's (exchange-resonance) energy transfer Table 6. Characteristics of transitions from the HESS of organic molecules. n Compound DEn0 /cm71 DEn, n71 /cm71 qfl n0 Solvent�isopentane ± methylpentane mixture, T=4 K 5000 Phenanthrene 7 7 1.361013 33 900 7 7 7 7 7 22 Solvent�methylcyclohexane, T =190 K 2600 Pyrene 3.561075 29 600 7 7 7 7 2.561013 36 300 6000 224 761076 Solvent�methylcyclohexane, T =193 K 5580 3700 1,2-Benzanthracene 3 Fluoranthene 5.561076 1.161075 6000 Chrysene 1.961075 33 400 34 700 7 7 7 7 *2.061013 36 800 7 7 7 7 *5.961013 4433 Solvent�1,1,2-trichloro-1,2,2-trifluoroethane, T=243 K Anthraquinone 7000 3500 5700 36 500 40 000 35 000 1075 (see c) 7 7 7 1076 (see c) Xanthone 343 Notes.Decay times and rate constants of nonradiative transitions were obtained by triplet ± triplet annihilation technique. The values estimated from the linewidths in structured spectra are given in parentheses. a At 4 K, b in a supersonic jet, c experimental conditions: c=1075 mol litre71, deoxygenated. 483 mechanism.{ However, the lack of data on the T1?Sn absorption spectra of BPT precludes calculations of S2 generation. Note that exchange-resonance mechanism of energy transfer between T1 states should be common to processes of generation of both Sn and Tn states and that in the case of similar intensities of the spin- allowed and spin-forbidden transitions in the BPT molecule the T1?Sn spectra are indistinguishable from the T1?Tn ones.The HESS lifetimes were estimated from the broadening of the 0 ± 0 vibronic bands in the absorption spectra compared to the 0 ± 0 band of the long-wavelength transition. Another method used for estimating the tn values involved the determination of (i) the quantum yield ratios of Sn?S0 fluorand delayed S1?S0 fluorescence under the same conditions and (ii) the rate constants for radiative Sn?S0 transitions from the absorption integrals for the corresponding Sn/S0 transitions. The estimates obtained are listed in Table 6. Triplet ± triplet annihilation technique has a drawback con- sisting in uncertainty in the identification of the annihilation- generated HESS, since the excitation probability distribution of these states is determined by the overlap integral of the T1?S0 emission and Sn/T1 absorption spectra normalised to a unit band area.Unfortunately, if the phosphorescence spectra are, as a rule, known, almost no information on the Sn/T1 absorption spectra is available. In the case of triplet ± triplet annihilation, intersystem crossing becomes allowed in the Sn/T1 transitions; however, the effect of the Franck ± Condon factors persists.131 Katoh and Kotani 132 studied fluorescence from the HESS of an anthracene crystal upon fusion of two singlet (S1) excitons.They also investigated 133 Sn?S0 fluorescence of an anthracene crystal using a two-step laser excitation technique. { Eisenberger et al.56 pointed out that triplet ± triplet annihilation of aromatic hydrocarbon molecules follows the exchange-resonance energy transfer mechanism. They used a much higher rate constant for the T1 ? S0 transition in BPT. Ref. kr n0 /s71 tn /ps knr (Sn?Sn71) /s71 144 146 (0.08)a (0.5)b 1.161013 3.86108 123 123 123 0.09 (50.04) 0.02 5.061013 46108 7.761012 3.861013 4.26107 46108 109 5.361013 121 121 121 121 121 0.13 0.026 (*0.05) 0.019 (*0.017) 7 7 7 124 7 124 7 124 7 7484 VI. Lifetimes of higher excited Sn states of molecules estimated from homogeneous broadening of absorption spectra The method of estimation of the HESS lifetimes from the data on homogeneous line broadening in absorption spectra is based on the uncertainty relation, which relates the lifetime of an excited electronic or vibronic state of a molecule under study to the homogeneous broadening of the corresponding energy level , tn=2pcDn 1 1=2 where, tn is the lifetime of the energy level (in seconds), c is the velocity of light (in cm s71) and Dn1/2 is the halfwidth of the energy level (in cm71).At first glance, tn values can be simply determined using this formula after measuring any absorption or emission line corresponding to the transition between the ground and excited states. Indeed, since the ground-state lifetime is infinitely long, this state contributes negligibly to the homogeneous broadening of the spectrum.However, there is also inhomogeneous broadening, which much exceeds the homogeneous broadening in the case of solutions, and the broadening due to phase relaxation.134 Subsequently several experimental methods for elimination or strong suppression of inhomogeneous broadening were devel- oped.135 ± 142 These are as follows: � low-temperature measurements of absorption and lumi- nescence spectra of crystals; � measurements of impurity spectra from crystals in those cases where the impurity molecule under study can be easily incorporated into the unit cell of the matrix crystal (e.g., a naphthalene molecule in durene crystal); �spectral line narrowing in Shpolskii matrices; � obtaining of narrow fluorescence spectra using the Per- sonov effect upon excitation to the region of the 0 ± 0 electronic transition by monochromatic radiation; � `hole burning' in absorption spectra using photochemical reactions or another mode-selective technique for removal of Table 7.HESS lifetimes of organic molecules estimated from vibronic bandwidths. T /K Matrix Compound n /cm71 Sn/S0 4.2 n-hexane Anthracene S1/S0 Coronene 4.2 4.2 n-heptane " S1/S0 S2/S0 26 234 " 77 S2 /S0 39 140 " 77 S3 /S0 45 080 23 450 28 440 " 77 S3 /S0 32 870 " 7 5/S0 7S 46 600 n-hexane 1,12-Benzoperylene 24 636 S1/S0 4.2 " 77 S2 /S0 25 800 " 77 S5 /S0 44 580 " 7 6/S0 7S 47 100 4 23 600 S2/S0 Zn-Tetraphenyl- porphyrin poly(methyl methacrylate) poly(vinyl butyral) 4 see b " 23 480 25 120 ? 20 S2/S0 S2/S0 S2/S0 16 500 23 040 19 884 4.2 4.2 7 THF±Et2O (1 : 1) mixture see b S2/S0 S3/S0 S2/S0 Zn-Tetratolyl- tetrabenzo- porphyrin Tetrabenzopor- phyrin derivative Free porphyrin base a From fluorescence yields, b vapour in a supersonic jet.V L Ermolaev `inhomogeneously broadened' molecules with specified electronic energy; � obtaining of extremely narrow absorption or fluorescence spectra in free supersonic jets. Narrowing of spectral lines broadened due to phase relaxation is achieved by cooling the samples under study. Therefore, the above-mentioned experimental methods are, as a rule, applicable at temperatures close to the liquid helium temperature. Ignoring phase relaxation requires preliminary measurements and analysis of the temperature dependence of linewidths.Hochstrasser and Marzzacco 134 studied the Tn/S0 and Sn/S0 absorption spectra of a number of aromatic compounds in the crystalline phase at 4.2 K. Presumably, they were the first who considered substantial broadening of vibronic bands in the absorption spectra upon excitation to the second and upper electronic states. For instance, marked broadening of vibronic lines was found in the spectra of pyrazine and its derivatives, phenazine, 9,10-diazaphenanthrene, naphthalene and its halogen derivatives in the region of the T2/S0 transition. According to Kasha's rule, excitation of a molecule to the second triplet state should be followed by fast internal conversion to the lowest triplet state.A similar broadening was observed in the singlet ± singlet absorption spectra of naphthalene and quinoxaline molecules incorporated into durene crystals, of naphthalene and quinoxaline vapours and of naphthaldehyde. These spectra exhibited narrow S1/S0 absorption lines which were substantially broadened on approaching the lines corresponding to the S2/S0 transition. Studies of vibronic absorption linewidths of different sing- let ± singlet transitions of anthracene, coronene and 1,2-benzper- ylene in Shpolskii matrices revealed 135 noticeable narrowing of the Sn/S0 (n51) lines with a decrease in the temperature from 300 to 77 K.Further reduction of temperature down to 4.2 Kwas found to have little effect on the linewidths. Narrowing of the S1/S0 line was followed down to 4.2 K. The linewidths for a number of compounds are listed in Table 7. However, the S1/S0 linewidths were found to be strongly overestimated since the S1 state lifetimes of the molecules under study estimated from these linewidths are nearly three orders of magnitude shorter than the Ref. Method tn /ps Homogene- ous linewidth /cm71 135 135 135 135 135 135 135 135 135 135 135 141 0.08 0.019 0.013 2.6 1.3 0.018 0.015 2.2 0.013 0.004 0.007 1.4 a 6.2 280 4002.1 4.1 300 3502.4 420 1400 8007.8 Shpolskii effect the same """""""""hole burning technique 141 142 141 1.6 1.3 1.7 6.5 4.0 6.1 the same supersonic jet the same 140 140 142 0.2 0.48 0.48 25 11 11 hole burning technique the same supersonic jetUltrafast nonradiative transitions between higher excited states in organic molecules measured lifetimes.Analysis of the results obtained in the above- mentioned studies 136 ± 139 shows that these linewidths much exceed the natural linewidths even in the Shpolskii matrices; determination of the natural linewidths requires the Personov holes 98, 99 to be `burnt' in the absorption lines. The linewidths of transitions to upper excited singlet states also seem to be overestimated.A series of studies on free porphyrin base dissolved in the Shpolskii matrices has been reported.136 ± 139 The homogeneous linewidths were determined using a photochemical hole burning technique based on the tautomeric transformations involving inner hydrogen atoms of the porphyrin molecule.64, 136 Samples were irradiated by a tunable single-mode laser with a frequency jitter of *10 MHz. The contribution of phase relaxation was determined from the temperature dependences of the 0 ± 0 line- widths in the range from 1.5 to 4.2 K. The homogeneous width of the 0 ± 0 line for the S1/S0 transition was found to be *9 MHz (0.0003 K,137 which corresponded to the experimen- tally observed S1/S0 fluorescence decay time of free porphyrin base in n-octane at 4.2 K (17 ns). This indicates that the hole burning technique does allow measurements of homogeneous line broadening, thus providing a means for determination of the lifetimes of the excited states of organic molecules.tautomeric transformation involving inner hydrogen atoms of porphyrin molecule. The hole halfwidth was found to be 10 cm71 for the S3/S0 transition (lmax=434 nm), which corresponded to the S3 level lifetime, t3, of 0.53 ps. The lifetime of the same state estimated from the S3?S0 fluorescence quantum yield and the ratio of the radiative rate constants for the S3?S0 and S1?S0 transitions (these were determined from the absorption spectrum) was 1.2 ps. The discrepancy between the two t3 values is not too large; however, it seems to be greater than the experimental error.Position of the S2 level of the tetrabenzoporphyrin derivative has also been determined.140 The corresponding band in the absorp- tion spectrum is overlapped with vibronic bands of the S1/S0 transition. However, hole burning measurements allowed the observation of the 0 ± 0 band of the S2/S0 transition at 606 nm and estimation of the S2 state lifetime at 0.38 ps. The lifetime of the vibronic sublevel of the S1 state with an energy of 1108 cm71 was also determined: the hole halfwidth was found to be 0.18 nm, which corresponded to a vibronic level lifetime of 1.1 ps. Hole burning technique was used for studying internal con- version in Zn-tetraphenylporphyrin molecules incorporated into polymeric matrices at 4 K.141 The homogeneous broadening in the Soret band was close to that found in the preceding case 140 (see Table 7).Supersonic jet measurements have led to similar results.142, 143 The lifetimes of vibronic levels for the S1 state of free porphyrin base in n-octane at 4.2 K (Table 8) were determined using the same procedure.139 The homogeneous holewidth of the 0 ± 0 band was found to be *9 MHz, whereas those of the vibronic bands for the same transition were varied from 5 to 150 GHz (0.1 ± 7 cm71). Both hole burning and measurements of vibronic bandwidths in the spectra corresponding to transitions to upper electronic excited levels were performed using a tunable pulsed laser with a bandwidth of*1 cm71.The lifetimes of several vibrational levels of the S1 state of free porphyrin base at 4.2 K vary from 1 to 30 ps (Table 8). Totally symmetrical modes in the frequency range from 150 to 1160 cm71 are characterised by substantial variations of the lifetimes. For the modes with frequencies lying between 1340 and*1580 cm71, the lifetimes of vibrational levels were estimated at 1.0 to 1.6 ps irrespective of the vibrational mode symmetry. This is in reason- able agreement with the results of measurements of the lifetimes of vibrational levels of organic molecules obtained by other exper- imental methods.116, 117 Note that the rates of vibrational relaxa- tion of most of higher excited states are much lower than those of electronic relaxation (see Tables 2 ± 5).The S3 state lifetimes of a tetrabenzoporphyrin derivative [a solution in THF± Et2O (1 : 1) mixture] determined from hole burning experiments at 4.2 K and estimated from the S3?S0 fluorescence quantum yield have been compared.140 As in the preceding case, holes were burnt using photochemically induced The HESS lifetimes were estimated by the photophysical (optical) hole burning method.144 Previously,145 this experimental technique was used in a study on the temperature dependence of phase relaxation of the S1 state of Zn-porphyrin molecule. To burn holes in the S2/S0 absorption band of phenanthrene (a solution in isopentene ± 2-methylpentane mixture at 4.2 K), mol- ecules were excited to a triplet state by irradiating samples by the light of a pulsed dye laser with frequency doubling.Absorption was measured using a high-resolution double monochromator. A phosphoroscope was constructed to provide alternate illumina- tion of a sample placed in a cryostat by the laser light and by the transmitted light of a Xe lamp. The homogeneous halfwidth of hole burnt was found to be 67 cm71, which corresponded to the S2(0) state lifetime of phenanthrene, t2, of 0.079 ps. The lifetime of the S2 vibronic state (400 cm71), also determined by both reso- nance and off-resonance hole burning techniques, was found to be nearly the same. On the other hand, homogeneous linewidth measurements of isolated phenanthrene molecules in a supersonic jet have led to a value of 0.5 ps for t2(0).146 The reason for such a large difference between the estimates of the lifetimes of the second excited state of phenanthrene molecules is still to be clarified.Indeed, since fast internal conversion occurs intramolec- ularly, one would expect similar rates of internal conversion for the molecule in the condensed phase and in the vapour phase. Unfortunately, comparable information on the rate constants for internal conversion from higher excited states in the condensed phase and in supersonic jets for the same molecules is scarce. Table 8. Lifetimes of vibrational levels of the S1 state of free base porphyrin molecules (a solution in n-octane) at 4.2 K.139 Mode symmetry t /ps n /cm71 Homogeneous width /GHz Ag Ag Ag Ag B1g 0.017 9.46103 82 1.9 50 3.2 5 32 44 96 Ag 0 a 155 307 710 783 943 970 B1g 1035.4 Ag Ag B1g Nickel and Wick 147 studied the vibrational structure of the S0±S2 absorption band of coronene in poly(methyl methacrylate) at 1.8 K by the photophysical hole burning technique.The lifetime of the vibronic S2 states, estimated from the homogeneous width of resonance transitions, was found to be somewhat shorter than 2 ps. This is in reasonable agreement with the results obtained by different authors 148, 149 who observed fluorescence from the S2 state of coronene under conditions of direct single- quantum excitation and with the results of measurements of the homogeneous width of fluorescence excitation lines in the Shpol- skii matrices and in a supersonic jet.Such a relatively long lifetime of the S2 state [with the energy gap DE(S27S1) ^ 5000 cm71] seems to be due to high molecular symmetry and forbiddenness of the S2(1B1u)?S1(1B2u) radiative transitions. Unlike phenan- threne, the homogeneous linewidth of coronene is nearly the same both in a supersonic jet and in the low-temperature matrix. 1161 1340 1355 1583 3.6 1.7 1.5 30 1.0 1.3 1.6 150 120 100 Ag a The 0 ± 0 transition. Alarge body of information on the linewidths in the molecular spectra (including those of porphyrins) obtained in supersonic jet studies has been reported.142, 143, 146 485486 VII. Lifetimes of higher excited Sn states estimated from picosecond and femtosecond spectroscopy data In the last decade, direct measurements of the HESS lifetimes of organic molecules were reported.To measure the decay of very weak emission from these states with femtosecond resolution is a complicated problem. Conventional experimental techniques involve consecutive stepwise excitation by two picosecond or femtosecond light pulses with a small pulse delay. The second pulse is used for measuring the kinetics of either the induced absorption spectrum or the absorption spectrum reconstruction for the lower-lying state (laser photolysis with picosecond or femtosecond resolution). Excitation by the second pulse also allows photoionisation of a molecule followed by mass-spectro- metric or another type of detection of the cation formed or electron detached.Recently, a large number of studies concerning measurements of ultrafast luminescence decay kinetics by lumi- nescence up-conversion, i.e., `mixing' of weak luminescence with an optically delayed femtosecond pulse in a nonlinear crystal, were reported. When comparing different methods for studying the dynamics of internal conversion from HESS, preference should be given to those which allow the observation of fluorescence kinetics due to transition from these states. Studies of molecular absorption spectra induced by laser pulses are complicated by three factors, viz., transition-state absorption, ground-state `bleaching' and stimulated emission from the excited state. The results of second- pulse induced photoionisation studies are also ambiguous.A stepwise photoionisation study of the lifetime of the S2 state (35 900 cm71) of naphthalene in a supersonic jet has been reported.150 However, only the upper bound of the lifetime of this state (10711 s) was estimated due to instrumental limitations on the pulse duration (6 ps). Quenching of the S2 fluorescence of Zn(TPP) in acetonitrile due to the addition of dichloromethane has been studied by an up-conversion technique.151 The lifetime t(S2) of ZnTPP was found to be 3.5 (in acetonitrile) and 0.75 ps (in dichloromethane). Since the S2 state lifetime is very short, the concentration dependence of fluorescence quenching can be described using the Perrin model: ln I0 I a VNAaCH2Cl2a, where I0 is the radiation intensity without a quencher, I is the radiation intensity, V = (4/3) pR3 is the action range of the quencher (R=15 A) and NA is the Avogadro constant.It was found 151 that quenching of fluorescence from the S1 state of Zn(TPP) with dichloromethane is inefficient. The authors sug- gested that quenching of the S2 fluorescence is due to electron transfer. Internal conversion in complexes of Zn-porphyrin (ZnP) derivatives and free porphyrin base (H2P) dissolved in benzene has been studied by femtosecond fluorescence spectroscopy.152 For ZnP, two fast components of the S2(B) ? S0 fluorescence decay were associated with the S2 ? S1 internal conversion with a time constant of 150 fs (spectral changes in the region of the S2 ? S0 transition) and with vibrational relaxation to the S1(Q) state with a time constant of 600 fs (spectral changes in the region of the S1 ? S0 transition).Free porphyrin base exhibits ultrafast relaxation of the S3(B) state (t<40 fs), fast relaxation of the S2(Qy) state with a time constant of 90 fs and a slower vibrational relaxation of the S1(Qx) state. A fluorescence up-conversion study of fluorescence decay from the S3(1Bb) state of tetracene in solution using 280 fs pulses was carried out.153 The decay time, t, of the (S3?S0) fluorescence of tetracene in the UV region was estimated to be*120 fs. Usual fluorescence from the S1(1La) level was found to arise with the same rate constant. Decay kinetics of the S2 and S1 states of Malachite Green dye in different solvents (methanol, water, ethylene glycol and water ¡¾ glycerol mixture) has been studied by femtosecond fluo- V L Ermolaev rescence up-conversion spectroscopy (pulse separation Dt=90 fs).154 The decay times in an aqueous solution were estimated at 0.27 ps for t(S2) and at 0.54 ps for t(S1).If the molecule is excited to the second singlet state, the rise of the S1 fluorescence and the decay of the S2 fluorescence occur with the same rate constant. Femtosecond time-resolved photoelectron spectroscopy (Dt<200 fs, l = 252 nm) was used for studying ultrafast internal conversion in 1,3,5-hexatriene, which is one of the simplest representatives of polyenes.155 The S1 state lifetimes were estimated at 730 and 270 fs for cis- and trans-hexatriene, respectively. The authors of this study suggested that relaxation process occurs as a cascade, viz., S2?S1?S0.The estimate of the S2 state lifetime, t(S2)450 fs, is consistent with that obtained earlier 156 from the homogeneous width of the S2 / S0 absorption line at 250 nm in supersonic jets. The S2 state decay times were estimated at 50 fs for trans-hexatriene and at 20 fs for cis- hexatriene. This is close to the S2 state lifetime of trans-hexatriene vapours (40 fs) obtained from the data on resonant Raman scattering.157 Femtosecond up-conversion studies of the effect of molecular structure and solvent on the S2 state lifetimes of carotenoids have been reported.158, 159 They revealed violation of Kasha's rule 8 and Ermolaev ¡¾ Sveshnikova's rule 160, 161 (the sum of the fluorescence and triplet-state generation quantum yields equals unity for solutions of fluorescent compounds) in the case of linear polyenes including carotenoids.It was pointed also out that the main emission of molecules with C2h symmetry originates from the S2 level (the S2 ? S1 radiative transition is forbidden). If the molecular symmetry is broken, the S1 emission becomes stronger. The t2 values for the three compounds studied lie between 100 and 200 fs. More recently, in studies of neurosporene (a linear carotenoid) it has been shown 162 that excitation to the S2 state with an excess vibrational energy of *3000 cm71 leads to the appearance of a time-dependent S2 ? S0 fluorescence spectrum.This was explained by the vibrational energy redistribution in the S2 state from the Franck ¡¾ Condon active modes to other (dark) modes with a time constant of 35 ¡¾ 40 fs, followed by the S2 ? S1 internal conversion occurring in 250 fs. The effect of hydrogen bond formation/dissociation on the relaxation rate of the S3 and S2 states of trans-retinal molecules (a solution in butan-1-ol ¡¾ cyclohexane mixture) on the subpico- second time scale has been first studied recently.163 The S3[Bau (p,p*)] and S2[A¢§g (p,p*] state lifetimes of the molecules under study in neat butan-1-ol (200 fs and 1.8 ps, respectively) were found to be much longer than the corresponding lifetimes in neat cyclohexane (30 fs and 0.7 ps, respectively).In the S2 state, all the trans-retinal molecules are hydrogen-bonded. The increase in the fluorescence and the isomerisation quantum yields and the decrease in the triplet quantum yield upon hydrogen bond formation were explained by the prolonged lifetime of the S2 state of trans-retinal. The authors cited were the first to study the formation and cleavage of a hydrogen bond in a subpicosecond time. Hirata et al.164, 165 studied the dependence of the rate constant for the S2 ? S1 internal conversion on the energy gap width in the molecules of (i) diphenylacetylene and its derivatives and (ii) trans- a,o-diphenylpolyenes with different chain length (solutions in n-hexane and THF were used in both cases).The S2 state lifetimes of diphenylpolyenes with the molecular chains of length from 3 to 8 monomer units and the energy gaps of width from 1600 to 4800 cm71 varied insignificantly (500100 fs) and increased only slightly with the decrease in the temperature. The rate constant for the S1 ? S0 nonradiative transition increased by three orders of magnitude for the same diphenylpolyenes. A femtosecond laser flash photolysis study on fast photo- induced reactions of trans-azobenzene excited to the S2 state (in solutions at 303 K, *200 fs resolution) has been reported.166 Analysis of the induced absorption decay curve measured in the region 370 ¡¾ 450 nm revealed its two-exponential structure andUltrafast nonradiative transitions between higher excited states in organic molecules allowed determination of the time constants which appeared to be different for different solvents.The time constant for the fast exponent, t1, was found to be 0.9 ps for hexane, cyclohexane and hexadecane and 1.2 ps for acetonitrile. That for the slow expo- nent, t2, was found to be 13 ps for the hydrocarbons and 16 ps for acetonitrile. The authors suggested that the decay of the fast exponent is due to deactivation of the S2(p,p*) state. The appear- ance of the slow exponent was explained by the presence of an intermediate excited singlet state lying between the S2 and S1 states on the energy scale, corresponding to the twist-conformer of the excited trans-azobenzene molecule. The dynamics of the S2 and S1 states of trans-4-n-butyl-40-methoxyazabenzene in hexane has been studied 167 by a similar experimental technique with the same resolution.The time constants were estimated at &250 fs for t(S2) and at 2.3 ps for t(S1). Femtosecond time-resolved photoelectron spectroscopy was used for measuring the Sn state lifetimes of pyrazine molecules in a supersonic jet.168 In particular, the lifetime of the S2 state, t(S2), was estimated at 20 fs. A supersonic jet study of the S2 state of benzene ± ammonia complexes has been reported.169 The S2 state lifetime was estimated at*100 fs. Fluorescence of coumarin 481 in cyclohexane at 293 K, which appeared after excitation by 100 fs pulses at l=267 nm has been studied in detail.170 The molecules were excited to an Sn state lying 12 000 cm71 higher than the S1 state and then the dynamics of the rise of the S1 ? S0 fluorescence was observed using up-conversion technique.The time constant of the rise of the fluorescence in the region 419 ± 475 nm was found to be 280 to 220 fs. This time was interpreted as that required for completion of the Sn ? S1 internal conversion. It was also suggested that intramolecular redistrib- ution of vibrational energy in the S1 and S0 states occurs with strongly different rates.170 However, there is also another concept of the redistribution of excess vibrational energy in the ground and first excited states after optical excitation.20 The authors came to conclusion that vibrational excitation is redistributed in the S0 and S1 states of dye molecules with the same rate constant.Based on analysis of their original studies and published data, they also suggested that in most cases redistribution of excess vibrational energy involves two stages. The first stage is fast, with a time constant t1 estimated at 30 to 200 fs, while the second stage is slow, with a time constant t2 estimated at 200 to 1000 fs. It was assumed that the fast stage is due to the interaction of isoenergetic vibrational levels, whereas the slow stage is due to coupling of levels with different energies. Yet another assumption is that the time required to `cool' a vibrationally `overheated' molecule via transfer of excess energy to the environment (solvent molecules) should be much longer (from 15 to 40 ps for different solvents).The dynamics of IR 125 dye in ethylene glycol solution was studied by pump-probe experiments.171, 172 The pump pulse duration was 30 fs (l=425 nm) while the probe pulse width was 20 fs (l=850 nm). It was found that a dye molecule excited to an Sn (n>1) state by a pulse at l=425 nm arrives at the S1 state within 50 fs; however, the time constant of S1 population (accu- mulation of molecules at the minimum of the PES of this state) is much larger (*1 ps). A number of papers concerning nonradiative transitions between higher excited electronic levels of the molecules com- prised of ten or smaller number of atoms was reported. In particular, the S2 state lifetimes of acetone and its deuterated analogue in the vapour phase were studied using femtosecond photoionisation (lexc=194 nm, lion=295 nm).173 It was found that t(S2)=13.5 ps for acetone-d6 and 4.7 ps for acetone-h6.In this case, the S1 state lifetimes of both molecules (<150 fs) are shorter than the corresponding S2 state lifetimes. This is similar to azulene and b-carotenes. 487 VIII. Ultrafast nonradiative transitions involving triplet states Asalient feature of aromatic and dye molecules is the possibility of ultrafast nonradiative transitions involving triplet states. Natu- rally, the question arises of the rates of such transitions. Liu and Edman 174 studied photoisomerisation reactions of various alkenes sensitised by anthracene and its derivatives.All the reaction pathways were known to pass through triplet states. The authors assumed that the photoreactions can proceed only via excited molecules of anthracene and its derivatives. Since the lowest triplet levels of anthracenes lie much below those of alkenes, it was suggested that the reactions under study proceed due to energy transfer from the second triplet level (T2) of the anthracenes and that the T2 level is populated via intersystem crossing from the S1 state. [The T2 level of the sensitiser molecules lies somewhat lower than their first excited singlet level, E(T2)=25 000 ± 26 000 cm71]. Among all the alkene reactions studied, the highest quantum yield (0.042) was found for cis ± trans isomerisation of piperylene in benzene (0.2 M) at 293 K sensitised by 9,10-dibromoanthracene. Assuming that the rate constant for triplet ± triplet energy transfer from the T2 level is equal to the diffusion rate constant, the T2 state lifetime of 9,10-dibromoan- thracene was estimated at 1.3610710 s.175 An alternative explanation for the results obtained by Liu and Gale 175 has been put forward in a paper 176 concerning fluores- cence quenching of haloanthracenes by naphthalene and phenan- threne.Taking into account a correlation between the rate constants for quenching and the spin-orbit coupling constants of anthracene derivatives, the authors concluded that a spin-forbid- den, singlet ± triplet energy transfer occurs in this case. Similarly to the earlier results,175 the most pronounced effect was observed for 9,10-dibromoanthracene. It was also assumed 176 that this process should also lead to population of triplet states of the energy acceptor and can induce alkene isomerisation. Only a few observations of the fluorescence due to trip- let ± triplet transitions have been reported.For instance, T2±T1 fluorescence was detected upon excitation of certain anthracene derivatives dissolved in methylcyclohexane to the S1 state.80, 177 Fluorescence spectra were measured in air at room temperature. The 0 ± 0 bands in the T2±T1 fluorescence spectra of these compounds lie in the region from 840 to 960 nm (12 000 ± 10 400 cm71) and are nearly mirror images of the long- wavelength Tn / T1 absorption bands. The T2 ? T1 fluores- cence quantum yields were estimated at 1076 for 9,10-dibromoan- thracene [t(T2)=2610712 s] and at 561078 for the unsubstituted anthracene [t(T2)=10713 s].These values are in poor agreement with indirect estimates of the T2 state lifetime of anthracene, obtained from the results of photochemical sensitisa- tion studies.178 Keller 179 was the first to report on the reverse intersystem crossing from higher excited triplet states of aromatic molecules to the excited singlet levels. At 77 K, aromatic molecules in a glassy matrix were excited by UV irradiation to populate the lowest triplet state. Then the molecules were excited by visible light into the Tn / T1 absorption bands and their UV fluorescence accom- panying the S1 ? S0 transition (Fig.9) was observed. The fluo- rescence intensity was proportional to the intensity of visible light and the fluorescence lifetime was identical to the lifetime of the lowest triplet T1 state. The quantum yield of the triplet ± singlet intersystem crossing process (the ratio of the number of Tn ? S1 transitions to the number of photons absorbed by the T1 state) is small, namely, 561077 for naphthalene, 261076 for quinoline and 1076 for fluorene. Kobayashi et al. 180 also observed the S1 ? S0 fluorescence of anthracene and its 9-methyl and 9-phenyl derivatives due to reverse intersystem crossing. The quantum yield of the Tn ? S1 intersystem crossing was estimated at 561076 for anthracene and at 861075 for 9-phenylanthracene.More recently 181 some arguments were adduced in favour of assump- tion that the S1 state of 9,10-dibromoanthracene molecule is488 kisc T3 knr S3 knr kisc T2 S2 knr knr hnIII hnII S1 kisc T1 hndfl hnfl hnI S0 Figure 9. An energy level diagram illustrating the appearance of delayed fluorescence (ndfl) after excitation of a molecule to a higher excited triplet state and intersystem crossing from this state to the singlet manifold. populated via absorption of a light quantum by the T1 state and is also due to the Sm / T1 transition. Population of the fluorescent state of perylene vapours via excitation of the lowest triplet state by consecutive irradiation by two laser pulses has been reported.182 A version of Keller's technique 179 which uses stepwise excita- tion by nanosecond laser pulses has been proposed in studies of xanthene dyes (eosin, erythrosin and Rose Bengal) in polymeric matrices.183, 184 The authors simultaneously observed both the delayed fluorescence induced by the second delayed pulse and the decrease in the optical density of the triplet ± triplet absorption, which was rationalised by the occurrence of the Tn ? Sm reverse intersystem crossing.Thus, a competition of nonradiative inter- system crossing with internal conversion over the triplet manifold is implied. This is quite probable since the molecules studied contain heavy Br or I atoms. However, too high quantum yields of the process 183, 184 are suspicious because the spin-forbidden transition competes with the spin-allowed process of internal conversion over the triplet manifold. Similar approach was used in studies of the S1 ? S0 fluores- cence of Rose Bengal in aqueous solutions (a phosphate buffer, pH 7.0).185 Fluorescence appeared after stepwise excitation of the dye molecules from the lowest triplet state to excited triplet states, followed by reverse intersystem crossing to the singlet manifold and emission.An intense absorption band at lmax=1063 nm was detected by near-infrared laser flash photolysis measurements and assigned to the T2 / T1 transition.185 It was found that excitation of the dye by the second harmonic of a Nd laser (l=532 nm, tpulse=134 ps) to the S1 state leads to population of the T1 triplet state with a probability of 0.98 and to fluorescence emission with a quantum yield of 0.02.The second pulse of the fundamental frequency of the Nd laser excited the solution with a delay of 34 ns to populate the T2 state since the fundamental frequency of neodymium penetrates the long-wavelength triplet ± triplet absorption band. Population of the T3 state was achieved by irradiating the solution by the second pulse (l=632 nm) with the same delay. Then, the dependence of the intensity of two-step, laser-induced delayed fluorescence on the intensity of laser pulses was measured and analysis using a simplified kinetic scheme was performed. As a result, the quantum yield of the T2 ? S1 reverse intersystem crossing, q(T2,S1), was estimated at 0.014 and the T2 state lifetime, t(T2), was found to be 5.8 ps.185 No S1 ? S0 fluorescence was observed upon two-step laser-induced excitation at l2=632 nm via the T1 state.Therefore, only the upper bound of the reverse V L Ermolaev Table 9. Quantum yields of reverse intersystem crossing for excited triplet levels of Rose Bengal dye molecule.185 E /cm71 q(Tn?Sn) Energy level DE(Tn?Sn) /cm71 DE(Tn?Tn71) /cm71 0.014 <0.06 0.8 9435 7580 1370 4030 2820 640 23 550 31 130 32 500 T2 T3 T4 intersystem crossing yield from the T3 state, q(T3 ? Sn), was estimated at 0.06. These results strongly disagree with those obtained in earlier studies,183, 184, according to which q(T3 ? Sn)=0.74. Previously,183 this value was estimated from the decrease in the T1 state population of the dye after irradiation by the second pulse at l2=632 nm.Previously (see, e.g., Refs 186 and 187), it was assumed that the reverse intersystem crossing occurs mainly from the T2 level. However, the results of measurements of the quantum yield q(Tn ? Sn) for Rose Bengal rule out this assumption. Table 9 lists the quantum yields of reverse intersystem crossing, positions of triplet levels and the energy gaps between them and the nearest lower-energy triplet and singlet levels of the Rose Bengal mole- cule. As can be seen, the quantum yields of reverse intersystem crossing are satisfactorily described by the `energy gap law'; however, it should be remembered that internal conversion over the triplet manifold leads to the energy gap narrowing with increase in the level number.This casts some doubts on the too high yield of reverse intersystem crossing from the T4 level, though the strength of the spin-orbit coupling between particular triplet and singlet states plays a significant role in this case.185 It was found that the addition of LiBr in combination with stepwise excitation to HESS leads to an increase in the quantum yield of triplet rhodamine generation in solution.188 IX. Conclusion The results obtained using different methods of determination of the rate constants for ultrafast nonradiative transitions, consid- ered in this review are, with rare exception, in reasonable agree- ment with one another.Those obtained by recently developed but rather widely used direct picosecond and femtosecond methods for measuring the fluorescence decay have for the most part confirmed indirect estimates based on the fluorescence yields from HESS under conditions of single-quantum or two-quantum excitation to corresponding levels and found from the homoge- neous absorption linewidths. It was established that in most cases the HESS lifetimes of organic and dye molecules are of the order of tens of femtoseconds. The only exception is the second excited singlet state (S2), the lifetime of which for a number of molecules is much longer, so that it can be considered as a specific state similar to the fluorescent state (S1). More complex is the problem of theoretical interpretation of experimental results.Correspond- ence between the observed rate constants and predictions of modern theories of nonradiative transitions is fragmentary. In some instances even the `physical' or `chemical' mechanism responsible for the nonradiative transition remains unclear. Undoubtedly, the development of femtosecond measuring devices will provide the possibility of obtaining new information on transitions between HESS. For instance, only in the very recent past has it been possible to observe non-exponential fluorescence decay due to transitions from the HESS, which is interpreted as superimposed `hot' fluorescence. `Physical' ultrafast nonradiative transitions can occur by tunneling and as conical intersections of the PES of excited states.Comprehensive theory of nonradiative transitions, based on calculations of possible fast processes due to such conical inter- sections in complex molecules is now at an early stage. The state- of-the-art in multidimensional theory of nonadiabatic transitionsUltrafast nonradiative transitions between higher excited states in organic molecules in the case of conical intersections of the PES has been recently considered in a special issue of the Chemical Physics journal.189 The author expresses his gratitude to Dr. B Nickel for submitting materials and valuable discussions and to Prof. V A Benderskii for valuable critical comments. This review has been written with the financial support of the Russian Foundation for Basic Research (Project No.00-03-32208) and the Ministry of Industry, Science and Technology of the Russian Federation. References 1. A L Buchachenko Usp. Khim. 68 99 (1999) [Russ. Chem. Rev. 68 85 (1999)] 2. A H Zewail Femtochemistry: Ultrafast Dynamics of the Chemical Bond (Singapore: World Sci., 1994) 3. V Sundstrom (Ed.) Femtochemistry and Femtobiology (Singapore: World Sci., 1998) 4. S I VavilovPhilos. Mag. 43 307 (1922) 5. S I Wawilow Z. Phys. 42 311 (1927) 6. A N Terenin Acta Phys. Chim. USSR 18 210 (1943) 7. G Lewis, M Kasha J. Am. Chem. Soc. 66 2100 (1944) 8. M Kasha Discuss. Faraday Soc. 9 14 (1950) 9. R Englman Nonradiactive Decay of Ions and Molecules in Solids (Amsterdam: North-Holland, 1979) 10. E S Medvedev, V I Osherov Teoriya Bezyzluchatel'nykh Perekhodov v Mnogoatomnykh Molekulakh (The Theory of Nonradiate Transi- tions in Polyatomic Molecules) (Moscow: Nauka, 1983) 11.E S Medvedev, V I Osherov Radiationless Transition in Polyatomic Molecules (Ser. Chemical Physics) Vol. 57 (Berlin: Springer, 1995) 12. A Nitzan, J Jortner, P M Rentzepis Proc. R. Soc. London, A Math. Phys. Sci. 327 367 (1972) 13. J Jortner, S A Rice, R M Hochstrasser, in Advances in Photo- chemistry (Eds J N Pitts Jr, G S Hammond, W A Noyes Jr) (New York; London: Wiley, 1969) Vol. 7, p. 149 14. E B Sveshnikova, V L Ermolaev Opt. Spektrosk. 30 379 (1971) a 15. E N Bodunov Opt. Spektrosk. 40 942 (1976) a 16. V L Ermolaev, E B Sveshnikova, E N Bodunov Usp. Fiz. Nauk 166 279 (1996) b 17. V L Ermolaev, E N Bodunov, E B Sveshnikova, T A Shakhverdov Bezyzluchatel'nyi Perenos Energii Elektronnogo Vozbuzhdeniya (Nonradiate Transfer of Electronic Excitation Energy) (Leningrad: Nauka, 1977) 18.V L Ermolaev, E B Sveshnikova Chem. Phys. Lett. 23 349 (1973) 19. V L Ermolaev, E B Sveshnikova Acta Phys. Pol. A 95 299 (1999) 20. T Elsaesser,W Kaiser Annu. Rev. Phys. Chem. 42 83 (1991) 21. G B Tolstorozhev, S A Tikhomirov Zh. Prikl. Spektrosk. 65 635 (1998) 22. M Beer, H C Longuet-Higgins J. Chem. Phys. 23 1390 (1955) 23. G Viswanath, M Kasha J. Chem. Phys. 24 574 (1956) 24. B D Wagner, D Tittelbach-Helmrich, R P Steer J. Phys. Chem. 96 7904 (1992) 25. D Huppert, J Jortner, P M Rentzepis J. Chem. Phys. 56 4826 (1972) 26.J B Birks Photophysics of Aromatic Molecules (London; New York; Sydney; Toronto: Wiley-Interscience, 1969) 27. P M Rentzepis, J Jortner, R P Jones Chem. Phys. Lett. 4 599 (1970) 28. D Huppert, J Jortner, PMRentzepis Chem. Phys. Lett. 13 225 (1972) 29. G D Gillispie, E C Lim J. Chem. Phys. 65 4314 (1976) 30. G D Gillispie, E C Lim J. Chem. Phys. 68 4578 (1977) 31. D Klemp, B Nickel Chem. Phys. Lett. 130 493 (1986) 32. Y Hirata, E C Lim J. Chem. Phys. 69 3292 (1978) 33. M Fujii, T Ebata, N Mikami, M Ito Chem. Phys. 77 191 (1983) 34. G Eber, S Schneider, F Dorr Chem. Phys. Lett. 52 59 (1977) 35. S Murata, C Iwanaga, T Toda, H Kokubun Ber. Bunsenges. Phys. Chem. 76 1176 (1972) 36. E P Ippen, C V Shank, R L Woerner Chem. Phys. Lett.46 20 (1977) 37. C V Shenk, E P Ippen, O Teschke, R L Fork Chem. Phys. Lett. 57 433 (1978) 38. EW-GDiau, SDe Feyter,A H Zeweil J. Chem. Phys. 110 9785 (1999) 39. M J Bearpark, F Bernardi, S Clifford, M Olivucci, M A Robb, B R Smith, T Vreven J. Am. Chem. Soc. 118 169 (1996) 40. G Eber, F Gruneis, S Schneider, F Dorr Chem. Phys. Lett. 29 397 (1974) 489 41. R Klann, R J Bauerle, F Laermer, T Elsaesser, M Niemeyer, W Luttke Chem. Phys. Lett. 169 172 (1990) 42. A Falchi, C Gellini, P R Salvi, K Hafner J. Phys. Chem. A 102 5006 (1998) 43. A Falchi, C Gellini, P R Salvi, K Hafner J. Phys. Chem. A 104 1078 (2000) 44. M J Bearpark,M A Robb J. Phys. Chem. A 104 1075 (2000) 45. U P Wild,H J Griesser, V D Tuan, J F M Oth Chem. Phys. Lett.41 450 (1976) 46. B Nickel, J Hertzberg Chem. Phys. 132 219 (1989) 47. J R Huber, M Mahaney Chem. Phys. Lett. 30 410 (1975) 48. M H Hui, P de Mayo, R Suan,W R Ware Chem. Phys. Lett. 31 257 (1975) 49. M Mahaney, J R Huber Chem. Phys. 9 371 (1975) 50. A Maciejewski, R P Steer J. Am. Chem. Soc. 105 6738 (1983) 51. A Maciejewski, R P Steer Chem. Rev. 93 67 (1993) 52. A Maciejewski, R P Steer J. Phys. Chem. 92 6939 (1988) 53. M Szymanski, R P Steer J. Phys. Chem. 96 8719 (1992) 54. M Szymanski, A Maciejewski, J Kozlowski, J Koput J. Phys. Chem. A 102 677 (1998) 55. H Eisenberger, B Nickel J. Chem. Soc., Faraday Trans. 92 733 (1996) 56. H Eisenberger, B Nickel, A A Ruth, R P Steer J. Chem. Soc., Faraday Trans. 92 741 (1996) 57. B Nickel, H Eisenberger, M T Wick, R P Steer J.Chem. Soc., Faraday Trans. 92 1101 (1996) 58. MSzimanski, AMaciejewski,R P Steer J. Phys. Chem. 92 2485 (1988) 59. M Szimanski, A Maciejewski, R P Steer J. Photochem. Photobiol. A 57 405 (1991) 60. A Maciejewski, M Milewski, M Szimanski J. Chem. Phys. 111 8462 (1999) 61. B Nickel J. Lumin. 44 1 (1989) 62. J Cai, E C Lim J. Chem. Phys. 95 3014 (1991) 63. L Bajema,M Gouterman, C B Rose J. Mol. Spectrosc. 39 421 (1971) 64. I E Zalesskii, V N Kotlo, A N Sevchenko, K N Solov'ev, S F Shkirman Dokl. Akad. Nauk SSSR 210 312 (1973) c 65. V N Kotlo, K N Solov'ev, S F Shkirman Izv. Akad. Nauk SSSR, Ser. Fiz. 39 1972 (1975) 66. A V Nikolaitchik,M A J Rodgers J. Phys. Chem. A 103 7597 (1999) 67. L A Martarano, Ch-P Wong, W DeW Horrocks Jr, A M Goncalves J.Phys. Chem. 80 2389 (1976) 68. Y Kurabayashi, K Kikuchi, H Kokubun, Y Kaizu, H Kobayashi J. Phys. Chem. 88 1308 (1984) 69. M P Tsvirko, G F Stelmakh, V E Pyatosin, K N Solovyov, T F Kachura Chem. Phys. Lett. 73 80 (1980) 70. V L Ermolaev, I P Kotlyar, V A Lyubimtsev, V G Maslov Opt. Spektrosk. 52 3 (1982) a 71. N Mataga, Y Shibata, H Chosrowjan, N Yoshida, A Osuka J. Phys. Chem. B 104 4001 (2000) 72. V L Ermolaev, V A Lyubimtsev Tr. Gos. Opt. Inst. im. S I Vavilova 65 (199) 20 (1987) 73. A Janovski, J Rzeszotarska J. Lumin. 21 409 (1980) 74. M Yoshizawa, K Suzuki, A Kubo, S Saikan Chem. Phys. Lett. 290 43 (1998) 75. T A Gregory, S Lipsky J. Chem. Phys. 65 296 (1976) 76. F Hirayma, T A Gregory, S Lipsky J.Chem. Phys. 58 4696 (1973) 77. T A Gregory, F Hirayma, S Lipsky J. Chem. Phys. 58 4697 (1973) 78. P O Andersson, S M Bachilo, R-L Chen, T Gillbro J. Phys. Chem. 99 16199 (1995) 79. R Katoh, S Fujiyoshi, M Kotani Chem. Phys. Lett. 292 621 (1998) 80. G D Gillispie, E C Lim J. Chem. Phys. 65 2022 (1976) 81. L A Hallidy, M R Topp Chem. Phys. Lett. 46 8 (1977) 82. A Terenin, G Kobyshev, G Lialin Photochem. Photobiol. 5 689 (1966) 83. W E Gibbs Appl. Phys. Lett. 11 113 (1967) 84. A Muller, E Pfluger Chem. Phys. Lett. 2 155 (1968) 85. MD Galanin, Z A Chizhikova Kratkie Soobshch. po Fizike (4) 35 (1971) 86. M D Galanin, Z A Chizhikova Izv. Akad. Nauk SSSR, Ser. Fiz. 36 941 (1972) 87. E N Kaliteevskaya, T K Razumova Opt.Spektrosk. 36 1118 (1974) a 88. D Leupold, R Konig, B Voigt J. Lumin. 9 113 (1974) 89. V Rehak, A Novak,M Titz Chem. Phys. Lett. 52 39 (1974) 90. G C Orner, M R Topp Chem. Phys. Lett. 36 295 (1975) 91. H-B Lin,M R Topp Chem. Phys. Lett. 47 442 (1977) 92. H-B Lin,M R Topp Chem. Phys. Lett. 48 251 (1977)490 93. M R Topp, H-B Lin Chem. Phys. Lett. 50 412 (1977) 94. H-B Lin,M R Topp Chem. Phys. Lett. 67 273 (1979) 95. M R Topp, H-B Lin, K-J Choi Chem. Phys. 60 47 (1981) 96. B Nickel, G Roden Ber. Bunsenges. Phys. Chem. 81 281 (1977) 97. V A Lyubimtsev, Candidate Thesis in Physicomathematical Sciences, Leningrad State University, Leningrad, 1989 98. R I Personov, E I Al'shits, L A Bykovskaya Opt. Commun. 6 169 (1972) 99. B M Kharlamov, R I Personov, L A Bykovskaya Opt.Commun. 12 191 (1974) 100. V L Ermolaev, V A Lyubimtsev Izv. Akad. Nauk SSSR, Ser. Fiz. 47 1405 (1983) 101. V L Ermolaev, V A Lyubimtsev Opt. Spektrosk. 56 1026 (1984) a 102. V L Ermolaev, V A Lyubimtsev Opt. Spektrosk. 60 74 (1986) a 103. V L Ermolaev, V A Lyubimtsev Opt. Spektrosk. 60 970 (1986) a 104. V A Lyubimtsev, V L Ermolaev Opt. Spektrosk. 61 511 (1986) a 105. V A Lyubimtsev, V L Ermolaev Opt. Spektrosk. 61 766 (1986) a 106. V L Ermolaev, V A Lubimtsev Acta Phys. Pol. A 71 731 (1987) 107. A A Krasheninnikov, A V Shablya Opt. Spektrosk. 60 70 (1986) a 108. V P Klochkov, V L Bogdanov Zh. Prikl. Spektrosk. 43 5 (1985) 109. R I Personov, E I Al'shits, L A Bykovskaya Pis'ma Zh. Eksp. Teor. Fiz. 15 609 (1972) 110.A P Veduta,M D Galanin, B P Kirsanov, Z A Chizhikova Pis'ma Zh. Eksp. Teor. Fiz. 11 157 (1970) 111. V L Bogdanov, in Topics in Fluorescence Spectroscopy. Nonlinear and Two-Photon-Induced Fluorescence (Ed. J R Lakowicz) (New York; London: Plenum, 1997) Vol. 5, p. 211 112. V L Bogdanov, V P Klochkov Opt. Spektrosk. 48 34 (1980) a 113. V A Lyubimtsev Izv. Akad. Nauk SSSR, Ser. Fiz. 54 391 (1990) 114. T P Kravets Izv. Mosk. Inzh. Uchil., Ser. 2 (6) (1912) 115. S J Strickler, R A Berg J. Am. Chem. Soc. 37 814 (1962) 116. N M Cottfried, A Seilmeier, W Kaiser Chem. Phys. Lett. 111 326 (1984) 117. J Scherer, A Seilmeier,W Kaiser J. Chem. Phys. 83 3948 (1985) 118. J L Baudour, H Caillean,W B Yelon Acta Crystallogr., Sect. B, Struct. Crystallogr. 33 1773 (1977) 119. G V Maier, V P Shcherbina Opt. Spektrosk. 52 1063 (1982) a 120. B Nickel Chem. Phys. Lett. 27 84 (1974) 121. B Nickel Helv. Chim. Acta 61 198 (1978) 122. B Nickel Chem. Phys. Lett. 71 238 (1980) 123. B Nickel, G Roden Chem. Phys. 53 243 (1980) 124. B Nickel, G Roden Chem. Phys. Lett. 74 368 (1980) 125. G F Stel'makh, M P Tsvirko Opt. Spektrosk. 50 998 (1981) 126. B Nickel Chem. Phys. Lett. 68 17 (1979) 127. Y J Kray, B Nickel Chem. Phys. 53 235 (1980) 128. D Klemp, B Nickel Chem. Phys. Lett. 97 66 (1983) 129. B Nickel, G Roden Chem. Phys. 66 365 (1982) 130. J Hertzberg, B Nickel Chem. Phys. 132 235 (1989) 131. D L Dexter J. Chem. Phys. 21 836 (1953) 132. R Katoh,M Kotani Chem. Phys. Lett. 201 141 (1993) 133. R Katoh,M Kotani Chem. Phys. Lett. 300 734 (1999) 134. R M Hochstrasser, C Marzzacco J. Chem. Phys. 49 971 (1968) 135. J L Richards, S A Rice Chem. Phys. Lett. 9 444 (1971) 136. S Voelker, J H van der Waals Mol. Phys. 32 1703 (1976) 137. S Voelker, R M Macfarlane, A Z Genack, H P Trommsdorff, J H van der Waals J. Chem. Phys. 67 1759 (1977) 138. S Voelker, R M Macfarlane, J H van der Waals Chem. Phys. Lett. 53 8 (1978) 139. S Voelker, R M Macfarlane Chem. Phys. Lett. 61 421 (1979) 140. V G Maslov Opt. Spektrosk. 50 1091 (1981) a 141. M Vacha, Sh Machida, K Horie J. Phys. Chem. 99 13163 (1995) 142. U Even, J Jortner, J Friedman J. Phys. Chem. 86 2273 (1982) 143. U Even, J Magen, J Jortner J. Chem. Phys. 77 4384 (1982) 144. B Dick, B Nickel Chem. Phys. 110 131 (1986) 145. R M Shelby, R M Macfarlane Chem. Phys. Lett. 64 545 (1979) 146. AAmirav,MSonnenschein, J Jortner J. Phys. Chem. 88 5593 (1984) 147. B Nickel, M T Wick Chem. Phys. 168 111 (1992) 148. C-J Ho, J Babbitt,M R Topp J. Phys. Chem. 91 5599 (1987) 149. R Jefferson, J Babbitt, C-J Ho,M R Topp J. Phys. Chem. 92 2422 (1988) 150. A Hiraya,Y Achiba,N Mikami,K Kimura J. Chem. Phys. 82 1810 (1985) 151. H Chosrowjan, S Tanigichi, T Okada, S Takagi, T Arai, K Tokumaru Chem. Phys. Lett. 242 644 (1995) V L Ermolaev 152. S Akimoto, T Yamazaki, I Yamazaki, A Osuka Chem. Phys. Lett. 309 177 (1999) 153. N Sarkar, S Takeuchi, T Tahara J. Phys. Chem. A 103 4808 (1999) 154. C C Hayden, D W Chandler J. Phys. Chem. 99 7897 (1995) 155. D R Cyr, C C Hayden J. Chem. Phys. 104 771 (1996) 156. D G Leopold, R D Pandley, J L Roebber, R J Hemley, V Vaida J. Chem. Phys. 81 4218 (1984) 157. A B Myers, K S Pranata J. Phys. Chem. 93 5079 (1989) 158. M Mimuro, U Nagashima, S Nagaoka, S Takaichi, I Yamazaki, Y Nishimura, T Katoh Chem. Phys. Lett. 204 101 (1993) 159. M Mimuro, S Akimoto, S Takaichi, I Yamazaki J. Am. Chem. Soc. 119 1452 (1997) 160. V L Ermolaev Usp. Fiz. Nauk 80 3 (1963) b 161. V L Ermolaev, E B Sveshnikova Acta Phys. Pol. 34 771 (1968) 162. S Akimoto, I Yamazaki, S Takaichi,M Mimuro Chem. Phys. Lett. 313 63 (1999) 163. Sh Yamaguchi, H Hamaguchi J. Phys. Chem. A 104 4272 (2000) 164. Y Hirata, T Okada, N Mataga, T Nomoto J. Phys. Chem. 96 6559 (1992) 165. Y Hirata, K Mashima, H Fukumoto, K Tani, T Okada Chem. Phys. Lett. 308 176 (1999) 166. I K Lednev, N-Q Ye, R E Hester, J N Moore J. Phys. Chem. 100 13338 (1996) 167. J Azuma, N Tamai, A Shishido, T Ikeda Chem. Phys. Lett. 288 77 (1998) 168. V Stern, P Farmanara, W Radloff J. Chem. Phys. 112 4460 (2000) 169. W Radloff, Th Freudenberg, H-H Ritze, V Stern, K Weyers, F Noack Chem. Phys. Lett. 245 400 (1995) 170. K Ohta, T J Kang, K Tominaga, K Yoshihara Chem. Phys. 242 103 (1999) 171. T Hasche, S H Ashworth, E Riedle,M Woerner, T Elsaesser Chem. Phys. Lett. 244 164 (1995) 172. S H Ashworth, T Hasche,M Woerner, E Riedle, T Elsaesser J. Chem. Phys. 104 5761 (1996) 173. J C Owrutsky, A P Baronawski J. Chem. Phys. 110 11206 (1999) 174. R S H Liu, J R Edman J. Am. Chem. Soc. 90 213 (1968) 175. R S H Liu, D MGale J. Am. Chem. Soc. 90 1897 (1968) 176. V L Ermolaev, E B Sveshnikova Opt. Spektrosk. 28 601 (1970) a 177. G D Gillispie, E C Lim Chem. Phys. Lett. 63 355 (1979) 178. R O Campbell, R S H Liu J. Am. Chem. Soc. 95 6560 (1973) 179. R A Keller Chem. Phys. Lett. 3 27 (1969) 180. Sh Kobayashi, K Kikuchi, H Kokubun Chem. Phys. Lett. 42 494 (1976) 181. K Kikuchi, H Fukumura, H Kokubun Chem. Phys. Lett. 123 226 (1986) 182. N A Borisevich, L M Bolot'ko, A V Doronin, A A Sukhodola Opt. Spektrosk. 63 1036 (1987) a 183. G A Ketsle, L V Levshin, S N Letuta Opt. Spektrosk. 68 344 (1990) a 184. G A Ketsle, L V Levshin, Yu N Lantukh, S N Letuta Izv. Akad. Nauk SSSR, Ser. Fiz. 54 473 (1990) 185. J M Larkin,W R Donaldson, T H Foster, R S Knox Chem. Phys. 244 319 (1999) 186. S Reindl, A Penzkofer Chem. Phys. 211 431 (1996) 187. P C C Lee,M A J Rodgers Photochem. Photobiol. 45 79 (1987) 188. V V Ryl'kov, E A Cheshev Opt. Spektrosk. 63 1030 (1987) a 189. Y Yaas,M Klissinger, S Zilberg (Eds) Conical Intersection in Photochemistry, Spectroscopy and Chemical Dynamics; Chem. Phys. 259 (2 ± 3) (Spec. Issue) (2000) a�Opt. Spectrosc. (Engl. Transl.) b�Physics-Uspekhi (Engl. Transl.) c�Dokl. Chem. Technol., Dokl. Chem. (Engl. Tr

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC

摘要:

Russian Chemical Reviews 70 (6) 491 ± 508 (2001) Design of site-specific RNA-cleaving reagents V N Silnikov, V V Vlassov Contents I. Introduction II. General problems of the design of site-specific RNA-cleaving reagents III. Catalysts of RNA cleavage IV. Site-specific RNA-cleaving reagents Abstract. cleaving of capable compounds chemical on Data Data on chemical compounds capable of cleaving ribonucleic acids under physiological conditions are systematised ribonucleic acids under physiological conditions are systematised and of design the of possibilities The generalised. and generalised. The possibilities of the design of site-specific site-specific RNA-cleaving reagents are considered. The bibliography includes RNA-cleaving reagents are considered. The bibliography includes 245 245 references.references. I. Introduction Reagents capable of performing efficient and specific cleavage of nucleic acids (NA) are required in molecular-biological investiga- tions, used for the development of genetic-engineering techniques and employed in genetic engineering and gene therapy. Selective NA-specific cleaving reagents can be used for the preparation of biologically active gene-targeting compounds capable of sup- pressing or modulating expression of particular genetic programs. The design of such compounds offers possibilities of influencing the origin of pathological states.1 ±4 This line of investigation can revolutionise the strategy of treatment of virtually all diseases. The construction of NA-cleaving reagents for all the above- mentioned applications is a complicated problem because cleav- age should proceed under nearly physiological conditions. Prep- arations used as therapeutic drugs should act strictly selectively on particular sequences in the compositions of certain types of nucleic acids.Side processes must be excluded, and the efficiency of the reaction with the target nucleic acid should be high to suppress functioning of the latter as completely as possible. Since the design of gene-targeting chemical reagents is a topical problem, this line of investigation has been developed intensively in recent years. Presently, the following types of compounds are primarily used for DNA cleavage. 1. Compounds containing variable-valence metal complexes as reactive groups.The latter catalyse oxidation processes result- ing in nucleoside degradation.5 ±7 2. Enediynes capable, under certain conditions, of generating biradical species, which react with both DNA chains and cause their cleavage.8 V N Silnikov, V V Vlassov Novosibirsk Institute of Bioorganic Chemistry, Siberian Branch of the Russian Academy of Sciences, prosp. Akad. Lavrent'eva 8, 630090 Novosibirsk, Russian Federation. Fax (7-383) 233 36 77. Tel. (7-383) 233 37 62. E-mail: Silnik@niboch.nsc.su (V N Silnikov) Tel. (7-383) 233 33 28 (V V Vlassov) Received 11 July 2000 Uspekhi Khimii 70 (6) 562 ± 580 (2001); translated by T N Safonova #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n06ABEH000610 491 491 492 499 3.Alkylating reagents whose reactions weaken the glycoside bond in modified nucleosides resulting in their degradation.9, 10 Recently, reagents cleaving DNA according to a photooxida- tion mechanism11, 12 as well as reagents based on natural and synthetic metal-independent antibiotics (varacin, lissoclinotoxin, etc.) generating hydroxyl radicals 13 have gained wide acceptance. Mechanisms of DNA hydrolysis, which occurs through interac- tions of radicals with various nucleic-acid elements, were consid- ered in detail in reviews.5, 14, 15 Unlike DNA, RNA are substantially more stable to oxidative cleavage 16 (in particular, under the action of metal-dependent antibiotics, such as bleomycin,17, 18 and enediyne antibiotics, such as neocarzinostatin, esperamycin, etc.19) due to higher stability of the glycoside bond in ribonucleotides compared to that in deoxyribonucleotides. At the same time, hydrolytic cleavage of RNA proceeds much more readily.Recently, considerable study has been given to the mecha- nisms ofRNAcleavage with various compounds and to the design of catalysts for directed RNA cleavage. Data accumulated in this field and the construction of the first chemical RNA-cleaving agents exhibiting higher specificity compared to that of natural enzymes call for generalisation and analysis of the results of the investigations. A very interesting line of investigation aimed at designing RNA-targeting agents involves the construction of ribozymes, viz., catalysts based on RNA or DNA.Different classes of ribosimes, procedures for their construction and the mechanisms of RNA cleavage under the action of these catalysts were considered in detail in reviews.20 ± 24 The results of studies in the field of construction of artificial catalysts whose active groups catalysing hydrolysis of phospho- diester bonds are involved in conjugates with oligonucleotides, peptides and intercalates have not been systematised. The present review is intended to close this gap and is devoted to the main classes of compounds capable of catalysing RNA depolymerisa- tion under physiological conditions and to their use in the preparation of `synthetic ribonucleases' for directed cleavage of particular sequences of ribonucleic acids.II. General problems of the design of site-specific RNA-cleaving reagents Compounds capable of performing selective site-specific RNA cleavage must, like enzymes, contain two domains, viz., an RNA- binding domain, which can form a specific complex with partic- ular RNA regions (target regions), and an RNA-hydrolysing domain performing hydrolysis of phosphodiester bonds adjacent to the binding sites or located in their vicinity (Fig. 1). Some492 B O 1 OH O O7 O P B O OOH O O7 O P B O OOH O O7 O PO B, heterocyclic base residue. Figure 1. Overall structure of synthetic ribonucleases; (1) an RNA-hydrolysing fragment or a reactive group, (2) a linker, (3) an RNA-binding fragment.peptides forming complexes with particular nucleotide sequences or elements of the three-dimensional RNA structure as well as low-molecular-weight organic compounds, such as intercala- tors,25 polycations 26 and aminoglycosides,27 can provide selectiv- ity of RNA binding. Selected aspects of interactions of low- molecular-weight ligands with nucleic acids have been considered previously.28 ¡¾ 30 Oligonucleotides and their analogues are the most promising affine structures for directed action on RNA. A complementary oligonucleotide can be directed at any nucleotide sequence of interest. An additional advantage of these reagents is the fact that all these compounds can be synthesised with the use of a universal technology.Nucleotide sequences consisting of 15 ¡¾ 20 units are unique even in genomes with complex organisa- tion, this length of complementary oligonucleotides being suffi- cient to provide recognition and efficient binding with an RNA target under physiological conditions. Recently, a series of oligo- nucleotide analogues, which are resistant to cell enzymes and can form more stable complexes with nucleic acids than those of natural oligonucleotides, have been designed.31 ¡¾ 36 Natural RNA-hydrolysing compounds, viz., enzymes and antibiotics, can be used as RNA-cleaving groups. Thus com- pounds, which can perform selective RNA cleavage in the vicinity of sequences involved in binding of conjugates, were synthesised based on oligodeoxyribonucleotides and natural enzymes, viz., ribonucleases A, S 37, 38 and H.39, 40 However, enzymes contain their own RNA(DNA)-binding affine structures along with catalytic centres, which is necessarily reflected in the specificity of such conjugates and in the effects caused by these conjugates in biological systems.Due to this fact as well as to the complexity of the synthesis of these conjugates, enzymes and natural antibiotics show little promise as reactive groups. RNA-hydrolysing fragments or reactive groups whose prop- erties have little effect on the recognition of nucleic acid by a conjugate are agents of choice. III. Catalysts of RNA cleavage The presence of 20-hydroxyl groups in RNA makes possible hydrolytic cleavage of these biopolymers according to the mech- anism of intramolecular transesterification. RNA cleavage by this mechanism is widespread in nature, for example, in reactions 2 3 V N Silnikov, V V Vlassov catalysed by ribonucleases and ribozymes.It is reasonable to employ processes occurring in active centres of ribonucleases and ribozymes in the construction of RNA-cleaving chemical reagents. It should be taken into account that high efficiency of natural enzymes results from functioning of complex active centres that arise in the course of evolution, while reactive groups forming a catalytic centre taken alone are generally weak catalysts of the corresponding processes. In the construction of artificial catalysts for the action on biopolymers, it is very important to provide the necessary spatial arrangement of all participants of the catalytic process by using the corresponding linkers.1. Mechanism of hydrolytic cleavage of ribonucleic acids Many enzymes and ribozymes cleave ribonucleic acids according to the mechanism of intramolecular transesterification through the nucleophilic attack of the oxygen atom of the 20-hydroxyl group at the phosphorus atom. This reaction can proceed by two stereochemical mechanisms. One mechanism is realised when the angle a (Fig. 2) is close to 180 8 and the distance d between the B O OO7 O H O P B O d aO OH O 7O O PO B, heterocyclic base residue. Figure 2. Principal parameters responsible for the reactivity of phospho- diester bonds in RNA; d is the 20-O7P distance (A) and a is the 50-O7P730-O angle.reacting groups <3 A (the linear mechanism), whereas another mechanism takes place when the angle a is close to 90 8 (the related mechanism). In the former case, the AE mechanism (an analogue of the SN2 mechanism) is realised. The latter mechansim requires an additional stage, viz., migration of the leaving group from the equatorial to apical position (pseudorotation).41 Pseudorotation was postulated for hydrolysis of a number of model com- pounds.42 ¡¾ 44 It was assumed that 30! 20-isomerization in ribo- nucleotides also involves pseudorotation.45 However, hydrolysis proceeded according to the linear mechanism in all cases for which the mechanism of catalytic RNA cleavage under the action of ribozymes or enzymes was established.Possible mechanisms of hydrolytic RNA cleavage under various conditions were consid- ered in detail in a number of reviews and original studies.45 ¡¾ 47 The structural parameters of the regions of some RNA in which spontaneous hydrolysis is highly probable were analysed.48 The dimensionless parameter F varying from 0 to 1 (for notations of a and d, see Fig. 2) was used as an integral value characterising stability of phosphodiester bonds in RNA sequences. F a a ¢§ 45 O3AU3 . d 3 180 ¢§ 45 It appeared that the parameter F characterising the mutual spatial arrangement of the oxygen atom of the 20-hydroxyl group, the phosphorus atom and the oxygen atom at the 50-terminus ofDesign of site-specific RNA-cleaving reagents the leaving nucleoside correlates well with stability of the inter- nucleotide bonds.The structural elements with F=0.5 ± 0.8 correspond to sites of spontaneous hydrolysis, whereas stable regions are characterised by the parameter F of 0.1 ± 0.15. Hydro- lytically labile sites with low values of the parameter F are rare exceptions. Generally, heterocyclic bases in these sites form either noncanonical GA pairs or CpA(UpA) sequences in single- stranded regions in which short-lived structures (which are able to undergo cleavage according to the linear mechanism, but which are not detected by NMR spectroscopy and X-ray diffraction analysis) can, apparently, arise. Low values of the parameter F are also typical of double- stranded RNA regions in which the phosphodiester bonds are characterised by high stability.It should be noted that experimental data on the high hydro- lytic lability of some sequences in RNA are available. However, no explanation for these data was offered. Thus, cleavage of the phosphodiester bonds in the CpA(UpA) sequences in single- stranded RNA occurs most readily, which is not clearly under- stood.49 These sequences in short single-stranded oligoribonu- cleotides undergo cleavage in the presence of various proteins (T7 RNA polymerase, T4 polynucleotide kinase, lysozyme or trypsin) and some polymers, such as polyethylene glycol, dextran or polyvinyl pyrrolidone.50 ± 52 The efficiency of cleavage depends on the length of the oligonucleotide and on the nature of the flanking nucleotides.Thus, the ratio between the rates of hydrol- ysis of the bonds in the UA sequence for GUAC, GUAU and CGUA is 1 : 6 : 16, whereas the corresponding ratio for GUAA, CGUAA and UCGUAA is 1 : 4 : 6. The change of the nucleotide residue X at the 30-terminus in the XUAA tetranucleotide has virtually no effect on the rate of hydrolysis, whereas the efficiency ofGUAXcleavage decreases in the seriesX=G>A>U>Cby a factor of three.53 It was also demonstrated that the UpA and CpA sequences in the loop regions of RNA are cleaved in the presence of nonionic [Triton X-100 (1) or Genapol X-100 (2)] or zwitterionic [zwitter- gent 3 ± 10 (3) or an amide derivative of cholic acid (4)] detergents at concentrations corresponding to the critical concentrations of micelle formation.54 Efficiency of hydrolysis is independent of the micelle size.The maximum rate of hydrolysis is achieved in the case of the loop consisting of five nucleotide units.50 Anionic detergents, such as sodium dodecyl sulfate or deoxycholic acid, suppress cleavage. H ButCH2CMe2 O (CH2)2O 10 1 + H (CH2)2O (CH2)5O 3 10 (CH2)3NR2(CH2)3SO¡ 3 2 Me Me HO Me CHCH2CNH(CH2)á3 NCH2CHCH2SO¡3 Me O OH Me 4 HO It can be suggested that the compounds 1 ± 4 influenceRNAin such a way that in some of its regions, viz., in the pyrimidine ± purine sequences, the oxygen atom of the 20-hydroxyl group of ribose and the internucleotide phosphate are brought into prox- imity and the angle a is increased, the resulting geometry being most favourable for spontaneous cleavage of the phosphodiester bonds.However, the mechanism of this reaction as well as the reasons for high lability of the CpA and UpA fragments still remain to be elucidated. 2. Hydrolysis of ribonucleic acids involving metal ions and metal complexes Presently, abundant data on hydrolysis of ribonucleic acids involving various metal ions are available. Destruction of ribonu- cleic acids under the action of the Zn2+, Ni2+, Ca2+, Co2+, Cu2+, Mg2+, Mn2+, Pb2+, Al3+ and Ln3+ ions and ions of other lanthanides has received the most study. Some of these ions can cause RNA cleavage under physiological conditions.Metal ions are involved in active centres of many enzymes hydrolysing phosphodiester bonds, such as phosphomono-,55 phosphodi- 56, 57 and phosphotriesterases.58 In some cases, these ions serve as cofactors necessary for functioning of ribozymes taking part in chemical catalysis of cleavage of phosphodiester bonds.59 Below are considered the schemes of cleavage of phos- phodiester bonds in RNA with the participation of metal ions. Metal ions can be involved in catalytic processes as Lewis acids by reacting directly with the oxygen atom of the phosphate group (A),60 with a nucleophile (for example, with the hydroxide anion) (B) 61 or with the oxygen atom of the leaving group (C). In addition, metal ions can be indirectly involved in catalysis.Thus, these ions can either coordinate a water molecule and act as intramolecular general acid catalysts (D) or coordinate the hydroxide anion and act as intramolecular general base catalysts (E or F).62, 63B OOH O O O P O7 Mn+ B O OOH A O B O OH Mn+ O P 7O P O H OH O O B OH D Apparently, hydrolytic RNA cleavage with different metal ions can proceed by one or another mechanism or according to a combination of them. Two main models of the participation of metal ions in hydrolysis of phosphodiester bonds were proposed. According to the model G, the metal ion and the hydroxide anion coordinated to this ion are involved in the catalytic reaction.64 ± 68 The modelHis based on the mechanism accepted for such natural enzymes as alkaline phosphatase andDNApolymerase I 57, 69 and assumes the participation of two metal ions one of which coordinates the hydroxyl group and the second ion activates the oxygen atom of the leaving group.Most of the results obtained in studies of RNA hydrolysis catalysed by metal ions can be interpreted based on both the models G and H.70 X-Ray diffrac- tion data for some ribozymes provide evidence in favour of the reaction mechanism involving one metal ion.64, 71 However, information on substantial acceleration of RNA hydrolysis 493 B O B OOH O O OH 7OH O7 O P P 7 Mn+ O B O O O O B Mn+ OH OH C B B O O B OH O 7 O OH O 7 7 Mn+7 O O P OH OH O H Mn+ H O O B O B O OH OH F E494 under the action of the Mg2+±Pb2+ (see Ref. 72), Nd3+±Pb2+ (see Refs 73 and 74), La3+±Fe3+ and La3+±Sn4+ pairs 75 can be interpreted only within the framework of the model H.B B O OOH OH O O 7 7 OH OH O7 O7 O P O P Mn+ Mn+ O O Mn+ H G Using the ApA dinucleotide as an example, it was demon- strated that the phosphodiester bond can be hydrolysed under the action of both the hydroxy complex of La3+ (see Ref. 76) and the binuclear hydroxy complex [La2(OH)]á5 .77 The rate of hydrolysis was increased by four orders of magnitude on going from the mono- to the binuclear complex.78 It can be suggested that hydrolysis can proceed involving either one or two metal ions depending on the reaction conditions. Within the frameworks of all mechanisms proposed, metal ions, which are stronger Lewis acids and have a larger coordination sphere, should be more efficient catalysts of hydrolytic cleavage.In this connection, considerable study should be given to cleavage of phosphodiester bonds under the action of the Ce4+ (see Ref. 79) and Zr4+ ions.80, 81 The rate of hydrolysis of phosphodiester bonds in RNA catalysed by metal ions depends on the nature of the flanking nucleosides. Thus, hydrolysis of the phosphodiester bond in ApA in a solution of the Tm3+ salt (0.01 mol litre71) proceeded 4.8 ± 5.1 times faster than that in UpU.82 In studies of hydrolysis of six different nucleotide pairs (UpG, CpC, GpA, ApU, UpAand ApA) in solutions of the Lu3+ salt (0.01 mol litre71), the reaction rate for the UpG dinucleotide, which underwent hydrolysis most rapidly, was found to be 2.2 times higher than that for ApA, which is the most stable under these conditions.82 The presence of the 30-terminal phosphate group leads to substantial acceleration of hydrolysis of the internucleotide phosphate bond in dinucleotides due, apparently, to more efficient cation binding.Thus, hydrolysis of the phosphodiester bond in ApAp in the presence of Zn2+ ions proceeded more rapidly than that in ApA. At the same time, the presence of the 20-terminal phosphate group or the 20,30-cyclo- phosphate group exerts only a slight effect on the rate of hydrolysis.83 Analogous results were obtained in other studies as well.84, 85 It was shown that the 30-terminal phosphate group in oligodeoxynucleotides containing a ribonucleotide unit, viz., Up(Tp)nTp or Up(Tp)nT (n=0 ± 4 or 7), at the 50-terminus accelerates hydrolysis even if it is spaced at seven nucleotide units.86 Apparently, the observed 5 ± 20-fold acceleration of hydrolysis of poly(U) compared to UpU under the action of the Mg2+, Co2+, Ni2+, Cd2+ or Zn2+ ions is attributable to the polyelectrolyte effect of the total negative charge of poly(U).87 The ability of metal ions to catalyse hydrolysis varies over a wide range depending on the nature of the substrate.Thus, the activity of metal ions in the hydrolysis of ApA decreases in the series Tm = Lu > Y > Nd = Eu = Sm >Ce = Sc = Gd = Tb> Pr= Dy > Ho = Er > Yb > La.76 At the same time, the activity of metal ions in hydrolysis of 30,50-cycloadenosinemono- phosphate (which is one of the intermediate hydrolysis products of ApA) under analogous conditions decreases in the series Ce Pr>Nd=La>Y=Sm=Dy>other lanthanides.88 Evidently, the differences in the rate of hydrolysis in the dunucleotide pairs as well as in more complex substantces arise from the different complex-forming ability of various fragments of nucleic acids.Problems of complex formation of metal ions with nucleic acids were considered in detail in a review. 89 Hydrolysis of RNA in the presence of various metal ions is widely used for revealing the cation-binding sites in nucleic acids.90 ± 94 The magnesium, lead and lanthanide ions can catalyse selective cleavage of phosphodiester bonds in the regions of some specific elements of the three-dimensional RNA structure.95 ± 99 The sensitivity of these reactions to the orientation of the ligands was used in studies of the RNA structure in solution and for comparison of the three-dimensional structures of mutant tRNA and 5SRNA.100, 101 It was demonstrated 102 that sequences of ribonucleic acids in complexes with oligodeoxynucleotides do not undergo hydrolysis in the presence of Eu3+ or La3+ com- plexes, apparently, because the conformational flexibility of RNA, which is necessary for efficient ion binding and for trans- esterification, is limited upon formation of the duplex.In imper- fect complexes, loop single-stranded regions are cleaved more rapidly.103 The effects of the nucleotide sequence and the size of the loop region on the efficiency of RNA hydrolysis in the presence of the Pb2+ and Zn2+ ions as well as in the presence of their complexes were examined in detail.104 The efficiency of hydrolytic cleavage was demonstrated to be increased as the size of the loop increases and as the hydrolysed phosphodiester bond moves away from the stalk fragment.The results obtained confirmed once again the necessity of conformational flexibility of the ribosephosphate fragment subjected to hydrolysis. Due to the ability of complexes of lanthanides,105, 106 cop- per,107 zinc and some other metals 108 to catalyse hydrolysis of phosphodiester bonds, they can be considered as potential hydro- lytically active groups in the synthesis of conjugates with organic compounds and oligonucleotides.The lanthanide ions, which are strong Lewis acids 109 and have a flexible coordination sphere,110 are believed to be the most promising catalysts. The construction of highly stable metal complexes, which are simultaneously characterised by the high acidity of the central atom, presents a contradictory problem. However, a number of metal complexes (for example, compounds 5 ± 8) possessing high stability with retention of high activity in hydrolysis were prepared.85, 106, 111, 112 It was demonstrated that catalytic RNA cleavage can proceed in the presence of zinc(II) 85 and lanthanum(III) 106 complexes.The rate of RNA hydrolysis in the presence of different metal complexes depends on pH. This suggests that, as in the case of free metal ions, hydrolysis of phosphodiester bonds occurs, apparently, with the participation of two molecules of the metal complex. In this case, one molecule acts as a Lewis acid, whereas the second molecule coordinates the hydroxide anion and acts as a base.113, 114 Me Me N N N N NLn3+ N Me Me 5 Ln=La, Eu, Tb, Lu. N n n Zn2+N N n 7 n=2, 3. 3. Reagents modelling active centres of metal-dependent enzymes catalysing phosphoester-bond cleavage Hydrolytic centres of metal-dependent enzymes, which are involved in hydrolysis of phosphoester bonds in various sub- V N Silnikov, V V Vlassov NH2 O H2NO N N La3+N N O O NH2 H2N 6 Me Me N 2+ N N Zn N 8 HDesign of site-specific RNA-cleaving reagents strates, often contain two complex-forming domains consisting of the amino acid residues His, Asp(Asn), Glu, Arg and Lys (in different combinations) and one of hydroxyl-containing amino acids, viz., Ser or Tyr.The distances between the metal ions (Ca2+, Zn2+, Mn2+, Mg2+ and Fe3+) in the active centres are in the range of 3 ± 5 A. In addition, the hydrolytic centres may contain one or several such amino acids as Glu, Arg, Lys, His or Ser. The functional groups of these amino acids are not directly bound to the metal ions and act as additional base, acid or nucleophilic catalysts (see reviews 115 ± 118).Active centres of some enzymes contain one (staphylococcus nuclease 119 and nuclease Sm from Serratia marcescens 120) or more than two metal ions (nuclea- se P1).121 Below are considered the principal structural elements of the active centres of natural metal-dependent enzymes, which hydrolyse phosphoester bonds in various substrates and contain from one to three metal ions, viz., staphylococcus nuclease (9), phosphorylase C from Bacillus cereus (10) and purpuric acid phosphatase (11). Owing to success achieved in the synthesis of stable metal complexes and progress made in the determination of the struc- tures of active centres and elucidation of the mechanisms of action of enzymes, which hydrolyse phosphoester bonds in various substrates, a broad range of synthetic catalysts, which serve as functional and structural models of natural enzymes, have been synthesised in recent years.122 ± 124 Binuclear metal complexes based on the Zn2+ (see Refs 125 ± 127), Cu2+ (see Refs 128 ± 130), Co3+ (see Refs 131 and 132) and La3+ ions 133, 134 were prepared.The efficiency of O Arg87 Arg35 NH NH R1 O HN HN NH O Asp40 HNH O HNH O P HN HN Asp21 O O R2 O O NH O Ca2+ O O O H H O Glu146 H OH H H O O H O O HN HN Glu43 Tyr113 His286 9 O R O NH O His325 O NH P NH NH O O N N HNH NH OH O Zn2+ Asn201 Fe3+ NH O O O Asp135 O O O N O HN OH O HN HN O Asp164 His323 HN Tyr167 11 O 495 hydrolysis in the presence of binuclear metal complexes appeared to be much higher than that in the case of their mononuclear analogues.This is particularly essential for the Zn ions, which are weak Lewis acids and whose mononuclear complexes do not catalyse hydrolysis.129, 135 Compounds 12 ± 16 proved to be the most efficient catalysts among all polynuclear metal complexes modelling active centres of natural enzymes. In particular, hydrolysis of the phosphoester bond in the model substrate in the presence of the complex 12, which is a structural and func- tional model of D-fructose 1,6-bis(phosphate-1) phosphatase, proceeded 103 times more rapidly than that in the presence of similar cobalt complexes.132 Analogous results were obtained for the copper complex 13.In the presence of the latter the rates of hydrolysis of phosphoester bonds in various compounds are increased by a factor of (1 ± 2.7)6104.129 It should be noted that in the case of hydrolysis of natural substrates in the presence of polynuclear metal complexes, metal ions can interact not only with phosphate residues, but also with heterocyclic bases as a result of which these compounds acquire particular substrate specificity. Thus, the relative rates of cleavage upon hydrolysis of ribonucleoside-20,30-cyclophosphates of adenosine, guanine, uri- dine and cytosine in the presence of the binuclear copper complex 15 are 1.5, 8.5, 4.3 and 1, respectively, whereas the relative rates in the presence of the complex 16 are 31.1, 1, 2.9 and 5.1, respec- tively.130 The ratio between the rates of hydrolytic cleavage of the GpG dinucleotide, which is most readily hydrolysed, and the most stable ApA dinucleotide in the presence of the trinuclear zinc complex 14 was 163 : 1.136 His128 O NH NH NH NH His142 N NZn2+ O OR1OR2 O O P H Asp55 NH H O O O O HN H H O O His69 O O NH O NH H2N N Zn2+ 2+Zn O O O Trp1 O NH O HN N N His118 NH HN NH HN O Asp122 His14 10496 OR O P H O H O7OH N N O 3+Co Co3+ N O N N N 12 Me R=4-NO2C6H4 .O NH N H N N O H Cu2+ Cu2+ N N H O N H 15 Below are listed compounds 17 ± 22 modelling hydrolytic centres of enzymes, which act as acid catalysts (20 ± 22), base catalysts (17, 18 and 22) or nucleophilic catalysts (19) and whose active centres contain functional groups that are not bound to metal ions.137 ± 140 O O 3+ N Eu (CH2)nCO2H O O 17 n=1±3.H 7O N C + N 2+Zn N N Me2HNCH2C H 19 H NMe2 Me2N N 2+ M2+ M + + NMe Me2HN NHMe2 RO OR OR OR 21 Me O H O O P 7 O O HO Cu2+ Cu2+ N N H2O N N HN HO NH NH RO OR OR OR 13 (R=(CH2)2OEt) O HN N H N N H O Cu2+ Cu2+ N H O O H H N H 16 Modelling of the active centres of such enzymes requires a more rigid mutual spatial arrangement of the reactive groups compared to that in compounds containing only metal ions with flexible and bulky coordination spheres. In most cases, the compounds 17 ± 22 proved to be only insignificantly more efficient than the basis metal complexes.N SR N N Zn2+ 18 HNHN . R=H, N Cyclodextrin N N C Cu2+ + CCH2NHMe2 20 Probably, this is attributable to the unfavourable spatial arrange- ment of the reactive centres. The efficiency of cleavage is increased by a factor of 103 ± 104 when the mutual arrangement of all participants of the catalytic process is optimum.141, 142 Models of the active centres of natural enzymes based on metal complexes containing additional functional groups, which are not involved in complex formation, were discussed in detail in a review.143 N 4. Hydrolysis of ribonucleic acids with amino compounds and cationic peptides Me N R=(CH2)2OEt Hydrolysis of RNA is catalysed by various amino compounds.144 In the studies of hydrolysis of bis(4-nitrophenyl) and bis(2,4- dinitrophenyl) phosphates with diamines of the general formula Me2N(CH2)nNMe2 (23a ± c) [n=1(a), 2 (b) or 3 (c)], two possible mechanisms of cleavage of the phosphodiester bonds were pro- posed.One of these mechanisms (A) is realised in the case of hydrolysis involving diamines as the corresponding monocations. The second mechanism (B) assumes the participation of neutral diamines. Thus hydrolysis in the presence of the diamines 23a ± c as monocations was accelerated by a factor of 870, 74 and 190, respectively, compared to hydrolysis in the presence of trimethyl- amine. In addition, the isotope effect, which was expected for V N Silnikov, V V Vlassov NH2 H2N NO2 N N M2+ M2+ R2 NH NH NH OH R1O OR1 OR1 OR1 14 NH2 M2+ H2C ; N N H R1=(CH2)2OEt; R2=H, M =Zn, Cu.NH2 N Zn2+ NH2NH N N 22Design of site-specific RNA-cleaving reagents general acid catalysis, was observed when H2O was replaced by D2O (kH : kD=1.6 for the diamine 23a). The diamines 23a ± c in the neutral form exhibit substantially higher catalytic activity (1400 ± 2040 times higher than that of trimethylamine), the isotope effect being absent. Apparently, in the latter case, the positive charge that arises in the course of the nucleophilic attack at the phosphorus atom (the transition state B) is electrostatically stabilised by the free electron pair of the second nitrogen atom.145 OR 7 7 OR O O H O7 P O7 P d+ d+ Me2N Me2N +OR OR NMe2 Me2NA B R=4-NO2C6H4 , 2,4-(NO2)2C6H3 .Hydrolysis of poly(A), poly(C) and poly(U) in solutions of polyamines (1 ± 2 mol litre71), such as ethylenediamine, triethy- lenetetramine, etc., was investigated.146 It appeared that ethyl- enediamine possesses the highest catalytic activity, the efficiency of hydrolysis (50 8C, 2 days) in a solution of ethylenediamine (1 mol litre71) being decreased in the series poly(A)>poly(U)> poly(C) (33%, 28% and 16%, respectively). Under these con- ditions, poly(G) was not hydrolysed.147 Apparently, the weak hydrolytic activity of bleomycin (metal- dependent antibiotic) in the absence of metal ions is explained by the presence of aliphatic amino groups.148 Regular peptides containing arginine and(or) lysine can accelerate hydrolysis of the phosphodiester bonds in ApAp and poly(A).149 ± 151 The regular peptides poly(Leu-Lys-Lys -Leu) and poly(Leu-Lys) capable of forming a-helices and b-folds, respectively, exhibit the highest activity in hydrolysis. In their complexes with RNA, the ratio Lys+: PO¡4 =2 : 1 was achieved.Distortion of the polypep- A OO O H H2N 7 O O P NH + A O 3 OOH O 7O P O + + O H3N NH3 A O A O O O O + P H H3N 3N + O 7O NH2 O 7O P NH + A HO A O 3 O7O OOH OH O O 7O P O P O 7O NH + H H NH + O O 3 3 3N + 3N + A, heterocyclic base residue. Figure 3. The assumed mechanism of hydrolytic RNA cleavage under the action of polycharged peptides.497 tide structure by introducing proline (Pro-Leu-Lys-Leu-Lys) or with the use of the D,L-isomers (D,L-Leu-D,L-Lys) leads to a sharp decrease in the catalytic activity of peptides. It should be noted that hydrolysis is not accelerated in the presence of poly(His), poly(Arg-Glu-Glu) and a number of other regular peptides containing various functional groups. No activity was also observed for the Ac-Leu-Lys-NHEt dipeptides. Apparently, the fact that the distances between the phosphate residues in the ribonucleotide (6.2 A) coincide with the distances between the positive charges in poly(Leu-Lys-Lys-Leu) (6 A) and poly(Leu- Lys) (6.9 A) plays a decisive role in the manifestation of the catalytic activity exhibited by the above-mentioned peptides in RNA hydrolysis (Fig.3).152, 153 On the whole, both polyamines and regular peptides ehxibit low hydrolytic activity. Thus, the degree of depolymerisation of ApAp under the action of the most active regular peptide, viz., poly(Leu-Lys), at 50 8C (pH 8.0) was 85% after 7 days.150 After 2 days, the degree of depolymerisation of poly(A), poly(U) and poly(C) under the action of the most active polyamines at 50 8C (pH 8.0) varied from 16% to 69%.147 5. Reagents modelling active centres of metal-independent ribonucleases Of metal-independent ribonucleases, ribonucleasesAand T1 have received the most study. Both enzymes were found to cleave phosphodiester bonds in RNA according to the mechanism of intramolecular transesterification under conditions of acid ± base catalysis.The scheme of the `classical' two-step mechanism of acid ± base catalytic hydrolysis of phosphodiester bonds in ribo- nucleic acids under the action of RNase A involves transester- ification to form cyclic phosphodiester 24 followed by its hydrolysis. The abstraction of the proton from the 20-hydroxyl group of the ribose residue under the action of the base 25 and the transfer of the proton from the fragment 26 to the oxygen atom of the leaving group are the key stages of this mechanism (Scheme 1).154, 155 The details of interactions of the amino acid residues in the active centre both with each other and with the substrate have consistently been refined.156 ± 159 Breslow and coworkers 160 proved that the active centres of ribonulceases can, in principle, be modelled by small molecules.The authors found that RNA hydrolysis was catalysed by a neutral concentrated (1 ± 2 mol litre71) imidazole buffer and then synthesised a conjugate of cyclodextrin with molecules containing imidazole fragments. This conjugate catalysed hydrol- ysis of a model substrate, viz., tert-butylcatechol 1,2-cyclophos- phate.161 Hydrolysis ofRNAwith imidazole conjugates or an imidazole buffer proceeded most efficiently at pH 7, which corresponds, as in the case of ribonuclease A, to the mechanism of general acid ± base catalysis, i.e., one protonated and one nonprotonated imidazole fragment were involved in hydrolysis.Breslow exam- ined three possible bisimidazole conjugates of cyclodextrin and found that only one of them, which is characterised by the spatial arrangement of the imidazole rings most favourable for the cleavage of the phosphodiester bond, exhibited catalytic activity in hydrolysis of the model phosphodiester.162 Some conjugates of cyclodextrin containing various combinations of the imidazole and amino groups, cyclic and acyclic ethers and metal complexes also possess phosphoesterase activity with respect to model substrates.163 ± 166 Prakash et al.167 proposed a structure modelling ribonuclease A in which the same imidazole fragment acts as a base and a proton donor. The model proved to be suitable in the presence of zinc ions, which stabilise the transition state and cause an increase in the electrophilicity of the phosphorus atom (struc- ture 27, Scheme 2). The authors believed 167 that hydrolytic cleavage of the phosphodiester bond in the diribonucleotide proceeds according to the related transfer mechanism (Section III.1).It was assumed that the leaving group in the structure 27 is located rather close to the 20-hydroxyl group. The proton transfer from the 20-hydroxyl group of the substrate to the498 His119 + 7 N H O O H N Asp121 26 His119 7 N O O H N Asp121 B, heterocyclic base residue. O HNON HN NH H2N N N O N N HO OH 27 H2N N N N HO imidazole fragment and the stage of pseudorotation are followed by the cleavage of the phosphodiester bond involving a proton of the same imidazole fragment (intermediate 28).This mechanism requires an additional stage of pseudorotation; however, the process can involve the same histamine fragment as proton donor and proton acceptor. More recently, hydrolytic RNA cleavage under the action of imidazole and its conjugates with cationic and intercalating compounds has been studied in detail.168, 169 It was demonstrated that these compounds can be used as structural probes in studies of the secondary structures ofRNAand RNA±deoxyoligonucleo- tide complexes in solutions.168, 169 B O O O H N O 7O P 25 B O OOH OP O 7O O B O His12 + NH H N O O P O 7O O H H OH O N O O H P O 7 O O Zn2+ O HO NH N N + O O H NH O7 P Zn2+ 7 N O O O 28 OH His12 B O His119 NH 7 O O H N N + O O P 7O Asp121 24 O His119 + 7 NH O O H N + O 7O P Asp121 OH O OH HN O N O +NH HN O O NH H2N N O P 7 N O O N O N Zn2+ HO OH O NH NH N O NH H2N N N O OH N N HO OH Hydrolysis with both an imidazole buffer and imidazole conjugates proceeds predominantly in single-stranded regions of RNA.Apparently, this is associated with the lower conforma- tional flexibility of nucleotides in double-stranded regions and the less favourable spatial arrangement of the 20-hydroxyl group with respect to the phosphorus atom of the internucleotide phosphate group.It should be noted that the rate of hydrolysis with an imidazole buffer is independent of the nucleotide sequence in the region subjected to cleavage (Fig. 4), whereas imidazole conju- gates have a pronounced tendency for cleavage at the pyrimi- dine ± purine regions.168 V N Silnikov, V V Vlassov Scheme 1 B HO O His12 + + OH O + NH HN O 7O P O O B His12 OH + NH N O Scheme 2 Stage of pseudorotation O HO O NH N + O O O P O7 ODesign of site-specific RNA-cleaving reagents ACCG G C C G C G G C C G G C A A A U U U U AA C C U U G A G A A U U T C D GGD C A G C C C C C G G G G A G GG C A G C C G G U C GCG C UG U C Figure 4.Structure and major cleavage sites of tRNAAsp. The most probable and less probable sites of cleavage under the action of an imidazole buffer are indicated by single and dashed arrows; the sites of cleavage under the action of imidazole conjugates are indicated by doubled arrows. Modelling of the active centre of staphylococcus nuclease (nuclease S) is also of interest. The structure and the mechanism of action of nuclease S were established based on X-ray diffraction data, results of kinetic studies and data on substitutions of amino acids in the active centre of the enzyme.170 ± 172 It is assumed that due to coordination by the guanidinium groups, the phosphate ion adopts the conformation favourable for hydrolysis with nuclease S.Simultaneously, the five-coordi- nate intermediate state of the phosphorus atom is stabilised. Although being a metal-dependent enzyme, nuclease S, like other compounds containing guanidinium groups, can be bound to the phosphate anions 173, 174 and accelerate hydrolysis of the phosphodiester bonds in the absence of metal ions.175 Hydrolysis of phosphodiester bonds under the action of compounds contain- ing the guanidinium fragments is accelerated in the presence of external or internal nucleophiles.176, 177 Yet another type of hydrolytic centre combining elements of the active centre of RNase A (general base catalysis with imidazole) and nuclease S (coordination of the phosphate anion and stabilisation of the five- coordinate state of the phosphorus atom) was proposed by Oost and Kalesse.178 The compounds synthesised contained the imida- zole fragment and one or two guanidinium groups bound to a steroid core through spacers of different lengths.Most of the compounds synthesised exhibited low hydrolytic activity with respect to the model substrate, viz., 20-hydroxypropyl 4-nitro- phenyl phosphate. The efficiency of hydrolysis depended on both the number of guanidinium groups (two groups were better than one group) and their positions with respect to the steroid rings as well as on the length of the linker. The most efficient construction, viz., 29, also exhibited a low hydrolytic activity with respect to UpU (k=2.461074 h71, 60 8C, pH 7.5). At the same time, simpler compounds containing only guanidinium groups [1,3- bis(guanidinium)propane] or simultaneously guanidinium and imidazole fragments {1,3-bis(guanidinium)-2[2-(1H-imidazol-4- yl)ethylamino]propane} appeared to be hydrolytically inactive.499 + + NH2 NH2 NH HN NH2 H2N NH N HN 29 HO Therefore, functional models of the active centres of natural ribonucleases can be, apparently, obtained by constructing mol- ecules which contain basic and acidic catalytic centres adopting particular orientations. Most often, synthetic ribonucleases contain imidazole resi- dues (an analogue of His12 in RNase A) or the carboxylate anion (an analogue of Glu58 in RNase T1) as a base and the protonated imidazole fragment as a proton donor (analogously to His119 and His92, respectively).Apparently, efficiency of catalysis can be improved by introducing positively charged groups (amino or guanidinium) into the catalytic centre to achieve a particular orientation of the phosphate ion with respect to the catalytic centres and to stabilise the five-coordinate intermediate state of the phosphorus atom. Probably, the introduction of the nucleo- philic group (for example, the hydroxyl group simulating the 20-hydroxyl group of ribose) into the reactive centre will allow one to construct catalysts hydrolysing phosphodiester bonds in DNA.179 IV. Site-specific RNA-cleaving reagents Site-specific RNA cleavage can be performed with the use of reagents containing groups which form complexes with particular nucleotide sequences or elements of the three-dimensional RNA structure. The most evident method for solving this problem is the use of oligonucleotides complementary to the predetermined primary sequences of the RNA target. In addition, a number of low-molecular-weight ligands possessing affinity for particular spatial structural elements of RNA are available.1. Conjugates based on polycyclic compounds Catalytically active groups can be added to polyaromatic com- pounds capable of intercalating into double-stranded regions of nucleic acids via flexible linkers. Synthetic constructions contain- ing polycyclic intercalating compounds and metal complexes are widely used for DNA cleavage.180, 181 Efforts were made to prepare compounds which can hydrolyse RNA, by adding various combinations of guanidinium, amino and carboxyl groups to intercalating compounds.Some most active representatives of such conjugates were synthesised based on acridine (30 and 31),182 ± 184 phenazine (32),185 isoalloxazine (33) 186 and anthra- quinone (34a,b).187 N O NH NHCCH2 N N NH2 O HN 30 HN N HN HN NH N N N N NH HN N N 31500 O O N HN X7 Me N+NH(CH2)2 NH NH N NH N 32 (CH2)4NHC(O)CH(CH2)4NHC(O)CHNH2 NHC(O)CHNH2 O N N N N NH N HN O HN 33 OH HN O Me N RN O 34a,b O R=H(a), CH2CO2H (b). The substrates (tRNA, mRNA and synthetic oligonucleoti- des) employed did not enable one to perform a reliable compar- ison of the catalytic activity of the conjugates 30 ± 34.Generally, the catalytic activity is determined based on the disappearance of the initial nucleotide-containing substance without analysis of the hydrolysis products. This procedure does not allow one to reveal relationships between the structures of synthetic ribonucleases and their catalytic activity. We synthesised a series of compounds, which differ by both the structure of the linkers and the combina- tion of reactive groups in the catalytic centre.186 The phenazine conjugates whose RNA-hydrolysing centres contained two imi- dazole fragments, which were separated from the intercalator by the linker consisting of 8 ± 14 C7C or C7N bonds, possessed the highest activity. The compounds containing the imidazolyl frag- ment and the carboxyl group exhibited somewhat lower activity.The conjugates 30 ± 32 and 34a exhibited pronounced specif- icity with respect to cleavage of the CpA and UpA sequences in single-stranded RNA regions.185, 186 The exceptions are the iso- alloxazine conjugate 33 (see Ref. 186) and the anthraquinone conjugate 34b (see Ref. 187) containing two carboxyl groups. Hydrolysis of tRNAPhe in the presence of the anthraquinone derivative 34a containing one carboxyl group proceeded predom- inantly at the CpA regions. However, when one more carboxyl group was introduced into the RNA-hydrolysing centre (the compound 34b), the phosphodiester bonds in the GpA and GpG sequences were predominantly subjected to hydrolysis, and GpC was hydrolyzed to a lesser degree.The low activity of most of the conjugates based on polycyclic compounds is attributable to the fact that these conjugates are inserted between RNA turns thus stabilising its double-stranded structure. In this case, the catalytic groups are located in the RNA region most resistant to hydrolysis. The use of long linkers makes possible localisation of the catalytic groups in the region of more labile single-stranded RNA fragments, which, in turn, provides higher efficiency of hydrolysis.187 2. Conjugates based on polycations Unlike conjugates based on polycyclic compounds, which tend to be bound to double-stranded fragments of nucleic acids, conju- gates based on cationic compounds are bound to RNA through Coulomb interactions, thus providing their homogeneous distri- bution over all sterically accessible phosphodiester bonds of the RNA target.Conjugates 35 ± 39 based on synthetic and natural V N Silnikov, V V Vlassov polyamines containing the imidazole and 1,4-diaminobutane fragments as the RNA-hydrolysing groups were described.188 ± 192 HN H2N NH2 NH NH O NH (CH2)2 35 N NH2 HN H2N R O NH N 36a ± cNH2C(O)CH(NH2)(CH2)4NH2 HN H2N R O NH N 37a ± c NH O R N 2X7 N+ +N C14H29 (CH2)n HN 38a ± c O 2X7 2X7 NH EtN+ +NCH2CNH(CH2)3N+ +NCH2CNHCHRCH2 N O 39a ± c 35 ± 39: R = H (a), CO2H (b), CO2Me (c). As mentioned above, 1,4-diaminobutane and imidazole at concentrations of 1 ± 2 mol litre71 promote RNA hydrolysis.Synthetic ribonucleases 35 ± 39 based on conjugates of imidazole with polyamines efficiently catalyse hydrolysis at substantially lower concentrations (1073 ± 1075 mol litre71), which is indica- tive of the synergistic action of the individual elements of these constructions. A sharp decreases in the efficiency of catalysis on going from imidazole conjugates to N-methylimidazole conju- gates 188 and the manifestation of an isotope effect (kH/kD=2.28) upon hydrolysis of dinucleosidemonophosphate CpA in water andD2O(see Ref. 169) indicate that acid ± base catalysis occurs at the limiting reaction stage. The catalytic activity of the com- pounds 35 ± 39 was investigated in experiments with synthetic oligonucleotides, various tRNA and fragments of virus RNA. The conjugates 35 ± 39 were demonstrated to provide high rates of cleavage of the phosphodiester bonds in the 50-Py-A motifs in single-stranded RNA regions.188 ± 190 It should be noted that no preferential binding of the conjugates to phosphates of the above- mentioned motifs compared to other internucleotide phosphates was assumed.The conjugates studied contained different RNA- binding fragments, while the RNA targets had different struc- tures. Hence, high instability of pyrimidine ± purine sequences in RNA can be explained only by their characteristic features. The conjugates 35 ± 39, which are able to perform efficient and specific depolymerisation of RNA under mild conditions, can be consid- ered as an alternative for RNase A in studies of the secondary structure of RNA.168, 188 Menger and coworkers 193, 194 proposed to carry out hydro- lytic cleavage of phosphodiester bonds in RNA with the use of compounds 40b and 41b, which, unlike the conjugates 35 ± 39, are nucleophilic catalysts.Due to the electronegative substituent in the b position, the aldehyde groups in the compounds 40a and 41a are hydrogenated in weakly alkaline and neutral solutions to give the highly nucleophilic catalysts 40b and 41b. Owing to theDesign of site-specific RNA-cleaving reagents positive charge of the nitrogen-containing fragment, the nucleo- philic group is localised in the vicinity of the phosphodiester bond. The authors believed that the presence of the alkyl substituent leads to the enhancement of the efficiency of hydrolysis due to micellar catalysis (by analogy with the micellar analogues of serine proteases proposed previously 195).The catalysts 40 and 41 accelerated hydrolysis of the model compound, viz., of bis- (4-nitrophenyl) phosphate, by a factor of 1800 and 750, respec- tively, compared to the rate of hydrolysis in the absence of catalysts. It should be noted that micellar catalysis led to accel- eration by only a factor of 210.193, 194 Since examples of cleavage of natural RNA substrates in the presence of catalysts of this type are lacking in the literature, the efficiency of these catalysts cannot be compared with that of compounds functioning based on the general acid ± base catalysis.Me H2O + C12H25 NCH2CHO C12H25 7H2O 40a Me 2X7 H2O + + N NCH2CHO C16H33 7H2O C16H33 N+ +NCH2CH(OH)2 41b 41a Catalysts for cleavage of phosphodiester bonds based on polyamines were constructed using methods of combinatorial chemistry.196 A mixture of several carboxylic acids was activated by water-soluble carbodiimide in the presence of polyallylamine with a molecular weight of 30 000 ± 40 000. The ratio between the components was chosen so that the total degree of acylation of the amino groups of the polymer was 5%± 45%. A large library of water-soluble polymers with a statistical arrangement of func- tional groups was obtained by varying both the ratio between the carboxylic acids in the reaction mixture and the degree of modification of the initial polyamine. 7 O O O O O P P P 7 O O 27 O O O HN NR H +NH O H +NH + O H H N H +NH 7 O O O O O P P P 7 O O 27 O O O HN NR H H +N + O H HN H HH + + NH O N 42 .R= HN(CH2)3 NH2 NH2 NH2 NH2 NH2 R1=HO2C, 4-HOC6H4 , Me HOCH2 , HS(CH2)2, C7H15; R2=Et, R3=(CH2)3NMe2 . +NCH2CH(OH)2 40b Me 2X7 The catalytic activity of the compounds synthesised was determined from the degree of hydrolysis of bis(4-nitrophenyl) phosphate in the presence of metal ions (Mg2+, Zn2+ or Fe3+). This approach made it possible to reveal a series of polymers accelerating hydrolysis by a factor of 20 000 ± 30 000. Polyamine containing 7.5% of caprylic-acid residues, 15% of imidazolyl- acetic-acid residues and 10% of 4-hydroxybenzoic-acid residues in the presence of iron ions proved to be most active (kcat/k=31 400).197 The drawback of this approach is the lack of information on the structures of catalytically active sites.At the same time, methods of combinatorial chemistry make possible the rapid determination of functional groups necessary for particular catalytic processes to proceed efficiently. Synthetic constructions based on macrocyclic polyamines are of particular interest. In these compounds, the catalytic and affine centres merge together. The efficiency of binding of macrocycles withRNAcan be additionally enhanced by addition of the former to polyaromatic compounds.198 The cleavage of the phospho- diester bond in adenosine-50-triphosphate catalysed by macro- cyclic polyamine 42 is presented in Scheme 3.Compounds of the type 42 did not exhibit hydrolytic activity with respect to nucleic NH2 N N O O N N O 27 P O N +HNH HO OH H +NH NH2 N N O O N N O P 27 O N +HNH HO OH + + H H H N 501 R2N=C=NR3 +R1CO2H n NH NH NH NH NH R1 R1 R1 R1 R1 nCH2 N , , (CH2)2NHCNH2 , NH NH Scheme 3 NH2 N N O N 27 O O N O P O P O 7 O O NR HO OH O H +HN + O NHH NH2 N N N O O N O O O27 P P7 O O O NR HO OH O +NH H H N O502 acids.199 However, the construction of synthetic catalysts based on molecules whose functional groups have a limited degree of freedom shows considerable promise.3. Conjugates based on oligonucleotides In the construction of oligonucleotide conjugates, virtually all types of the above-described catalytic groups were used (see reviews 200 ± 204). These groups were introduced into the midpoint of the oligonucleotide address (Fig. 5 a) or were added to the terminal fragments of the nucleotides (Fig. 5 b). The first mode of construction of conjugates seems to be attractive because subse- quent RNA cleavage under the action of these conjugates leads to a substantial decrease in stability of the complexes of the con- jugate with the substrate, which should be favourable for hydrol- ysis under the catalytic conditions. At the same time, the catalytic group in the complex of RNA with such a conjugate is located in the region of the duplex structure characterised by high stabil- ity.102 When located at the terminus of the oligonucleotide, the catalytic group in the resulting complex is oriented toward the more labile single-stranded structure.However, the stability of the complex afterRNA cleavage remains virtually unchanged. There- fore, exchange of the conjugate between depolymerised RNA and intact RNA is hindered, which should retard efficient functioning of the catalysts. Conjugates containing catalytic groups attached to the 50-ter- minal phosphate group differ substantially in efficiency of RNA depolymerisation. For the Lu3+ complex, which was prepared in situ based on oligo-DNA containing the iminodiacetic-acid resi- due (43) at the 50-terminus, the degree of polymerisation was 17% O RPO(CH2)6NHC(O)CH2N O7 43 R, oligonucleotide residue.(37 8C, 8 h). In the case of conjugates based on macrocyclic Eu3+ complexes 44a,b, the degree of depolymerisation was 30% (37 8C, a 1 2 4 3 2 2 Figure 5. Main approaches to the construction of superspecific synthetic ribonucleases based on oligonucleotide conjugates; (a) introduction of catalytic groups at the midpoint of the oligonucleotide address, (b) the addition to the terminal fragment; (1) is an oligonucleo- tide, (2) is an RNA target, (3) is a catalytic group, (4) is a linker. OC CH2 OH OH C CH2 O b 1 2 3 4 2 2 V N Silnikov, V V Vlassov 18 h) and 80%± 90% (37 8C, 16 h), respectively.206 The introduc- tion of the same metal complexes at position 5 of thymidine located at the midpoint of the oligonucleotide address gave rise to inactive 205 or weakly active 206 conjugates.The catalytic activity of oligonucleotide conjugates 45 containing copper complexes at position 5 of thymidine also appeared to be low. Thus the degree of cleavage upon incubation of RNA with such conjugates at 45 8C for 72 h was at most 18%± 25%.207 Me N Me N N R2 N Eu3+ R1 N N N Me NMe 44a,b O O P OR3 (a); R1=H,R2=NHCNH(CH2)6O O7 O O P OR3, R2 = H (b); R1=NHCNH(CH2)6O O7 R3, oligonucleotide residue. O (CH2)2C(O)NHCH2S HN N O N R N Cu2+ N 45 R, oligonucleotide residue. High efficiency of cleavage is achieved when hydrolytic groups are located in the region of loop structures of the RNA target (Fig.6 a). As mentioned above, phosphodiester bonds in loop regions of nucleic acids are characterised by high sensitivity with respect to cleavage under the action of reagents of different nature.72, 103, 104, 208, 209 The efficiency of RNA cleavage with conjugates forming loop structures with RNA was as high as 92%, whereas that in analogous reactions involving conjugates forming perfect complexes was at most 7%.205 In the case of addition of a metal complex at the 50-terminal phosphate group of the oligonucleotide address, hydrolysis of phosphodiester bonds occurs at a distance from one to seven nucleotide units depending on the type of the linker.204, 210 In the case of formation of an artificial loop in the vicinity of the 50-terminus, the cleavage is virtually quantitatively reoriented from the single-stranded region to the loop, the total degree of hydrolysis being increased.205 It was demonstrated that synthetic ribonucleases based on 12-unit 20-modified oligoribonucleotides, which were constructed on the above-described principle, can execute up to 40 events of RNA cleavage in the course of incubation at 37 8C during 64 h (at a reagent concentration of 161073 mol litre71) in the presence of a 50-fold excess of RNA.211 Yet another way of increasing the flexibility of the sugar- phosphate core in the target RNA region is based on the replace- ment of one nucleotide unit in the oligonucleotide address by a non-nucleotide insertion.212Design of site-specific RNA-cleaving reagents a 6 1 2 2 3 4 7 5 8 1 2 2 Figure 6.Principles of construction of synthetic ribonucleases based on conjugates containing noncomplementary regions (a) and based on module conjugates (b); (1) is oligonucleotide, (2) is an RNA target, (3) is a linker, (4) is a noncomplementary region, (5) RNA cleavage, (6) are oligonucleotides containing components of the catalytic centre, (7), formation of the catalytic centre at the RNA target, (8) RNA cleavage. O Reactive group Linker HN R O R=H, Me. O Reactive group Linker O O The ability of these oligonucleotide conjugates to cleave RNA under catalytic conditions was exemplified by conjugates 46a containing macrocyclic Dy3+ complexes.213 R3 Me O N N OXO R1P Y3+ N O7 R2 N N 46a ± d R3 Me R1 is the oligonucleotide residue; R2=H, OMe; R3=(CH2)3OH, Et; X=(CH2)3 , (CH2)6 , (CH2)6NHCOCH2; Y=Dy (a), La (b), Lu (c), Eu (d).A promising approach to the construction of RNA-cleaving oligonucleotide conjugates involves the design of module (binary) b 6 1 2 Et Et 503 reagents (see Fig. 6 b). The approach is based on separation of the catalytic centre into two fragments and the synthesis of conjugates of these fragments with oligonucleotides complementary to the adjacent regions of the RNA target. The catalytic centre arises only in the case of formation of a proper ternary complex in which two inactive structures are recombined to form a catalytically active centre.This centre can be formed either by individual reactive groups involved in natural ribonucleases (for example, two imidazole fragments) or by poorly reactive metal complexes forming more reactive binuclear catalytic centres. Undoubtedly, this approach imposes much more rigid limitations on both the structural parameters of the fragments of the reactive centre and the geometry of the conjugates as a whole. However, this approach has a number of important advantages, viz., it makes possible the improvement in specificity of the reaction with the target sequence and allows one to exclude side reactions of the conjugates with non-target biopolymers.The most efficient conjugates of 46a ± d contain complexes of rare-earth metals (La3+, Lu3+, Eu3+ or Dy3+) as catalytic groups. Dissociation of these ions from the metal complexes can lead to nonspecific cleavage of other biopolymers resulting in their high toxicity in biological systems. Recently, binuclear zinc complexes have been proposed for solution of this problem. Under physiological conditions, zinc ions by themselves cannot efficiently catalyse depolymerisation of biopolymers, whereas the efficiency of their binuclear complexes is rather high.214 Reagents based on oligonucleotide conjugates with catalytic groups, which do not contain metal ions, seem to be more promising because they allow one to exclude completely any cofactors.In combination with high specificity of the oligonucleo- tide address, the above fact may enable one to prepare compounds capable of performing efficient cleavage of the target RNA sequences leaving other biopolymers intact. However, some groups (amino and carboxyl groups, heterocyclic fragments) exhibit substantially lower catalytic activity than metal ions and metal complexes. High efficiency of hydrolysis in the presence of organic compounds is achieved only in the case of the optimum spatial arrangement of all groups involved in the catalytic process. Oligonucleotide derivatives containing ethylene- 215 or propylene- diamine 216 at the 50-terminus are the simplest RNA-cleaving conjugates, which do not contain metal ions.Cleavage of tRNAPhe under the action of a 19-unit oligonucleotide conjugate containing the diamine residue was reported.217 The efficiency of this reaction was 10% after incubation at 50 8C for 4 h. Similar efficiency was observed upon cleavage of a 30-unit model oligo- ribonulceotide under the action of an analogous conjugate.218 In both cases, the 50-CpA sequences in the region of the reagent localisation were subjected to hydrolysis. The introduction of oligoamines through a non-nucleotide insertion at the midpoint of the conjugate leads to a decrease in efficiency of hydrolysis under analogous conditions to 3%.219 At the same time, the degree of hydrolysis performed at 40 8C for 4 h with the use of a 10-unit oligopeptidylnucleotide containing diethylenetriamine at the 50-terminus instead of the 19-unit oligodeoxyribonucleotide was 30%, and hydrolysis was quantitative after 24 h.220 The target used in the latter study differed from that employed in the study 217 by the single substitution G?C as a result of which the CpApCpA fragment was present in the region of location of the hydrolytic group.Both bonds in CpA were subjected to hydrol- ysis, but to a different degree. In the absence of metal ions, conjugates containing fragments, which bear the 1,4-diaminobutyl, N-propylimidazolium or hista- mine residues at the 30-terminal phosphates of oligonucleotides, do not exhibit catalytic activity in hydrolysis of the model oligoribonucleotide. The histamine-containing compound exhib- ited weak activity upon introduction of zinc ions into the reaction mixture.221 In a latter study, the polyuridyl sequence was used as the target.The synthesis of oligonucleotide conjugates containing the (Leu-Arg)nGlyNH2 oligopeptides (n=2 ± 4) at the 50-terminal504 phosphate was reported.222, 223 Incubation of a model 14-unit oligoribonucleotide with the complementary six-unit oligodeoxy- ribonucleotide whose oligopeptide fragment consists of alternat- ing Leu-Arg pairs led to cleavage of theRNA target. In the case of the conjugate containing three Leu-Arg pairs, prolonged incuba- tion (48 h, 37 8C) resulted in insignificant cleavage (*5%) of the target at the CpU site in the vicinity of the localisation site of the oligopeptide.In experiments with conjugates bearing two or four Leu-Arg pairs, the efficiency of cleavage under these conditions reached 30% and 70%, respectively. In the latter case, two GpA sequences and, to a lesser extent, UpU located in the double- stranded region of the complex were the major cleavage sites. In the case of incubation of E. coli tRNAPhe with conjugates based on the pTCAATC oligonucleotide, which is complementary to two sequences in the anticodon and TCC loops, the phospho- diester bonds in the region of the D loop, in the TCC loop and, to a lesser extent, in the anticodon region of this tRNA were cleaved. Cleavage of the phosphodiester bonds in the non-target region of the D loop was related to the possible spatial proximity to the TCC loop possessing the partial binding site of the reagent.222, 223 Unlike the model 14-unit oligoribonucleotide, the site of tRNA subjected to hydrolysis under the action of the reagent (n=3 or 4) was independent of the length of the peptide residue, although the degree of hydrolysis of the target was also substantially decreased on going from n=4 to n=3.The maximum efficiency of hydrolysis was 80% (n=4, 48 h, 20 8C). We prepared oligonucleotide conjugates containing the bis- imidazolium fragments at the 50- or 30-terminal phosphate.224 The activity of the conjugates was tested using yeast phenylalanine tRNA as well as pre- and mini-exons of mRNA from Leishmania amazonensis. In the case of tRNAPhe, the C(61)A(62) and C(63)A(64) sequences located in the vicinity of the hydrolytically active groups were subjected to cleavage, in all cases the highest efficiency of cleavage being observed at the CpA bond separated from the binding site of the reagent by three nucleotide units.The reagent containing the catalytic group at the 30-terminal phos- phate appeared to be somewhat more active than its 50-analogue. After incubation at 37 8C for 8 h, the total efficiency of tRNA cleavage at three sites reached 60%.225 In the case of the mini-exon of mRNA from Leishmania amazonensis, the efficiency of hydro- lysis for the 17-unit oligonucleotide containing the bisimidazo- lium fragment at the 30-terminus was 100% after 5 h. Cleavage proceeded primarily at the CpA sequences (80%) in the region of the expected localisation of the reactive groups.Similar results were obtained for oligonucleotide conjugates whose 50-terminal phosphates bear synthetic fragments containing the imidazole fragment and the aliphatic amino group.226 An increase in the length of the spacer, which links the bisimidazole fragments with the oligonucleotide address, as well as the introduction of reactive groups through a non-complementary nucleotide residue led to substantial enhancement of the efficiency of hydrolysis. Incuba- tion of tRNAPhe with such conjugates at 37 8C for 3 ± 5 h resulted in quantitativeRNAcleavage (Fig. 7). Hydrolysis was completely inhibited in the presence of an oligonucleotide complementary to the cleavage sites (indicated by arrows, the oligonucleotide A, see Fig. 7).227 Dendrimeric oligonucleotide conjugates containing from 2 to 24 imidazole residues were proposed.228 This approach seems to be attractive for the synthesis of libraries of oligonucleo- tide conjugates containing not only imidazole fragments, but also other functional groups typical of the active centres of natural enzymes.Methods of combinatorial chemistry can be used not only for the construction of reactive groups, but also for the choice of the RNA region most favourable for hydrolysis. The synthesis of a library of 8-unit 20-O-modified oligoribonucleotides with random sequences was reported.229 Each oligonucleotide contains the Eu3+ complex 44 at the 50-terminus. Four hydrolysis sites were found after incubation of this library with 39-unit RNA (1 mmol litre71 of the conjugates, 50 nmol litre71 of RNA, 37 8C) for even 2 h.This result was unexpected because the D GA DG A G G R= HN HN Figure 7. Metal-independent catalytic groups R used in the construction of synthetic ribonucleases based on oligonucleotide conjugates. Sites of tRNAPhe cleavage with bisimidazole oligonucleotide conjugates are indicated by arrows; (1) oligonucleotide A, (2) oligonucleotide B. concentration of each particular conjugate in the reaction mixture was no higher than 15 pmol litre71. Two radically different approaches to the synthesis of oligo- nucleotide conjugates with various ligands, viz., the introduction of ligands in one of the stages of the oligonucleotide synthesis and the post-synthetic modification of the oligonucleotide, are of considerable current use.230 Several alternative procedures for performing both strategies are available. The first approach most often involves the automated synthesis of oligonucleotides based on amidophosphites of protected nucleosides and ligands. Mono- meric blocks containing fragments of various metal com- plexes 231 ± 235 and imidazole-containing constructions 236 ± 240 were proposed for the preparation of synthetic ribonucleases based on oligonucleotide conjugates. The successive synthesis of the oligonucleotide and catalytic fragments of the conjugate at the same solid-phase carrier is a promising variation of this approach.221, 241 However, this approach is unavailable for the majority of potential consumers of these conjugates because of the necessity of synthesising unique amidophosphites in each partic- ular case and the problems associated with the choice of protective groups and procedures for their removal in the case of the successive synthesis of oligonucleotide conjugates.On the con- trary, the synthesis of oligonucleotides containing functional groups along with 50- or 30-terminal phosphates and amino, hydroxyl and carboxyl groups was well developed by many biochemical companies due to which the post-synthetic modifica- tion of oligonucleotides shows considerable promise. Variations V N Silnikov, V V Vlassov 1 T A G C G C T A G C C G G C A U A U 2 R A T T A T A C GA G GCGAAUUU T G A C U C m 2G A G A G C2mG A G 7mG U C65A C63C U1m AGCCC T A G C U A C A GC 5mC U G G A C 5mC C G T C C G A U G A CA 5mC W A A U2mGO O HN N NH , NH O N O 2 OO O NH O N N .O O N NH O NH N O 2 ODesign of site-specific RNA-cleaving reagents O7 O7 1) Ph3P±Py2S2, MeIm R P O7 R P NHCH2 2) O O CH2NH2 O O O O HN N HN O NH N O7 NH R PO R is deoxyoligonucleotide; MeIm is methylimidazole. of the functionalisation of the above-mentioned groups were considered in the studies.230, 242 The latter approach was widely used in the synthesis of ribonucleases containing metal com- plexes,103, 205, 210, 211, 229 polyamines 215, 219 or imidazole deriva- tives.226 The soft procedure for activation of the terminal phosphate group 243 makes possible the preparation of oligonu- cleotide conjugates with molecules bearing aliphatic amino groups without protection of the imidazole 224, 225, 244 or guanidi- nium 222, 223 fragments.Beloglazova et al.245 proposed an ingen- ious approach to the preparation of oligonucleotide conjugates with molecules containing virtually any functional groups neces- sary for the design of highly efficient catalytic centres. The authors attached the catalytic centre to the terminal phosphate group of deoxyoligonucleotide by a Diels ± Alder reaction between the anthracene fragment and the double bond of a maleimide group (Scheme 4).* * * Intensive studies on the construction of synthetic catalysts for the specific cleavage of ribonucleic acids have resulted in the prepa- ration of promising compounds belonging to a new family of structural probes for investigation ofRNAconformations and for the design of conjugates of antisense oligonucleotides. Low- molecular-weight synthetic catalysts capable of cleaving RNA with an efficiency and selectivity close to those typical of natural ribonulceases are as yet unavailable. However in many cases, this efficiency is not necessary. In particular, in spite of low efficiency, synthetic ribonucleases as structural probes have advantages over natural enzymes.For example, when binding with RNA, ribonu- clease A, which is an enzyme catalysing hydrolysis predominantly of pyrimidine ± purine sites in single-stranded regions, changes the RNA structure and even leads to untwisting of double-stranded regions, which limits the possibilities of its use in structural studies. Small synthetic molecules do not destroy the structures of nucleic acids and cleave the latter only within single-stranded regions existing in the native molecule, which allows one to obtain information on the biologically active RNA structure. Studies both aimed at examining processes occurring in the active centres of enzymes and directed towards modelling of these processes in small chemical nucleases are being continued.It is hoped that synthetic nucleases, which may be highly competitive with natural biocatalysts and even outperform them in efficiency, will be prepared in the coming years. Such compounds can help in designing a new generation of highly efficient antiviral and anticancer drugs. O O O HN N HN O NH N HN NH N References 1. D G Knorre, V V Vlassov, V F Zarytova, A V Lebedev, O S Fedorova Design and Targeted Reactions of Oligonucleotide Derivatives (Raton, FL: Boca Press, 1994) 2. D S Sigman, A Mazumder, D M Perrin Chem. Rev. 93 2295 (1993) 3. P D Cook, in Antisense Research and Applications (Eds S T Crooke, B Lebleu) (Raton, FL: Boca Press, 1993) 4. C Helene, J-J Tulme, in Oligodeoxynucleotides. Antisense Inhibitors of Gene Expression (Ed.J S Cohen) (London: Macmillan, 1989) 5. D G Knorre, O S Fedorova, E I Frolova Usp. Khim. 62 70 (1993) [Russ. Chem. Rev. 62 65 (1993)] 6. W K Pogozelski, T D Tullius Chem. Rev. 98 1089 (1998) 7. C J Burrows, J G Muller Chem. Rev. 98 1109 (1998) 8. J L Hue, Y C Xue, M Y Xie, R Zhang, T Otani, Y Minami, Y Yamada, M Casaazza Proc. Natl. Acad. Sci. USA 90 2822 (1993) 9. D G Knorre, V V Vlassov Prog. Nucl. Acids Res. Mol. Biol. 32 291 (1985) 10. K Kido, H Inoue, E Ohtsuka Nucl. Acids Res. 20 1339 (1992) 11. B Armitage Chem. Rev. 98 1171 (1998) 12. B Armitage, T Koch, H Frydenlund, H OÈ rum, H-G Batz, G B Schuster Nucl. Acids Res. 25 4674 (1997) 13. T Chatterji, K S Gates Bioorg. Med. Chem. Lett. 8 535 (1998) 14.G Pratviel, J Bernadou, B Meunier Angew. Chem., Int. Ed. Engl. 34 746 (1995) 15. J Bernadou, G Pratviel, B Meunier, in DNA and RNA Cleavers and Chemotherapy of Cancer and Viral Diseases (Ed. B Meunier) (Dordrecht: Kluwer Academic, 1995) p. 211 16. H H Thorp Chem. Biol. 7 33 (2000) 17. J Stubbe, J W Kozarich Chem. Rev. 87 1107 (1987) 18. R M Burger Chem. Rev. 98 1153 (1998) 19. J-M A Battigello, M Cui, S Roshong, B J Carter Bioorg. Med. Chem. 3 839 (1995) 20. T Tuschl, J B Thomson, F Eckstein Curr. Opin. Struct. Biol. 5 296 (1995) 21. L Jaeger Curr. Opin. Struct. Biol. 7 324 (1997) 22. N G Walter, J M Burke Curr. Opin. Chem. Biol. 2 24 (1998) 23. W G Scott Curr. Opin. Struct. Biol. 8 720 (1998) 24. D Sen, C R Geyer Curr. Opin. Chem.Biol. 2 680 (1998) 25. W D Wilson, L Ratmeyer, M Zhao, L Strekowski, D Boykin Biochemistry 32 4098 (1993) 26. L Frydman, P C Rossomando, V Frydman, C O Fernandez, B Frydman, K Samejima Proc. Natl. Acad. Sci. USA 89 9186 (1992) 27. K Michael, Y Tor Chem. Eur. J. 4 2091 (1998) 28. C S Chow, F M Bogdan Chem. Rev. 97 1489 (1997) 29. M G Wallis, R Schroeder Prog. Biophys. Mol. Biol. 67 141 (1997) 30. F Leclerc, R Cedergren J. Med. Chem. 41 175 (1998) 31. M Egli Angew. Chem., Int. Ed. Engl. 35 1894 (1996) 32. M J Soth, J S Nowick Curr. Opin. Chem. Biol. 1 120 (1997) 505 Scheme 4 O HN NH N506 33. C Apple, C Leumann Chem. Biol. 5 209 (1998) 34. A M Kawasaki, T P Prakash, A S Fraser, M Prhave, G B Inamati Nucleosides Nucleotides 18 1737 (1999) 35.T P Prakash, A M Kawasaki, G Vasquez, A S Fraser, M D Casper, P D Cook, M Manoharan Nucleosides Nucleotides 18 1381 (1999) 36. S K Singh, P Nielsen, A A Koshkin, J Wengel Chem. Commun. 455 (1998) 37. R N Zuckermann, P G Schultz J. Am. Chem. Soc. 110 6592 (1988) 38. R N Zuckermann, D R Corey, P G Schultz J. Am. Chem. Soc. 110 1614 (1988) 39. Y Uchiyama, H Inoue, E Ohtsuka, C Nakai, S Kanaya, Y Ueno, M Ikehara Bioconjugate Chem. 5 327 (1994) 40. C Nakai, A Konishi, Y Komatsu, H Inoue, E Ohtsuka, S Kanaya FEBS Lett. 339 67 (1994) 41. D Barton,W D Ollis (Eds) Comprehensive Organic Chemistry Vol. 2 (Oxford: Pergamon Press, 1979) 42. F H Westheimer Acc. Chem. Res. 1 70 (1968) 43. D-M Zhou, K Taira Chem. Rev. 98 991 (1998) 44.T Uchimaru, S Kawahara, S Tsuzuki, K Matsumura, K Taira J. Mol. Struct. 469 215 (1999) 45. M Oivanen, S Kuusela, H LoÈ nnberg Chem. Rev. 98 961 (1998) 46. S Mikkola, E Stenman, K Nurmi, E Yousefi-Salakdeh, R StroÈ mberg,H LoÈ nnberg J. Chem. Soc., Perkin Trans. 2 1619 (1999) 47. E MaÈ ki,M Oivanen, P Poijirvi, H LoÈ nnberg J. Chem. Soc., Perkin Trans. 2 2493 (1999) 48. G A Soukup, R R Breaker RNA 5 1308 (1999) 49. A Bibillo, K Ziomek,M Figlerowicz, R Kierzek Acta Biochim. Pol. 46 145 (1999) 50. H Hosaka, I Sakabe, K Sakamoto, S Yokoyama, H Takaku J. Biol. Chem. 269 20 090 (1994) 51. RKierzek Collect. Czech. Chem. Commun. (Spec. Issue) 61 161 (1996) 52. R Kierzek Nucl. Acids Res. 20 5079 (1992) 53. R Kierzek Nucl. Acids Res. 20 5073 (1992) 54.A Riepe, H Beier, H J Gross FEBS Lett. 457 193 (1999) 55. D Gani, J Wilkie Chem. Soc. Rev. 24 55 (1995) 56. A Lahm, S Volbeda, D J Suck J. Mol. Biol. 215 207 (1990) 57. L S Beese, T A Steitz EMBO J. 10 25 (1991) 58. M M Benning, J M Kuo, F M Raushel, H M Holden Biochemistry 34 7973 (1995) 59. W G Scott Curr. Opin. Chem. Biol. 3 705 (1999) 60. D R Jones, L F Lindoy, A M Sargeson J. Am. Chem. Soc. 105 7327 (1983) 61. P Hendry, A M Sargeson Prog. Inorg. Chem. 38 201 (1990) 62. T C Bruice, A Tsubouchi, R O Dempsy, L P Olson J. Am. Chem. Soc. 118 9867 (1996) 63. J K Bashkin, L A Jenkins J. Chem. Soc., Dalton Trans. 3631 (1994) 64. W G Scott, J T Finch, A Klug Cell 81 991 (1995) 65. S C Dahm, W B Derrick, O C Uhlenbeck Biochemistry 32 13040 (1993) 66.S Kuusela, H LoÈ nnberg J. Phys. Org. Chem. 5 803 (1992) 67. S Kuusela, H LoÈ nnberg J. Phys. Org. Chem. 6 347 (1993) 68. R S Brown, J C Dewan, A Klug Biochemistry 30 4785 (1991) 69. P S Freemont, J M Friedman, L S Beese, M R Sanderson, T A Steitz Proc. Natl. Acad. Sci. USA 85 8924 (1988) 70. B W Pontius,W B Lott, P H von Hippel Proc. Natl. Acad. Sci. USA 94 2290 (1997) 71. W G Scott, A Klug Trends Biochem. Sci. 21 220 (1996) 72. T Pan, O C Uhlenbeck Nature (London) 358 560 (1992) 73. N Sugimoto, T Ohmichi FEBS Lett. 393 97 (1996) 74. T Ohmichi, N Sugimoto Biochemistry 36 3514 (1997) 75. M Irisawa, M Komiyama J. Biochem. 117 465 (1995) 76. M Komiyama, K Matsumura, Y Matsumoto J. Chem. Soc., Chem. Commun.640 (1992) 77. K Matsumura,M Komiyama J. Biochem. 122 387 (1997) 78. P Hurst, B K Takasaki, J Chin J. Am. Chem. Soc. 118 9982 (1996) 79. S J Frandkin Curr. Opin. Chem. Biol. 5 201 (2001) 80. J Visscher, A W Schwartz Nucl. Acids Res. 20 5749 (1992) 81. R Ott, R KraÈ mer Angew. Chem., Int. Ed. Engl. 37 1957 (1998) 82. K Matsumura, Y Matsumoto,M Komiyama Nucl. Acids Symp. Ser. 27 11 (1992) 83. J J Butzow, G L Eichhorn Biochemistry 10 2019 (1971) 84. A George, P Draganac, W R Farkas Inorg. Chem. 24 3627 (1985) 85. V M Shelton, J R Morrow Inorg. Chem. 30 4295 (1991) V N Silnikov, V V Vlassov 86. S Kuusela,M Rantanen, H LoÈ nnberg J. Chem. Soc., Perkin Trans. 2 2269 (1995) 87. S Kuusela, H LoÈ nnberg J. Chem. Soc., Perkin Trans.2 2301 (1994) 88. J Sumaoka, M Yashiro,M Komiyama Nucl. Acids Symp. Ser. 27 37 (1992) 89. S A Kazakov, in Bioorganic Chemistry (Ed. S M Hecht) (New York: Oxford University Press, 1996) p. 224 90. L S Behlen, J R Sampson, A B DiRenzo, O C Uhlenbeck Biochemistry 29 2515 (1990) 91. J Wrzesinski, D Michalowski, J Ciesiolka, W J Krzyzosiak FEBS Lett. 374 62 (1995) 92. N A Kolchanov, I I Titov, I E Vlassova, V V Vlassov Prog. Nucl. Acids Res. Mol. Biol. 53 131 (1996) 93. J Ciesiolka, M Yarus RNA 2 785 (1996) 94. I Majerfeld, M Yarus RNA 4 471 (1998) 95. R S Brown, B E Hingerty, J C Dewan, A Klug Nature (London) 303 543 (1983) 96. R S Brown, J C Dewan, A Klug Biochemistry 24 4785 (1985) 97. D Michalowski, J Wrzesinski,W Krzyzosiak Biochemistry 35 10 727 (1996) 98. J Ciesiolka, J Gorski,M Yarus RNA 1 7538 (1995) 99.M Wech, I Majerfeld,M Yarus Biochemistry 36 6614 (1997) 100. B Dichtl, T Pan, A B DiRenzo, O C Uhlenbeck Nucl. Acids Res. 21 351 (1993) 101. J Ciesiolka, S Lorenz, V A Erdmann Eur. J. Biochem. 204 575 (1992) 102. K A Kolasa, J R Morrow, A P Sharma Inorg. Chem. 32 3983 (1997) 103. D Huesken, G Goodall,M J J Blommers,W Jahnke, J Hall, R Haener, H E Moser Biochemistry 35 16591 (1996) 104. I Zagorowska, S Kuusela, H LoÈ nnberg Nucl. Acids Res. 26 3392 (1998) 105. M Yashiro, A Ishikubo,M Komiyama J. Biochem. 120 1067 (1996) 106. J R Morrow, L A Buttrey, V M Shelton, K A Berback. J. Am. Chem. Soc. 114 1903 (1992) 107. L A Jenkins, J K Bashkin,M E Autry J.Am. Chem. Soc. 118 6822 (1996) 108. D P Weiner, T Wiemann, M M Wolfe, P Wentworth Jr, K D Janda J. Am. Chem. Soc. 119 4088 (1997) 109. S J Oh, K H Song, J W Park J. Chem. Soc., Chem. Commun. 575 (1995) 110. J Sumaoka, S Miyama, M Komiyama J. Chem. Soc., Chem. Commun. 1755 (1994) 111. U Baykal, E U Akkaya Tetrahedron Lett. 39 5861 (1998) 112. R Breslow, D Berger, D-L Huang J. Am. Chem. Soc. 112 3686 (1990) 113. P R Norman, R D Cornelius J. Am. Chem. Soc. 104 2356 (1982) 114. M K Stern, J K Bashkin, E D Sall J. Am. Chem. Soc. 112 5357 (1990) 115. N StraÈ ter,W N Lipscomb, T Klabunnde, B Krebs Angew. Chem., Int. Ed. Engl. 35 2024 (1996) 116. D E Wilcox Chem. Rev. 96 2435 (1996) 117. J A Cowan Chem. Rev. 98 1067 (1998) 118. R A Kovall, B W Matthews Curr.Opin. Chem. Biol. 3 578 (1999) 119. D J Weber, A K Meeker, A S Mildvan Biochemistry 30 6103 (1991) 120. S V Shlyapnikov, E V Blaganov, V M Levdikov, V Yu Lunin, V V Lunin, A M Mikhailov, Kh Betzel', K R Raiyashankar, M Perbandt Mol. Biol. 33 454 (1999) a 121. A Volbeda, A Lahm, F Sakiyama, D Suck EMBO J. 10 1607 (1991) 122. J K Bashkin Curr. Opin. Chem. Biol. 3 752 (1999) 123. E K van den Beuken, B L Feringa Tetrahedron 54 12985 (1998) 124. E Rizzarelli, G Vecchio Coord. Chem. Rev. 188 343 (1999) 125. R Breslow, S Singh Bioorg. Chem. 16 408 (1998) 126. W H Chapman Jr, R Breslow J. Am. Chem. Soc. 117 5462 (1995) 127. S Matsuda, A Ishikubo, A Kuzuya,M Yashiro,M Komiyama Angew. Chem., Int. Ed. Engl. 37 3284 (1998) 128.M Wall, R C Hynes, J Chin Angew. Chem., Int. Ed. Engl. 32 1633 (1993) 129. P Molenveld, J F J Engbersen, H Kooijman, A L Spek, D N Reinhoudt J. Am. Chem. Soc. 120 6726 (1998) 130. S Liu, Z Luo, A D Hamilton Angew. Chem., Int. Ed. Engl. 36 2678 (1997)Design of site-specific RNA-cleaving reagents 131. Y Chung, E U Akkaya, T K Venkatachalam, A W Czarnik Tetrahedron Lett. 31 5413 (1990) 132. J S Seo, N-D Sung, R C Hynes, J Chin Inog. Chem. 35 7472 (1996) 133. A Tsubouchi, T C Bruice J. Am. Chem. Soc. 116 11 614 (1994) 134. K G Rogunathan, H-J Schneider Angew. Chem., Int. Ed. Engl. 35 1219 (1996) 135. M Yashiro, A Ishikubo,M Komiyama J. Chem. Soc., Chem. Commun. 1793 (1995) 136. P Molenveld, J F J Engbersen,D N Reinhoudt Angew. Chem., Int.Ed. Engl. 38 3189 (1999) 137. A Roigk, O V Yescheulova, Yu V Fedorov, O A Fedorova, S P Gromov, H-J Schneider Org. Lett. 1 833 (1999) 138. S D Dong, R Breslow Tetrahedron Lett. 39 9343 (1998) 139. E KoÈ vaÈ ri, R KraÈ mer J. Am. Chem. Soc. 118 12704 (1996) 140. P Molenveld, J F J Engbersen, D N Reinhoudt Eur. J. Org. Chem. 3269 (1999) 141. E Kimura, Y Kodama, T Koike, M Shiro J. Am. Chem. Soc. 117 8304 (1995) 142. R KraÈ mer Coord. Chem. Rev. 182 243 (1999) 143. J Costamagna, G Ferraudi, B Matsuhiro, M Campos-Vallette, J Canales,M Villagra'n, J Vargas, M J Aguirre Coord. Chem. Rev. 196 125 (2000) 144. M Komiyama, K Yoshinari J. Org. Chem. 62 2155 (1997) 145. M Komiyama, K Yoshinari J. Chem. Soc., Chem. Commun. 1880 (1989) 146.K Yoshinari, M Komiyama Nucl. Acids Symp. Ser. 26 23 (1991) 147. K Yoshinari, K Yamazaki, M Komiyama J. Am. Chem. Soc. 113 5899 (1991) 148. M V Keck, S M Hecht Biochemistry 34 12029 (1995) 149. B Barbier, A Brack Orig. Life Evol. Biosph. 17 381 (1987) 150. B Barbier, A Brack J. Am. Chem. Soc. 110 6880 (1988) 151. B Barbier, A Brack Orig. Life Evol. Biosph. 20 139 (1990) 152. M Perello, B Barbier, A Brack Int. J. Protein Res. 38 154 (1991) 153. B Barbier, A Brack J. Am. Chem. Soc. 114 3511 (1992) 154. H Shindo,M B Hayes, J S Cohen J. Biol. Chem. 251 2644 (1976) 155. C N Pace, U Heinemann, U Hahn,W Saenger Angew. Chem., Int. Ed. Engl. 30 343 (1991) 156. G A Sowa,A C Hengge,W W Cleland J. Am. Chem. Soc. 119 2319 (1997) 157. B D Wladkowski, L A Svensson, L Sjolin, J E Landner, G L Gilliland J.Am. Chem. Soc. 120 5488 (1998) 158. L W Schultz, D J Quirk, R T Raines Biochemistry 37 8886 (1998) 159. D M Perreault, E V Anslyn Angew. Chem., Int. Ed. Engl. 36 432 (1997) 160. R Breslow,M Labelle J. Am. Chem. Soc. 108 2655 (1986) 161. R Breslow, J B Doherty, G Guillot, C Lipsey J. Am. Chem. Soc. 100 3227 (1978) 162. E Anslyn, R Breslow J. Am. Chem. Soc. 111 5972 (1989) 163. C G Ferguson, G R J Thatcher Org. Lett. 1 829 (1999) 164. J Yu, Y Zhao, M J Holterman, D L Venton Bioorg. Med. Chem. Lett. 9 2705 (1999) 165. M J Han, K S Yoo, J Y Chang, T-K Ha Angew. Chem., Int. Ed. Engl. 39 347 (2000) 166. J Yan, R Breslow Tetrahedron Lett. 41 2059 (2000) 167. T P Prakash, S S Kunta,K N Ganesh Tetrahedron 50 11699 (1994) 168.R Giege, B Felden, V N Silnikov, M A Zenkova, V V Vlassov Meth. Enzymol. 318 147 (2000) 169. M A Zenkova, N L Chumakova, A V Vlassov, N I Komarova, A G Ven'yaminova, V V Vlassov, V N Silnikov Mol. Biol 34 456 (2000) a 170. D W Hibler,N J Stolowich,M A Reynolds, J A Gerlt, J A Wilde, P H Bolton Biochemistry 26 6278 (1987) 171. E H Serpersu, J McCracken, J Peisach, A S Mildvan Biochemistry 25 68 (1986) 172. E H Serpersu, J McCracken, J Peisach, A S Mildvan Biochemistry 27 8034 (1988) 173. R P Dixon, S J Geib, A D Hamilton J. Am. Chem. Soc. 114 365 (1992) 174. F Chu, L S Flatt, E V Anslyn J. Am. Chem. Soc. 116 4194 (1994) 175. D M Perreault, L A Cabell, E V Anslyn Bioorg. Med. Chem. 5 1209 (1997) 176.M W GoÈ bel, J W Bats, G DuÈ rner Angew. Chem. 104 217 (1992) 177. V Jubian, A Veronese, R P Dixon, A D Hamilton Angew. Chem., Int. Ed. Engl. 34 1237 (1995) 178. T Oost,M Kalesse Tetrahedron 53 8421 (1997) 507 179. M-S Muche, M W GoÈ bel Angew. Chem., Int. Ed. Engl. 35 2126 (1996) 180. S A Ross, M Pitie', B Meunier Eur. J. Inorg. Chem. 557 (1999) 181. V Steullet, D W Dixon Bioorg. Med. Chem. Lett. 9 2935 (1999) 182. J Smith, K Ariga, E V Anslyn J. Am. Chem. Soc. 115 362 (1993) 183. A Lorente, J F Espinosa, M Ferna'ndez-Saiz, J -M Lehn, W D Wilson, Y Z Yi Tetrahedron Lett. 37 4417 (1996) 184. C-H Tung, Z Wei,M J Leibowitz, S Stein Proc. Natl. Acad. Sci. USA 89 7114 (1992) 185. M A Podyminogin, V V Vlassov, R Giege Nucl.Acids Res. 21 5950 (1993) 186. V N Silnikov, N P Luk'yanchuk, G V Shishkin, R Zhizhe, V V Vlassov Dokl. Akad. Nauk 364 690 (1999) b 187. M Endo, K Hirata, T Ihara, S Sueda, M Takagi,M Komiyama J. Am. Chem. Soc. 118 5478 (1996) 188. V V Vlassov, G Zuber, B Felden, J-P Behr, R Giege Nucl. Acids Res. 23 3161 (1995) 189. D A Konevets,M A Zenkova, V N Silnikov, V V Vlassov Dokl. Akad. Nauk 360 554 (1998) b 190. D A Konevetz, I E Beck, N G Beloglazova, I V Sulomenko, V N Silnikov, M A Zenkova, G V Shishkin, V V Vlassov Tetrahedron 55 503 (1999) 191. X H Li, R Wan, Q Zhang, Y F Zhao Chin. Chem. Lett. 9 333 (1998) 192. N S Zhdan, I L Kuznetsova, A V Vlassov, V N Silnikov, M A Zenkova, V V Vlassov Bioorg. Khim. 25 723 (1999) c 193. F M Menger, L G Whitesell J.Am. Chem. Soc. 107 707 (1985) 194. F M Menger, R A Persichetti J. Org. Chem. 52 3451 (1987) 195. M L Bender,M Komiyama, in Bioorganic Chemistry (Ed. E E van Tamelen) (New York: Academic Press, 1977) p. 19 196. F M Menger, Z X Fei Angew. Chem., Int. Ed. Engl. 33 346 (1994) 197. F M Menger, A V Eliseev, V A Migulin J. Org. Chem. 60 6666 (1995) 198. M P Mertes, K B Mertes Acc. Chem. Res. 23 413 (1990) 199. M W Hosseini, A J Blacker, J-M Lehn J. Am. Chem. Soc. 112 3896 (1990) 200. M Komiyama J. Biochem. 118 665 (1995) 201. R Hiner, J Hall Antisense Nucl. Acids Drug Dev. 7 423 (1997) 202. B N Trawick, A T Daniher, J K Bashkin Chem. Rev. 98 939 (1998) 203. V V Vlassov, V N Silnikov,M A Zenkova Mol. Biol. 32 62 (1998) a 204. M Komiyama, J Sumaoka Curr. Opin. Chem. Biol. 2 751 (1998) 205. D Magda, R A Miller, J L Sessler, B L Iverson J. Am. Chem. Soc. 116 7439 (1994) 206. J Hall, D Husken, R HaÈ ner Nucl. Acids Res. 24 3522 (1996) 207. J K Bashkin, E I Frolova, U Sampath J. Am. Chem. Soc. 116 5981 (1994) 208. T Pan, B Dichtl, O C Uhlenbeck Biochemistry 33 9561 (1994) 209. I J Brittain, X Huang, E C Long Biochemistry 37 12 113 (1998) 210. J Hall,D Husken,U Pieles,H E Moser, R Hiner Chem. Biol. 1 185 (1994) 211. R Hiner, J Hall, A PfuÈ tzer, D Hmsken Pure Appl. Chem. 70 111 (1998) 212. M A Reynolds, T A Beck, P B Say, D A Schwartz, B P Dwyer, W J Daily, M M Vaghefi, M D Metzler, R E Klem, L J Arnold Nucl. Acids Res. 24 760 (1996) 213. D Magda,M Wright, S Crofts, A Lin, J L Sessler J. Am. Chem. Soc. 119 6947 (1997) 214. S Matsuda, A Ishikubo, A Kuzuya,M Yashiro,M Komiyama Angew. Chem., Int. Ed. Engl. 37 3284 (1998) 215. M Komiyama, T Inokawa, T Shiiba, N Takeda, K Yoshinari, M Yashiro Nucl. Acids Symp. Ser. 29 197 (1993) 216. A Kuzuya, Y Azuma, T Inokava, K Yoshinari, M Komiyama Nucl. Acids Symp. Ser. 37 209 (1997) 217. M Komiyama, T Inokava J. Biochem. 116 719 (1994) 218. M Komiyama, T Inokawa, K Yoshinari J. Chem. Soc., Chem. Commun. 77 (1995) 219. M Endo, Y Azuma, Y Saga, A Kuzuya, G Kawai,M Komiyama J. Org. Chem. 62 846 (1997) 220. J C Verheijen, B A L M Deiman, E Yeheskiely, G A van der Marel, J H van Boom Angew. Chem., Int. Ed. Engl. 39 369 (2000) 221. J Hovinen, A Guzaev, E Azhayeva, A Azhayev, H LoÈ nnberg J. Org. Chem. 60 2205 (1995)V N Silnikov, V V Vlassov 508 222. D V Pyshnyi,M N Repkova, S G Lokhov, E M Ivanova, A G Ven'yaminova, V F Zarytova Bioorg. Khim. 23 497 (1997) c 223. D Pyshnyi,M Repkova, S Lokhov, E Ivanova, A Venyaminova, V Zarytova Nucleosides Nucleotides 16 1571 (1997) 224. V Silnikov, G Zuber, J-P Behr, R Gige, V Vlassov Phosphorus Sulfur Silicon Relat. Elem. 109 ± 110 277 (1996) 225. V Vlassov, T Abramova, T Godovikova, R Gige, V Silnikov Antisense Res. Dev. 7 39 (1997) 226. K Ushijima, H Gouzu, K Hosono, M Shirakawa, K Kogosima, K Takai, H Takaku Biochem. Biophys. Acta 1379 217 (1998) 227. N G Beloglazova, N N Polushin, V N Silnikov, M A Zenkova, V V Vlassov Dokl. Akad. Nauk 369 827 (1999) b 228. N N Polushin Collect. Czech. Chem. Commun., Coll. Symp. Ser. 2 145 (1999) 229. D HuÈ ken, A Deichert, J Hall, R HaÈ ner Nucleosides Nucleotides 18 1507 (1999) 230. A S Butorin, G N Grimm, K Elen Mol. Biol. 34 946 (2000) a 231. A S Modak, J K Gard, M C Merriman, K A Winkeler, J K Bashkin, M K Stern J. Am. Chem. Soc. 113 283 (1991) 232. J K Bashkin, J Xie, A T Daniher, U Sampath, J L-F Kao J. Org. Chem. 61 2314 (1996) 233. G Wang, D E Bergstom Tetrahedron Lett. 34 6721 (1993) 234. J Hovinen Bioconjugate Chem. 9 132 (1998) 235. H Inoue, M Shimizu, T Furukawa, T Tamura, M Matsui, E Ohtsuka Nucleosides Nucleotides 18 1503 (1999) 236. J K Bashkin, J K Gard, A S Modak J. Org. Chem. 55 5125 (1990) 237. J K Bashkin, S M Sondhi, U Sampath, A d'Avignon New J. Chem. 18 4604 (1994) 238. N N Polushin, B-ch Chen, L W Anderson, J S Cohen J. Org. Chem. 58 4606 (1993) 239. G Wang, D E Bergstom Tetrahedron Lett. 34 6725 (1993) 240. T H Smith, J V LaTour, D Bochkariov, G Chaga, P S Nelson Bioconjugate Chem. 10 647 (1999) 241. A Yu Karyagin, T V Abramova, V N Silnikov, G V Shishkin Izv. Akad. Nauk, Ser. Khim. 536 (2000) d 242. G N Grimm, A S Boutorine, C Helene Nucleosides Nucleotides 19 1943 (2000) 243. D G Knorre, P V Alekseev, Y V Gerassimova, V N Silnikov, G A Maksakova, T S Godovikova Nucleosides Nucleotides 17 397 (1998) 244. L Yurchenko, V Silnikov, T Godovikova, G Shishkin, J-J Toulme, V Vlassov Nucleosides Nucleotides 16 1721 (1997) 245. N G Beloglazova, V N Silnikov, M A Zenkova, V V Vlassov FEBS Lett. 481 277 (2000) a�Mol. Biol. (Engl. Transl.) b�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) c�Russ. J. Bioorg. Chem. (Engl. Transl.) d�Russ. Chem. Bull. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC

摘要:

Russian Chemical Reviews 70 (6) 509 ± 533 (2001) Affinity sorbents containing nucleic acids and their fragments I G Shishkina, A S Levina, V F Zarytova Contents I. Introduction II. Solid supports and linkers III. Chemical methods for the immobilisation of nucleic acids IV. Immobilisation of nucleic acids in the course of solid-phase synthesis V. Miscellaneous methods of immobilisation of nucleic acids VI. Immobilisation of nucleic acids using enzymes and avidin ± biotin interactions VII. The use of matrices containing nucleic acids VIII. Conclusion Abstract. the for methods main the on data published The The published data on the main methods for the preparation of polymeric supports containing nucleic acids (NA) preparation of polymeric supports containing nucleic acids (NA) or their fragments (oligonucleotides) are reviewed with special or their fragments (oligonucleotides) are reviewed with special emphasis on chemical immobilisation.Some physical and phys- emphasis on chemical immobilisation. Some physical and phys- icochemical immobilisation techniques, including those based on icochemical immobilisation techniques, including those based on the use of enzymes and avidin ± biotin interactions and prepara- the use of enzymes and avidin ± biotin interactions and prepara- tion of NA-containing supports by direct oligonucleotide syn- tion of NA-containing supports by direct oligonucleotide syn- thesis on these supports are considered. A special section is thesis on these supports are considered.A special section is devoted to the application of NA-containing sorbents for the devoted to the application of NA-containing sorbents for the isolation of individual NA and proteins as well as in hybridisation isolation of individual NA and proteins as well as in hybridisation analysis including those utilising DNA chips and DNA biosen- analysis including those utilising DNA chips and DNA biosen- sors. The bibliography includes 391 references sors. The bibliography includes 391 references. I. Introduction Polymeric supports containing nucleic acids (NA) and their frag- ments have long attracted the attention of specialists in different fields of biochemistry, gene engineering and molecular biology. In the past decade, such supports have successfully been used in hybridisation analysis for research purposes as well as in medicine for the diagnostics of infectious, cancer, genetic and other diseases.NA-Containing matrices are efficient tools for isolation of individual NA and NA-dependent enzymes. The synthesis of affinity sorbents is based on the following principle: the ligands (in our case, NA, oligonucleotides or polynucleotides) are immobi- lised on a polymeric support (a matrix) through a linker or a spacer, or without them. With the advent of automated synthetic techniques, oligonucleotides have become commercially readily available, which makes them especially attractive as ligands. Native NA are also immobilised as ligands for the solution of some practical tasks. I G Shishkina, A S Levina, V F Zarytova Novosibirsk Institute of Bioorganic Chemistry, Siberian Branch of the Russian Academy of Sciences, prosp.Acad. Lavrent'eva 8, 630090 Novosibirsk, Russian Federation. Fax (7-383) 233 36 77. Tel. (7-383) 239 62 75. E-mail: igshishkina@hotmail.com (I G Shishkina) Levina@niboch.nsc.ru (A S Levina) Tel. (7-383) 239 62 24. E-mail: Zarytova@niboch.nsc.ru (V F Zarytova) Received 15 March 2001 Uspekhi Khimii 70 (6) 581 ± 608 (2001); translated by R L Birnova #2001 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2001v070n06ABEH000640 509 509 513 524 524 524 525 528 All presently known methods for immobilisation of oligonu- cleotides and NA can be conventionally divided into chemical methods, which involve covalent binding of NA to the support, and other methods, which include all the procedures demanding special equipment or technology (e.g., synthesisers, irradiation, etc.) in addition to the support, NA and chemicals.This review is devoted to the description of the main methods for the preparation of NA-containing sorbents, their properties and applications. It should be noted that these methods can be used for immobilisation of not only oligo(poly)nucleotides, but also their analogues and peptide-nucleic acids (PNA) in which the sugar-phosphate backbone is replaced by the peptide chain. A separate section is devoted to the use of NA-containing matrices for the isolation of individual NA and proteins as well as in hybridisation analysis (including DNA chips and DNA biosen- sors).II. Solid supports and linkers 1. Supports Before we pass to the consideration of various procedures for immobilisation of NA, we shall dwell briefly on the requirements placed upon polymeric supports to be used in the design of `ideal' affinity matrices. These include: (1) insolubility (to avoid the loss of the ligand attached and contamination of the isolated compound with the dissolved support); (2) accessibility of the immobilised ligand for the formation of complexes with specific biomolecules; (3) zero adsorption capacity (to prevent non-specific interac- tions); (4) high reactivity essential for chemical modification of the support surface; (5) chemical stability under conditions employed for the attachment of ligands as well as sorption and desorption of the products and regeneration of the matrix; (6) high resistance to microbes and enzymes; (7) hydrophilicity (to reduce non-specific sorption due to hydrophobic interactions); (8) commercial availability.The requirements for the nature, shape, composition and mechanical strength of the solid support vary depending on the experimental task. The most popular supports can be divided into two groups, viz., natural and synthetic ones (Table 1). High-510 Table 1. Commercially available carriers. Manufacturer Commercial name Carrier Natural supports Agarose Cellulose Dextran Pharmacia, LKB Bio-Rad IBF Whatman Pharmacia Sepharose, Superose Bio-Gel A Ultrogel Cellulose Sephadex Synthetic supports Polyacrylamide supports Bio-GelP Bio-Rad Trisacryl GF-2000, Sephacryl IBF Biotechnics Ultrogel Azlactone beads 3M Emphaze biosupport Pharmacia IBF 3M Corporation Pierce Chemical Polymethacrylate TSK-Gel Toyopearl HW supports Tosoh Corporation EM Science, Merck TESSEK Ltd.Alltech Associates Rohm Pharma Fractogel TSK HEMA Separon, Spheron Eupergit Polystyrene supports PerSeptive Biosystem Pierce Chemical Trondheim Rapp Polymers Merck Porous supports Polystyrene balls Dyno particles TentaGel MerckoGel Glass supports CPG glass Glass beads Microscope glass slides Pierce Chemical Jenco, Supelco Fisher Scientific Menzel-Glaser pressure affinity chromatography demands hard materials, whereas in column chromatography performed at normal pres- sure hardness is not a critical factor.The relatively small specific surface area in non-porous supports can largely be compensated by the reduced size (down to the micron level) of the support particles with the preservation of their satisfactory hydrodynamic characteristics. Porous supports are suited to affinity chromato- graphy where the size of the analysed molecules is much smaller than that of the matrix pores. In the majority of cases, sorbents represent spherical granules with the size of 40 to 150 mm, whereas silica gel-based sorbents for high-performance affinity chroma- tography have a particle size of about 10 mm.Modern studies often utilise membranes as supports. They are usually manufactured from various polymeric materials, such as cellulose, polyamide (nylon), polyacrylonitrile, polysulfone, poly- (MeO)3Si(CH2)3NH(CH2)2NH2 (MeO)3Si(CH2)3OCH2 (EtO)3Si(CH2)3NH2 O 1 2 O H2N HN NH O O C C O N N N NH C O O HN O O NH H2N 5 I G Shishkina, A S Levina, V F Zarytova propylene, polyesters, polyethylene, etc. Cellulose, nylon and poly(ethylene terephthalate) are the most popular supports, since they allow easy incorporation of various functional groups. Cover glasses and polystyrene immunological assay plates are widely used in hybridisation analysis. Some supports incorporating various functional groups (Table 2) accessible for immobilisation of ligands are commer- cially available; however, they demand additional modification prior to application because they partly lose their activities upon storage.Therefore, we shall consider some synthetic procedures leading to such sorbents along with the principal known methods for the preparation of affinity sorbents containing NA or their fragments. 2. Linkers and spacers Oligonucleotides and NA can be immobilised directly on the surface of the solid support; however, this can result in steric hindrance to subsequent reactions with biomolecules, especially in the case of planar supports, such as membranes and slides. To prevent this, immobilisation is carried out with the use of linkers or spacers which hold the ligand at some distance away from the surface and thus make it more accessible for further interactions.In the absence of linkers, immobilisation can be carried out by virtually all physical methods (e.g., UV irradiation, sorption on the surface by drying) or by direct synthesis of oligonucleotides on glass 1 or on aminated polypropylene.2 However, immobilisation is usually performed after incorporation of spacers into the oligonucleotide and/or on the surface of the support.3± 40 The larger is the distance between the immobilised ligands and the surface, the more accessible they are for the reaction with the dissolved molecules. Most often, the role of linkers is played by linear aliphatic chains containing different numbers of hexame- thylenediamine, 6-aminocaproic acid or ethylene glycol units.In addition to the length, an important role is played by the nature of the linker, which may govern the extent of non-specific sorption. This can be overcome through the use of linker chains consisting of hydrophilic molecules, e.g., 1,3-diaminopropan-2-ol or amino acid residues. Hydrophilic linkers of any length can be designed from phosphoroamidites of various diols 8 or by stepwise addition of 1,3-diaminopropan-2-ol and bromoacetic acid.9, 10 The linkers are usually attached to the support by the same methods as ligands. Silica gel and glass are premodified by (3-aminopropyl)- triethoxysilane (1),11 ± 13 (3-glycidyloxypropyl)trimethoxysilane (2),3, 4, 13 ± 15 [N-(2-aminoethyl)-3-aminopropyl]trimethoxysilane (3), (diethylenetriaminopropyl)trimethoxysilane (4) 16 or other trialkoxysilanes.17 ± 21 The number of immobilisation sites can be increased by an order of magnitude using branched linkers, such as dendrimers (5).22 Oligonucleotides with linkers (3 ± 11 units) attached to their 50- (see Refs 7, 23 ± 28) or 30-terminal 15, 29 phosphates, to the apurinisation sites,30 to the amino group of 5-methyldeoxycyti- dine,31, 32 to C(5) of deoxyuridine 33 ± 36 or to C(8) of deoxyadeno- 3 (MeO)3Si(CH2)3NH(CH2)2NH(CH2)2NH2 4 O NH2 O O N H H N O N NH2 O C O O O NH2Table 2.Supports for the immobilisation of ligands. Functional groups of supports CONHNH2 OC(O)NH(CH2)2NHCO OCH2CHCH2SS N OH O C NCHCONHCHSS O C(O)NHCH2CO2H CO2HNH NH(CH2)3NHC(CH2)3SS CH2NHCOCH2I(Br) O C NH O O OCH2CH CH2 Ligand LCHO LSH HgCl LSH LSH N CO2H LSH NO2 LSH LNH2 LOH LNH2 Groups of supports formed after immobilisation of the ligand C(O)NHN=CHL OC(O)NH(CH2)2NHCO OCH2CH(OH)CH2SSL O C NCHCONHCHSSL O C(O)NHCH2CO2H CO2HNH NH(CH2)3NHC(CH2)3SSL CH2NHCOCH2SL O C NL O OH OCH2CHCH2XL X=O, NH Commercial name of the support Affi-Gel Hz gel, Affi-Prep Hz support Agarose adipic acid hydrazide Carboxymethylcellulose hydrazide 3M Emphaze hydrazide biosupport Hydrazide beads Polyacrylamide hydrazide Polyacrylhydrazidoagarose Immobilized p-chloromercuribenzoate gel Affi-Gel 501gel HgSL Thiopropyl-activated Agarose Sepharose 6B Thiolagarose Activated thiolsepharose 4B Immobilized TNB-thiol Ultralink iodoacetyl Bromoacetylagarose Bromoacetylcellulose Cyanogen bromide-activated Agarose, Sepharose 6M B Agarose CL-4B Sepharose 4B Epoxy-activated Agarose Sepharose 6B Eupergit C Oxirane acrylic beads Polymercarrier VA-epoxy POROS 20 EP Supelcosil epoxy 540 Toyopearl AF-epoxy-650M Manufacturer Bio-Rad Pharmacia Sigma, Supelco Pierce Pierce, Sigma, Supelco Sigma, Supelco Pierce Pierce Bio-Rad Sigma, Supelco Pharmacia, Sigma, Supelco ICN Pharmacia, Supelco Pierce Pierce ICN, Calbiochem Sigma Sigma, Supelco Fluka Fluka, Sigma, Supelco, Pharmacia Pierce, Sigma, Supelco Pharmacia, Sigma, Supelco Rohm Pharma, Supelco Sigma Riedel-de Haen PerSeptive Biosystem Supelco Supelco, Tosoh Co.Table 2 (continued).Functional groups of supports O CO2NO OCO2 NO2 OC(O)N N Me CH2SO2 CH2OSO2CH2CF3 OCH2CHO Ligand LNH2 LNH2 LNH2 LSH LNH2 LSH LNH2 LNH2 Groups of supports formed after immobilisation of the ligand OC(O)NHL OC(O)NHL OC(O)NHL CH2XL X=NH, S CH2XL X=NH, S OCH2CH2NHL Commercial name of the support N-Hydroxysuccinimide-activated Agarose CH-Sepharose Sepharose, Superose Affi-Gel, Affi-Prep support Nitrophenyl-activated Agarose Carbonyldiimidazole(CDI)-activated Agarose Trisacryl GF-200, TSK HW-65F Reacti-Gel (6X, GF-2000, HW-65F) Supports Tosyl-activated Agarose Dynabeads M280 Tresyl-activated Agarose Toyopearl AF-tresyl-650M Toyopearl AF-formyl-650M Aldehydeagarose Manufacturer Sigma, Supelco Fluka, Sigma, Supelco Pharmacia Bio-Rad Sigma, Supelco Sigma, Supelco IBF Biotechnics Pierce Pierce, Sigma, Supelco Dynal Inc.Pierce, Schleicher&Schuell, Sigma, Supelco Supelco, Tosoh Co. Tosoh Co. SigmaAffinity sorbents containing nucleic acids and their fragments S S OMe O P O TTTTTTTTTTTTTTTATAACCCGGAATCCT (CH2)6 O Si (CH2)3 H N C N H H N C N H OMe sine 37 are immobilised on commercially available supports (see Table 2) bearing relatively short linkers (2 ± 11 units). There is no general agreement as regards the dependence of hybridisation properties of immobilised oligonucleotides on the linker's length.It was noted 6, 16 that the hybridisation abilities of immobilised oligonucleotides are enhanced with increase in the linker's length. It is also suggested 8, 38, 39 that efficient hybrid- isation requires the optimum length of the spacer to be selected. The use of linkers (6 ± 8) containing different numbers of repeating units has also been documented.8 O O P O P O (CH2CH2O)2 (CH2)3O O7 O7 n n 7 6 OP O (CH2CH2O)3 O7 n 8 The maximum efficiency of hybridisation is achieved for n=8 ± 10 with all three types of units. This suggests that the yields of hybridisation products depend not only on the length of the linker, but also on its structure (in this case, probably, on the number of negative charges). The maximum increase in the yields of hybridisation products (150-fold in comparison with the yields in the absence of the spacer) was observed with the linkers 7 and 8 for n=8.With a further increase in the number of units, the hybridisation ability diminishes, and for n=25 ± 30 the yields of the hybridisation products are virtually the same as in the absence of the linker. However, it was shown 40 that the length of the linker between the polyacrylamide support and an oligonucleotide does not influence the stability of a duplex formed. This is consistent with the results of the above-cited studies and those devoted to the sequencing of DNA by hybridisation on oligonucleotide micro- chips.41 ± 47 Presumably, the optimum length of the linker depends on many factors, such as the nature of the support, the ligand and the linker itself, and should be chosen experimentally for each particular task. It is noteworthy that immobilisation on granu- lated supports can be made using short linkers or no linkers at all, whereas films and slides demand longer linkers.In order to increase the distance between the surface and an oligonucleotide to be immobilised, the latter sometimes is elon- gated by introducing additional chains consisting of thymidine units [e.g., oligo(dT)1 ± 400]17, 19, 48, 49 or non-essential oligonucleo- tide sequences (cf. 9).7, 18 III. Chemical methods for the immobilisation of nucleic acids Immobilisation of NA by covalent binding to polymeric supports is achieved by the reaction of available or incorporated functional groups of the oligonucleotide with reactive groups of the support.In this review, the chemical methods used for the formation of a covalent bond between the ligand and the polymer are divided into two large groups, viz., (1) immobilisation of functionalised and/or activated NA on nucleophilic supports, and (2) immobilisation ofNAon supports containing electrophilic groups. It should be noted that the functional groups of the natural nucleic acids (hydroxy groups of ribose, phosphoric acid residues and heterocyclic bases 50) possess nucleophilic properties and 513 OO7 9 react with the electrophilic groups of the polymer.These reactions usually yield numerous binding sites of the ligands to the supports, which results in the loss of the full-value interactions of immobi- lised NA with biomolecules. To prevent this, NA are modified by introduction of additional functional groups which are more reactive than the corresponding functional groups of the natural NA. 1. Immobilisation of functionalised and activated nucleic acids on nucleophilic supports Immobilisation of NA and their fragments on solid supports containing surface nucleophilic groups (e.g., hydroxy, thiol or amino groups) can be achieved by preliminary activation of the functional groups of the ligand (Fig. 1). Ligand Functionalised ligand Activated ligand Electrophilic support Nucleophilic support Carrier ± Ligand Figure 1.Chemical methods for the immobilisation of ligands. The introduction of functional groups into poly- and oligo- nucleotides has been the subject of several reviews.51 ± 53 Most often, the post-synthetic introduction of additional functional groups involves terminal phosphate groups using reagents which affect other functional groups of NA to a much lesser extent. Carbodiimides, arenesulfonyl chlorides, 2,4,6-trimethylbenzoyl chloride, etc., are commonly used for this purpose.54 ± 58 The reactive oligonucleotide derivatives formed are unstable and used for the introduction of additional functional groups into oligonucleotides or for direct immobilisation on solid supports immediately following preparation. a.Activation of nucleic acids by carbodiimides The addition of NA using carbodiimides 10a ± c has been one of the earliest approaches to the preparation of affinity sorbents and has continued to be a very popular procedure.50 ± 64 This method is based on the ability of carbodiimides to activate either the carboxy groups of polymers or the terminal phosphate group of NA. Carbodiimide-activated oligonucleotides react readily with amino or thiol groups of solid supports. Reactions with hydroxyl- containing supports proceed less effectively. The carbodiimide method is used for the immobilisation of mono-, oligo-, polydeoxyribonucleotides and DNA in both organic solvents using dicyclohexylcarbodiimide (10a) 50 ± 62 and aqueous media using water-soluble carbodiimides, e.g., 1-(3- dimethylaminopropyl)-3-ethylcarbodiimide (10b) 13, 18, 24, 25, 65, 66 or 1-cyclohexyl-3-(2-morpholinoethyl)carbodiimide metho-p-tol- uenesulfonate (10c).25, 67514 R1N C NR2 10a ± c R1=R2=cyclo-C6H11 (a); R1=Et, R2=Me2N(CH2)3 (b); + [SO3C6H4Me-4]7 (c). N(Me)(CH2)2 R1=cyclo-C6H11, R2= O The mechanism of activation of phosphate groups of oligo- nucleotides by carbodiimides has been studied in detail and some side reactions have been revealed.68 ± 73 Protonation of carbodi- imides and the formation of intermediate unstable reactive O-phosphorylisourea derivatives 11 is the limiting step of this reaction.74 This stage is also inhibited by strongly basic nucleo- philic groups of the support.Besides, protonated nucleophilic groups do not react with compounds 11. As a consequence, the yield of the final product 12 decreases if polymers with strongly basic groups are used. The substituted O-phosphorylisoureas 11 can also be transformed into side products, e.g., be hydrolysed with the liberation of the original phosphate group and urea or undergo consecutive conversions leading to pyrophosphates, polyphosphates and O-phosphorylisourea derivatives.75 The lat- ter, in turn, produce non-reactive N-pyrophosphorylurea deriva- tives 1375, 76 or N-phosphorylurea derivatives 14.66, 77 ± 79 O O H+ R3O P O C NR1 R1N C NR2 +R3O P O7 10a ± c NHR2 OH OH 11 O NuH Nu P OR3+R1NHCONHR2 O7 12 O O O R3O P N(R1)CONHR2 + R3O P O P OR3 O7 OH NR1CONHR2 14 13 R3, oligonucleotide residue; NuH=NH, SH, OH.In the activation of oligonucleotides by carbodiimides result- ing in compounds 11 ± 14, modification of internucleotide phos- phate groups 76, 79 and heterocyclic bases 69, 80 is observed. In the case of oligodeoxyribonucleotides, the reaction with the inter- nucleotide phosphate groups results in the urea derivatives 15,76 whereas the reaction with oligoribonucleotides leads to the isomerisation and cleavage of the internucleotide bonds.51, 81 (RO)2P N(R1)CONR2 O 15 The activation of the terminal phosphate group of the nucleo- tide by carbodiimide in alkaline media (pH58) can be accom- panied by nucleophilic addition of the anion formed from the heterocyclic base to the double bond of the carbodiimide.This reaction affects the bases containing aCONH group, i.e., the N(3) atom in the uracil and thymine residues and the N(1) atom in the guanine residues.69 O O O R2NH 7 OH7 10a ± c N HN N C R1N pH58 O O O NX NX NX X, oligonucleotide residue. The rate of the limiting stage can be accelerated, the yield of the target product 12 in the immobilisation of the oligonucleotide derivatives 11 can be increased and the modification of the I G Shishkina, A S Levina, V F Zarytova internucleotide phosphodiester groups and heterocyclic bases by carbodiimides can be overcome if the reaction is carried out in weakly acidic media (pH 4.25 ± 5.0).80 In this case, the carbodi- imide-activated terminal phosphate group of an oligonucleotide reacts predominantly with the amino, thiol or hydroxy groups on the surface of the support.The selectivity and efficiency of oligonucleotide immobilisa- tion increases slightly after functionalisation of their 50-phosphate groups by linkers containing a terminal carboxy group,18, 82 which reacts with the amino groups of the support after its activation by carbodiimide. An alternative approach consists in the activation of the carboxy groups of the polymer by carbodiimides (see Sec- tion III.2). In this case, immobilisation of nucleotides and oligo- nucleotides on supports involves the amino groups of the heterocyclic bases or linkers attached to the 30- (see Refs 66, 83) or 50- (see Refs 25, 84) terminal phosphates of the oligonucleo- tides.Immobilisation of NA and oligonucleotides by the carbodi- imide method is carried out on virtually all available solid supports, e.g., cellulose, Sepharose,59 ± 63, 72, 85 Sephacryl,23, 65, 67 Sephadex,23, 73, 85 SiO2-based supports 12, 18, 64, 65, 86, 87 and various membranes.66, 82, 83, 88 In addition to conventional supports, mag- netic carriers (e.g., polystyrene-coated magnetite particles) 25 and latex microspheres (polystyrene beads able to form colloidal suspensions) 24 are used. The resulting matrices are used in hybridisation analysis 12, 13, 18, 23, 25, 66, 67 and in affinity chroma- tography.63, 89 The efficiency of immobilisation of oligonucleotides on poly- meric supports can be increased by performing activation by carbodiimides in the presence of nucleophilic catalysts, e.g., imidazole, triazole or tetrazole, which generate the corresponding phosphoramidates;24, 25, 65, 73, 88 the latter manifest high reactiv- ities towards nucleophiles after protonation of the azole units.Phosphorimidazolidates 16 are commonly used.88, 90, 91 N O O NH N +H+ N RO P RO P O7 +10a ± c 11 7H+ O7 O7 O O NH2 NH + RO P HN N RO P O7 O7 16 R, oligonucleotide residue. The addition of compounds containing hydroxy groups with higher acidities, e.g., N-hydroxysuccinimide or N-hydroxybenzo- triazole, to the reaction mixture instead of nucleophilic catalysts results in the formation of the activated NA esters 17.73, 92 OH N O N N 11 RO P O7 +10a ± c O7 O O P OR O NH2 O7 N RO P HN N N O7 17 R, oligonucleotide residue.The main advantage of the last two approaches is the lack of side reactions, since the derivatives 16 and 17 are more stable than the O-phosphorylisoureas 11. The compounds 16 and 17 react with the amino groups of the support more efficiently, which increases the yield on immobilisation.Affinity sorbents containing nucleic acids and their fragments b. Activation of nucleic acids by 2,4,6-trimethylbenzoyl chloride High reactivity of the title reagent with respect to the terminal and internucleotide phosphate groups ensures rapid formation of the corresponding mixed anhydrides. In aqueous media, only the mixed anhydride 18 formed upon reaction with the terminal phosphate persists.93, 94 This is used for the immobilisation of deoxyribonucleic acids on polymers containing nucleophilic amino groups.In aqueous media, immobilisation is due to the formation of a phosphoramide bond between the terminal 50-phosphate group of the oligonucleotide and the primary amino group of the polymer (the product 19).57, 95 O7 P O NH(CH2)6NH2+2,4,6-Me3C6H2CO2 18 OR O7 NH(CH2)6NH P O OR 19 R, oligonucleotide residue. In the study by Shumyantseva and Khomutov,96 the activated oligonucleotide 18 was immobilised on aminooxybutylcellulose 20. The resulting affinity sorbent 21 is stable only at pH 7.0 due to the high lability of the alkoxyamide bond. In addition, binding of oligonucleotides to the polymer 20 occurred involving the amino groups of cytidine to give the product 22, most probably due to transamination in aqueous media.56, 94 O7 O(CH2)4ONH2+18 O(CH2)4ONH P O 20 OR 21 NH2 ONH 20 + N N O O NR NR 22 R, oligonucleotide residue.This method has not gained wide acceptance despite the fact that TMP, AMP, CMP, d(pT)3, d(pA)3 and other oligonucleo- tides 57, 95, 96 were immobilised on polysaccharide supports in fairly good yields (0.05 ± 0.15 mmol of the nucleotide per 1 g of the support). The significant disadvantages of this method are the lengthy immobilisation step (>3 days) and side reactions (e.g., the formation of pyrophosphates of oligonucleotides, reactions involving internucleotide phosphates and the heterocyclic bases of oligonucleotides).This method is inappropriate for the immobi- lisation of ribooligonucleotides and RNA, since mixed anhydrides favour the formation of cyclophosphates in the activation step, subsequent isomerisation and cleavage of the internucleotide bonds.93 c. Activation of the terminal phosphates of oligonucleotides by the triphenylphosphine ± dipyridyl disulfide redox couple The triphenylphosphine ± dipyridyl disulfide redox couple (Ph3P±Py2S2) was first recommended for use as a condensation reagent for peptide and oligonucleotide syntheses by Mukaiyama et al.97, 98 The following reaction scheme was proposed:98O R1O P OH S7 + O7 N Ph3P+Py2S2 7 N Ph3PS S N + OP OR1 7O N Ph3PS O7 + S7 O N P OR1 O Ph3P O7 R1, nucleotide or oligonucleotide residue; R2=Alk, Ar.However, the reaction of an oligonucleotide with highly basic amino groups in the presence of the Ph3P ± PyS2 couple occurs at a slow rate and with low yields, which places certain restrictions on its practical application. A detailed study of this system by Zarytova et al.99 ± 102 revealed that the Ph3P±Py2S2 couple selec- tively activates the terminal phosphate group of an oligonucleo- tide in the presence of nucleophilic catalysts with the formation of the zwitterionic oligonucleotide derivatives 23. + S7 O Nu N OR P O Ph3P O7 O + Nu P OR+Ph3P O+ O723 O + Nu P OR NH2+ O7 NMe2 NMe , ; Nu= N N R, nucleotide or oligonucleotide residue.It should be emphasised that heterocyclic bases and internu- cleotide phosphate groups are not involved in the activation, which precludes the formation of side products. Therefore, this approach can be used for the activation of both deoxyoligonu- cleotides and their analogues and, which is especially important, of ribooligonucleotides without migration of the C(3)7C(50) internucleotide bond. The activated derivatives 23 react rapidly and virtually quan- titatively with primary amines to yield oligonucleotide phosphor- amides.100 The activated oligonucleotide 23 can easily be immobilised provided the primary amino group is localised on its surface.101, 102 The immobilisation is completed within 2 h in aqueous alkaline media or in organic solvents.Up to 97% of the starting oligonucleotide material binds to the support, which enables one to control and vary the capacities of affinity sorbents. Any material can be used as a support, but silica gel 103 ± 105 and polymethacrylate gels 101, 106 are especially widely used. The matrices with immobilised oligonucleotides were used for affinity chromatography of NA-dependent proteins,101, 106 nucleic acids and their analogues.102 ± 105 The Ph3P±Py2S2 couple can be recommended for the immobilisation of ribooligonucleotides and RNA, since in contrast with the known condensation reagents (e.g., carbodiimides, arenesulfonyl chlorides, azoles, 2,4,6-trime- thylbenzoyl chloride), it does not react with internucleotide phosphates and nucleophilic centres of heterocycles and no corresponding side products are formed. 515 O R2OH P OR1 R2O O7O R2NH2 P OR1 R2NH O7 N S OP OR NH O7516 d.Activation of nucleic acids by bifunctional reagents Activation and subsequent immobilisation of NA and their frag- ments can be achieved through the use of bifunctional reagents. The reactions with bisoxiranes, disuccinimidyl carbonate, carbon- yldiimidazole and glutaraldehyde will be considered in detail in Section III.2. The bifunctional reagents also include 2,4,6-trichloro-1,3,5- triazine (24), which reacts with the primary amino group of a spacer introduced into the oligonucleotide in the course of its synthesis.27Cl Cl H2N(CH2)5CH2OR N N N N Cl Cl Cl N NH(CH2)5CH2OR 25 N24 R, oligonucleotide residue The triazine derivatives of oligonucleotides 25 are stable for at least one week at 4 8C and can be immobilised on any supports containing amino (see Refs 6, 27) or thiol groups.6 The resulting sorbents are characterised by low non-specific sorption of proteins and NA and are stable under hybridisation conditions.27, 107 e.Reductive amination with nucleic acids and oligonucleotides containing aldehyde groups Reductive amination is based on the well-known reaction of aldehydes and ketones with primary amines resulting in the formation of Schiff's bases. The reduction of Schiff's bases by sodium borohydride affords stable secondary amines. NaBH4 R1N CHR2 R1NH2+O CHR2 R1NHCH2R2 Immobilisation based on this method requires that NA or oligonucleotides contained aldehyde groups. This requirement is easy to meet in the case of ribooligonucleotides, RNA and deoxyoligonucleotides carrying a 30-terminal ribose moiety.The cis-diol group of the terminal ribose is selectively oxidised by sodium periodate. The aldehyde groups formed (compound 26) react with supports containing surface amino 108 ± 111 and hydra- zide 112 ± 114 groups. O O B B NH2 RO RO P P OCH2 OCH2 O O IO¡4O7 O7 CH HC OH OH O O 26O O B B RO RO P P OCH2 OCH2 NaBH4 O O O7 + O7 CH HC N OH OH N N 27O O B RO P OCH2 B RO P OCH2 O O O7 O7 + CH2 H2C N HN NH R, ribooligonucleotide, RNA or deoxyoligonucleotide; B, heterocyclic base.In this reaction, the dialdehyde can react both with one amino group of the support (to give the morpholine derivative 27 50, 52) and with two amino groups leading to bisadducts. Since the I G Shishkina, A S Levina, V F Zarytova derivatives of the type 27 are stable only at pH>9 and phosphate undergoes b-elimination at pH<8, they should be reduced with sodium borohydride or cyanoborohydride immediately after preparation. Adecrease in the yield of the immobilisation products can also be due to the cleavage of the 50-C7O bond in the periodate oxidation product 26 in alkaline media (b-elimination) resulting in the splitting of the modified nucleoside from the 30-end of RNA.50 O O B B RO P O pH>7 OCH2 O H2C + RO P O7 O7 CH HC CH HC O7 O O 26 O O Immobilisation on supports containing hydrazide groups occurs with higher yields, since hydrazones do not require treat- ment with reductants, being more stable than azomethines.29 It should be noted, however, that the binding of an oligonucleotide to the support through the aliphatic amino group formed upon reduction is stronger 29 than that formed through the hydrazone group.Reductive amination was used for efficient immobilisation of all ribonucleoside mono-, di- and triphosphates 114 ± 118 on cellu- lose and agarose (Sepharose) modified by amino or hydrazide groups. The sorbents formed were used in affinity chromatogra- phy of proteins, such as myosins 118 and the elongation factors Tu and Ts.110 Reductive amination is the main technique in immobi- lisation of ribonucleotides.Polyribonucleotides [poly(A) and poly(U)] following oxidation of the cis-glycol groups were attached to polysaccharide supports.111, 115, 119 Among numerous studies aimed at immobilisation of RNA by this method,112, 116, 119 ± 121 special mention should be made of those studies 115, 117, 122, 123 in which ribosomal RNA were covalently attached to hydrazide-agarose and used for the isolation and purification of ribosomal proteins. tRNAs immobilised on poly- saccharide templates 111, 115, 124 were used for the purification of aminoacyl-tRNA synthases.125 ± 128 Recent developments in oligonucleotide synthesis have stimu- lated studies aimed at the synthesis of oligodeoxyribonucleotides containing a terminal ribonucleoside unit. Its 30-cis-glycol group is oxidised and used for immobilisation of oligodeoxynucleotides on solid supports.29, 64, 129, 130 The affinity matrices thus prepared are widely used in hybridisation analysis and sequencing of DNA.42, 43, 45, 131 ± 133 Immobilisation of NA and their fragments by reductive amination can also be achieved by the introduction of the aldehyde groups into the supports.One can synthesise polymers of virtually any type containing reactive aldehyde groups. The methods of their synthesis will be discussed in Section III.2.i. f. Immobilisation of nucleic acids containing thiol groups The introduction of chemically reactive thiol groups into NA and their fragments 53 allows specific immobilisation of such ligands by conventional methods, e.g., on supports containing epoxy groups (Section III.2.b) or active halogen atoms (Section III.2.e).134 Special procedures can also be used for this purpose.Thiol groups can be introduced in the last step of the phosphor- amidite oligonucleotide synthesis, e.g., using commercial phos- phorylating reagents, such as phosphoramidites of S-protected thio alcohols.49, 134, 135 OMe Ph3CS(CH2)nOP O N n=3, 6. The removal of all protective groups (including the trityl group) by silver nitrate leads to an oligonucleotide containing aAffinity sorbents containing nucleic acids and their fragments 50-terminal thiol group.49, 135 Thiol groups can either be intro- duced into the 30-end of an oligonucleotide in the initial stage of the synthesis,16, 136 or incorporated into the isolated oligonucleo- tide by an enzymic reaction, viz., using ATP containing a g-thiol group.137 An alternative approach consists in preliminary incorporation of an easily modifiable aliphatic amino group into the terminal 50-phosphate group of an oligonucleotide.90 The reaction of the spacer amino group of the deblocked oligonucleotide with 3,30- dithiobis-N-(propionyloxy)succinimide in the presence of a disul- fide bond reductant, e.g., dithioerythritol (DTE), dithiothreitol (DTT), cysteine or mercaptoethanol, affords an oligonucleotide containing a thiol group.O O7 pH 7.7 H2NCH2CH2O P O+ NOCOCH2CH2S DTE OR 2 O O7 P O HSCH2CH2CONHCH2CH2O OR R, oligonucleotide residue; DTE, dithioerythritol. Oligonucleotides containing thiol groups (SH-oligonucleoti- des) can be selectively immobilised on solid supports containing thiol groups, i.e., through the formation of a disulfide bond.Prior to immobilisation, the thiol groups of oligonucleotides can be activated by 2,20-dipyridyl disulfide.6 If the supports and oligonucleotides contain non-activated thiol groups, the formation of the disulfide bond between NA and the matrix occurs under the action of oxidants (I2 or H2O2).49 Polysaccharides, glass (CPG),90 magnetite particles coated with a silicon-containing polymer (Biomag),49 polyacrylamide (Tri- sacryl, Sephacryl) 28 and polymethacrylate6 are widely used as supports.Owing to the ease of the formation and the cleavage of the disulfide bonds by oxidising and reducing agents, thiol matrices are often used for reversible immobilisation of oligonu- cleotides containing thiol groups. Immobilisation of purified SH-oligonucleotides proceeds rap- idly and specifically. The resulting matrices manifest high stability and purity in hybridisation analysis 28, 49, 90 and in the purification of DNA-dependent proteins.137 It should be noted, however, that the susceptibility of the thiol groups to oxidation creates storage problems. In the course of time, intermolecular disulfide bonds and side products can be formed. Immobilisation of SH-oligonu- cleotides on activated or functionalised supports will be consid- ered below (Section III.2.f).2. Immobilisation of nucleic acids on electrophilic supports Native non-modified NA and their fragments are immobilised on functionalised and activated supports. Such supports can also be used for immobilisation of NA containing primary aliphatic amino or thiol groups introduced either in the course of or following the synthesis. The methods for the incorporation of the functional groups on the surface of supports able to react with the corresponding groups of the ligand will be discussed below. a. Activation of supports by cyanogen bromide A method of the activation of supports containing hydroxy groups by cyanogen halides was first proposed in 1967 for immobilisation of proteins and peptides.138 The activation is usually carried out by cyanogen bromide at pH 10 ± 11.139 ± 141 The reactive cyanate 28 formed upon activation reacts with the free hydroxy group of the support to give the reactive cyclic imidocarbonate 29 and the inactive carbamate 30.142, 143 A sig- nificant contribution to the development of this method was made by Cuatrecasas 144 who has selected the conditions for the immo- bilisation of various ligands including NA.145 In the early 1970's, this method became a routine procedure for the immobilisation of NA.84, 139, 140, 146 ± 150 The attachment of NA occurs at the amino groups of heterocyclic bases of a polynucleotide.In this way, polyuridylic 147, 149 and polyade- nylic 139 acids and RNA84, 141, 150 have been immobilised.After immobilisation of NA, the remaining reactive groups of the support are blocked by treatment with 1 M 2-aminoethanol. OH BrCN OC N OH OH 28NH2 RNH2 O 29 C NHR O H2O O C 29 NH2 OH O 7NH3 It was shown that the isourea derivatives 31 were the main products formed in the immobilisation of a single-stranded DNA (0.2 mol litre71 of 2-morpholinoethanesulfonic acid, pH 8.0) in formamide 139, 145; the N-substituted imidocarbonates 32 and carbamates 33 are formed in lesser amounts. This method is most commonly used for the immobilisation of single-stranded DNA139, 140, 146, 151, 152 and RNA.84, 139, 141, 150 It was possible to prepare polymeric supports of high capacity and specificity despite possible reaction of the amino groups of hetero- cycles with activated supports.However, the affinity properties of the sorbents containing short oligonucleotides are weak, since the NH2 groups of the heterocyclic bases involved in the covalent binding to the polymer cannot take part in the complex formation. Immobilisation of poly- and oligonucleotides on cyanogen bromide-activated supports is much more selective when the linkers containing aliphatic primary amino groups are attached to either the 50- or the 30-end of the oligonucleotide chain 54 ± 58, 153 or the internucleotide phosphate group.154 However, in this case, too, partial immobilisation of the ligand occurs through hetero- cyclic bases.155 Nucleoside mono-, di- and triphosphates, such as AMP,156, 157 ADP,158, 159 ATP,160 UMP,58, 161 GTP58 and dCp4 (see Ref.162), were immobilised on cyanogen bromide-activated Sepharose (all the nucleotides contained linkers with primary aliphatic amino groups incorporated into the phosphate groups or the heterocyclic bases). These sorbents were used for the purification of ribonu- cleotide reductase of phage T4,160 the cap-binding protein 163 and deoxynucleotide kinase of Lactobacillus.162 The disadvantages of this method are as follows. First, immobilisation of the ligand must be carried out immediately after the activation of the support because of the low stability of the reactive groups formed upon treatment with cyanogen bro- mide. Second, cyanogen bromide is a toxic substance.Third, the basic isourea groups of the support 31 (pKa=9.5) formed by the reaction with the primary amine appear to be positively charged at physiological pH values, which confers undesirable anion- exchanging properties on such affinity sorbents. Moreover, this reaction yields numerous side products, while the binding sites of 517 O NH C O 29 OCONH2 OH30 NH OCNHR OH31O NR O C32 7NH3 OCONH2 OH O 30 OCNHR RNH2 O C O O OH33518 nucleotides cannot be determined with certainty. On storage, the ligand is gradually split off from the polymeric support.113, 146, 164 It was shown 153 that affinity matrices prepared by immobilisation of oligothymidylates containing 50-terminal amino groups on cyanogen bromide-activated cellulose retain their stabilities in aqueous media during one week at 22 8C and for only 24 h at 42 8C.Most probably, this is due to the instability of the N- substituted isourea derivatives 31.143 NH NH2 H+ O OCNHR C NHR O OH31 OH OCNHR H2O pH>10 H2NCNHR NH2 OH O NH2 8<pH<10 H2NCNHR OCNHR NH3 NH2 NH OH NH2 O OH7 C +NH2R OH 5<pH<10 O However, despite these drawbacks this method is widely employed for the immobilisation of DNA, RNA and synthetic oligonucleotides 165 ± 170 aiming at the preparation of affinity matrices which are further used for the isolation and purification of DNA-dependent proteins. b. Activation of supports by oxiranes Bifunctional oxiranes were first recommended for the incorpora- tion of reactive groups into polymeric supports and the stabilisa- tion of gels by cross-linking.171 The hydroxy groups of supports are usually activated by the diglycidyl ether of 1,4-butanediol at pH>7.0.37, 172, 173 This treatment affords a polymer containing reactive epoxy groups, which react with the nucleophilic groups of the ligands.O O OH+ H2C CHCH2O(CH2)nOCH2CH CH2 OH O NH2R CHCH2O(CH2)nOCH2CH CH2 OCH2 OH OH OCH2CHCH2O(CH2)nOCH2CHCH2NHR Epichlorohydrin is also used for the preparation of supports containing epoxy groups.78, 107, 114 At present, epoxy groups are often incorporated on silica gel 3, 174 and glass 15, 175 using trime- thoxysilanes 2 rather than polysaccharides. O OH+(MeO)3Si(CH2)3OCH2CH CH2 2 OMe O OSi(CH2)3OCH2CH CH2 OMe As a rule, immobilisation of ligands containing amino, hydroxy or other nucleophilic groups on the supports containing epoxy groups occurs smoothly at pH 9 ± 13 (8 ± 20 h, 20 ± 25 8C).The unconsumed epoxy groups are blocked by treatment with 2-aminoethanolamine (1mol litre71). This method was used for the covalent binding of polynucleotides and some DNAs.37, 172, 173, 176 I G Shishkina, A S Levina, V F Zarytova The efficiency of binding of polynucleotides changed in the series: poly(dT)>poly(dC)=poly(dA)>poly(dG).172 As nucleotides usually contain many functional groups, the exact binding site for polynucleotides is difficult to establish and therefore this is not always specified.172, 173 For example, it is believed 37 that the immobilisation of the C(8)-hydroxyethylthio derivative of cyclo- AMP on a polymeric support containing epoxy groups is achieved by its binding through the terminal hydroxy group of the linker. NH2 N N SCH2CH2OH O N H2C CHCH2O N CH2 O O O OH P O OH NH2 N N SCH2CH2OCH2CHCH2O N OH N CH2 O O OH O P O OH It was shown that the immobilisation of oligodeoxyribonu- cleotides in concentrated (2.7 mol litre71) potassium phosphate occurs predominantly through purine bases.177 Despite its obvious advantages over other methods, this method has only one disadvantage, viz., the uncertainty of the binding site of nucleic acids.The use of an extended bisoxirane reagent ensures the introduction of a long uncharged hydrophilic arm between the ligand and the polymer surface, which is especially convenient for chromatography of large biomolecules. The affinity sorbents prepared on such supports are rather stable, since the ligand is attached through an ether bond, and the immobilisation is not accompanied by non-specific sorption of biopolymers.15, 171 Immobilisation of biologically active ligands on mechanically stable silica gels containing epoxy groups allows preparation of matrices for high-performance affinity chromatography.It should be remembered, however, that some supports are unstable in alkaline media where such immobilisation takes place. c. Supports containing N-hydroxysuccinimide groups The idea to use N-hydroxysuccinimide esters as reactive groups for the formation of amide bonds with the primary amino groups of ligands was first formulated and realised in studies devoted to peptide synthesis.178, 179 High solubility of N-hydroxysuccinimide in water and organic solvents makes it a convenient reagent for the activation of various polymeric surfaces.The supports containing esterified N-hydroxysuccinimide 34 are obtained by treating the polymers containing terminal carboxy groups with N-hydroxysuccinimide in the presence of dicyclohex- ylcarbodiimide (DCC) in dry solvents.26, 180 O O DCC CH2CH2COH+HON CH2CH2CON O O O O 34 DCC = N C N . Bifunctional reagents, e.g., N,N0-disuccinimidyl carbonate (35a) or N-succinimidyloxycarbonyl chloride (35b), are used forAffinity sorbents containing nucleic acids and their fragments the activation of supports containing hydroxy or amino groups.22, 181 O OH+XCON O O 35a,b O X= NO (a), Cl (b).O The N-hydroxysuccinimide esters 34 and 36 react rapidly with oligonucleotides containing aminoalkyl groups at the 50-terminal phosphate (37a ± d) or in the heterocyclic bases in buffer solutions (pH 7.5 ± 9.0) resulting in the formation of stable amide bonds. O CH2CH2CON +NH2R 37a ± d O 34 O R=(CH2)6NH(pT)29 (a), (CH2)2SS(CH2)2NH(pN)29 (b), (CH2)12O(pCTT)7 (c), (CH2)2O(pT)18 (d). It was shown 65 that immobilisation of a 29-membered oligo- nucleotide occurs predominantly through the amino group of 6-aminohexyl linker attached as 50-phosphoramidates (37a).Immobilisation through the cystaminyl linker (37b) is less effi- cient, while binding through the amino group of the heterocyclic bases proceeds to a smaller degree. Unfortunately, the capacity of the affinity sorbents thus obtained does not exceed 1 nmol g71 at best, apparently, due to rapid hydrolysis of the activated ester 34. The efficiency of immobilisation is higher in non-aqueous media where competitive hydrolysis is reduced to a minimum. Hydrolysis of the reactive surface groups of the support 34 yields carboxy groups which confer ion-exchange properties on the affinity sorbent, while unconsumed groups of the activated support 36 are gradually hydrolysed to regenerate the starting groups.+ Cl7 NCH2OCH2 OH+ O CH CH2+HX X=NH, S. NH2+ClOC NHOC38c O NH2+ NOCOCH2 O NHC(O)CH2 38d O OCON O O 36 CH2CH2CONHR NO2 OCH2 NH2 pH>9 CH CH2X OH NHOC NO2+ 7 N2Cl NHC(O)CH2 NO2 + 7 N2Cl 519 An interesting approach to the immobilisation of double- stranded fragments of DNA on N-hydroxysuccinimide-activated Sepharose has been proposed.31, 182 A 39-membered oligonucleo- tide was synthesised with a hairpin and a 1,6-diaminohexane unit was introduced into a heterocyclic base of the hairpin loop. The incorporated amino group was used for the attachment of the oligonucleotide to the polymer. Any double-stranded DNA frag- ments containing the corresponding sticky ends can be bound then to the immobilised oligonucleotide using DNA-ligase. T T Tm5C NH(CH2)6NHCO TCGAGCTCCGGAATTCGA CGAGGCCT TAAGCT T T T Supports with immobilised poly- and oligonucleotides were used in hybridisation analysis 21, 22, 183 and affinity chromatogra- phy.7, 26 Thus the amino-containing oligonucleotide (CTT)7 37c was immobilised on Sepharose and this sorbent was used for the isolation of plasmid pXL2563 containingGAA sites due to triplex formation.7 Column chromatography on silica gel containing immobilised 5 0-aminoethyloctadecathymidylate (37d) was suc- cessfully used for the separation of a mixture of adenylates A12±A18.26 d.Diazotised supports The azo coupling reaction has long attracted the attention of investigators engaged in immobilisation of nucleotides, since the introduction of aromatic amino groups into the support can be performed under mild conditions. Different approaches to the synthesis of polymers containing aromatic amino groups are shown in Scheme 1.144, 184 ± 186 Such supports are usually prepared by the reduction of the corresponding aromatic nitro com- pounds.187 Diazotisation of aromatic amino groups is carried out with sodium nitrite in acidic media.Immobilisation of native nucleic acids on diazotised supports, such as compounds 38a ± d (see Scheme 1), involves exclusively heterocyclic bases.25, 67, 184 The introduction of spacer groups into oligonucleotides, e.g., the introduction of phenol residues into 5 0-phosphate groups, results in the predominant (up to 87%) immobilisation at this particular group.188 Scheme 1 7 +N2Cl NH2 NO2 [H] NaNO2 OCH2 OCH2 HCl 38a 7 +N2Cl NH2 NaNO2 CH CH2X HCl OH 38b [H] NaNO2 NHOC NH2 NO2 HCl [H] NaNO2 NH2 NHC(O)CH2 NO2 HCl520 The azo coupling reaction consists in electrophilic substitution where the diazonium ion plays the role of an attacking agent.This reaction can be visually monitored by following colouration of the support. Immobilisation of 50-AMP (39) at the C(8) atom of the heterocyclic base on agarose containing diazo groups is an example.189 NH2 N N OH N + 7 HO P N2Cl+ N OCH2 O O OH OH 39 NH2 N N N N OH N HO O N P OCH2 O OH OH Azo coupling can be accompanied by the reaction of diazo- nium salts with the amino groups of heterocyclic bases, which affords a diazoamino compound.190 + 7 NR N2Cl+H2N N O NH N N N R, NA residue.In addition to reactions of diazonium salts with heterocyclic bases of nucleic acids, side reactions with water (to give phenols) and aromatic amino groups which are always present on the surface of such supports can also take place. If azo couping occurs at a slow rate, the formation of phenol and the cross-linking of the polymeric support can predominate.185 Aromatic diazo compounds are rather unstable, they decom- pose at elevated temperatures and upon illumination. Therefore, diazotised supports are used immediately after their synthesis. They can also be treated with NaBF4 or NaPF6 to give more stable tetrafluoroborates or hexafluorophosphates.185 Denatured DNA and RNA immobilised by this method were used mostly in hybridisation analysis.184, 185, 191 Good results were obtained with the use of supports 38b (see Scheme 1) containing diazophenylsulfide groups.140 Immobilised cyclic DNA was used as a template in an asymmetric polymerase reaction.192 The efficiency of immobilisation on diazotised supports is differently estimated by different authors, viz., from 795 mg of the nucleotide per 1 g of the support 146 to trace amounts.25 Affinity sorbents prepared by azo coupling are usually unstable.Up to 50% of immobilised DNA is lost within the first 2 ± 3 days at 45 8C irrespective of the nature of the matrix.146 Low non-specific sorption of biopolymers on diazotised supports is their main advantage.Cellulose or paper filters containing diazo groups are most commonly used for immobilisation of DNA; however, Sephacryl proved to be the best polymer for such immobilisation.67 Mag- netic particles were also tested, however, with little effect.25 e. Supports containing active halogen atoms Activation of oligonucleotides by 2,4,6-trichloro-1,3,5-triazine (24) was considered in Section III.1.d. In this section, we shall consider the preparation of supports activated by 2,4,6-trichloro- NR O I G Shishkina, A S Levina, V F Zarytova 1,3,5-triazine and their use for the immobilisation of ligands. The reactions of polynucleotides and NA with non-modified polysac- charides (cellulose and dextran) in the presence of 2,4,6-trichloro- 1,3,5-triazine (24) 107 which permit consecutive substitution of chlorine atoms will also be discussed.Cl Cl N N N O N OH +Cl N N O Cl 24 OR N ROH N O N ONHR N RNH2 N O N O R, NA or oligonucleotide residue. When the activation is carried out in an alkaline medium, two chlorine atoms react with the hydroxy groups of the supports resulting in the binding of triazine to the polymers. The third chlorine atom remains intact and can be substituted by the amino or hydroxy groups of the ligand to be attached.193, 194 It was shown that immobilisation of DNA using 2,4,6-trichloro-1,3,5- triazine involves predominantly the amino groups of the hetero- cyclic bases.107 As a result, the multipoint binding of DNA and oligonucleotides to the polymer makes the ligand less accessible for the interaction with the complementary polynucleotides in solution.Immobilisation of oligonucleotides bearing an amino group-containing linker at the 5 0-end occurs selectively through the introduced amino group.107, 194 However, even in this case the yields of the addition products are rather low due to the low accessibility of the chlorine atom in the triazine-containing sup- port. In order to increase the efficiency and selectivity of the method, it was suggested to activate the supports by 2-amino- 4,6-dichloro-1,3,5-triazine (40).142, 193 Cl Cl N N RNH2 O OH+Cl N N N N NH2 NH2 40 NHR N N O N NH2 R, oligonucleotide residue.The presence of an amino group in the triazine 40 decreases the reactivity of one of the chlorine atoms with respect to nucleophilic attack and thus prevents cross-linking of the poly- mer. The remaining chlorine atom reacts exclusively with strong nucleophiles, e.g., with the aliphatic amino group of the spacer introduced into an oligonucleotide.193 Triazine-activated polymers are fairly stable at elevated tem- peratures and can be stored for several weeks at 4 8C. The use of matrices with immobilised DNA or poly(U) made it possible to isolate both proteins and the complementary chains of NA.193 However, application of this method is restricted by the high toxicity of trichlorotriazine.521 Affinity sorbents containing nucleic acids and their fragments The carboxy groups of polyethylene terephthalate films can easily be converted into acid halide groups upon treatment with PCl3, PCl5, POCl3, SOCl2 or (COCl)2 in the presence of DMF. respectively.SH-containing ligands can also be immobilised using other methods. Oligonucleotides containing thiol groups can selectively be immobilised on solid supports containing thiol groups owing to the formation of disulfide bonds. The supports are pretreated with p-chloromercuribenzoate, di(3-ethoxycarbonyl-4-nitrophenyl) disulfide, 2,20-dipyridyl disulfide or 2,20-dithiobis(5-nitropyri- dine).90, 144 O2N DMF SH+ The polymers 41 containing acid halide groups react with the nucleophilic groups of the ligands. Immobilisation carried out under anhydrous conditions involves predominantly primary amino groups of the spacer incorporated into the 50-terminal phosphate groups of oligonucleotides.195 The stability of the sorbents 42 is provided by the stability of the amide bond formed.The films containing immobilised oligonucleotides obtained by this method were successfully used in hybridisation analysis.196 S N 2 RNH2 COCl COOH NO2 41 RSH S S R S S N CONHR R, oligonucleotide with an aliphatic linker. 42 R, NA or oligonucleotide residue. High thiophilicity of mercury ions allows the attachment of SH-containing oligonucleotides to modified polymers (e.g., to agarose) containing HgCl groups.90 RSH A chloromethylated styrene ± divinylbenzene copolymer can also be used for the immobilisation of ligands.This approach was used for the attachment of 50-ATP through the triphosphate HgCl OCONH(CH2)2NHCO pH 8.5 residue.197 O O O HO O P P O POCH2 A O HgSR OCONH(CH2)2NHCO O7 O7 O7 CH2Cl + R, oligonucleotide with an aliphatic linker. OH OH O O O O P P O POCH2 CH2O A O O7 O7 O7 The reaction of the thiol groups with the maleimide residues present on the polymer surface is yet another approach to the immobilisation of SH-containing oligonucleotides. Maleimide can be incorporated into virtually all materials containing primary amino groups using succinimidyl 4-(maleimidophenyl)butyrate (43) or maleic anhydride.11, 16, 136 OH OH A, adenine residue.O O N NH2+ NOC(CH2)3 O O O 43 O RSH N NHC(CH2)3 O O O N NHC(CH2)3 Halogenoacetyl agarose has become especially popular among commercially available supports containing active halogen atoms (see Table 2). This sorbent is resistant to hydrolysis; there- fore, the immobilisation procedure can be carried out in aqueous buffers. The best results were obtained in the case of immobilisa- tion of oligonucleotides containing thiol groups. To prevent attachment at heterocyclic bases, the SH-containing oligonucleo- tide obtained with the use of T4 kinase and g-thiol-ATP was hybridised with the complementary chain.137 The duplex formed was attached to bromoacetyl agarose. The immobilisation occurred exclusively at the 50-terminal thiolphosphate group of the oligonucleotide. If required, the complementary chain could be removed.SR O O O HOR2 g-thio-ATP7 S HOR1 P OR1 T4 kinase O7 O O NHCCH2Br 7 P OR1 .R2OH S The nucleophilic addition of SH-containing oligonucleotides to the double bonds of the support 44 in a buffer solution occurs rapidly (2 h at 4 8C) resulting in the formation of a stable bond between the support and an oligonucleotide. O O O7 RSH O O NH CO2H NH2+O S P OR1 .R2OH NHCCH2 44 O O7 R1,R2, complementary oligonucleotide residues. O CH2CO2H NH SR R, oligonucleotide residue. f. Activated supports for the immobilisation of SH-containing oligonucleotides Immobilisation of NA and their fragments containing thiol groups on supports containing epoxy groups or active halogen atoms 134 has been described in Sections III.2.b and III.2.e,522 g. Activation of supports by carbonyldiimidazole Hydroxy and carboxy groups of supports can easily be activated by N,N0-carbonyldiimidazole (45).This reagent earlier proposed for peptide synthesis 198 is now successfully used for the immobi- lisation of ligands containing primary amino groups.199, 200 NCON COOH+ N N 45 RNH2 CONHR CON N RNH2 OCONHR OH +45 OCON N 46 R, oligonucleotide residue. The activation of polymers is performed in non-aqueous media; the optimum conditions for the immobilisation of ligands must be selected for each particular case. The high efficiency of the ligand attachment in dry solvents is due to the lack of competitive hydrolysis. Affinity matrices prepared by this method are rather stable, since the amide bond formed between the support and the ligand is very strong.The support 46 is stable in dry solvents for several years. The half-life time of the hydrolysis of imidazolyl carbamate 46 is several hours. Thus activated agarose loses completely its activity in aqueous buffers only after 30 h at pH 8.5 ± 9.0.200 Following elimination of CO2 and imidazole, the support 46 is converted into the original hydroxyl-containing carrier devoid of charged groups, which excludes the manifesta- tion of ion-exchange properties of the support and non-specific sorption of biomolecules. It is noteworthy that the half-lifetime of the hydrolysis of cyanogen bromide- or hydroxysuccinimide- activated supports is several minutes.However, despite the advantages of the above-described imidazole-containing supports the immobilisation of, e.g., 6-ami- nohexylphosphoryl-dCTP and 50-aminoalkyloligonucleotides, occurs with low yields.3, 201 This method has not found wide application for the immobilisation of NA and its fragments. h. Activation of supports by sulfonyl chlorides Organic sulfonyl chlorides, such as p-toluenesulfonyl chloride (TsCl), 2,2,2-trifluoroethanesulfonyl chloride (TrsCl) and meth- anesulfonyl chloride (MsCl), are potent acylating agents, which convert hydroxy groups of a carrier into reactive sulfonates to afford the supports 47 ± 49.202 ± 204 TsCl OSO2C6H4Me-4 47 TrsCl OH OSO2CH2CF3 48 MsCl OSO2Me 49 TsCl =ClSO2C6H4Me-4, TrsCl =ClSO2CH2CF3, MsCl=ClSO2Me.Immobilisation of oligonucleotides on the supports 48 205 and 49 29 occurs exclusively through the amino group of a spacer attached to the 50- or 30-end of the oligonucleotide. In the absence of aminospacers, immobilisation through heterocyclic bases and terminal phosphate groups of oligonucleotides proceeds with lower yields. Immobilisation on the support 49 occurs with low yields,29 whereas polymers 48 are used for the preparation of I G Shishkina, A S Levina, V F Zarytova affinity sorbents with high capacities; these are used for purifica- tion of DNA-dependent proteins.205 The sulfonate residues are good leaving groups and the reactivities of the activated supports depend on the nature of the sulfonate and decrease in the following order: Trs>Ts>Ms.Thus, the support 49 is the most stable and the least reactive. The supports 47 ± 49 are stable in non-aqueous media and can be stored for several weeks at 4 8C in aqueous HCl (1 mmol litre71) without any loss in their activities. Immobilisation of NA containing thiol or primary amino groups on sulfonyl chloride-activated supports occurs as the nucleophilic substitution of the sulfonate residue with the formation of a stable bond between the nucleophile and the carbon atom of the original support. The attachment of the ligand can be carried out in both aqueous and organic media.Immobi- lisation on the more reactive polymers 48 is more efficient at neutral pH values at 4 8C, whereas that on the carriers 47 is better performed at pH 9 ± 11.204 i. Supports containing aldehyde groups (reductive amination) Immobilisation ofNAand their fragments by reductive amination can be carried out using two approaches. The first of them consists in the introduction of the aldehyde groups into NA and their subsequent immobilisation on supports containing hydrazide or primary aliphatic amino groups (see Section III.1.e). The second approach includes the incorporation of aldehyde groups into supports and their subsequent application for the immobilisation of NA through the amino groups of heterocyclic bases or through introduced spacer groups.The latter version is considered in this section. Aldehyde groups can be introduced into supports by different methods. Polysaccharide supports containing diol groups are oxidised by sodium periodate to give the modified polymer 50 with high content of the aldehyde groups. The supports devoid of hydroxy groups are modified by glutaraldehyde (51), glycidol (52), an oligomer containing aldehyde groups (53) or compounds which are further oxidised by periodate. Some examples of preparation of polymers of virtually any type containing aldehyde groups are shown in Scheme 2. The aldehyde groups of the polymer react readily with ligands containing primary amino groups to give Schiff's bases. The latter are further reduced by NaBH4, NaBH3CN or the pyridine ± bor- ane complex (Py .BH3).Unlike sodium borohydride, the last two reagents rapidly reduce the Schiff's bases, but not the aldehyde groups.29, 206 It should also be noted thatNaBH4 is decomposed in acidic media, whereas NaBH3CN functions effectively at pH 4 ± 10. It was recommended to replace highly toxic NaBH3CN by the Py .BH3 complex.29 NaBH3CN CH NR CH2NHR CHO+RNH2 Oligonucleotides with incorporated aminospacers are immo- bilised in good yields on supports containing aldehyde groups to form strong bonds.11, 12, 29, 30 However, this is accompanied by the addition of oligonucleotides through the amino groups of the heterocyclic bases. Especially strong non-specific immobilisation was observed in the case of supports activated by glutaraldehyde (51).29 The glutaraldehyde 51 is used to prepare affinity sorbents with a high content of NA.This reagent brings about cross-linking of the polysaccharides with nucleotides and NA (DNA, RNA207, NAD208) through primary amino groups of the modified carriers (aminoethylcellulose,207 aminohexylsepharose 208) and the hetero- cyclic bases. High concentrations of the ligand on the carrier results from additional intramolecular cross-linking of the ligand, which is often a drawback, especially in affinity chromatography.Affinity sorbents containing nucleic acids and their fragments CH2OH O NaIO4 O OH O O n OH Polysaccharide support NaBH3CN NH2+HOC(CH2)3CHO 51 O OH7 XH+H2C CHCH2OH 52 OH OH NaIO4 XCH2CH CH2 X=O, NH.HOC HOC(CH2)2 CONH2+ HOC (CH2)2 (CH2)3CHO CONHCHCH(CH2)2 CHO Me Me Polymerisation O O Acrylamide + (CH2)4NHCOCH CH2 O Me CONH(CH2)4 Me 2) NaIO4 O CONH(CH2)4CHO j. Supports bearing isothiocyanate groups The isothiocyanate group is yet another functional group which rapidly reacts with aliphatic primary amino groups to give stable thiourea derivatives. 1,4-Phenylene diisocyanate can be used for the activation of virtually all polymers containing primary amino groups.17, 19, 22 Immobilisation of oligonucleotides on isothiocya- nate-containing supports occurs with good yields17, 19 and is not accompanied by immobilisation of non-modified oligonucleoti- des.19, 22 The remaining isothiocyanate groups of the activated NC+MeCHO + NH NH CH2 A, adenine residue.Scheme 2 CH2OHO O CH HC n O O 50NH(CH2)4CHO XCH2CHO+CH2O (CH2)3CHO 53 CHO (CH2)2CHO 1) H+ CH2O N POCH2 O O7 NHCOCH OH Me 523 carrier undergo gradual hydrolysis to amino groups, which imparts ion-exchange properties to the affinity matrices. NCS NH2+SCN S NH2R NCS NHCNHS S NHCNHR NHCNH 3-Isothiocyanatopropyltriethoxysilane 209 or 1-(3-trimethoxy- silylpropyl)-3-(4-isothiocyanatophenyl)thiourea 19 can be used for the modification of glass surfaces. This one-step procedure is used for the preparation of glasses bearing reactive isothiocyanate groups. k. The four-component condensation The Ugi reaction is a multicomponent condensation leading to N-alkylamides under mild conditions 171, 210 ± 212 and involves a-addition of electrophilic compounds containing C=N bonds to isocyanides 54.The immonium salt 55 formed in the reaction of a carboxylic acid with an amine and an aldehyde or a ketone is added to isocyanide. Intramolecular acylation affords the final product 56. R1CHO+R2NH2+R3COOH R4NC (54) R4N C OCOR3 [(R1CH NHR2)+ R3CO¡2 ] R1CHNHR2 55R1 R4NHCOCHNCOR3 R2 56 The three soluble reactants may be varied depending on the nature of the functional groups (isocyanide,201 carboxy, aldehyde or amino groups 212) present on the support surface. The reaction of 50-AMP with Sepharose modified by 4,40- diaminodiphenylmethane in the presence of cyclohexyl isocyanide and acetaldehyde is an example of the Ugi reaction employed for the preparation of modified supports.171, 212, 213 In this case, the role of an anion of the immonium salt 55 is played by the phosphoric acid residue of 50-AMP and not by the carboxy group (Scheme 3).The best results (39 mg g71) were obtained in the immobilisa- tion ofRNAon Sepharose containing amino groups.212 However, this method has not found wide application due to the side reactions between the functional groups of the nucleotides and the components of this complex system. Scheme 3 O 7 A POCH2 O O NH2+ O7 OH A524 IV. Immobilisation of nucleic acids in the course of solid-phase synthesis The development of automated methods for oligonucleotide syn- thesis offered new opportunities for the preparation of affinity matrices bearing covalently bound oligonucleotides.The carriers on which oligonucleotides are synthesised can be used as affinity sorbents. This obviates the necessity of removal of the nucleotide material from the carrier and its further immobilisation on the support. However, one essential step, viz., purification of the newly synthesised oligonucleotide and the estimation of the quality of the immobilised material, is lost. The last step in any solid-phase oligonucleotide synthesis includes the removal of protective groups from the heterocyclic bases and splitting of the nucleotide material from the support by treatment with concentrated ammonia. Elimination of the nucleo- tide material under deblocking conditions can be excluded if oligodeoxythymidylates 214 devoid of protective groups are syn- thesised or if supports are used containing primary aliphatic amino 8, 22, 215 or hydroxy 1, 216 ± 218 anchor groups or the hydroxy groups of preliminarily attached nucleosides.31, 219 Oligonucleo- tide synthesis on such supports results in the formation of an ether bond resistant to ammonia.Monomers containing non-conven- tional protective groups of heterocyclic bases are also used in the synthesis of affinity sorbents; their removal requires milder con- ditions,220 which do not cause the destruction of the ester bond between the support and the nucleotide residue. In addition to macroporous glass 8, 31, 214, 216, 217, 220 and glass slides,22 polyacrylmorpholide,31 polypropylene,215 silica gel 31 and Teflon copolymers 216 are used as supports for oligonucleotide synthesis. The affinity matrices based on them manifest good hybridisation properties 8, 215, 217 and can be used for the isolation of mRNA214 and DNA-specific proteins.216 The advantages and disadvantages of different supports in the synthesis of oligonucleotide libraries (combinatorial synthesis) have been reviewed in detail.218 TentaGel (polyoxyethylene- coated polystyrene beads, Rapp Polymere GmbH) and Mercko- Gel [partially hydrolysed poly(vinyl acetate), Merck KgaA] were recognised as the best granulated carriers.Their capacities with respect to the first fragment (nucleoside) are 120 and 700 mmol g71, respectively (cf.30 ± 50 mmol g71 for conventional supports); the elongation of the oligonucleotide chain occurs also with high yields. At present, new procedures and equipment for oligonucleo- tide synthesis on films and glass slides used as DNA chips and DNA sensors are being developed (see Section VII.2.b). V. Miscellaneous methods of immobilisation of nucleic acids Physical and physicochemical methods have been used for the immobilisation of NA earlier than other methods. Adsorption, UV irradiation and mechanical incorporation into different gels are especially popular nowadays. Adsorption of DNA on cellulose is a very simple and rather widespread procedure which consists in freeze-drying 221 ± 223 or conventional drying 222, 224 of cellulose impregnated with solu- tions of NA.The DNA-cellulose thus prepared is successfully used for purification of DNA-dependent proteins. However, such DNA-cellulose is not always suitable for hybridisation of NA because of instability of the nucleotide material at elevated temperatures and its easy desorption in buffers of low ionic strength. Nevertheless, non-covalent sorption of DNA fragments on nitrocellulose filters is a very popular procedure for the preparation of samples for hybridisation analysis 225 despite its obvious disadvantages, such as the loss of the nucleotide material during multistage hybridisation assays, fragility of cellulose after vacuum treatment, etc. The development of polymerase chain reaction (PCR) and the advent of new methods of hybridisation assay brought forth the I G Shishkina, A S Levina, V F Zarytova need for rapid immobilisation of NA fragments on the surface of microplates.Adsorption of oligonucleotides on such surfaces is carried out in the presence of high salt concentrations (NaCl, tetramethylammonium chloride) and/or cationic detergents (cetyltrimethylammonium bromide, dimethyloctylamine hydro- chloride).88, 226 ± 229 Presumably, in this case immobilisation occurs exclusively by virtue of hydrophobic interactions between the oligonucleotides and the support surface.88 It was noted 22 that adsorption is selective as regards surfaces which possess hydro- philic properties.226 Although the mechanism of such immobilisa- tion is not completely understood, the samples prepared by this method are widely used in hybridisation assays.88, 227, 228 Yet another procedure consists in the application of polyelec- trolyte films onto glass surfaces and the sorption of DNA or polynucleotides occurs owing to ion-exchange interactions.230, 231 Nucleic acids and their fragments can be immobilised on different supports upon irradiation with UV light.155, 222, 223, 232 ± 235 Side reactions occurring in the ligands are not significant and do not prevent the application of these systems in affinity chromatography of proteins and enzymes.222, 232, 234 The use of specially prepared photosensitive supports, e.g., cellulose 236 or microporous Nylon 6 237 containing aryl azide residues, increases significantly the maximum capacity and effi- ciency of photoimmobilisation of NA.The sensitivity of hybrid- isation assays with these matrices is 2- to 4-fold higher than that obtained with matrices prepared by conventional sorption on nitrocellulose. This may be ascribed to mild immobilisation conditions which do not affect the structure of NA. Nucleic acids or their fragments can be immobilised on polyacrylamide,238, 239 agar 240 or on glass slides bearing the acrylamide groups,20, 241 which allows the preparation of sorbents with high content of the nucleotide material. The advantage of this method is that the immobilisation can be carried out on virtually any surface.241, 242 To ensure higher selectivity of immobilisation, copolymerisa- tion is carried out with oligonucleotides functionalised with groups containing multiple bonds (e.g., pyrrolyl,242 allyl 20 or acrylamide 241).CHC(O)NH(CH2)nR Polymerisation NH2C(O)CH CH2+CH2 C(O)NH2 CH2CHCH2CH2C(O)NH(CH2)nR R, oligonucleotide residue. VI. Immobilisation of nucleic acids using enzymes and avidin ± biotin interactions Recently, streptavidin- or avidin-containing sorbents have gained wide acceptance for the attachment of biotin-modified DNA or oligonucleotides. In these matrices, immobilisation of oligonu- cleotides is achieved due to the formation of a stable avidin ± bio- tin complex (Kd&10715 mol litre71) rather than a covalent bond. Such sorbents are easily and rapidly prepared by mechan- ical mixing of a streptavidin carrier and a biotin-modified oligo- nucleotide and used for the isolation of proteins and enzymes,35, 36, 243 ± 246 DNA sequencing 34, 247 and hybridisation assays.248, 249 However, these sorbents possess a number of disadvantages, such as non-specific sorption of proteins which is more pronounced in the case of avidin-containing matrices,250, 251 than in the case of streptavidin supports.252 The remaining avidin residues are blocked by bovine serum albumin or methyl manno- side;253 buffers of high ionic strength are used for chromatogra- phy.Short and homogeneous oligonucleotides are more accessible and can be immobilised more easily than long NA fragments.The sorbents containing oligo(dT) [especially oligo(dT)25-Dynabeads] and oligo(dA) have now become especially popular.254 ± 256 Immo-Affinity sorbents containing nucleic acids and their fragments bilisation of large fragments of NAon these or any other supports containing short oligonucleotides (10 ± 20 nucleotide residues) can be carried out enzymically using ligase, by attaching specially synthesised or isolated long fragments of NA,24, 257 or using DNA-polymerase, by completing the mRNA258, 259 or the DNA template 260 which forms complexes with the short oligonucleotide on the polymer.VII. The use of matrices containing nucleic acids 1. Affinity chromatography The solution of the key problems of molecular biology and biochemistry related to the structure and interactions between biopolymers demands the use of special techniques allowing the isolation of individual biopolymers from complex multicompo- nent mixtures.Thus proteins, enzymes, NA, oligonucleotides and their analogues are necessary in a pure state for the studies of their structures as well as for protein-NA,DNA-DNA andRNA-DNA recognitions. These biomolecules can be isolated by affinity chromatography on sorbents containing immobilised NA or oligonucleotides. This method is based on the ability of ligands immobilised on insoluble supports to form highly specific com- plexes with macromolecules, which dissociate under definite conditions (e.g., either upon change in the ionic strength or pH of the eluent or upon increase in temperature). a.Isolation of individual nucleic acids and their fragments The principle of affinity chromatography of NA was first applied to the isolation of oligo- and polynucleotides having homoge- neous sequences. The possibility of fractionation of oligonucleo- tides on supports containing immobilised polynucleotides has been noted in the early papers.60, 61, 261 A series of papers devoted to affinity chromatography of poly(A)-mRNA,214, 254, 262, 263 NA or oligonucleotides containing homogeneous sequences were published in the late 1980's ± early 1990's owing to the develop- ment of methods of synthesis of oligonucleotides and the accessi- bility of novel carriers.26, 264 However, this method was not successful for the isolation of individual oligonucleotides or longer sequences devoid of homogeneous inserts.Present-day studies widely employ commercial supports carrying immobilised homogeneous oligonucleotides, most commonly, oligo(T) sup- ports. The use of durable materials, such as silica gel, glass and some methacrylate-based copolymers as supports allows high- performance affinity chromatography and complete separation of oligonucleotides differing in length by one base.105 Affinity chromatography of heterogeneous oligonucleotides and NA has become possible following development of efficient methods for immobilisation of heterogeneous oligonucleotides on insoluble supports. Matrices containing relatively short (10 ± 15- mers) immobilised oligonucleotide fragments complementary to a Reaction mixture after oligonucleotide synthesisHybridisation pCATTAGTTCTGGGTGCC 50 50 TGGAGAGGTGGAAGTAAGTAATCAAGACCCACGG pCATTAGTTCTGGGTGCC 50 H2O, 65 8C H2O, 65 8C Homogeneous 34-membered oligonucleotide Figure 2.Affinity chromatography. 525 region of an oligonucleotide to be purified have been used for the isolation of extended (more than 30-mers) synthetic oligonucleo- tides which can hardly be purified by conventional chromato- graphic methods.102, 104 This approach was applied for the isolation of a synthetic 34-membered oligonucleotide representing a deoxy copy of an RNA fragment of the Russian tick-borne encephalitis virus on LiChrosorb-NH2 bearing an immobilised 17-membered oligonucleotide complementary to the 50-end of the 34-membered fragment (Fig.2).102 Apart from synthetic oligonucleotides, individual NA and DNA fragments obtained upon scission of DNA with restriction enzymes can be isolated by affinity chromatography from com- plex mixtures (see Fig. 2).102, 104 This approach was used to obtain highly enriched tRNAPhe from human placenta 103 utilising two successive affinity chromatography steps on matrices containing immobilised oligonucleotides complementary to the CCA-end of tRNA (the first step) and to a site in the anti-codon loop of tRNAPhe (the second step). Both single- and double-strandedNAcan be isolated from cell extracts on sorbents carrying immobilised specific oligonucleo- tides. In the latter case, oligonucleotides should be able to form triplexes with definite regions of NA, as can be exemplified by purification of the DNA plasmids pTS2 (Ref.265) and pXL2563 (Ref. 7) from cell extracts. b. Separation of diastereomeric mixtures of non-ionic oligonucleotide analogues Considerable interest in oligonucleotide analogues containing modified internucleotide phosphate groups, viz., methylphospho- nates and phosphotriesters, can be explained by their physico- chemical and biological properties, e.g., resistance to nucleases, sorption on cell membranes, penetration inside the cell and ability to form complexes with the complementary regions of NA. Since these analogues contain an asymmetric internucleotide phospho- rus atom, they represent mixtures consisting of a large number of diastereomers which differ in the stabilities of the complementary complexes formed. Individual diastereomers are necessary for detailed investigations of physicochemical properties and bio- logical behaviour of oligonucleotide analogues in cellular systems.For the first time, chromatographic separation of diastereo- mers of nonathymidilyluridine octaethyl ester into fractions was carried out on a column with poly(A) cyanogen bromide-activated Sepharose.266 However, individual diastereomers of non-ionic analogues of oligonucleotides could be obtained only by high- performance affinity chromatography. The silica gel-based sorbent LiChrosorb bearing the octanu- cleotide pTGTTTGGC immobilised through the 50-terminal phosphate group was used for the isolation of individual diaster- eomers of the heterogeneous octanucleotide GCCAAACA ethyl esters at a constant column temperature.105 This allowed separa- Reaction mixture after DNA scission pTGACCCTCTTCCCATT 50 50 CGTGGGAGAAGGGTAA pTCACCCTC TTCCCATT 50Homogeneous 302-membered DNA fragment526 400 800 600 200 0 ml Figure 3.A chromatographic separation profile of a mixture of four diastereomers of the octanucleotide GpCpCpApOEtApOEtApCpA ethyl esters on an affinity sorbent (LiChrosorb) containing pTGTTTGGC immobilised through the 50-terminal phosphate group.105 tion of stereoisomers of the octanucleotide GCCAAACA con- taining one, two or three phosphotriester groups (Fig. 3). The same affinity sorbent was successfully used to obtain individual diastereomers of the heptanucleoside methylphosphonate CpMeCpMeApMe ApMeApMeCpMeA; the absolute configurations of these diastereomers have been established.267 c.Isolation of NA-dependent enzymes At present, affinity chromatography on cellulose and Sepharose containing calf thymus DNA has become a routine procedure for obtaining pure preparations of NA-dependent proteins which are non-specific as regards the nucleotide sequence. Purification of site-specific proteins is carried out on sorbents with immobilised oligonucleotides having the corresponding recognition sites. Most of the proteins isolated represent transcription and replication protein factors (for a comprehensive description, see the reviews 181, 268, 269).The list given in these reference can be supple- mented by several other papers 36, 137, 270, 271 describing the meth- ods for the preparation of affinity sorbents used for the purification of these proteins. Affinity chromatography was also used for the isolation of some receptors.62, 168, 252 Isolation of site-specific restrictases is of the greatest interest, since affinity matrices in this particular case should be resistant to enzymes to be isolated. The restriction endonucleases EcoRI 272 and SphI 273 hydrolyse DNA only in the presence of Mg2+ ions, whereas in the absence of Mg2+ ions they bind to the site-bearing DNA. This property is used for their isolation. For the isolation of IIS-type restrictases (FokI, HgaI, SfaNI), which hydrolyse DNA at a site other than the recognition site, a duplex which incorpo- rates the recognition sites for these enzymes but contains no sites necessary for DNA hydrolysis has been synthesised.101 The structure of the duplex is given below (the corresponding recog- nition sites are enclosed in boxes).One of the oligonucleotides of the duplex was covalently linked to the methacrylate-based polymeric support. p T T G G A TGA C G C A T C T T p T T G G A TGA C G C A T C T T AAC C T A C T GCGTAGAA FokI SfaNI HgaI Sorbents containing immobilised oligonucleotide duplexes, which, in turn, contain non-canonical base pairs, are used for the isolation of DNA repair enzymes responsible for mismatch correction.Thus thymine-DNA glycosylase 274, 275 was isolated on an affinity DNA sorbent containing a dG:U mismatch, however, in only trace amounts. Highly purified biologically active mismatch-specific DNA glycosylases (MutM and MutY) responsible for the removal of the oxidised form of guanine from I G Shishkina, A S Levina, V F Zarytova DNA as well as for mismatch correction were isolated by affinity chromatography.106 This study utilised a support (Toyopearl-AF) carrying an immobilised oligonucleotide duplex containing the non-canonical pairs N8oxo-dC or rA-dG (where N8oxo is 8- oxonebularine, an analogue of the oxidised form of guanine). Affinity chromatography on oligonucleotide-containing sorb- ents allows both the isolation of NA-dependent proteins and studies of their interactions with immobilised NA frag- ments.106, 276 2.Hybridisation assay Molecular hybridisation of NA is based on the fundamental property of complementary chains ofNAto form stable duplexes. This method allows rapid testing of mutated viral NA and viroids and is distinguished by high specificity, sensitivity and reproduci- bility. The main stages of the hybridisation analysis have been described in detail.277 ± 282 The sensitivity of this method is determined by the choice of procedures for the detection of oligonucleotide probes carrying radioactive ([32P]), fluorescent and other labels. The selection of a support and the procedure for the immobi- lisation of NA or oligonucleotides is the key problem in any solid- phase hybridisation assay, which is determined by a specific goal in each particular case.Dot-blot hybridisation includes immobilisation of a dena- tured portion of total NA to be tested on nitrocellulose.283 ± 289 Polystyrene immunological microwell plates are employed for the automation of serial tests of clinical samples.290, 291 Immobilisa- tion of NA targets or specific cloned fragments is carried out by physicochemical methods, e.g., by sintering at 80 8C or by UV irradiation. Blot-hybridisation in which a NA preparation is hydrolysed by restriction enzymes subjected to fractionation (e.g., separation according to the molecular mass) by gel electrophoresis and further transferred to a support from the gel is a more efficient method of analysis.After immobilisation on the support, NA are hybridised with labelled probes (the so-called Southern blot DNA assay) 225 and Northern blot (RNA assay).292 Currently, the Southern blot is used for the detection of PCR products and determination of the reaction specificity.293 ± 295 Nitrocellulose filters 225, 292, 296, 297 and nylon membranes 298, 299 (less frequently, p-azidobenzoylcellulose 236) or dried gels following fractionation of NA fragments 300 are commonly used for the fixation of the nucleotide material. Sandwich blot is a promising approach to the analysis of NA in non-purified specimens containing proteins, polysaccharides and other components.301 This method is based on the use of two different oligo- or polynucleotide fragments complementary to different sites of the NA to be assayed. One of these fragments is immobilised on a support, whereas the other is soluble and carries a reporter group (e.g., a radiolabel, a biotin residue, etc.).After hybridisation and removal of admixtures and the excess of the second (labelled) probe, theNAthat has formed a complementary complex with the nucleotide immobilised on the support is detected by the label bound. A vast number of solid supports and immobilisation procedures have been used for sandwich blot. Initially, nitrocellulose filters have been used,302, 303 which were replaced by mechanically stronger nylon membranes 304 and more handy polypropylene and polystyrene immunological microwell plates.32, 305 ± 307 Hybridisation of NA with oligonucleotide probes immobilised on the surface of micro- granulated supports is a way of acceleration of the hybridisation. Different versions of this approach using Sephacryl,67 polyacryl- amide,28 nylon 27 and polystyrene beads 304, 308, 309 and magnetic microparticles 25, 219, 304, 310 have been described.The use of mag- netic microparticles has an additional advantage, since they can easily be concentrated from suspensions, which significantly simplifies manipulations with samples. `Reverse dot blot' 48 implies the introduction of a label into the NA or its fragment obtained by PCR. After hybridisation of a labelled NA with an oligonucleotide immobilised on a solidAffinity sorbents containing nucleic acids and their fragments support, the complementary hybrid formed is detected by virtue of the label bound to the support.Nylon 48, 66, 83, 311 and poly- (ethylene terephthalate) 195, 312, 313 membranes, immunological microwell plates 314, 315 and glass 17, 316 have been used as supports for the reverse dot blot. The use of cover glasses in hybridisation assays is especially convenient, since glass is readily available, inexpensive and has a well-studied surface which is easy to modify and is suitable for fluorescent detection. Immobilisation of oligonucleotides is usually carried out by photochemical or various chemical methods. 3. DNA chips and DNA sensors The development of effective methods for the immobilisation of oligonucleotides as well as array-based mapping methods have led to the design of microchips with immobilised oligonucleotide sets (arrays).Such DNA chips which represent slides (usually, glass slides) containing an immense variety of immobilised oligonucleo- tides (tens of thousands of DNA fragments) find ever growing application in molecular biology, medicine, biotechnology, etc. Initially, microchips were proposed for the sequencing of nucleic acids (the so-called SBH method, sequencing by hybrid- isation).132, 317 ± 319 It was assumed that the use of arrays with a complete set of immobilised octanucleotides (65536=48) would make it possible to determine the unknown sequences of nucleic acids by hybridisation assays. However, it was found that serious problems are inherent in the SBH method and not much success in this area has been achieved yet. These studies have shown, however, that microchips can find wide application in the analysis of gene expression and mutations as well as in the diagnostics of genetic, infectious and cancer diseases.320 ± 334 There are two main approaches to the preparation of DNA microchips, viz., the synthesis of oligonucleotides directly on surfaces 4, 5, 8, 175, 215, 217, 218, 321, 333 ± 341 and immobilisation of pre- synthesised oligonucleotides on slides of glass or other materi- als.17, 19, 22, 30, 131, 133, 248, 322, 323, 330 ± 332, 342 ± 344 The former approach has been realised in such methods as combinatorial photolithography on glass slides or other supports using photoreactive protective groups.321, 337 ± 341 Irradiation of slides through a cover mask results in the removal of photo- reactive protective groups from the oligonucleotide chains at the predetermined sites so that the next monomeric unit is coupled to the deblocked oligonucleotide chain.By shifting the mask and by iterating the irradiation and coupling procedures, one can obtain ultimately surfaces containing a large number of oligonucleotides of predetermined sequences at specified points. Despite its obvious economic expediency, direct synthesis of oligonucleotides on surfaces suffers from one serious disadvantage, since it does not provide 100% yields in coupling steps, which can result in errors in further analysis.The latter approach to the manufacture of DNA microchips entails the use of isolated and purified oligonucleotides. Thorough pre-washing and etching prior to immobilisation is a very important stage. These are necessary for the removal of admixtures (grease, dirt, dust, etc.) and for the generation of surface hydroxy groups.345, 346 The washing procedure makes use of organic solvents (most commonly, methanol) or potent oxi- dants (e.g., hydrogen peroxide). The activation of the glass surface is achieved by inorganic acids or alkalis (HF, HCl, HNO3, H2SO4 or NaOH). This is followed by the introduction of functional groups on the slide surface for subsequent reactions using silanes of the general formula (R1O)3Si(CH2)nR,2 where R1=Me or Et and R2 is a functional group (see compounds 1 ± 4 in Sec- tion III.2).The reactions considered in the foregoing sections are used for covalent attachment of oligonucleotides onto the surface of glass slides. Here we shall cite a paper 344 which describes a novel approach. Oligonucleotides or cDNA with linkers contain- ing 50-terminal SH or CH2=CH groups were preincubated with the corresponding trimethoxysilanes, after which the conjugates obtained were immobilised on non-modified glasses. 527 RSH+HS(CH2)3Si(OMe)3 RSS(CH2)3Si(OMe)3 RSH+ [S(CH2)3Si(OMe)3]2 O ICH2CONH(CH2)5C(O)ONO RSH+H2N(CH2)3Si(OMe)3 RSCH2CONH(CH2)5CONH(CH2)3Si(OMe)3 Copoly- merisation RCOC(Me) CH2+H2C C(Me)COO(CH2)3Si(OMe)3 RCOC(Me)CH2CH2C(Me)COO(CH2)3Si(OMe)3 R, oligonucleotide or cDNA.The nature and the lengths of the linkers binding the oligonu- cleotide to the surface are important characteristics for effective hybridisation. Steric accessibility of oligonucleotides for comple- mentary interaction can be achieved through the use of long-chain linkers. The efficiency of hybridisation increases severalfold through the use of spacers of appropriate lengths 8, 38, 320, 331 ± 333 (see Section III.2). The amount of the immobilised oligonucleo- tide referred to the unit of surface area (the surface capacity) is yet another important characteristic of DNA chips; this should not exceed a certain optimum value in order to avoid steric hindrances during subsequent hybridisation. The capacity is not very high (0.01 ± 1.00 pmol mm72) 8, 15 ± 17, 19, 20, 242, 333 but sufficient for successful assays.The optimum capacity of 0.25 ± 0.35 pmol mm72 for the immobilisation of polynucleotides (150 ± 350 b.p.) has been recommended.17 Obviously, the sensitivity of a hybridisation assay increases with an increase in capacity. This is achieved using several approaches, viz., by attachment of oligonucleotides to branched linkers (dendrimers) 22 and by formation of acryl- amide,20, 29, 30, 45, 347 ± 349 agarose 183 or gelatine 350 layers on glass surfaces, which ensures `three-dimensional' attachment of oligo- nucleotides. In this case, the capacities of oligonucleotides referred to the unit of surface area is much higher than in two-dimensional immobilisation and can amount to 30 pmol mm72,45 while steric accessibility of oligomers for subsequent hybridisation is pre- served. In early studies, radiolabelled probes were used for the detection of signals in hybridisation assays.However, other labels incorporated into oligonucleotide probes, such as biotin, digox- igenin, chemiluminophores, fluorescent and enzymic labels, also gained wide acceptance.302, 308, 351 Physical methods of detection have come into use owing to the recent progress in DNA chip technology. The DNA sensors (biosensors), like other biosensors, provide rapid and direct detection both with labels and without them. The use of biosen- sors allows kinetic studies of hybridisation processes.The top- ology of immobilised oligonucleotides in films applied onto the surfaces of slides and fine details of molecular interactions at the film-solution interface are also investigated.337 ± 341, 352 ± 356 Thus the effects of the shape, length and composition of the double helix formed on the efficiency of hybridisation have been studied.4, 357, 358 DNA sensors can utilise piezoelectric technologies for meas- uring the changes in the oscillation frequencies of crystals bearing immobilised oligonucleotides during the sorption of the comple- mentary DNA chain,359 ± 362 surface plasmon resonance technol- ogy (in this case, the refractive index of the metal surface changes upon binding of the soluble ligand to the ligand immobilised on the metal film),363 ± 366 scanning microscopy 352, 367 and X-ray photoelectron spectroscopy (XPS).353 Electrochemical detection makes use of electrodes bearing immobilised oligonucleotides which change their electrical characteristics in the course of hybridisation.365, 368 ± 370 Optical sensors able to detect changes528 in the optical properties (e.g., refractive index, interference, etc.) of films are also widely used for the detection of hybridisation of DNA chains on the surfaces containing immobilised oligonucleo- tides.230, 371 ± 373 These methods ensure high sensitivity, i.e., the detection of nano- and picogram quantities of compounds.The design of DNA chips and DNA sensors is an explosively developing field of science and biotechnology.Advances in this area will be a breakthrough in the diagnostics of genetic, cancer and infectious diseases as well as in medicine and pharmacology. 4. Enzymic synthesis on NA templates Immobilised synthetic oligonucleotides and NA as templates or primers for enzymic synthesis of oligo- and polynucleotides can repeatedly be used for the synthesis of complementary oligonuc- leotide chains, which simplifies the separation of the final product from the templates. The latter circumstance is especially impor- tant, since along with the template or the primer the reaction mixture contains the enzyme and a large excess of nucleoside triphosphates (NTP) used as substrates for NA biosynthesis. DNA polymerases of bacteriophage T7,34, 374 polymerase I of E.coli, the Klenow fragment 85, 257, 260, 375, 376 and Taq polymer- ase 377 ± 380 are used in solid-phase enzymic synthesis of DNA or DNA fragments.cDNAs synthesised on oligo(dT)-bearing supports (e.g., cellulose,258, 259, 381 latex 382 ± 384 or magnetic par- ticles 374) by transcription of the corresponding poly(A) ±mRNAs in the presence of RNA-dependent DNA polymerase (reverse transcriptase) are most often used as templates. Some authors recommend to immobilise plasmids 34 or bacteriophages 192, 376 containing incorporated exogeneous DNAs (including cDNAs) 192 and utilise them as templates for subsequent enzymic synthesis on supports. Using primers complementary to the 30-terminal region of the immobilised template, one can obtain full complementary DNA.85, 192, 257, 380 Enzymic template syn- thesis permits one to introduce various functional groups and labels into DNAs and their fragments 257, 375 using primers, dNTP or terminating dideoxyribonucleoside triphosphates (ddNTP) carrying the corresponding functional groups.The addition of one or all the four ddNTP with different labels to the reaction mixture along with four dNTP makes it possible to obtain series of labelled products and to determine the oligonucleotide sequence of the immobilised DNA.34, 378, 379 The chemical synthesis of oligoribonucleotides and RNA is more complicated than that of oligodeoxyribonucleotides, hence, the interest in enzymic methods is obvious. Biosynthesis of RNA is carried out using DNA-dependent RNA polymerases, e.g., RNA polymerases of T7 129, 385, 386 and E.coli.381, 387, 388 Immobi- lised oligodeoxyribonucleotides 129, 385, 386, 388 and cDNA are used as templates for oligoribonucleotide and mRNA syntheses, respectively.381, 387 The amount of the final products can be increased by means of PCR technology, while modern magnetic particles (Dynabeads) ensure easier separation of the products synthesised.Enzymic synthesis on immobilised NA templates is used for the replication, transcription and sequencing of nucleic acids as well as for the introduction of functional groups and labels and the construction of genomic and cDNA libraries. VIII. Conclusion The work considered in this review demonstrates the immense diversity of methods for immobilisation of NA on solid supports.Immobilisation of NA on activated supports are especially widely used. In this case, native NA is immobilised non-specifically through many points. However, this results in the preparation of affinity matrices with higher capacities. The attachment of nucleotides and oligonucleotides containing additional reactive groups occurs much more selectively, i.e., without affecting the heterocyclic bases and with the preservation of their abilities for affinity interactions. I G Shishkina, A S Levina, V F Zarytova Physical immobilisation methods are irreproducible and non- specific. The advantage of chemical approaches is the possibility to use the immobilisation methods corresponding to a specified experimental task.The stability of affinity sorbents under various conditions and the procedure used for the removal of the ligand from the support can be determined in advance provided that the nature of the bond between the ligand and the support is known. Reactions of ligands containing amino or thiol groups with the reactive groups of the support have become especially wide- spread among other methods of immobilisation of NA and their fragments. The attachment of oligonucleotides containing amino groups to supports containing aldehyde groups is a common procedure. The method utilising the Ph3P±Py2S2 couple as an activating reagent has proved to be efficient in the binding of oligonucleotides with terminal phosphate groups to granulated polymers.In the case of planar supports, e.g., films and glasses, it is more expedient to use methods implying the activation of the functional groups of the support or utilise bifunctional cross- linking reagents, e.g., 1,4-phenylene diisocyanate. It should be noted, however, that no universal immobilisation procedure exists and both the support and the immobilisation procedure should be chosen for each particular case. NA-containing carriers have found wide acceptance for both research and applied purposes. A number of reviews devoted to this topic has been published in recent years.326, 328, 347, 389 ± 391 The supports with immobilised oligonucleotides are used as sorbents in affinity chromatography in order to obtain highly purified individualNAor their fragments.Affinity chromatography seems to be the only efficient procedure for the separation of diastereo- meric mixtures of non-ionic oligonucleotide analogues. Affinity oligonucleotide-containing matrices have proved to be highly efficient for the isolation of nucleotide-dependent enzymes. Such supports are also used in solid-phase enzymic synthesis of DNA, RNA and their fragments. Special mention should be made of NA-containing carriers for hybridisation assays, which underlie the diagnostics of many diseases. It is used for detecting mutations, viral and bacterial NA, etc. This method is gaining increasing popularity for genetic expertise in medical and forensic practice. The trend which includes the design and application of DNA microchips is developing explosively in recent years.These tools have proved to be useful in various fields of biology and medicine. Their unique properties are based on the fact that a small area of glass slides contains a large amount of immobilised oligonucleo- tides, which allows a great number of assays to be conducted simultaneously. The use of immobilised oligonucleotides as biosensors adds novel unique capabilities to the arsenal of modern analytic techniques. They permit one to obtain information more rapidly, simply and less expensively than in conventional hybridisation techniques. The novel devices, viz., DNA biosensors, hold tre- mendous potential in many areas, e.g., clinical analyses for forensic expertise, monitoring of the environment, food quality, etc.References 1. G Cohen, J Deutsch, J Fineberg, A Levine Nucl. Acids Res. 25 911 (1997) 2. R S Matson, J B Rampal, P J Coassin Anal. Biochem. 217 306 (1994) 3. K Ernst-Cabrera, M Wilchek Anal. Biochem. 159 267 (1986) 4. U Maskos, E M Southern Nucl. Acids Res. 20 1679 (1992) 5. E M Southern, S C Case-Green, J K Elder, M Johnson, K U Mir, L Wang, J C Williams Nucl. Acids Res. 22 1368 (1994) 6. H Hakala, H LoÈ nnberg Bioconjugate Chem. 8 232 (1997) 7. P Wils, V Escriou, A Warnery, F Lacroix, D Lagneaux, M Ollivier, J Crouzet, J-F Mayaux, D Scherman Gene Ther. 4 323 (1997) 8. M S Shchepinov, S C Case-Green, E M Southern Nucl. Acids Res. 25 1155 (1997)Affinity sorbents containing nucleic acids and their fragments 9.J Turkova Affinity Chromatography (Amsterdam: Elsevier, 1978) 10. V Jirku, J Turkova Methods Enzymol. 135 341 (1987) 11. MYang,RYC Kong,NKazmi,AKC Leung Chem. Lett. 257 (1998) 12. M Zammatteo, L Jeanmart, S Hamels, S Courtois, P Louette, L Hevesi, J Remacle Anal. Biochem. 280 143 (2000) 13. L Henke, P A E Piunno, A C McClure, U J Krull Anal. Chim. Acta 344 210 (1997) 14. B S Sproat, D M Brown Nucl. Acids Res. 13 2979 (1985) 15. J B Lamture, K L Beattie, B E Burke, M D Eggers, D J Ehrlich, R Fowler,M A Hollis, B B Kosicki, R K Reich, S R Smith, R S Varma,M E Hogan Nucl. Acids Res. 22 2121 (1994) 16. L A Chrisey, G U Lee, C E O'Ferrall Nucl. Acids Res. 24 3031 (1996) 17.Z Guo, R A Guilfoyle, A J Thiel, R Wang, L M Smith Nucl. Acids Res. 22 5456 (1994) 18. B Joos, H Kuster, R Cone Anal. Biochem. 247 96 (1997) 19. D Chen, Z Yan, D L Cole, G S Srivatsa Nucl. Acids Res. 27 389 (1999) 20. A V Vasiliskov, E N Timofeev, S A Surzhikov, A L Drobyshev, V V Shick, A D Mirzabekov Biotechniques 27 592 (1999) 21. N P Gerry, N E Witowski, J Day, R P Hammer, G Barany, F Barany J. Mol. Biol. 292 251 (1999) 22. M Beier, J D Hoheisel Nucl. Acids Res. 27 1970 (1999) 23. T R Gingeras, D Y Kwoh, G R Davis Nucl. Acids Res. 15 5373 (1987) 24. J N Kremsky, J L Wooters, J P Dougherty, R E Meyers, M Collins, E L Brown Nucl. Acids Res. 15 2891 (1987) 25. V Lund, R Schmid, D Rickwood, E Hornes Nucl. Acids Res. 16 10861 (1988) 26.T A Goss, M Bard, H W Jarrett J. Chromatogr. 508 279 (1990) 27. J Van Ness, S Kalbfleisch, C R Petrie,M W Reed, J C Tabone, N M J Vermeulen Nucl. Acids Res. 19 3345 (1991) 28. E Fahy, G R Davis, L J DiMichele, S S Ghosh Nucl. Acids Res. 21 1819 (1993) 29. E N Timofeev, S V Kochetkova, A D Mirzabekov, V L Florentiev Nucl. Acids Res. 24 3142 (1996) 30. D Proudnikov, E Timofeev, A Mirzabekov Anal. Biochem. 259 34 (1998) 31. S Pochet, T Huynh-Dinh, J Igolen Tetrahedron 43 3481 (1987) 32. J A Running,M S Urdea Biotechniques 8 276 (1990) 33. M Shimkus, J Levy, T Herman Proc. Natl. Acad. Sci. USA 82 2593 (1985) 34. T Hultman, S Stahl, E Hornes,M Uhlen Nucl. Acids Res. 17 4937 (1989) 35. M Leblond-Francillard, M Dreyfus, F Rougeon Eur.J. Biochem. 166 351 (1987) 36. O S Gabrielsen, E Hornes, L Korsnes, A Ruet, T B éyen Nucl. Acids Res. 17 6253 (1989) 37. W Weber, C-W Vogel, H Hilz FEBS Lett. 99 62 (1979) 38. Y Zhang,M Y Coyne, S G Will, C H Levenson, E S Kawasaki Nucl. Acids Res. 19 3929 (1991) 39. P A E Piunno, U J Krull, R H E Hudson,M J Damha, H Cohen Anal. Chim. Acta 288 205 (1994) 40. A V Fotin, A L Drobyshev, D Y Proudnikov, A N Perov, A D Mirzabekov Nucl. Acids Res. 26 1515 (1998) 41. Yu P Lysov, A A Chernyi, A A Balaev, F N Gnuchev, K L Bitti, A D Mirzabekov Mol. Biol. 29 104 (1995) a 42. M A Livshits, V L Florentiev, A D Mirzabekov J. Biomol. Struct. Dyn. 11 783 (1994) 43. D Guschin, G Yershov, A Zaslavsky, A Gemmell, V Shick, D Proudnikov, P Arenkov, A Mirzabekov Anal.Biochem. 250 203 (1997) 44. A Kunitsyn, S Kochetkova, E Timofeev, V Florentiev J. Biomol. Struct. Dyn. 14 239 (1996) 45. S Parinov, V Barsky, G Yershov, E Kirillov, E Timofeev, A Belgovskiy, A Mirzabekov Nucl. Acids Res. 24 2998 (1996) 46. S Dubiley, E Kirillov, Y Lysov, A Mirzabekov Nucl. Acids Res. 25 2259 (1997) 47. I B Ivanov, GMErshov, V E Barskii, A I Bel'govskii, E V Kirillov, E Ya Kreindlin, S V Parinov, N V Mologina, A D Mirzabekov Mol. Biol. 31 159 (1997) a 48. R K Saiki, P S Walsh, C H Levenson, H A Erlich Proc. Natl. Acad. Sci. USA 86 6230 (1989) 49. Ph J R Day, P S Flora, J E Fox, M R Walker Biochem. J. 278 735 (1991) 529 50. N K Kochetkov, E I Budovskii, E D Sverdlov, N A Simukova, M F Turchinskii, V N Shibaev, in Organicheskaya Khimiya Nuklei- novykh Kislot (Organic Chemistry of Nucleic Acids) (Moscow: Khimiya, 1970) 51.D G Knorre, V F Zarytova, A G Badashkeeva, O S Fedorova Biotekhnologiya (Itogi Nauki i Tekhniki) [Biotechnology (Advances in Science and Engineering Series)] (Moscow: Izd. VINITI, 1991) Vol. 37 52. S Agrawal Methods Mol. Biol. 26 93 (1994) 53. J A Fidanza, H Ozaki, L W McLaughlin Methods Mol. Biol. 26 121 (1994) 54. B Chu, L E Orgel Methods Mol. Biol. 26 145 (1994) 55. DGKnorre,VVVlassov,VFZarytova,AVLebedev,OS Fedorova, in Design and Targeted Reactions of Oligonucleotide Derivatives (Boca Raton, FL: CRC Press, 1994) p. 101 56. V V Shumyantseva, R M Khomutov Bioorg. Khim. 3 1471 (1977) b 57. V V Shumyantzeva, N I Sokolova, Z A Shabarova Nucl.Acids Res. 3 903 (1976) 58. V N Shibaev, Yu Yu Kusov, N A Kalinchuk, N K Kochetkov Bioorg. Khim. 3 120 (1977) b 59. E K F Bautz, B D Hall Proc. Natl. Acad. Sci. USA 48 400 (1962) 60. P T Gilham J. Am. Chem. Soc. 84 1311 (1962) 61. P T Gilham J. Am. Chem. Soc. 86 4982 (1964) 62. S C Gross, S A Kumar, H W Dickerman J. Biol. Chem. 257 4738 (1982) 63. T Y Shih,M A Martin Biochemistry 13 3411 (1974) 64. WO PCT 9 009 393; Chem. Abstr. 149 247 682 (1990) 65. S S Ghosh, G F Musso Nucl. Acids Res. 15 5353 (1987) 66. M G Ivanovskaya, I A Kozlov, N A Naryshkin, Z A Shabarova Bioorg. Khim. 21 535 (1995) b 67. J A Langdale, A D Malcolm Gene 36 201 (1985) 68. H G Khorana Some Recent Developments in the Chemistry of Phos- phate Esters of Biological Interest (New York: Wiley, 1961) 69. Z A Shabarova, A A Bogdanov Khimiya Nukleinovykh Kislot i Ikh Komponentov (The Chemistry of Nucleic Acids and Their Components) (Moscow: Khimiya, 1978) 70.J P Riehm, H A Scheraga Biochemistry, 5 99 (1966) 71. A Williams, I T Ibrahim Chem. Rev. 81 589 (1981) 72. P T Gilham Biochemistry 7 2809 (1968) 73. D Rickwood Biochim. Biophys. Acta 269 47 (1972) 74. G T Babkina, D G Knorre, L G Olenchich Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Khim. Nauk (1) 120 (1972) 75. V F Zarytova, E M Ivanova, A V Lebedev Bioorg. Khim. 2 189 (1976) b 76. D G Knorre, V F Zarytova Nucl. Acids Res. 3 2709 (1976) 77. M G Ivanovskaya, N A Naryshkin, Z A Shabarova Bioorg. Khim. 21 454 (1995) b 78. D G Knorre, T N Shubina Kinet.Katal. 5 637 (1964) c 79. V F Zarytova, E M Ivanova, A V Lebedev Bioorg. Khim. 2 1196 (1976) b 80. M G Ivanovskaya,M B Gottikh, Z A Shabarova Nucleosides Nucleotides 6 913 (1987) 81. V F Zarytova, V K Rait, T S Chernikova Bioorg. Khim. 3 1626 (1977) b 82. M G Ivanovskaya, I A Kozlov, I V Lebedeva, Z A Shabarova Mol. Biol. 28 1176 (1994) a 83. Y Zhang,M Y Coyne, S G Will, C H Levenson, E S Kawasaki Nucl. Acids Res. 19 3929 (1991) 84. M V Berridge, A I Aronson Anal. Biochem. 53 603 (1973) 85. A Panet, H G Khorana J. Biol. Chem. 249 5213 (1974) 86. W H Scouten, in Affinity Chromatography. Bioselective Adsorption on Inert Matrices (New York: Wiley, 1981) Vol. 59, p. 349 87. V I Latich, V P Varlamov, N N Semenova, S V Rogozhin Biokhimiya 45 1597 (1980) d 88.T T Nikiforov, Y H Rogers Anal. Biochem. 227 201 (1995) 89. V G Allfrey, A Inoue Methods Cell Biol. 17 253 (1978) 90. R Bischoff, J M Coull, F E Regnier Anal. Biochem. 164 336 (1987) 91. B C F Chu, G M Wahl, L E Orgel Nucl. Acids Res. 11 6513 (1983) 92. M B Gottikh,M G Ivanovskaya, E A Skripkin, Z A Shabarova Bioorg. Khim. 16 514 (1990) b 93. V L Drutsa, V F Zarytova, D G Knorre, A V Lebedev, N I Sokolova, Z A Shabarova Nucl. Acids Res. 5 185 (1978)530 94. V L Drutsa, V F Zarytova, D G Knorre, A V Lebedev, N I Sokolova, Z A Shabarova Dokl. Akad. Nauk SSSR 233 595 (1977) e 95. V V Nosova, I S Vel'moga, N I Sokolova, Z A Shabarova, M A Prokof'ev Dokl. Akad. Nauk SSSR 219 1134 (1974) e 96.V V Shumyantseva, R M Khomutov Bioorg. Khim. 4 1477 (1978) b 97. T Mukaiyama, R Matsueda,M Suzuki Tetrahedron Lett. 1901 (1970) 98. T Mukaiyama,M Hashimoto J. Am. Chem. Soc. 94 8528 (1972) 99. V F Zarytova, T S Godovikova, I V Kutyavin, L M Khalimskaya, in Reactive Oligonucleotide Derivatives as Affinity Reagents and Probes in Molecular Biology (Eds K S Bruzik,W J Stec) (Amsterdam: Elsevier, 1987) p. 149 100. T S Godovikova, V F Zarytova, L M Khalimskaya Bioorg. Khim. 12 475 (1986) b 101. S Kh Degtyarev, P A Belavin, I G Shishkina, V F Zarytova, L P Gavryuchenkova, S N Morozov Bioorg. Khim. 15 358 (1989) b 102. V F Zarytova, I G Shishkina Anal. Biochem. 188 214 (1990) 103. V F Zarytova, G G Karpova, D A Mundus, I G Shishkina Bioorg.Khim. 16 1069 (1990) b 104. T V Abramova, N V Amirkhanov, V V Gorn, V F Zarytova, E I Frolova, I G Shishkina Bioorg. Khim. 17 232 (1991) b 105. I A Vtorushina, Yu A Gorbunov, V F Zarytova, N I Komarova, A V Lebedev, S G Lokhov, A V Manaenko, A N Sinyakov, I G Shishkina Bioorg. Khim. 18 92 (1992) b 106. I G Shishkina, N V Bulychev, A P Grollman, F Johnson Int. J. Bio-Chromatogr. 3 329 (1997) 107. B M Kurinenko, F F Shagi-Mukhamedova, A M Nuzhina Bioorg. Khim. 1 1624 (1975) b 108. P T Gilham Adv. Exp. Med. Biol. 42 173 (1974) 109. P T Gilham Methods Enzymol. 21 191 (1971) 110. G R Jacobson, J P Rosenbusch FEBS Lett. 79 8 (1977) 111. A A Nedospasov, R M Khomutov Bioorg. Khim. 4 645 (1978) b 112. D L Robberson, N Davidson Biochemistry 11 533 (1972) 113.A Weissbach,M Poonian Methods Enzymol. 34 463 (1974) 114. M Wilchek, R Lamed Methods Enzymol. 34 475 (1974) 115. M B Ustav, Ya L Remme, A Ya Lind, R L-E Villems Bioorg. Khim. 5 365 (1979) b 116. N Ulbrich, I G Wool, E Ackerman, P B Sigler J. Biol. Chem. 255 7010 (1980) 117. H R Burrell, J Horowitz FEBS Lett. 49 306 (1975) 118. A Oplatka, O Muhlrad, R Lamed J. Biol. Chem. 251 3972 (1976) 119. A S Butorin, S K Vasilenko Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Khim. Nauk (9) 143 (1976) 120. E Junowicz, S E Charm Biochim. Biophys. Acta 428 157 (1976) 121. N V Belitsina, S M Elizarov, M A Glukhova, A S Spirin, A S Butorin, S K Vasilenko FEBS Lett. 57 262 (1975) 122. M Ustav, R Villems, A Lind FEBS Lett. 82 259 (1977) 123.J O Langland, S M Pettiford, B L Jacobs Protein Exp. Purif. 6 25 (1995) 124. D G Knorre, S D Myzina, L S Sandakhchiev Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Khim. Nauk (3) 135 (1964) 125. C M Joyce, J R Knowles Biochem. Biophys. Res. Commun. 60 1278 (1974) 126. P Remy, C Birmele, J P Ebel FEBS Lett. 27 134 (1972) 127. S Bartkowiak, J Pawelkiewicz Biochim. Biophys. Acta 272 137 (1972) 128. J J Befort, N Befort, G Petrissant, P Remy, J-P Ebel Biochimie 56 625 (1974) 129. A A EÈ lov, E M Volkov, T G Reintamm, T S Oretskaya, Z A Shabarova Bioorg. Khim. 15 159 (1989) b 130. S Agrawal, C Christodoulou, M J Gait Nucl. Acids Res. 14 6227 (1986) 131. Yu P Lysov, V P Florent'ev, A A Khorlin, K R Khrapko, V V Shik, A D Mirzabekov Dokl.Akad. Nauk SSSR 303 1508 (1988) e 132. K R Khrapko, Y P Lysov, A A Khorlin, V V Shik, V L Florentiev, A D Mirzabekov FEBS Lett. 256 118 (1989) 133. K R Khrapko, Y P Lysov, A A Khorlin, I B Ivanov, G MErshov, S K Vasilenko, V L Florent'ev, A D Mirzabekov DNA Sequence 1 375 (1991) 134. R Blanks, L W McLaughlin Nucl. Acids Res. 16 10283 (1988) 135. B A Connolly, P Rider Nucl. Acids Res. 13 4485 (1985) I G Shishkina, A S Levina, V F Zarytova 136. L A Chrisey, C E O'Ferrall, B J Spargo, C S Dulcey, J M Calvert Nucl. Acids Res. 24 3040 (1996) 137. S-H Kang, X Xu, O Heidenreich, S Gryaznov,M Nerenberg Nucl. Acids Res. 23 2344 (1995) 138. R Axen, J Porath, S Ernback Nature (London) 214 1302 (1967) 139. M S Poonian, A J Schlabach, A Weissbach Biochemistry 10 424 (1971) 140.D J Arndt-Jovin, T M Jovin, W Bihr, A-M Frischauf, M Marquardt Eur. J. Biochem. 54 411 (1975) 141. A F Wagner, R L Bugianesi, T Y Shen Biochem. Biophys. Res. Commun. 45 184 (1971) 142. B A Klyashchitskii, V S Shapot Usp. Biol. Khim. 17 234 (1976) 143. J Kohn, M Wilchek Biochem. Biophys. Res. Commun. 107 878 (1982) 144. P Cuatrecasas J. Biol. Chem. 245 3059 (1970) 145. S C March, I Parikh, P Cuatrecasas Anal. Biochem. 60 149 (1974) 146. H Bmnemann, P Westhoff, R G Herrmann Nucl. Acids Res. 10 7163 (1982) 147. U Lindberg, T Persson Eur. J. Biochem. 31 246 (1972) 148. G Vassart, H Brocas, P Nokin, J E Dumont Biochim. Biophys. Acta 324 575 (1973) 149. J N Ihle, K-L Lee, F T Kenney J. Biol. Chem. 249 38 (1974) 150.J Kempf, N Pfleder, J M Egly J. Chromatogr. 147 195 (1978) 151. S G Siddell Eur. J. Biochem. 92 621 (1978) 152. J C Schabort J. Chromatogr. 73 253 (1972) 153. L Clerici, F Campagnari, J F M de Rooij, J H van Boom Nucl. Acids Res. 6 247 (1979) 154. N I Sokolova, I V Ponomarenko, Z A Shabarova, M A Prokof'ev Dokl. Akad. Nauk SSSR 253 1395 (1980) e 155. C R Lowe, P D G Dean, in Affinity Chromatography (London: Wiley, 1974) p. 200 156. P D G Dean, D B Craven, M J Harvey, C R Lowe Adv. Exp. Med. Biol. 42 99 (1974) 157. H Guilford, P-O Larsson, K Mosbach Chem. Scr. 2 165 (1972) 158. P Brodelius, P-O Larsson, K Mosbach Eur. J. Biochem. 47 81 (1974) 159. I P Trayer, H R Trayer, D A P Small, R C Bottomley Biochem. J. 139 609 (1974) 160.O Berglund, F Ecstein Eur. J. Biochem. 28 492 (1972) 161. R Barker, K W Olsen, J H Shaper, R L Hill J. Biol. Chem. 247 7135 (1972) 162. S Ikeda, R P Swenson, D H Ives Biochemistry 27 8648 (1988) 163. N R Webb, R V J Chari, G DePillis, J W Kozarich, R E Rhoads Biochemistry 23 177 (1984) 164. G I Tesser, H-U Fisch, R Schwyzer Helv. Chim. Acta 57 1718 (1974) 165. J T Kadonaga, R Tjian Proc. Natl. Acad. Sci. USA 83 5889 (1986) 166. Z Hostomsky, J Smrt, L Arnold, Z Tocik,V Paces Nucl. Acids Res. 15 4849 (1987) 167. D Staiger, F Becker, J Schell, C Koncz, K Palme Eur. J. Biochem. 199 519 (1991) 168. G G F Mason, A-M Witte, M L Whitelaw, C Antonsson, J McGuire, A Wilhelmsson, L Poellinger, J A Gustafsson J. Biol. Chem. 269 4438 (1994) 169.A Denizli, E Pikin J. Chromatogr. B., Biomed. Sci. Appl. 666 215 (1995) 170. G S Cox, D W Gutkin, M J Haas, D E Cosgrove Biochim. Biophys. Acta 1396 67 (1998) 171. J Porath Biochimie 55 943 (1973) 172. L G Moss, J P Moore, L Chan J. Biol. Chem. 256 12655 (1981) 173. P A Lazo Biochem. J. 217 331 (1984) 174. P O Larsson,M Glad, L Hansson, M-O Mansson, S Ohlson, K Mosbach Adv. Chromatogr. 21 41 (1983) 175. E M Southern, U Maskos, J K Elder Genomics 13 1008 (1992) 176. H Potuzak, P D G Dean Nucl. Acids Res. 5 297 (1978) 177. J B Wheatley, M H Lyttle,M D Hocker, D E Schmidt Jr J. Chromatogr. A 726 77 (1996) 178. G W Anderson, J E Zimmerman, F M Callahan J. Am. Chem. Soc. 86 1839 (1964) 179. S Sakakibara, N Inukai Bull. Chem. Soc.Jpn. 38 1979 (1965) 180. M Wilchek, K L Knudsen, T Miron Bioconjugate Chem. 5 491 (1994) 181. M Wilchek, T Miron, J Kohl Methods Enzymol. 104 3 (1984)Affinity sorbents containing nucleic acids and their fragments 182. S Pochet, B Arcangioli, T Huynh-Dinh Nucl. Acids Res. 16 1619 (1988) 183. V Afanassiev, V Hanemann, S WoÈ lfl Nucl. Acids Res. 28 e66 (2000) 184. B E Noyes, G R Stark Cell 5 301 (1975) 185. B Seed Nucl. Acids Res. 10 1799 (1982) 186. J K Inman, H M Dintzis Biochemistry 8 4074 (1969) 187. D Barton, W D Ollis (Eds) Comprehensive Organic Chemistry Vol. 2 (Oxford: Pergamon Press, 1979) 188. M Fontanel, H Bazin, R Tzhoule Bioconjugate Chem. 4 380 (1993) 189. V V Korshak, M I Shtil'man, in Polimery v Protsessakh Immobili- zatsii i Modifikatsii Prirodnykh Soedinenii (Polymers in Processes of Immobilisation and Modification of Natural Compounds) (Moscow: Nauka, 1984) 190.E N Moudrianakis, M Beer Proc. Natl. Acad. Sci. USA 53 564 (1965) 191. J C Alwine, D J Kemp, G R Stark Proc. Natl. Acad. Sci. USA 74 5350 (1977) 192. P L Ashley, R J MacDonald Anal. Biochem. 140 95 (1984) 193. J F M De Rooij, G Wille-Hazeleger, A B J Vink, J H van Boom Tetrahedron 35 2913 (1979) 194. J H van Boom, J F M de Rooij, L Clerici, F Campagnari Nucl. Acids Res. (Spec. Publ.) 4 85 (1978) 195. T S Godovikova, T N Orlova, E Yu Dobrikova, V A Shamanin, V F Zarytova, N V Vorob'eva, N A Serdyukova, N Yu Shamanina, I O Petruseva, N D Pitsenko Bioorg. Khim. 20 1196 (1994) b 196. Russ.P. 1 808 014; Byull. Izobret. (13) 227 (1993) 197. M A Harpold, M Calvin Nature (London) 219 486 (1968) 198. R Paul, G W Anderson J. Org. Chem. 27 2094 (1962) 199. G J Bartling, H D Brown, S K Chattopadhyay Nature (London) 243 342 (1973) 200. M T W Hearn Methods Enzymol. 135 102 (1987) 201. G S Bethell, J S Ayers,W S Hancock , M T W Hearn J. Biol. Chem. 254 2572 (1979) 202. K Nilsson, K Mosbach Eur. J. Biochem. 112 397 (1980) 203. K Nilsson, K Mosbach Biochem. Biophys. Res. Commun. 102 449 (1981) 204. K Nilsson, K Mosbach Methods Enzymol. 135 65 (1987) 205. H-C Kirch, H KruÈ ger, H S Holthausen Nucl. Acids Res. 19 3156 (1991) 206. R F Borch,M D Bernstein, H D Durst J. Am. Chem. Soc. 93 2897 (1971) 207. V V Romanov, V K Starostina Biotekhnologiya 7 618 (1987) 208.H Ukeda, K Ono, M Imabayashi, P Matsumoto, Y Osajima Agric. Biol. Chem. 53 235 (1989) 209. F B Anspach, H-J Wirth , K K Unger, P Stanton, J R Davies, M T W Hearn Anal. Biochem. 179 171 (1989) 210. A Freeman, M Sokolovsky, L Goldstein Biochim. Biophys. Acta 571 127 (1979) 211. I Ugi Angew. Chem. 74 9 (1962) 212. P Vretblad, R Axen Acta Chem. Scand. 27 2769 (1973) 213. R Axen, P Vretblad, J Porath Acta Chem. Scand. 25 1129 (1971) 214. T Chow, C Juby, L Yuen Anal. Biochem. 175 63 (1988) 215. R S Matson, J Rampal, S L Pentoney Jr, P D Anderson, P Coassin Anal. Biochem. 224 110 (1995) 216. C H Duncan, S L Cavalier Anal. Biochem. 169 104 (1988) 217. U Maskos, E M Southern Nucl. Acids Res. 20 1675 (1992) 218.H Seliger, R Bader,M Hinz, B Rotte, F Eisenbeiss, S Gura, B Nitzan, S Margel React. Funct. Polym. 43 325 (2000) 219. C Albretsen, K-H Kalland, B-I Haukanes, L-S Havarstein, K Kleppe. Anal. Biochem. 189 40 (1990) 220. Y Hayakawa, S Wakabayashi, H Kato, R Noyori J. Am. Chem. Soc. 112 1691 (1990) 221. B M Alberts, F J Amodio, M Jenkins, E D Gutmann, F J Ferris Cold Spring Harbor Symp. Quant. Biol. 33 289 (1968) 222. H Potuzak, U Wintersberger FEBS Lett. 63 167 (1976) 223. R M Litman J. Biol. Chem. 243 6222 (1968) 224. R F Baker Anal. Biochem. 78 569 (1977) 225. E M Southern J. Mol. Biol. 98 503 (1975) 226. T T Nikiforov, R B Rendle, P Goelet, Y-H Rogers, M L Kotewicz, S Anderson, G L Trainor, M R Knapp Nucl. Acids Res. 22 4167 (1994) 227.P Cros, R Kurfmrst, P Allibert, N Battail, N Piga, V Roig, N T Thuong, B Mandrand, C Hzhlzhne Nucl. Acids Res. 22 2951 (1994) 531 228. M M DeAngelis, D G Wang, T L Hawkins Nucl. Acids Res. 23 4742 (1995) 229. V Balladur, A Theretz, B Mandrand J. Colloid Interface Sci. 194 408 (1997) 230. J Piehler, A Brecht, G Gauglitz,M Zerlin, C Maul, R Thiericke, S Grabley Anal. Biochem. 249 94 (1997) 231. B I Sukhorukov, G B Sukhorukov, L I Shabarchina,MMMontrel' Biofizika 45 40 (2000) f 232. R J Britten Science 142 963 (1963) 233. S Fukui, K Sonomoto, N Itoh, A Tanaka Biochimie 62 381 (1980) 234. E W Khandjian Mol. Biol. Rep. 11 107 (1986) 235. S M Kalachikov, V A Adarichev,G M Dymshits Bioorg. Khim. 18 52 (1992) b 236. L A Frumgarts, S M Kalachikov, G M Dymshits, M I Dobrikov, G V Shishkin Izv.Sib. Otd. Akad. Nauk SSSR, Ser. Khim. Nauk (1) 81 (1988) 237. M I Dobrikov, T A Prikhod'ko, I V Safronov, G V Shishkin Sib. Khim. Zh. 2 18 (1992) 238. A Wada, A Kishizaki Biochim. Biophys. Acta 166 29 (1968) 239. L I Gus'kova, V P Demushkin Usp. Khim. 43 1241 (1974) [Russ. Chem. Rev. 43 574 (1974)] 240. E T Bolton, B J McCarthy Proc. Natl. Acad. Sci. USA 48 1390 (1962) 241. F N Rehman, M Audeh, E S Abrams, P W Hammond, M Kenney, T C Boles Nucl. Acids Res. 27 649 (1999) 242. T Livache, A Roget, E Dejean, C Barthet, G Bidan, R Tzhoule Nucl. Acids Res. 22 2915 (1994) 243. J G Morgan, G M Dolganov, S E Robbins, L M Hinton, M Lovett Nucl. Acids Res. 20 5173 (1992) 244. H Ji, L M Smith Anal.Chem. 65 1323 (1993) 245. J Ostrowski, K Bomsztyk Nucl. Acids Res. 21 1045 (1993) 246. G Gadaleta, D D'Elia, L Capaccio, C Saccone, G Pepe J. Biol. Chem. 271 13537 (1996) 247. T Hultman, S Stahl, T Moks,M Uhlen Nucleosides Nucleotides 7 629 (1988) 248. D O'Meara, P Nilsson, P-AÊ Nygren,M Uhle n, J Lundeberg Anal. Biochem. 255 195 (1998) 249. S W Lucas, M M Harding Anal. Biochem. 282 70 (2000) 250. R C Duhamel, J S Whitehead Methods Enzymol. 184 201 (1990) 251. G S Wood, R Warnke J. Histochem. Cytochem. 29 1196 (1981) 252. R Alon, E A Bayer,M Wilchek Biochem. Biophys. Res. Commun. 170 1236 (1990) 253. G Bussolati, P Gugliotta J. Histochem. Cytochem. 31 1419 (1983) 254. K S Jakobsen, E Breivold, E Hornes Nucl. Acids Res.18 3669 (1990) 255. I R Rodriguez, G J Chader Nucl. Acids Res. 20 3528 (1992) 256. W H Scouten, P Konecny Anal. Biochem. 205 313 (1992) 257. T Goldkorn, D J Prockop Nucl. Acids Res. 14 9171 (1986) 258. M P Vitek, S G Kreissman, R H Gross Nucl. Acids Res. 9 1191 (1981) 259. T G Wood, J B Lingrel J. Biol. Chem. 252 457 (1977) 260. L R Solomon, L R Massom, H W Jarrett Anal. Biochem. 203 58 (1992) 261. P T Gilham, W E Robinson J. Am. Chem. Soc. 86 4985 (1964) 262. H Tsurui, A Suyama, A Wada Nucl. Acids Symp. Ser. 19 49 (1988) 263. A S Butorin, V V Vlasov, E M Ivanova, A S Rait, I G Shishkina, L V Yurchenko, L A Yakubov Biopolim. Klet. 5 71 (1989) 264. U Baumann, U Lehmann, K Schwellnus, J H van Boom, H Kuhn Eur. J. Biochem. 170 267 (1987) 265.T Schluep, C L Cooney Nucl. Acids Res. 26 4524 (1998) 266. G G Karpova, L O Kozlova, N P Pichko Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Khim. Nauk (7) 96 (1983) 267. E V Vyazovkina, E V Savchenko, S G Lokhov, J W Engels, E Wickstrom, A V Lebedev Nucl. Acids Res. 22 2404 (1994) 268. J T Kadonaga Methods Enzymol. 208 10 (1991) 269. S Eisenberg, S C Francrsconi, Ch Civalier, S S Walker Methods Enzymol. 182 521 (1990) 270. H Kawaguchi, A Assai, Y Ohtsuka, H Watanabe, T Wada, H Handa Nucl. Acids Res. 17 6229 (1989) 271. C DiRusso, R P Rogers, H W Jarrett J. Chromatogr. A 677 45 (1994) 272. C Pozidis, G Vlatakis, V Bouriotis J. Chromatogr. 630 151 (1993) 273. G Vlatakis, V Bouriotis Anal. Biochem. 195 352 (1991) 274. P Neddermann, J Jiricny J.Biol. Chem. 268 21218 (1993)532 275. P Neddermann, J Jiricny Proc. Natl. Acad. Sci. USA 91 1642 (1994) 276. G Schnapp, H P Rodi, W J Rettig, A Schnapp, K Damm Nucl. Acids Res. 26 3311 (1998) 277. J Meinkoth, G Wahl Anal. Biochem. 138 267 (1984) 278. H G Pereira BioEssays 4 110 (1986) 279. J A Matthews, L J Kricka Anal. Biochem. 169 1 (1988) 280. U Landegren, R Kaiser, C T Caskey, L Hood Science 242 229 (1988) 281. G S Sayler, A C Layton Annu. Rev. Microbiol. 44 625 (1990) 282. A G Badashkeeva, D G Knorre. Mol. Biol. 25 309 (1991) a 283. A H Al-Hakim, R Hull Biochem. J. 251 935 (1988) 284. J J Leary, D J Brigati, D C Ward Proc. Natl. Acad. Sci. USA 80 4045 (1983) 285. H Yokota, K Yokoo, Y Nagata Biochim. Biophys. Acta 868 45 (1986) 286.C R Petrie III, A D Adams,MStamm, J van Ness, SMWatanabe, R B Meyer Jr Bioconjugate Chem. 2 441 (1991) 287. A Oser,W K Roth, G Valet Nucl. Acids Res. 16 1181 (1988) 288. V A Shamanin, A G Pletnev, S G Rubin, V I Zlobin J. Gen. Virol. 71 1505 (1990) 289. K Yamakawa, H Oyamada, O Nakagomi Mol. Cell Probes 3 397 (1989) 290. A Oser, G Valet Nucl. Acids Res. 16 8178 (1988) 291. P Hurskainen, P Dahlen, J Ylikoski,M Kwiatkowski, H Siitari, T LoÈ vgren Nucl. Acids Res. 19 1057 (1991) 292. P S Thomas Proc. Natl. Acad. Sci. USA 77 5201 (1980) 293. D Y Kwoh, G R Davis, K M Whitfield, H L Chappelle, L J DiMichele, T R Gingeras Proc. Natl. Acad. Sci. USA 86 1173 (1989) 294. V Deubel, M Laille, J P Hugnot, E Chungue, J L Guesdon, M T Drouet, S Bassot, D Chevrier J.Virol. Methods 30 41 (1990) 295. M Shibata, H Takano, T Hironaka, K Hirai J. Virol. Methods 46 279 (1994) 296. P Frossard, K L Sisco, D L Rucknagel Anal. Biochem. 134 265 (1983) 297. T K Christopoulos, E P Diamandis, G Wilson Nucl. Acids Res. 19 6015 (1991) 298. G M Church,W Gilbert Proc. Natl. Acad. Sci. USA 81 1991 (1984) 299. S Beck Anal. Biochem. 164 514 (1987) 300. M Purrello, I Balazs Anal. Biochem. 128 393 (1983) 301. P J Nicholls, A D Malcolm J. Clin. Lab. Anal. 3 122 (1989) 302. M Ranki, A Palva,M Virtanen,M Laaksonen, H Soderlund Gene 21 77 (1983) 303. A C Syvanen Med. Biol. 64 313 (1986) 304. D V Morrissey,M Lombardo, J K Eldredge, K R Kearney, E P Groody,M L Collins Anal. Biochem. 181 345 (1989) 305.P Dahlen, A C Syvanen, P Hurskainen, M Kwiatkowski, C Sund, J Ylikoski, H Soderlund, T Lovgren Mol. Cell Probes 1 159 (1987) 306. G H Keller, D-P Huang,M M Manak Anal. Biochem. 177 27 (1989) 307. H Kohsaka, A Taniguchi, D D Richman, D A Carson Nucl. Acids Res. 21 3469 (1993) 308. M S Urdea, J A Running, T Horn, J Clyne, L Ku, B D Warner Gene 61 253 (1987) 309. D V Morrissey,M L Collins Mol. Cell Probes 3 189 (1989) 310. A Jungell-Nortamo, A C Syvanen, P Luoma, H Soderlund Mol. Cell Probes 2 281 (1988) 311. Y Tsuruta, T Yoshioka, R Suzuki, T Sakata J. Immunol. Methods 169 17 (1994) 312. T I Vener, M F Turchinsky, V P Zubov, T S Godovikova, T N Orlova, V A Shamanin, V F Kulikova Nucl. Acids Symp. Ser. 24 267 (1991) 313.T S Godovikova, T N Orlova, V F Zarytova, V A Shamanin, N V Vorobjova, N A Serdjukova, A G Romashchenko Nucl. Acids Symp. Ser. 24 284 (1991) 314. S Kawai, S Maekawajiri, A Yamane Anal. Biochem. 209 63 (1993) 315. M Shindo, A M Di Bisceglie, J Silver, T Limjoco, J H Hoofnagle, S M Feinstone J. Virol. Methods 48 65 (1994) 316. U Maskos, E M Southern Nucl. Acids Res. 21 2267 (1993) 317. W Bains, G Smith J. Theor. Biol. 135 303 (1988) 318. R Drmanac, I Labat, I Brukner, R Crkvenjakov Genomics 4 114 (1989) 319. C R Cantor, A Mirzabekov, E Southern Genomics 13 1378 (1992) I G Shishkina, A S Levina, V F Zarytova 320. D J Duggan,M Bittner, N Chen, P Meltzer, J M Trent Nat. Genet. 21 10 (1999) 321. R J Lipshutz, T Fodor, T R Gingeras, D J Lokhart Nat.Genet. 21 20 (1999) 322. D D L Bowtell Nat. Genet. 21 25 (1999) 323. A Vente,M Korn, G Zehetner, A Poustka, H Lehrach Nat. Genet. 21 22 (1999) 324. C M Niemeyer, D Blohm Angew. Chem. 38 2865 (1999) 325. T Pastinen, A Kurg, A Metspalu, L Peltonen, A-C Syvanen Genome Research 7 606 (1997) 326. A Metspalu, J M Shumaker, in Microsystem Technology. A Power Tool for Biomolecular Studies Vol. 10 (Eds J M Kohler, T Mejevaia, H P Saluz) (Basel; Boston; Berlin: Birkhauser Verlag, 1999) p. 371 327. Z Guo, Q Liu, L M Smith Nat. Biotechnol. 15 331 (1997) 328. G Ramsay Nat. Biotechnol. 16 40 (1998) 329. S Granjeaud, F Bertucci, B R Jordan BioEssays 21 781 (1999) 330. M Schena, D Shalon, R W Davis, P O Brown Science 270 467 (1995) 331.V G Cheung,M Morley, F Aguilar, A Massimi, R Kucherlapati, G Childs Nat. Genet. 21 15 (1999) 332. D J Hacia Nat. Genet. 21 42 (1999) 333. J S FruÈ chtel, G Jung Angew. Chem. 108 19 (1996) 334. R J Lipshutz, D Morris,M Chee, E Hubbell,M J Kozal, N Shah, N Shen, R Yang, S P A Fodor Biotechniques 19 442 (1995) 335. A C Pease, D Solas, E J Sullivan,M T Crovin, C P Holmes, S P A Fodor Proc. Natl. Acad. Sci. USA 91 5022 (1994) 336. G Lowe Chem. Soc. Rev. 24 309 (1995) 337. G H McGall, A D Barone, M Diggelmann, S P A Fodor, E Gentalen, N Ngo J. Am. Chem. Soc. 119 5081 (1997) 338. M C Pirrung, L F G MacGall J. Org. Chem. 63 241 (1998) 339. S P Fodor, J L Read,M C Pirrung, L Stryer, A T Lu, D Solas Science 251 767 (1991) 340. M Beier, J D Hoheisel Nucl. Acids Res. 28 e11 (2000) 341. E Gentalen,M Chee Nucl. Acids Res. 27 1485 (1999) 342. Yu-H Rogers, P Yiang-Baucom, Z-J Huang, V Bogdanov, S Anderson, M T Boyce-Jacino Anal. Biochem. 266 23 (1999) 343. T Strother, R J Hamers, L M Smith Nucl. Acids Res. 28 3535 (2000) 344. A Kumar, O Larsson, D Parodi, Z Liang Nucl. Acids. Res. 28 e71 (2000) 345. L C Shriver-Lake, in Immobilised Biomolecules in Analysis (Eds T Cass, F S Liger) (Oxford: Oxford University Press, 1998) p. 1 346. J J Cras, C A Rowe-Taitt, D A Nivens, F S Ligler Biosens. Bioelectron. 14 683 (1999) 347. A D Mirzabekov Trends Biotechnol. 12 27 (1994) 348. G Yershov, V Barsky, A Belgovsky, E Kirillov, E Kreindlin, I Ivanov, S Parinov, D Gushin, A Drobyshev, S Dubiley, A Mirzabekov Proc. Natl. Acad. Sci. USA 93 4913 (1996) 349. A L Drobyshev, A S Zasedatelev, G M Yershov, A D Mirzabekov Nucl. Acids Res. 27 4100 (1999) 350. E Ermantraut, K Wohlfart, S WoÈ lfl, T Schulz, M Kohler, in Microreaction Technology. Proceedings of the 1st International Conference on Microreaction (Ed. W Ehrfeld) (Berlin; Heidelberg: Springer, 1997) p. 332 351. L-J van Doorn, B Kleter, J Voermans, G Maertens, H Brouwer, R Heijtnik, W Quint J. Med. Virol. 42 22 (1994) 352. L T Mazzola, S P Fodor Biophys. J. 68 1653 (1995) 353. T M Herne,M J Tarlov J. Am. Chem. Soc. 119 8916 (1997) 354. R Levicky, T M Herne,M J Tarlov, S Satija J. Am. Chem. Soc. 120 9798 (1998) 355. A B Steel, T M Herne,M J Tarlov Anal. Chem. 70 4670 (1998) 356. S O Kelley, N M Jackson,M G Hill, J K Barton Angew. Chem. 111 991 (1999) 357. U Maskos, E M Southern Nucl. Acids Res. 21 4663 (1993) 358. J C Williams, S C Case-Green, K U Mir, E M Southern Nucl. Acids Res. 22 1365 (1994) 359. N C Fawcett, J A Evans, L C Chien, N Flowers Anal. Lett. 21 1099 (1988) 360. H Su, K M R Kallury,M Thomson, A Roach Anal. Chem. 66 769 (1994) 361. C Nicolini, V Erokhin, P Facci, S Guerzoni, A Ross, P Pashkevitsch Biosens. Bioelectron. 12 613 (1997)533 Affinity sorbents containing nucleic acids and their fragments 362. J Wang, P-E Nielsen,M Jiang, X Cai, J R Femandez,D H Grant, M Ozsoz, A Begleiter,M Mowat Anal. Chem. 69 5200 (1997) 363. E F A de Vries, R B M Schasfort, J van der Plas, J Greve Biosens. Bioelectron. 9 509 (1994) 364. A A Kruchinin, Y G Vlasov Sensors Actuators B 30 77 (1996) 365. M Yang,M E McGovern, M Thomson Anal. Chim. Acta 346 259 (1997) 366. S Sawata, E Kai, K Ikebukuro, T Iida, T Honda, I Karube Biosens. Bioelectron. 14 397 (1999) 367. S Kossek, C Padeste, L X Tiefenauer, H Siegenthalter Biosens. Bioelectron. 13 31 (1998) 368. K M Millan, A Saraullo, S R Mikelsen Anal. Chem. 66 2943 (1994) 369. J Wang, G Rivas, X Cai, E Palecek, P E Nielsen, H Shiraishi, N Dontha, D Luo, C Parrado, M Chicharro, P A M Farias, F Valera, D Grant,M Ozsoz, M Flair Anal. Chim. Acta 347 1 (1997) 370. G Marazza, I Chianella, M Mascini Biosens. Bioelectron. 14 43 (1999) 371. G B Sukhorukov, M M Montrel, A I Petrov, L I Shabarchina, B I Sukhorukov. Biosens. Bioelectron. 11 913 (1996) 372. M Sauer, A Brecht, K Charisse,M Maier,M Gerster, I Stemmler, G Gauglitz, E Bayer Anal. Chem. 71 2850 (1999) 373. C Xiao, X-E Zhang, Y-Q Chai, W-P Hu, Z-P Zhang, X-M Zhang, A F G Cass Biosens. Bioelectron. 13 451 (1998) 374. T M Jovin, A Kornberg J. Biol. Chem. 243 250 (1968) 375. V V Didenko Anal. Biochem., 213 75 (1993) 376. C Hansen, F Libert, G Vassart, D Christophe Anal. Biochem. 162 130 (1987) 377. I Raineri, C Moroni, H P Senn Nucl. Acids Res. 19 4010 (1991) 378. A Rolfs, I Weber Biotechniques 17 782 (1994) 379. A Rosenthal, D S C Jones Nucl. Acids Res. 18 3095 (1990) 380. S Stamm, J Brosius Nucl. Acids Res. 19 1350 (1991) 381. C R Thrash, R T Schimke J. Biol. Chem. 252 5615 (1977) 382. K Kuribayashi-Ohta, S Tamatsukuri, M Hikata, C Miyamoto, Y Furuichi Biochim. Biophys. Acta 1156 204 (1993) 383. E Hara, T Kato, S Nakada, S Sekiya, K Oda Nucl. Acids Res. 19 7097 (1991) 384. K Kuribayashi, M Hikata, O Hiraoka, C Miyamoto, Y Furuichi Nucl. Acids Symp. Ser. 19 61 (1988) 385. A A EÈ lov, E M Volkov, Z A Shabarova Bioorg. Khim. 17 789 (1991) b 386. H A Marble, R H Davis Biotechnol. Prog. 11 393 (1995) 387. R H Davis Curr. Opin. Biotechnol. 6 213 (1995) 388. L Ya Denisova, S N Zagrebel'nyi, N M Pustoshilova, A G Ven'yaminova, V V Gorn, V F Zarytova,M N Repkova, I G Shishkina Bioorg. Khim. 15 562 (1989) b 389. H W Jarrett J. Chromatogr. 618 315 (1993) 390. V Delpech Ann. Biol. Clin. 58 29 (2000) 391. M J O'Donnell-Maloney, C L Smith, C R Cantor Trends Biotechnol. 14 401 (1996) a�Mol. Biol. (Engl. Transl.) b�Russ. J. Bioorg. Chem. (Engl. Transl.) c�Kinet. Catal. (Engl. Transl.) d�Biochemistry (Engl. Transl.) e�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) f�Biophysics (En

ISSN:0036-021X    

出版商:RSC    

年代:2001

数据来源: RSC