1. INTRODUCTIONBy W. D. Ollis(Department of Chemistry, University of S h W )and J. H . Ridd(Department of Chemhtry, Univem&y College, London, W.C.1)SEVERAL changes in the coverage of this Report have been made this yearin the hope of meeting the demands of current trends in organic chemistry.Two new sections have been introduced and several of the traditional sectionshave been modified. The new section on Organometallic Compounds isdesigned to cover primarily the chemistry of organic derivatives of the Don-transition metals, with particular reference to those compounds containingmetal-carbon bonds, The behaviour of these non-transition-metal deriva-tives 8s organic compounds is emphasised and this section is intended to becomplementary to the discussion of the organic chemistry of transition-metalcomplexes which is retained in the Inorganic Chemistry Report.A further change is the introduction of a new section in Biosynthesis.This is clearly an important and rapidly developing area of natural productchemistry which is of considerable general interest.In the past separate sections have usually been allocated to steroids,terpenoids, and carbohydrates, but this year steroids and terpenoids havebeen included in the section on Alicyclic Compounds.The chemistry ofmonosaccharides of general interest is discussed in the Eeterocyclic sectionand polysaccharides are reviewed in the Biological Chemistry Report.Previously amino-acids, peptides, and nucleic acids have been reviewed attwo- or even three-yearly intervals, but clearly the breadth, importance, andgeneral interest of these areas of research demand frequent summary andappraisal.In this Report these sections deal only with results obtainedduring 1965. The inclusion of a section on Nucleic Acids in the OrganicChemistry Report is debatable. However, as nucleic acids were not to bediscussed specifically in the Biological Chemistry Report, an attempt hasbeen made t o review progress in this field in a language which is intelligibleto the non-specialist, an objective which is surely one of the major functionsof Annual Reports.It is customary in the Introduction to summarise important advancesthat have been made during the preceding year, but it would be invidiousto attempt to provide a detailed commentary this year.The most obviousimpression gained by reading the following Reports is that organic chemistryis in an extremely virile state: important advances have been made on allfronts extending from the physical to the biochemical borders of the subject,Probably the most significant developments in theoretical organicchemistry involve the use of simple and satisfying arguments based onthe symmetry of molecular orbitals to predict the steric consequences o212 ORGANIC CHEMISTRYelectrocyclic processes. The relative facility of thermal and photochemicalintermolecular cyclo-additions can now be understood and these views havebeen extended to include some rearrangement reactions. The ideas ofWoodford and Hoffman promise to provide a refreshingly new approach anda new nomenclature to a large area of chemistry.Many interesting structural details are emerging from nuclear magneticresonance studies of carbonium ions at low temperatures.The nature ofthe intermediate carbonium ions involved in certain solvolyses continuesto be examined mainly on the basis of kinetic studies and product analyses.Relative reaction rates in electrophilic aromatic substitution are also pre-senting a number of problems. There is increasing evidence that loosen;-complexes can be intermediates, but the extent to which the formation ofsuch complexes influences the overall rate of substitut'ion is not clear.Progress in synthetic organic chemistry is clearly demonstrated, not onlyby the continuing discovery of new reagents and methods, but also by thesuccessful synthesis of a remarkable range of organic compounds during thepast year.These achievements include many novel compounds such ascyclobutadiene (it really can be made) and derivatives of tetrahedrane, thecalicenes, Dewar benzene, prismane, benzvalene, brendane, brexane, theasteranes, and congressane, now to be called diamantane. Clearly exoticcompounds are entitled to exotic names. The claim (1965) to have syn-thesised a derivative of tetrahedrane has since been retracted (1966).The use of physical methods, including mass spectrometry and X-raycrystallography, continues to be widely applied, particularly in the examina-tion of natural products. This year the first isolation of a naturally occurringcumulene has been recorded and fucoxanthin is shown to be an alleniccarotenoid. A gap in the terpenoid class is filled by the discovery of thesesterterpenes, ophiobolin and gascardic acid, and the recognition of naturalrepresentatives, androcymbine and melanthioidine, of the 1 -phenylethyliso-quinoline class of alkaloids is important in relation to the biosynthesis ofcolchicine. The determination of the complete sequence of seventy-sevennucleotide units in alanine transfer-RNA represents progress a t a differentlevel of molecular complexity as does the total synthesis of insulin.Inconnexion with the synthesis of natural macromolecules, solid-phase syn-thesis provides exciting possibilities. Elegance in the total synthesis ofnatural products is exemplified by a number of successful syntheses of ter-penoids and alkaloids including colchicine, and Iboga, Hunteria, and Aspi-dosperm type alkaloids.The study of the details of the biosynthesis ofalkaloids by Barton and Battersby and the elucidation of the subtle stereo-chemical details of steroid biosynthesis by Cornforth and Popjak indicatethe extent to which detailed description of biosynthetic pathways is nowpossible.The volume of the current literature to be summarised in the OrganicChemistry Report has placed challenging and very time-consuming demandsupon the Reporters. Omissions are inevitable, but we hope that the reasonsfor these will be appreciated2. PHYSICAL METHODS OF STRUCTURE DETERMINATIONBy I).H. Williams (A, B, and 0 )(University Chemical Lahora&y, LensJEeH Road, Cambridge)and A. Horsfield (D)(Varian Research Laboratory, Molesey Road, Walton-on- T h a w )A. OPTICAL ROTATORY DISPERSION AND CIRCULAR DICHROISMTwo important books dealing with applications of optical rotatory dis-persion (0,r.d.) and circular dichroism (c.d.) in organic chemistry have beenpublished in 1965.1,2 Research advances have occurred in the field ofvariable-temperature studies 3, 4 (as low as - 192 "c) and measurementscan now be made fairly routinely down to 200 mp.Vmiable-tempemhe Studies.-Measurements a t very low temperaturesmay be made by use of EPA ether (5 vo1.)-isopentane (5 vo1.)-ethanol(2 vol.)as solvent? A temperature-variable c.d.curve may be afforded by 8 com-pound if it is conf'ormationally heterogeneous or if increased solvation occursat lower temperatures; the two phenomena can be distinguished by the useof various solvents and by the blue shift which accompanies increased sol-vation.5 Similarly, the presence of multiple Cotton effects arising from ann+n* transition of a carbonyl chromophore can be explained in terms of asolvation equilibrium involving differently solvated species and/or a confor-mational eq~ilibrium.~ Multiple Cotton effects cause complications in theinterpretation of a c.d. curve; it can be shown that whenever two Cottoneffects of similar amplitudes but opposite sign are superimposed with theirindividual maxima separated by 1-20 mp, a complex c.d.curve with twoextrema of opposite sign separated by ca. 30 mp will arise.5 The rotationalstrength of (+)-2-t-butylcyclohexanone ( I ) is temperature-dependent and,although a solvational equilibrium cannot be ruled out, the data have beeninterpreted in terms of a conformational equilibrium involving restrictedrotation in the t-butyl group. The l7b-acetyl side-chain of 20-oxo-steroidsexists predominantly in conformation (2) (view along the @-174-20 bond),although an additional contributor is indicated by temperature-dependentcircular dichroism.Chromophoric Derivatives.-If instruments are not available to facilitatemeasurements on amino-acids at low wavelengths, dimedone-amino-acidcondensation products ( 3 ) can be employed. * N-Neopentylidene deriva-tives (4) of a-amino-acid esters show anomalous dispersion curves with thefirst extremum in the 250-mp region; in the compounds studied, the deri-vatives with t'he L-codguration show a negative Cotton effect, and thosewith the D-configuration a positive 10 Other derivatives which mayprove useful include N-methylthionamides (5),11 azomethines in general,12N-chloroamines,l3 and thiolacetates.l4 N-Thiobenzoyl l5 and N-salicyli-dene l6 derivatives of amino-acids have also been investigated.(3) (4)SR - & N H M ~Amins-acids a d 1Peptides.-Low-wavelength measuremenfxs now permitthe absolute configurations of amino-acids to be obtained directly fromo.r.d.17-19 or c.d.curves.2o Amino-acids show a Cotton effect a t about216 mp; L-amino-acids show positive Cotton effects in acidic solution.17, 18In both neutral and acidic conditions, members of the L-series show asteeply descending negative '' tail " between 200 and 225 mp, and mem-bers of the D-series, a steeply ascending positive one.Additional htudieshave appeared on the 0.r.d. and c.d. of ~ e p t i d e s . ~ l - ~ ~ In an importantreview article,23 it has been reported that helical polypeptides show acomplex dichroism in the far-ultraviolet region, with three components(192, 204, 206 mp), characteristic of the helical array of peptide groups;these workers a3 have extended ad. measurements down to 185 mp.Amiraes and Alkaloids.-Saturated amines show steeply ascending ordescending 0.r.d. curves below 225 mp owing to a Cotton effect associatedwith the excitation of the non-bonding electrons on nitrogen;2* In amineisso far examined, possessing a single cc-asymmetric centre, the absolutecontiguration can be assigned from the direction of the dispersion curvebetween 200 and 225 mp, provided either that the amine is not protonated orthat the compound contains 8 chromophore absorbing above 200 mp.Benzyltetrahydroisoquinolines 24, 25 and bisbenzyltetrahydroisoquinolines 25show mult'iple Cotton effects which can be related to the stereochemistry ofthe compounds.Aporphine alkaloids exhibit a Cotton effect of high ampli-tude centred at 235-245 mp, the sign of which is diagnostic of the absoluteconfiguration of the asymmetric carbon atom adjacent to nitrogen.26Dienes and af3-Unsaturated Ketones : Non-planar Chromophores.-Using a very simplified model for the n+n* transition of or/?-unsaturatedketones, a relationship has been derived between the chirality of the non-planar chroniophores ( i e ., the helical sense of the four-carbon system) andthe sign of their c.d. Cisoid cyclohexenones give a positive (nega-tive) c.d. curve if the double bond lies in n positive (negative) octant.28The sign of the Cotton effect of a number of non-planar homoannular cisoiddienes can be predicted in every instance in terms of the chirality of theThe intensity of the absorption bands has been calculated as afunction of the skew angle of the diene system; where the skew angle isknown in one case (from X-ray data), the shape of the experimentallyobserved 0.r.d.curve is in excellent agreement with that calculated.30Mhcebeous.-The c.d. of 4/3-nitro-5a-cholestane and its G/?-analogueare enanti~meric;~~ models indicate that the nitro-groups are " fixed " bythe C-10 methyl groups and two axial hydrogens [see (6) and (7)].The absolute configuration of (+)-methyl n-butyl sulphoxide (8) is (8)and it has a positive Cotton effect;s2 absolute configuration of other methylalkyl sulphoxides can now be assigned by comparison of 0.r.d. curves withthat of (8). The cyclopropane and ethylene oxide rings of cyclopropylketones and epoxy-ketones make contributions to the Cotton effect at ca,290 mp, which are opposite in sign to those made by alkyl groups.33 In thefield of magneto-optical rotation spectra, zinc phthalocyanine and magnesiumphthalocyanine have the largest magnetic rotations observed to date.34a-Substituted acids have been extensively in~estigated.~~B. MASS SPECTROMETRYData, Analysis.-A computer programme has been written for theanalysis of high-resolution mass spectra.36 The computer print-out giveselemental compositions of all ions in the spectrum, which are listed accordingto their heteroatom content; the technique has been used to aid the structureelucidation of a new Pleiocarpa alkal~id.~' A programme has been devisedfor the construction of Tables of empirical formulze ordered according totheir masses.38 Data from the spectra of up to 150 pure compounds can beused to determine automatically both the qualitative a,nd quantitativecomposition of a sample mixture (if the spectra of all components are pre-sent in the spectral library) without requiring any advance information as tot'he sample compo~ition.~~Fragmentation Mechanisms.-It has been pointed out that a t lowenergies of bombarding electrons, some mass-spectral fragmentation pro-cesses show a striking similarity to photochemical decompositions.*0 Studiesof the McLafferty rearrangement [ (9) -+ (lo)] in some oxo-steroids indicatethat a critical intreratomic distance ( < l * S A) between the y-hydrogen atomand carbonyl oxygen atom is required before the rearrangement can- c H c H R'3.1V.E. Shashova, J. Amer. Chefit. Soc., 1965, 87, 4044.35 J. Cymerman Craig and S. I<. Roy, Tetrahedron, 1965, 21, 1847; A.Fredga,J. P. Jennings, W. Klyne, P. M. Scopes, B, Sjoberg, and S. Sjoberg, J. Chem. Soc.,1965, 3928; T. R. Emerson, D. F. Ewing, W. Klyne, D. G. Ncilson, D. A. V. Peters,L. H. Roach, and R. J. Swan, ibid., p. 4007.36 K. Biemann and W. McMurray, Tetrahedron Letters, 1965, 647.37 H. Achenbach and K. Biemann, Tetrahedron Letters, 1965, 3239.38 D. D. Tunnicliff, P. A. Wadsworth, and D. 0. Schissler, Analyt. Chem., 1965,3g D. D. Tunnicliff and P. A. Wadsworth, Analyt. Chem, 1965, 37, 1082.40 N. J. Turro, D. C. Neckers, P. A. Lsermakers, D. Seldner, and P. D'Angelo,*l C. Djerassi, G. von Mutzenbecher, J. Fajkos, D. H. Williams, and H. Budzikiewicz,37, 543.J . Amer. Chem. Soc., 1965, 87, 4097.J. Amer. Chem. Soc., 1965, 87, 817WILLIAMS AND HORSFIELD : STRUCTURE DETERMINATION 21'7A pair of epimeric steroidal tertiary alcohols [see partial structures (1 1)and (12)] eliminate water by distinct mechanisms upon electron impact ; (1 1)eliminates DHO and H,O to approximately equal extents, whereas (12), inwhich the ring hydrogens are more accessible to the OH-group, decomposesalmost exclusively by loss of H,O.42 Quaternary ammonium halides frag-ment in three main ways: (a) by a Hoffmann degradation, ( b ) by formation ofa tertiary amine and an alkyl halide, and (c) by formation of a tertiary aminewith an associated substitution by the halogen atom.43 In the retro-Diels-Alder fragmentation, stabilisation of the positive charge in the molecular ionis important in determining whether an ionised ene or diene fragment will bef0rrned.4~ Thus, (13) gives a spectrum in which the relative abundance ofM - 54 to m/e 54 is much greater than in the spectrum of (la), because theformation of a is more favourable when R = Me (tertiary carbonium ion).Bicyclopentane (59) reacts with dicyanoacetylene to give compound(60) .g3 Tetracyanoethylene oxide (61) undergoes some remarkable cyclo-additions.It reacts with ethylene a t 140" to give tetracyanotetrahydro-furan (62) in >80% yield. Under the same conditions, the benzofuran (63)is formed (35%) with benzene.94 The mechanism involves pre-equilibriumof (61) with an activated species.95 Linn and Benson note that (63) does8(59)NC CNG oNC CNnot react with tetracyanoethylene or maleic anhydride ; 94 probably, anelectron-rich dienophile is needed to achieve a Diels-Alder addition withinverse electron demand.Spontaneous dimerisations of ketens (with theexception of lieten itself) appear to give cyclobutane-l,3-diones as themajor primary products.96More than ten different papers dealing with 1,3-dipolar cycloadditionshave appeared from Huisgen's The reactions of A2-triazolineswith isocyanates, however, do not fall into this mechanistic category.981,Q-Dipolar additions appear to be relatively rare.99Sigmatropic Rearrangements.-In a further illuminating paper, R. B.Woodward and R. Hoffmann have defined an i,j-sigmatropic rearrange-ment as the '' migration of a o-bond, flanked by one or more n-electronsystems, to a new position whose termini are i - 1 andj - 1 atoms removedfrom the original bonded loci, in an uncatalysed intramolecular process."Thus, the Claisen rearrangement and its all-carbon analogue (the Coperearrangement) are 3,3-sigmatropic shifts.I n cases where a free choice oftransition state geometry is possible the Cope rearrangement [(64) -+ (65)]has been known to proceed more readily through the chair-like four-centreDsP. G. Gassman and I(. Mansfield, Chem. Comm., 1965, 391.94 W. J. Linn and R. E. Benson, J . Amer. Chem. Soc., 1965, 87, 3657.O6 D. G. Farnum, J. R. Johnson, R. E. Hess, T. B. Marshall, and B. Webster,J . Amer. Chem. SOC., 1965, 87, 5191.9 7 Part 12, G. Grashey, R. Huisgen, K. K. Sun, and R. M. Moriarty, J . Org.Chenz.,1965, 30, 74; Part 23, R. Huisgen, R. Knorr, L. Mobius, and G. Szeimies, Chem. Ber.,1965, 98, 4014; see also E. Grigat, R. Putter, and E. Miihlbauer, ibid., p. 3777.J. E. Baldwin, G. V. Kaiser, and J. A. Rornersberger, J . Amr. Chem. SOC.,1965, 87, 4114; R. Huisgen, R. Grashey, J. M. Vernon, and R. Kunz, Tetrahedron,1965, 21, 3311.laoR. B. Woodward and R. Hoffmann, J . Amer. Chem. SOC., 1965, 87, 2511.W. J. Linn, J . Amer. Chem. SOC., 1965, 87, 3665.9 9 R . Huisgen and K. Herbig, Annalen, 1965, 688, 98HOFFMANN : REACTION MECHANISMS 247tramition state (66) than through the boat-like six-centre alternative (67) .lolThis conformational preference, which is small, has been rationalised onthe basis of molecular-orbital symmetry relationships.lo2 It has also beenpredicted 102 that the 5,5-sigmatropic shift of a cis,cis-decatetraene willoccur most readily through a chair-like transition state (68).A chair-liketransition state seems also favoured in the Claisen rearrangement, althoughthe evidence here is less definitive.103 On heating the ketone (69) to 110"for 5 min., (70) is formed as an intermediate which is stable at room tem-perature; in the presence of dilute acid or base, however, aromatisation to0 2Me I! ' But (70)the thermodynamically more stable phenol (7 1) occurs instantaneously. Ifregarded as the penultimate step of a para Claisen rearrangement, then theconversion of (69) into (70) appears to be the first example of such arearrangement, in which the final 1,5-hydrogen shift occurs more slowlythan the preceding 3,3-sigmatropic shift .lo* Presumably, steric repulsionbetween the adjacent ally1 and t-butyl groups in (71) raises the barrier ofthe transition state leading to compound (71).Penta-l,4-dien-3-~1 vinyl ether (72) rearranges to the aldehyde (73)below room temperature in a remarkably ready reaction.105 3,S-Sigmatropicshifts which are facilitated by cyclopropane ring-opening are now welldocumented.Compound (74) is the first cis-1,2-divinylcyclopropane deriva-tive that has been isolated; it rearranges readily to bicyclo[3,2, lloctadienelol W. von E. Doering and W. R. Roth, Tetrahedron, 1962, 18, 67; Angew. Chem.,1963, 75, 27.loa R. Hoffmann and R. B. Woodward, J . Amer. Chem. SOC., 1966, 87, 4389; seealso K.Fukui and H. Fujimoto, Tetrahedron Letters, 1966, 261.lo3 E. N. Marvell, J. L. Stephenson, and J. Ong, J . Amer. Chem. SOC., 1965, 87,1267; for further studies of the Claisen rearrangement, see E. N. Marvell, B. J. Burreson,and T. Crandell, J . Org. Chem., 1965,30, 1030; E. N. Marvell, B. Richardson, R. Ander-son, J. L. Stephenson, and T. Crandell, ibid., p. 1032; H. L. Goering and W. I. Kimoto,J . Amr. Chem. Soc., 1965, 87, 1748.lo* B. Miller, J . Anaer. Chem. SOC., 1965, 8'4, 5515.lo6 S. F. Reed, jun., J . Org. Chem., 1965, 80, 1663248 ORGANIC CHEMISTRY(74a) as expected.lo6 A full paper has appeared on the related rearrange-ments of 2-alkenylcyclopropyl isocyanates ; lo7 similarly, double ScW-bases(75) rearrange to (76) on heating.108For 1 ,+sigmatropic shifts, two different transition-state geometries havebeen envisaged: in the suprufaciaE route " the migrating atom is associateda t all times with the same face of the n-system; in the second, anturafacialprocess the migrating atom is passed from the top face of one carbonterminus t o the bottom of the other." The reader is referred to the originalpaper 100 for a summary of symmetry-allowed 1,j-sigmatropic shifts and anabstract of literature results.The homovariant to 1 ,&hydrogen shifts, which is well established formedium-sized rings, has now been observed in acyclic compounds als0.~0~The most simple system is cis-hexa-1,4-diene (77), which on heating to 400"isomerises to (79); a rapid pre-equilibrium of (77) and (78) a t somewhatlower temperature is followed by irreversible formation of the product(79) .lo9 A similar hydrogen shift has been reported for allylacetophenone,which reacts in its enol form (80); this shift represents one step of analiphatic analogue to the '' abnormal Claisen rearrangement ." 110 Deriva-bives of cyclopropylacetic acid undergo decarboxylative ring-opening (81 ) ;e.g., above its melting point compound (82) is converted into (83) withintroduction of an angular methyl group.lllSince a direct 1,3-shift has been ruled out as a sequel to the nitrous aciddeamination of n-propylamine28 and is also unlikely to be involved in the6,2-migration of hydrogen in the norbornyl there is now, to theknowledge of the writer, no established example of a thermal urmtulysed1,3-sigmatropic shift within carbonium ions.l12 Therefore, a number oflo6 J.M. Brown, Chem. Comm., 1965, 226 [structure (111) in this paper should beidentical to (74a)l; see also P. K. Freeman and K. G. Kuper, Chem. and Ind., 1965,424.lo' E. Vogel, R. Erb, G. Lenz, and A. A. Bothner-By, Annulen, 1965, 682, 1.lo* H. A. Staab and F. Vogtle, Chm. Ber., 1965, 98,2681, 2691, 2701; H. A. StaabloS W. R. Roth and J. Konig, Annulen, 1965, 688, 28.R. M. Roberts, R. N. Greene, R. G. Landolt, and E. W. Heyer, J . Amer. Chem.ll1 T. Hanafusa, L. Birladeenu, and S. Winstein, J. Amer. C'hem. SOC., 1965, 87,112 For a catalysed 1,5-intramolecular hydride shift, see R. K. Hill and R. M. Carlson,and C. Wunsche, ibid., p.3479.Soc.. 1965, 87, 2282.3510; see also J. J. Sims, ibid., p. 3511.J. Amer. Chem. Xoc., 1965, 87, 2772CHALLIS : R E A C T I O N MECHANISMS 249reports (quoted in ref. 8) which have suggested this type of rearrangement,need further checking. Intramolecular 1,3-shifts within carbonium ionshave occasionally been proposed for the Jacobsen rearrangement 113 also.Part (ii). By B. C. Challis(Department of Chemistry, St. Salvator’s College, St. Andrews, Fife)Acidity Functions.-This topic was last reported in 1961.l Recentreviews cover investigations up until 1963 and therefore only developmentsof the last two years will be reported here.There can be no doubt that the acidity dependence of protonationequilibria for indicators of the same charge type also depends on theirspecific structure. This has led to a proliferation of acidity functions,particularly for acidic solvents, and those for amide (HA),3 indole (HI),*tertiary amine (H;”),5 and azulene (HAz) 6 indicators have been measuredand are different from H o and HR (GJ,,) values in the same solutions.Values of H , have been redetermined for perchloricusing only primary arrine indicators, and also for several other acidicsystem^.^ Considerable attention has been paid to reasons for the differentbehaviour of the various indicators: 3, 10 solvation of the conjugate acidappears to be a dominant factor and, generally speaking, the larger thenumber of acidic protons available for solvation, the smaller is the acidityfunction for a given acidity.Also, the effect of neutral salts on the H ,function can be closely predicted from their concomitant effect on aHz0.l1Other unknown factors must also be important, as is evident from reportsthat protonation of pyridine and pyridine l-oxide l2 both closely follow H ,and that the acidity-function values for secondary amides are larger thanthose for their tertiary c~unterparts.~a However, solvation effects appearto be least for the ionization of triphenylmethanol indicators 10, 13 and itwould therefore seem that, currently, the HR function (based on t,heseindicators) represents the best available approximation to the variation ofaH+ with stoicheiometric acidity. Similar arguments can be advanced113 References may be traced through a paper by E.N. Marvel1 and B. M. Graybill,J . Org. Chem., 1965, 30, 4014.J. H. Ridd, Ann. Reports, 1961, 58, 153.J. T. Edwards, Trans. Roy. SOC. Canada, 1964, 2, 313; E. M. Arnett, ProgPhys. Org. Chem., 1963, 1, 223.( a ) K. Yates and J. B. Stevens, Canad. J . Chem., 1965, 43, 529; ( b ) K. Yatesand J. C. Riordan, ibid., p. 2328; (c) K. Yates, J. B. Stevens, and A. R. Katritzky,ihid., 1964, 42, 1957; ( d ) R. B. Homer and R. B. Moodie, J . Chem. SOC., 1965, 4399.R. L. Hinman and J. Lang, J . Amer. Chem. SOC., 1964, 86, 3796.E. M. Arnett and G. W. Mach, J . Amer. Chenz. SOC., 1964, 86, 2671.F. A. Long and J. Schulze, J . Amer. Chem. SOC., 1964, 86, 337.&I. J. Jorgenson and D. R. Hartter, J . Amer. Chem. SOC., 1963, 85, 878.J. G. Dawber, J .Chem. SOC., 1965,4111; W. J. Zajac and R. B. Pu’owicki, J . Phys.Chem., 1965, 69, 2649; A. I. Gel’bshtein, R. P. Airapetova, G. G. Shchelova, and M. 1.Temkin, Zhur. neorg. Khim., 1964, 9, 1502; E. M. Arnett and C. F. Douty, J . Amer.Chem. SOC., 1964, 86, 409.lo E. M. Arnett and R. D. Bushick, J . Amer. Chem. SOC., 1964, 86, 1564.l1 C. Perrin, J . Amer. Chem. SOC., 1964, 86, 256.l2 C. D. Johnson, A. R. Ihtritzky, B. J. Ridgewell, N. Shakir, and A. M. White.Tetrahedron, 1965, 21, 1055.l3 E. D. Jensen and R. W. Taft, J . Amer. Client. SOC., 1964, 86, 116.and sulphuric acids’ K. Yates and H. Wai, J . Anaer. Chem. SOC., 1964, 86, 5409.250 ORGANIC CHEMISTRYfavouring the HR’ and HAZ functions, but data on these are a t presentincomplete.As a consequence of this strong dependence of equilibrium protonationon molecular structure, Kresge and his co-workers l4 have pointed out thatdifferent kinetic acidity dependence can be consistent with a single mechan-ism, if changes in reaction rate for different substrates are correlated witha single acidity function.Although Bunnett and Olsen l5 have shown howchanges in rate or equilibrium can be treated by means of linear free-energyrelationships, their treatment still requires an understanding of why somecompounds protonate in accordance with H , and others do not, before itcan be used as a mechanistic criterion. It therefore seems that acidity-function correlations must be interpreted with caution unt,il a clearer under-standing of protonation phenomena is available.Measurements of the H - function in several basic media have beenreported16 and in a t least one instance this function is dependent on theindicator, being different for amines 16b and substituted phenols.lGa Thisdifference, however, is less marked than that not’ed above for acidic solutions.As reported last year, several potential indicators, mainly polynitro-aromaticcompounds, complex with the lyate ion in basic solutions rather thanundergoing simple ionisation.17 The formation and structure of theseMeisenheimer complexes,ls and also the related Janovsky[lsb3 91 complexesformed with ketonic anions, have been studied further with an emphasison structural and mechanistic implications.The ionisation of 2,4-dinitro-aniline and the addition of methoxide ion to 2,4-dinitroanisole in methanoliesodium methoxide have been correlated with the H - and J- functions.16bSeveral reactions have been studied recently, involving nucleophilicdisplacement of a substituent by methoxide ion.20 These are thought tobe Bl reactions involving rapid formation of an intermediate (l), followedby slow elimination of the substituent.The reaction rates correlate withMe0 Ythe H - function; however, to a first approximation a t least, the concentra-tion of the intermediate should be proportional to the J , function. Ittherefore seems that the same problems confront kinetic acidity-functioncorrelations in basic media as those noted by Kresge and his co-workersl4 A. J. Kresge, R.A. More 0’ Ferrall, L. E. Hakka, and V. P. Vitullo, Chem. Cmm.,1965, 47.l5 6. F. Bunnett and F. P. Olsen, Chenz. Contm., 1965, 601.l6 ( a ) C. H. Rochester, J . Chem. Soc., 1965, 676; ( b ) C. H. Rochester, ibid., p. 2404;( c ) K. Bowden, Canad. J . Chem., 1965, 43, 2624; ( d ) R. Stewart and J. P. O’Donnell,ibid., 1964,42, 1681; ( e ) K. Bowden and R. Stewart, Tetrahedron, 1965,21,261; (f) V. I.Lazarev and Yu. V. Moiseev, Zhur. Jiz. Khirn., 1965, 39, 445.l7 B. Capon and C . W. Rees, Ann. Reports, 1964, 61, 277.l6 (a) M. R. Crampton and V. Gold, Chenz. Comm., 1965, 256; ( b ) R. Foster andIs R. J. Pollitt and B. C. Saunders, J. Chem. SOC., 1965, 4615.2o F. Terrier, Compt. Tend., 1965, 261, 1001; R. Schaal and J.-C. Latour, Bull.C . A. Fife, Tetrahedron, 1965, 21, 3363.SOC.chim. France, 1964, 2 177CHALLIS : REACTION MECHANISMS 251under acid conditions.14 Measurements of the strength of several extremelyweak organic acids, up to pK, = 34, have been possible with the availabilityof strongly basic media.16e, 21Deuterium-isotope Eff ects.-These are conveniently discussed under threesub-headings : primary effects in reactions where the isotopically substitutedbond undergoes fission ; secondary effects associated with isotopic substitu-tion in a bond not undergoing fission; and solvent effects arising fromdifferences between water and deuterium oxide as reaction media. Generalaspects of the subject have been discussed from the standpoint of usingstatistical-mechanical calculations to predict both kinetic and equilibriumisotope ratios,22 and Bourns 23 has reviewed kinetic isotope effects in electro-philic aromatic substitution.Primary isotope eflects.The interpretation of low k d k D ratios hasbeen discussed by Bell.24 Although evidence from the ionisation rates ofpseudo-acids indicates that the k d k D ratio varies with the symmetry ofthe transition 25 as suggested by Westheimer,26 other factors, in-cluding bending vibrations and non-equilibrium solvation of the transitionstate, can reduce the isotope effect. Proton tunnelling is thought 24 to beless important than suggested by earlier calculations, Nevertheless, interestcontinues in possible contributions from this source and barrier dimensionsfor several reactions thought to involve tunnelling have been compared.27The Arrhenius parameters for proton abstraction from both di-jsopropyllcetone28 and acetone 29a are quite normal and independent of isotopicsubstitution; thus the curved Arrhenius plots observed for these reac-tions 2 8 y 29 probably arise from steric or medium effects rather than fromnon- classical behaviour.The Arrhenius plot for acetone is apparently quitelinear a t higher temperature^.^^ Swain and his co-workers 31 have marshalledarguments against any primary isotope effect for proton transfers from oneoxygen (or nitrogen) to another, accompanying cleavage of a carbon bond,as in the cyclisation of chlorohydrins (see below). These arguments probablyaccount for the absence of an isotope effect in the thermal decarboxylationof malonic acid,38 hitherto expected on the basis of a cyclic intermediate.Secondary isotope eflects.As is evident from Halevi’s 33 excellent reviewof this subject, several factors influence these effects, depending on theposition of isotopic substitution and the kind of reaction. The n.m.r.21 E. C. Steiner and J. M. Gilbert, J. Amer. Chern. SOC., 1965,87,382; A. Straitwieser,J. I. Braumn, J. H. Hammons, and A. H. Pudjaatmaka, ibid., p. 384.22 M. J. Stern and M. Wolfsberg, J. Pharm. Sci., 1965, 54, 849; cf. M. Wolfsbergand M. J. Stern, Pure Appl. Chem., 1964, 8, 225.23A. N. Bourns, Trans. Roy. SOC. Canada, 1964, 2, 277.2 4 R. P. Bell, Discuss. Faraday SOC., 1965, 39, 16.25 R. P. Bell and J. E. Crooks, Proc.Roy. SOC., 1965, A, 286, 285.26 F. H. Westheimer, Chem. Rev., 1961, 61, 265; cf. J. Bigeliesen, Pure Appl,2 7 E. F. Caldin and M. Kasparian, Discuss. Farday SOC., 1965, 39, 25.28 J. R. Jones, Trans. Faraday SOC., 1965, 61, 2456.29 J. R. Hulett, J. Chern. SOC., 1965, ( a ) p. 1166; ( b ) p. 430.30 J. R. Jones, Trans. Faraday SOC., 1965, 61, 95.31 C. G. Swain, D. A. Kuhn, and R. L. Schowen, J . Amer. Chem. SOC., 1965, 87,82A. T. Blades and M. G. H. Wallbridge, J . Chem. Soc., 1965, 792.33 E. A. Halevi, Prog. Phys. Org. Chem., 1963, 1, 109.Chern., 1964, 8, 217.1553252 ORGANIC CHEMISTRYproton shifts induced by a-deuteration of butan-2-ones 34 and the effect ofdeuterium substitution on the stability of silver ion-olefin complexes 35sustain arguments for greater inductive electron donation by deuterium.However, the Taft 19F chemical shifts for m- and p-fluorotoluene indicatethat the inductive effect of the methyl group is not altered by deuterationand that the resonance effect decreases by only one-tenth of that predictedby chemical studies.36 There have been more reports 37 of appreciable ratereductions (about 12% per D atom) accompanying a-deuteration in limiting(SN1) solvolysis reactions.The exact cause of these reductions is still notclear,38 although further evidence of their diminution by neighbouring-group participation,39 and in direct displacement (&2) reactions,4* supportsthe working hypothesis advanced earlier by Streit~ieser.~l On the otherhand, Wu and Robertson’s 42 recent finding that the a-deuterium isotopeeffect in displacement reactions of trimethylsulphonium ion changes from“ normal ” (EH/kD > 1) to “ inverse ” with only modest increase in thenucleophilicity of the reagent suggests other factors may be involved.Rate retardations arising from /?-deuteration have frequently beenattributed to the larger hyperconjugative release of electrons by hydrogen.However, participation by trans-p-hydrogen in the form of a non-classicalunsymmetrically hydrogen- bridged ion rather than simple hyperconj ugationis suggested by Shiner a,nd Jewett 43a to explain the large but non-cumulativeeffects of 2- and 6-deuteration in the acetolysis of cis-4-t-butylcyclohexylp-bromobenzenesulphonate (2).The effects of B-deuteration on the solvo-OBslysis of the corresponding trans-isomer (3) 43b and the cyclohexyl toluene-p-sulphonates 44 have been interpreted as evidence for a skew-boat ratherthan a chair transition state.The absence of appreciable /?-deuteriumeffects in the SNl solvolysis of several cyclopropyl and cyclobutyl saltssuggests reaction via non-classical carbonium ions in which hyperconjugativeeffects are minimi~ed.~~s4 0. S. Tee and J. Warkentin, Canad. J. Chem., 1965, 43, 2424.35 R. J. Cvetanovid, F. J. Duncan, W. E. Falconer, and R. S. Irwin, J. Amr. Chetn.36 D. D. Traficante and G. E. Maciel, J. Amer. Chem. Soc., 1965, 87, 4917, 2508.s 7 V. Belanid-Lipovac, S. Bordid, and D. E. Sunko, Croat. Chem. Acta, 1965, 37,61; C. C. Lee and E. W. C.Wong, Canad. J. Chenz., 1965, 43, 2254; A. Streitwieser andH. S. Klein, J. Arner. Chem. SOC., 1964, 86, 5170.38P. Geneste and G. Lamaty, Tetrahedron Letters, 1965, 4633.39 C. C. Lee and E. W. C. Wong, Tetrahedron, 1965, 21, 539; J. P. Schaefer and40 B. btman, J. Am;; Chem. Soc., 1965, 87, 3163.Solvolytic Displacement Reactions,” McGraw-Hill, New4 2 C. Y. Wu and R. E. Robertson, Chem. and lnd., 1964, 1803.43V. J. Shiner and J. G. Jewett, J. Amer. Chem. SOC., 1965, 87, ( a ) p. 1382; (b)44 W. H. Srtunders and K. T. Findley, J . Amer. Chem. SOC., 1965, 87, 1384.45 M. Nikoletid, S. BorEid, and D. E. Sunko, Proc. Nut. Acad. Sci. U.S.A., 1964,Soc., 1965, 87, 1827.D. S. Weinberg, Tetrahedron Letters, 1965, 2491.York, 1962, p. 172.Cf. A. Streitwieser,p. 1383.52, 893CHALLIS : REACTION MECHANISMS 253Further temperature-independent p-deuterium-isotope effects have beenreported.46 In one case 46a these are viewed as additipid evidence for theimportance of non- bonded interactions arising frQm different potentialbarriers for the rotation of CH, and CD, greaps.However, the verydifferent behaviour of isopropyl47 and t-butyl salts 48 is difficult to accountfor in this way.48 An alternative explanation, proffered recently by Wolfs-berg and is that temperature independence arises from a fortuitouscancellation of zero-point energy differences. Whether secondary isotopeeffects are temperature-dependent or not may be a criterion for themechanism of solvolyses.48Secondary isotope effects have proved particularly useful in elucidatingthe mechanism of N-nitration 50 and of the Diels-Alder reaction.51Solvent isotope e#ects.The importance of initial rather than transition-state solvation in determining solvent isotope ratios in solvolyses continuesto be debated, and Laughton and Robertson52 have replied to recentcriticism of their arguments for this effect. Other evidence53 also showsthat a t least part of the activation energy in tertiary halide solvolysis isrequired to break down the initial solvent shell. On the other hand, Swainand his co-workers 31 have concluded that differences in transifion-statesolvation can fully account for the kinetic solvent isotope effects in a numberof cyclisations. A simple method of calculating these effects is given andsome mechanistic implications are discussed.Despite some evidence suggesting that in protein synthesis m-RNA is readfrom the 3’-hydroxyl to the 5’hydroxyl end,65 very clear-cut experimentsnow indicate that it is read from the 5’-end.6s, 67 For example,66 poly-nucleotides of the type ApApAp . .. pApApC with about twenty nucleotideresidues, directed the synthesis of a group of peptides of varying chain-lengthcontaining lysine (the codon for which is AAA) and one asparagine (codon,AAC), which was carboxyl-terminal, Le., lys-(lys),-aspN. Hence, if thepolypeptide is assembled from the amino- to the carboxyl-terminal end, thepolynucleotide message must be read from the 5’- to the 3‘-hydroxyl end.A study of the direction of chain growth in enzymatic RNA synthesis fromthe nucleoside 5‘-triphosphate with RNA polymerase has shown unequi-vocally that it also proceeds from tthe 5’- to the 3‘-end.@It has been known for some time that streptomycin interferes withprotein synthesis in some way.69 It has now been found 70 that this anti-biotic has an effect on the m-RNA-ribosome-aminoacyl-s-RNA complex.The coding ambiguity observed in the presence of streptomycin with poly U(which then directs the incorporation of ileu in addition to the normal phe)is explained by the finding that the binding of phe-s-RNA to the complex isinhibited, whereas the binding of ileu-s-RNA is stimulated by streptomycin.Work has continued on natural m-R,NA itself.The thread of RNA whichconnects the ribosomes in reticulocyte polysomes has been isolated.It hasa molecular-weight of about 150,000, which is the size predicted for messen-gers for each hsmoglobin chain if the messengers are monoci~tronic,~~ i.e.,if each in-RNA makes one protein. The isolation of an m-RNA of about36 nucleotide residues for Gramicidin S synthesis has been reported.72 Thisantibiotic contains 10 amino-acids. Determination of the m-RNA sequenceshould indicate whether initiation and termination codons exist. Work onpure species of RNA should be aided by the discovery that superior frac-tionation is achieved in a sucrose gradient in a zonal ultracentrif~ge.~~ The invitro synthesis of infectious bacteriophage (bacterial virus) RNA, using isolatednatural m-RNA and the appropriate enzymes, provides a clear demonstrationof the correctness of current ideas regarding nucleic acid metabolism.7*Secondary Structure.-Of crucial importance to an understanding of themechanism of metabolic reactions involving nucleic acids is knowledge oftheir secondary structure. All the evidence regarding the secondary struc-ture of s-RNA mentioned in a previous report 75 has proved to be wrong.Thus, the ‘‘ crystalline s-RNA ” whose structure was investigated by X-raycrystallography has now been found to have consisted of degraded P R N A , ~ ~and the idea that the minor constituents of s-RNA (methylated bases,pseudouridine, etc.) are concentrated in a centra.1 region is not supported bythe sequence studies mentioned above.48, 54 s-RNA and r-RNA are knownto consist of single polynucleotide chains.Most of the evidence previouslyadvanced in favour of the existence of double-helical regions in these rnole-cules (by the molecule folding back onto itself) has since been shown to beambiguous (see below), and a t present the secondary structure of thesesubstances is quite uncertain. Holley’s structure far alanine s-RNA does notpermit extensive base-pairing t o form double- helical regions.Closely related to the question of the nature of the secondary structureitself is a consideration of the stability of helical forms. Ideas concerningthe factors stabilising helical polynucleotides have undergone a radicalchange in recent years. It was originally supposed that the specific hydrogenbonds between the heterocyclic bases was responsible for the stability of theDNA double helix, and this was apparently confirmed by the fact that, thehigher the GC content, the more stable the DNA (sincs the GC pair can formthree hydrogen bonds, and the AT pair only two).Considerable evidenceindicates, however, that hydrogen-bonding cannot account for the stabilityof helical DNA. For example, classical hydrogen- bond breaking agentssuch as urea are relatively ineffective in denaturing DNA. [Denaturation ofdouble-helical DNA involves two processes: ( 1 ) separation of the strands,and (2) helix+coil transition in the separated strands]. Much more effective71A. Burmy and G. Marbaix, Biochem. Biophys. Acta, 1965, 103, 409.74 J.B. Hall, J. W. Sedat, P. R. Adiga, I. Uemara, and T. Winnick, J. Mol. BioZ.,73 J. R. B. Hastings, J. H. Parish, K. S. Kirby, and E. S. Klucis, Nature, 1965,7 4 S . Spiegelman, I. Haruna, I. B. Holland, G. Beaudreau, and D. Mills, Proc.75 T. L. V. Ulbricht, Ann. Reporh, 1963, 59, 378.7 6 M. Spencer and F. Poole, J. Mot. Biol.. 1965, 11, 314.1965, 12, 162; 5. W. Sedat and J. B. Hall, ibid., p. 174.208, 645.Nut. Acad. Sci., U.S.A., 1965, 54, 91941 2 ORGANIC CHEMISTRYare certain anions, such as C104-, which alter the structure of water and whichare regarded as hydrophobic bond breaking agents.77 (The concept ‘‘ hydro-phobic bond,” which originated in protein chemistry, means in this instancethat the attraction of the bases for each other is greater than their attractionfor water molecules).That the main factor stabilising helical polynucleo-tides is interaction between the bases stacked parallel on top of each otheris also supported by recent theoretical studies.78The mast important recent evidence regarding secondary structure andhelix stability has come from a study of oligo- and poly-nucleotides usingoptical rotatory dispersion (0.r.d.). In the course of the last year this tech-nique has become much more widely used in this field, and significantadvances have been made in interpretation. On comparing a mononu-cleotide with its oligo- and poly-nucleotide homologues, the biggest changein the 0.r.d. is found on going from the monomer to the dimer.T9, 8 0 Tinocoand co-workers 80 calculated that the interaction between the two bases indiadenylic acid leads to a splitting of the adenine 260 mp absorption band.The two bands thus produced must have Cotton effects (C.E.s) of oppositesign and of approximately equal magnitude.The troughs of the twoC.E.s overlap, leading to the observed curve, which consists of twopeaks and a trough, the latter being near the U.V. absorption maximum.What is striking is the close resemblance of the overall shape of the 0.r.d.curves of simple dinucleotides to the curves given by DNA and RNA,81 andhomologous polynucleotides,~2-84 which all give two peaks and a trough inthe 230-290 mp region, with the exception of poly I, which gives two troughsand one peak.84 An examination of a wide variety of di- and tri-nucleotides 85has indicated that the shape of the 0.r.d.curves depends not only on basecomposition but also on base sequence.The effect of chain-length on secondary structure and helix stability hasbeen studied by comparing the physical properties of a series of oligoadenylicacids. The results using o.r.d.86 and circular dichroism 87 complement eachother. They confirm previous work which indicated that poly A has a single-stranded helical structure at neutral pH 81, 82 and a double-stranded structureat acid pH (<5). The mean residue amplitude of the first C.E. increasesgradually with chain-length at neutral pH, levelling off and reaching amaximum at about 30 residues, whereas in acid there is a transition at about77 K. Hamaguchi and E. P.Geidushek, J. Arner. Chem. SOC., 1962, 84, 1329.78H. DeVoe and I. Tinoco, jun., J. MoZ. Biol., 1962, 4, 500; 0. Sinanoglu andS. Abdulnur, Photochem. Photobwl., 1964, 3, 333; D. M. Crothers and B. H. Zimm,J. Mol. Biol., 1964, 9, 1.7 9 T. R. Emerson, R. J. Swan, A. M. Michelson, and T. L. V. Ulbricht, AbstractsFed. Europ. Biochem. SOC., 1965, e 59.M. W. Warshaw, C. A. Bush, and I. Tinoco, jm., Biochem. Biophys. Res. Comm.,1965, 18, 633.81 T. Samejima and J. T. Yang, J . BioZ. Chem., 1965, 240, 2094.8 2 G. D. Fasman, C. Lindblow, and L. Grossman, Biochemistry, 1964, 3, 1015.83P. K. Sarkar and J. T. Yang, J. Biol. Chem., 1965, 240, 2099.K. Sarkar and J. T. Yang, Biochemistry, 1965, 4, 1238.s 5 M. W. Warshaw and I. Tinoco, jun., J.MoZ. Biol., 1965, 13, 54; C. R. Cantor86 A. M. Michelson, T. L. V. Ulbricht, T. R. Emerson, and R. J. Swan, Nature,87 J. Brahms, A. M. Michelson, and K. E. Van Holde, J. MoZ. Biol., 1966, 15, 467.and I. Tinoco, jun., ibid., 1965, 13, 65.1966, 209, 873ULBRICHT: NUCLEIC ACIDS 4137 residues (pH 4.5) or 12 residues (pH 4.86) to the double-stranded structure,with a further marked increase in rotation.88, 87 Optical rotatory dispersionstudies have also confirmed that poly C exists as a single-stranded helix atneutralPoly G has been synthesised enzymatically a t l a ~ t . 8 ~ It forms a 1 : 1complex with poly C which is very stable;g0 the extent of the interactionbetween the two polynucleotide strands a t different pH values is shown veryclearly by the changes in ~ .r . d . ~ l The 0.r.d. of yeast s-RNA indicates that ithas a highly ordered secondary structure which is probably Addi-tion of ethylene glycol to the solution causes a loss of secondary structure.Since ethylene glycol stabilises helices held together by hydrogen- bondingbut decreases hydrophobic interactions, the latter are responsible for thesecondary structure of S-RNA.~~ This is in agreement with the structure ofalanine s-RNA and is in marked contrast to models of s-RNA with over80 o/o of ba~e-pairing.~~If the stability of helical polynucleotides is mainly due to base interaction,why is helix stability related to GC content? It has been shown that, of thefour RNA bases, U interacts least strongly with its neighbour~.~~ Conse-quently, the DNA with the lowest T content (since T is very similar to V),that is, the highest GC content, should be the most stable, as found.More-over, this should be measurable by the amplitude of the first Cotton effect(the size of which seems to be closely related to the extent of base-interac-tion) and, indeed, there is a linear correlation between the molecular rotationat the 290 mp. peak of DNA and GC content.a1The denaturation of single-stranded helical polynucleatides (such aspolp A, poly C ) is accompanied by a hyperchromic effect and a reduction inrotation, just as is the denaturation af double helical DNA. Similar changesaccompanying the denaturation of RNA used to be interpreted as being dueto the separation of double-helical regions of the molecule, but it is clear thatthese changes indicate a helix-+coil transition but do not in themselvesindicate whether the helix is single- or double-stranded.This explains thepresent uncertainty regarding the secondary structure of s-RNA.A thorough examination of the 0.r.d. of a large number of nucleosideshas made it possible to correlate the sign and magnitude of the C.E. with theconformation of nucleosides in solution. 94 These studies indicate that thepreferred conformation is the same as that found in DNA and in crystallinenucleosides. 9588 V. Luzzati, A. Mathis, F. Masson, and J. Witz, J. 1clol. Biol., 1964, 10, 28;J. Witz and V. Luzzati, ibid., 1965, 11, 620; D. N. Holcomb and I. Tinoco, &n., Bio-polymers, 1965, 3, 121; K.E. van Holde, J. Brahms, and A. M. Michelson, J. Mol.Biol., 1965, 12, 726.s9 M. N. Thang, $1. Graffe, and 3%. Grunberg-Manago, Biochirn. Biopl~ys. Acta,1965, 108, 125.F. %chon and A. M. Michelson, Proc. iVat. Acad. Sci., U.S.A., 1965, 53, 1425.91 T. L. V. Ulbricht, R. J. Swan, and A. M. Michelson, Chem. Comni., 1966, 63.92 G. D. Fasman, C. Lindblow, and E. Seaman, J. Mol. Biol., 1965, 12, 630.O3 G. Felsenfeld and G. Sandeen, J. Mol. Biol., 1962, 5, 587; G. Falsenfeld and G. L.Cantor, Proc. Nut. Acad. Xci., U.S.A., 1964,51,518; S. W. Englander and J. J. Englander,ibid., 1965, 370.BP T. L. V. Ulbricht, T. R. Emerson, and R. J. Swan, Biochena. Biophys. Res. Conam.,1965, 19, 643.95 M. Sundaralingam and L. H. Jensen, J. Mol.Biol., 1966, 13, 914, 93013. BIOSYNTHESISBy G). W. Kirby(Department of Chemistry, Imperial College, London, S. W.7)IN recent years organic chemists have shown increasing interest in the experi-mental investigation of natural product biosynthesis. Particular attentionhas been paid to the complex ‘‘ secondary metabolites ” produced by plantsand, to a lesser extent, by animals. Many earlier biogenetic theories havenow been verified and extended by in, vivo experiments with labelledprecursors. For reasons of space the present report must be highly selective.Emphasis is placed on experimental work with compounds of special interestto the organic chemist. Topics have been chosen to illustrate, as far as pos-sible, the scope of current investigations.Aromatic Alkaloids.-.IndoZe alkaloids.It is well established that theindolylethane unit of alkaloids such as vindoline (1) is derived biologicallyfrom trypt~phan.~ At the beginning of 1965 the origin of the remainingb(4)A R. ..II (3) **IJ. D. Bu’Lock, “ The Biosynthesis of ?tural Products,” McGraw-Hill, London,1965; J. H. Richards and J. B. Hendrickson, The Biosynthesis of Steroids, Terpenes,and Acetogenins,” Benjamin, New York, 1964; “ Biogenesis of Natural Compounds,”ed. P. Bernfeld, Pergamon Press, Oxford, 1963.R. Robinson, “ The Structural Relations of Natural Products,” Clarendon Press,Oxford, 1955.E. Leete, A. Ahmad, and I. Kompis, J . Amer. Chem. SOC., 1965, 87, 4168, andreferences citedKIRBY: BIOSYNTHESIS 415Cl0 (or C,) fragment of the indole alkaloids was still obscure.Leete and hiscolleagues had earlier presented evidence that this fragment was constructedfrom acetic and malonic acids together with a C,-unit derived from formate.However, more recent work from three other laboratories cast doubt on thishypothesis which has now been withdrawal During the last few monthsconvincing proof of the terpenoid origin of the C,, fragment has appeared.Scott and his colleagues 5 showed that [2-14C]mevalonic lactone (2) servedas a precursor for vindoline (1) in Catharanthw roseus (Vinca rosea). De-gradation established that 22.0% of the total activity was at C-22. Inde-pendent studies by Goeggel and Arigoni,6 with the same plant, gave essen-tially the same result. Clearly mevalonate, unlike acetate and formate: wasincorporated in a specific manner into the alkaloid, approximately one quar-ter of the activity appearing in one carbon atom.These results are consistentwith a hypothesis developed independently by Thomas and Wenkert,'i.e., that mevalonate is converted into a monoterpenoid derivative (3),which can give rise to the typical indole alkaloid fragments (4), ( 5 ) , and (6).To explain the labelling pattern in vindoline the methyl carbons of theisopropylidene groups in (3) must become equivalent a t some stage in thebiosynthesis and thus each acquire, effectively, one quarter of the totalactivity. Excellent precedence for this is available from the work by Yeowelland Schmid 8 on the monoterpenoid plumiericin.The unrearranged fragment (4) is present in many indole alkaloids.Goeggel and Arigoni showed 6 that reserpinine (7), a representative of thisclass, was derived from [ 2-WJmevalonate.Again, approximately onequarter (26%) of the activity was present in one carbon atom [as (7)].Recent independent studies by Battersby and his colleagues have confirmedand extended these findings. Their experiments with Rhuzia strictct provedespecially fruitful. [ 2-14C]Mevalonate was incorporated into dehydroaspi-dospermidine (8) and degradation located two-thirds (65 & 6%) of theactivity at C-8. Since a labelled carbon atom (originally attached to (2-3) islost during the biosynthesis of this alkaloid, their result is in completeagreement with theory.Further support came from the degradation ofdehydroaspidospermidine derived from [ 3-14C)mevalonate : approximatelyhalf the alkaloid's activitywas found at C-20. This group further showed thatcatharanthine (9), which contains the hypothetical fragment (6), is alsoderived from [2-14C]mevalonate in the predicted manner.BenzyZisoquinoZine group. Full details have appeared 1 0 of work by two*A. R. Battersby, R. Binks, W. Laurie, G. V. P q , and B. R. Webster, Proc.Chem. Soc., 1963, 369; H. Goeggel and D. Arigoni, Experientia, 1965, 21, 369; D. H. R.Barton, G. W. Kirby, R. H. Prager, and E. M. Wilson, J . Chem. Soc., 1965, 3990.T. Money, I. G. Wright, F. McCapra, and A. I. Scott, Proc. Nat. Acad. Sci.U.S.A., 1965, 53, 901; F. McCapra, T. Money, A.I. Scott, and I. G. Wright, Chem.Comm., 1965, 537.eH. Goeggel and D. Arigoni, Chem. C m m . , 1965, 538.R. Thomas, Tetrahedron Letters, 1961, 544; E. Wenkert, J . Amer. Chem. SOC.,1962, 84, 98.8D. A. Yeowell and H. Schmid, Experientia, 1964, 20, 250.* A . R. Battersby, R. T. Brown, R. S. Kapil, A. 0. Plunkett, and J. B. Taylor,Chem. Comm., 1966, 46.lo D. H. R. Barton, G. W. Kirby, W. Steglich, G. M. Thomas, A. R. Battersby,T. A. Dobson, and H. Ramuz, J . Chem. SOC., 1965, 2423416 ORGANIC CHEMISTRYgroups on the biosynthesis of the morphine alkaloids in Papaver somniferum.Feeding experiments with multiply-labelled precursors, supported by appro-priate degradations, established that reticuline (lo), salutaridine ( I 1 ), andone of the epimeric dienols (12), were incorporated intact into thebaine (13).All three methyl groups of reticuline were retained in the final alkaloid.Radio-dilution studies showed that salutaridine was a transient intermediatein the conversion of norlaudanosoline (trisnor-reticuline) or tyrosine into themorphine alkaloids.For stereochemical reasons ( - )-reticuline (10) mustbe the immediate precursor of the dienone (11). However, Battersby andhis colleages 11 have shown that both ( -)- and (+ )-reticuline are efficientprecursors of morphine. Incorporation of the ( +)-isomer, however, involvedconsiderable (though not complete) loss of tritium from material labelled asshown [asterisk in (lo)]. This is understandable if reticuline is racemised inthe plant vicc its 192-dehydro-derivative. A separate experiment confirmedthat lY2-dehydroreticuline was indeed efficiently converted into morphine.Moreover, partially racemic reticuline has now been isolated from opium.12OMeA preliminary account of work on the biosynthesis of sinomenine (14) inXinomenium acutum has appeared.13 The enantiomer of salutaridine (1 1 ),sinoacutine,l* labelled with tritium at position 1, was incorporated intosinomenine without “ scrambling ” of the label.( j-)-Reticuline was alsoincorporated, though much less efficiently.but not (-)-retidine, was a good precursor for berberine in Hydrastiscanadensis and for protopine in Bicentra spectabibis and Argzmone species.15Clearly, at least one step in the biosynthesis of these optically inactive mole-cules from reticuline must be stereospecific.Elegant experiments l7 withmultiply-labelled, optically active, specimens of reticuline have elucidatedthe main steps in the construction of protopine, stylopine (18), and cheli-donine (19) in Chebidoniurn majus. Again it was shown that (+)-reticulinewas the true precursor of all three alkaloids. Narcotine (20) can also beadded to the now extensive list of alkaloids derived from this versatileprecursor.l* Conversion of the simpler benzylisoquinoline ( + )-coclaurinelabelled with tritium as shown (21), into the correspondingly labelled dienone,crotonosine (22), has been observed l9 in Croton linearis. (-))-Coclaurineand isococlaurine [methoxyl and hydroxyl groups in ring A of (21) inter-changed] were not incorporated into crotonosine.Monkovi6 and Spenser 20have recently reported the specific incorporation of ( &)-noradrenaline intoberberastine (5-hydroxyberberine) in Hydrastis canadensis.An auspicious start to the study of aporphine biospthesis has been madeby Battersby and his colleagues.21 ( +)-Orientdine (23) was incorporatedinto isothebaine (24) in Papaver orientale. Again, experiments with doubly-labelled [see (23) and (24)] and optically active precursors established beyonddoubt that conversion had taken place in a specific manner. The overalltransformation, which proceeds via the dienone (25) and a correspondingdienol, has also been achieved chemically.Colchicine and phenethylisoquinoline alkaloids.It is well established thatthe entire carbon skeleton of colchicine (26) is derived from phenylalanine(C,-C, unit) and tyrosine (expanded C,-C, unit).22, 23 Leete has now re-ported 23 the specific incorporation of ( &)-[4-1PC]tyrosine (27) into colchi-cine in Colcchicum byxantinum, thus confirming the origin of the tropolonering. Battersby and Leete's results suggest therefore a biosynthetic inter-mediate of the type (28). The discovery 24 of a new alkaloid, androcymbineMe0Me0, <NM~Me0 "'"or>. , HMe0(29), from a botanically related plant, Androcyrnbium melanthiodes, providesstriking support for this idea. The same plant also contains melanthioidine,a bis-l-phenethylisoquinoline alkaloid :25 it appears then that colchicine maybe the first known member of an interesting new group of phenethyliso-quinoline derivatives.Pyridine and quinoline alkaloids.The biosynthesis of nicotine (30) hasbeen more intensively studied than that of any other alkaloid,26 but new22 A. R. Battersby, R. Binks, J. J. Reynolds, md D. A. Yeowell, J . Chem. SOC.,1964, 4257; R. D. Hill and A. M . Unrau, Canad. J . Chem., 1965, 43, 709.23 E. Leete, Tetrahedron Letters, 1963, 333.2 4 A. R. Battersby, R. B. Herbert, L. Pijewska, and F. gantavy, Chew. Cmnz.,25 A. R. Battersby, R. B. Herbert, and F. Santavy, Chem. Comm., 1965, 415.26 E. Leete, Science, 1965,147, 1000; E. Ramstad and S. Agurell, Ann. Rev. Plant.1965, 228.Physiol., 1964, 15, 143KIEBY: BIOSYNTHESIS 419results are still forthcoming.Experiments with whole tobacco plants(Nicotiana tizbacum L.) have demonstrated that quinolinic acid, like nicotinicacid, is an efficient precursor for the alkal~id.~' Thus [2,3,7,8-14C]quinolinicacid (31) was converted into nicotine which contained all its activity in thepy-ridine ring. A parallel experiment with [2,3,7-14C]nicotinic acid gaveradioactive nicotine, which was degraded t o establish the labelling pattern(30) expected from earlier work. Recent experiments by Rapoport and hiscolleagues28 concern the origin of the pyrrolidine ring. Nicotine, isolatedfrom Nicotiancr glutinosa plants which had been exposed to [14C]carbondioxide, was degraded to establish the activity at each carbon atom in thepyrrolidine ring. The activities at C-2', C-3', and C-5' were similar, but t'hatat C-4' was approximately 4 times a,s large.This result is inconsistent withI (32) ( 3 3 )the obligatory participation of a symmetrical intermediate [as (32)], derivedfrom acetate 29 or ornithine (33). The same group confirmed that ornit'hinewas incorporated in the usual way in N . gbutinosa, but concluded that theincorporation must involve a minor or '' aberrant " pathway.Preliminary studies 30 on the biosynthesis of a group of quinoline deriva-tives, (34; R = alkyl), from a Pseudomonas species, have been reported.[carb~xyl-~*C]Anthranilic acid was incorporated intact, and acetate appearedto be an efficient precursor for the fatty side chain, R. The alkaloid peganine(vasicine) (35) is also derived, in Adhctboda vmica, from [c~rboxyZ-~~CIanthra-nilic a~id.~1 Degradation of the labelled peganine gave a,nthranilic acid ofthe same molar activity.The biological conversion of shikimic and anthra-nilic acids into damascenine (36) has also been reported.32Lsoprenoids.-Valuable comprehensive reviews of steroid and terpenoidbiosynthesis have appeared.33 The stereochemistry of the reactions leadingfrom mevalonate t o squalene and other higher terpenoids is now understood27 K. S. Ymg, R. K. Gholson, and G. R. Wder, J . Amer. Chem. Soc., 1965, 87,28A. A. Liebmann, F. Morsingh, and H. Rapoport, J. Amer. Chem. Soc., 1965,28 P.-€3. L. Wu and R. V. Byerrum, Biochemistry, 1965, 4, 1628.30 M. Luckner and C. Ritter, Tetrahairon Letters, 1965, 741.31 D.Groger, S. Johne, and K. Mothes, Experientia, 1965,21,13; see also L. Skurskf,32 D. Munsche and I(. Mothes, PhytochRmistry, 1965, 4, 705.33 R. B. Clayton, &wart. Rev., 1965, 19, 168, 201; F. Lynen, Angew. Chem., 1965,4184; see also K. S . Yang and C. R. Waller, Phytochemistry, 1965, 4, 881.87, 4399.Coll. Czech. Chem. Comm., 1965, SO, 2080.77, 929; I(. Bloch, ibid., 1965, 77, 944420 ORGANIC CHEMISTRYin impressive detail (see p. 428). However, one investigation 34 stemmingfrom this stereochemical work will be outlined here. [4R-3H]Mevalonic acidwas mixed with [ 2-14C]mevalonic acid to give the doubly-labelled specimen(37). This was converted by a rat liver homogenate into lanosterol (38) andcholesterol (39) via squalene (40). Measurement of 3H : 14C ratios showedthat lanosterol and cholesterol retained respectively 5 and 3 of the 6 tritiumatoms in the intermediate squalene.Biological degradation of the labelledcholesterol revealed one tritium atom a t the l7a-position and left little doubtthat the remaining two were a t (3-20 and (3-24. The results were entirely ineHO( 3 9)AT I A’ OHagreement with the mechanism for squalene cyclisation proposed earlier.35Lederer and his colleagues have shown 36 that biological C-methylation withmethionine, for example in the biosynthesis of ergosterol in Neurosporacrassa, can involve loss of one hydrogen atom from the original S-methylgroup.A full account has appeared of the studies by Leete and his group on thebiosynthesis of digitoxigenin (41 ) in DigitaZis p~rpurea.~’ The labelledcardenolide (41) derived from [2J4C Jmevalonate (42) contained no activityin the butenolide ring [cf.the 14C labelling pattern in cholesterol (39)].Microbial degradation located radioactivity a t C-1, C-7, and C-15, showingthat the steroidal nucleus was constructed in the usual way. However, onethird of the activity derived from [3’-14C]mevalonate (42) was found a t C-21;also [1-14C]acetate gave material labelled at C-20 and (3-23 as well as atother positions in the nucleus. These results strongly suggest that thedigitoxigen is formed by the condensation of a C,, pregnane derivative withone molecule of acetic acid. Independent work by Tsche~che,~~ Rei~hstein,~~and their respective colleagues had led to the same conclusion. Tschescheand Brassat have now reported 40 related investigations on the bufadienolideglycoside, hellebrin (43), in Helleborus atrorubens.A5-Pregnen-3/l-ol-20-one,labelled with I4C at (2-21, was fed as an aqueous solution of its glucoside.Degradation of the derived hellebrin showed that none of the radioactivitywas located in the steroidal nucleus.An important intermediate, 2-decaprenylphenol (44; R = H), in thebiosynthesis of ubiquinone- 10 (45) has been isolated from Rhodospirillumr u b r ~ m . ~ l This is apparently the first known precursor containing bothan aromatic ring and a polyisoprenyl side-chain. Its biosynthesis probablyinvolves alkylation of p - hydroxybenzoic acid by a decaprenyl pyrophos-phate to give the acid (44; R = CO,H), which is then decarboxylated.Newresults concerning the transformation of the diterpene ( - )-lraurene intogibberellic acid (46) have been announced.42 Gibberellin A12, labelled with14C as shown (47), was added to a fermentation of Gibberella. fujikuroi. De-gradation of the derived [ 14C]gibberellic acid showed that all the radioactivity0MeIEH2*CH=C-CHd,,HHO(44)6 Gluc-Rham.0OH0Me00CH =Meresided as expected in the terminal methylene group. The [14C]-diol, ob-tained by the reduction of the acid (47) with lithium aluminium hydride,wa*s an even better precursor of gibberellic acid.Polyketides and Polyacety1enes.-A new approach to the chemical studyof poly-8-ketone cyclisations has been announced.43 The biosynthesis ofterrein (48) from acetate and malonate in Aspergillus terreus is interesting inthat the five-membered ring contains two linked '' carboxyl " carbons a t the6,7-p0sitions.~~ This is illustrated by the labelling pattern (48) produced from[l-lWIacetate. Terrein may possibly be derived from the contraction of aprecursor with a six-membered ring.The biosynthesis of auroglaucin (49)and fuscin (50) involves the condensation of a cyclised polyketide unit (thelabelling pattern from [1-14C]acetate is shown) with a C,-fragment derivedfrom 1nevalonate.45 Acetate was also incorporated into the mevalonatefragment but here, unexpectedly, the labelling (A and t in the formulae) wasnot uniform. Further progress is reported in the elucidation of tetracyclinebi0synthesis.4~ An intermediate accumulated by a blocked mutant ofStreptomyces uureofmiens has been identified as 4- hydroxy-6-methylpre-tetramid (51).Bentley and Gatenbeck 47 have shown that the entire carbon skeleton ofmollisin (52) is derived in Mollisia caesia from acetate or malonate.At least( 4 8 ) (49)I-CHMe,two polyketide chains must be involved in the biosynthesis: one possibilityis indicated (53). Orsellenic acid (54) is probably the true precursor offumigatin (55) in Aspergillus fumigatus.48 Good evidence has been obtained 49to suggest that the amino-acid orcylalanine (56), of Agrostemma githago, isderived by condensation of orsellenic acid and serine.The fern, Dryopteris marginalis, produces a family of methylene- bis-pholoroglucinol derivatives, for example, desaspidin (57).Penttila, Kapadia,and Fales 5O found that three methyl groups [asterisks in (57)] and thecentral methylene group were derived with comparable efficiency from the8-methyl group of methionine. The remainder of the molecule is constructed,presumably, from two polyketide acylphloroglucinol units. Evidence ispresented for a biosynthetic intermediate such as (58). Attention is drawnt o the recent work by Samuelson61 on the biosynthesis of prostaglandin El(59) in vesicular gland preparations. Eicosa-8,11,14,-trienoic acid was shownto be a precursor. Experiments with lSO, established that all three oxygenfunctions were derived from molecular oxygen. Of especial interest was theproof that both oxygen atoms on the five-membered ring came from the sameoxygen molecule : a plausible mechanistic explanation of this is presented.The investigation of polyacetylene biosynthesis has been especiallyfruitful this year. Bohlmann and his colleagues have studied the biologicalconversion of [ 1 -14C]dehydromatricaria ester (60) into the thioethers (61)and (62) in Anthemis t i ~ ~ t o r i ~ ~ . ~ ~ ~ 53 In each case an appropriate degradationof the end product was carried out. The production of an aromatic ringin the ether (62) is particularly interesting;53 the authors have proposed amechanism involving a 1,7-rnethyl migration. The incorporation of [2-W]-dehydromatricaria ester into the thiophen (63) in Chrysanthemum vulgarehas also been established.54 The biological formation of thiophen derivativesby the addition of H,S, or its e q ~ i v a l e n t , ~ ~ to polyacetylenes is of widespreadimportance. Feeding experiments with Echinops sphuerocephlus have ea-tablished a number of these transformation^.^^ For example, the tritium-labelled tridecenpentyne (64) was incorporated into the dithienyl (65).Shikimic Acid and Derived Aromatic Compounds.-A comprehensivereview of the chemistry and biochemistry of shikimic acid (66) has beenpublished.57 It has been shown 58 that chorismic acid (67) is an intermediatein the biological conversion of shikimic acid into prephenic acid (68). Thetransformation (67) +( 68) also proceeds nonenzymatically a t 70 O and pH 8.Full details of the chemical investigations by Edwards and Jackmannleading to the structure (67) are now available. These authors have estab-lished that a compound, isolated by Lingens and Luck,GO and assignedstructure (67) is in fact different from chorismic acid.with Rhus typhina have led to the tentative conclusion that gallic acid maybe derived, at least in part, from dehydroshikimic acid without the inter-vention of a cinnamic derivative.Biosynthetic experiments with labelled phenylalanine or coniferin haveproved useful in elucidating the structure of ligniaG2 For example, degrada-tion of spruce ligiiin derived from [ 2-14C]phenylalanine (69) gave radioactiveveratric acid. This suggests that radical coupling must occur to some extentin the sense shown [(70) and (71)], one of the coniferyl side-chains being losta t a later stage.The predicted 63 re1at)ionship between rotenone (72) and the isoflavonoidshas received experimental support. Crombie and Thomas 64 have observedthe incorporation of [%1*C]phenylalanine (69) into rotenone in Derriselliptica. Degradation showed that the radioactivity was confined to posi-tion 12a. Clearly an aryl migration, 6a--+12a, analogous t o that known 65to occur in isoflavanone biosynthesis, had taken place a t some stage in thebiosynt hetic sequence.Compounds," ed. W. D. Ollis, London, Pergamon Press, 1961, p. 59.6 5 H. Grisebach, " Recent Dcveloprnents in the Chemistry of Natural Phenoli