16 NUCLEIC ACIDS Part (ii) By M. J. Waring (Departmentof Pharmacology Downing Street Cambridge) RNA.-Double-helical complexes formed by mixtures of synthetic poly- nucleotides (poly-A plus poly-U ; poly-I plus poly-C ; poly-G plus poly-C) have been examined in detail by X-ray diffracti0n.l They reveal three distinct helical conformations of RNA the A-form with eleven base-pairs per turn; the A’-form with twelve ; and the A”-form which is non-integral and may be a family of closely related structures rather than a single molecular species. Transitions from one form to another are related to changes in salt concentra- tion. All three conformations resemble the A-form of DNA but no structure like the DNA B-form is seen. This could be explained by the T-hydroxy- group of ribose keeping the sugar ring in the C(3)-endo-conformation since when DNA takes up the B-helical form its sugar rings change to C(2)-endo.Such a restriction imposed by the ribose 2’-hydroxy-group could also explain the failure of a DNA-RNA hybrid to show an A - B transition.’r2 The new X-ray data favour an eleven-fold helix (rather than ten-fold) for the double- helical RNA of re0virus.l Electron microscopy of the replicative form (RF) and replicative intermediate (RI) of the RNA phage R17 reveals a translation per residue of 3.14 A which the authors3 consider more consistent with a ten- fold helix but would also be consistent with the twelve-fold A’-form. Ribosomal RNA.-Ribosomes from different sources yield r-RNAs of characteristically different size (see the 1967 Report).Occasionally the r-RNA shows evidence of in~tability,~ e.g. to heat,5*6 and in one instance (Euglena gracilis) more careful preparative techniques have revealed the occurrence as expected of a large and a small component where only a single species had previously been detected.6* A survey of r-RNA size in many different organisms has interesting implications for evolution procaryotic organisms have r-RNAs of molecular weight 1.09 and 0.56 million; all plants and animals have their smaller r-RNA of molecular weight 0.7 million ; and while the molecular weight of the larger r-RNA of plants is consistently 1.3 million that of the r-RNA of animals increases up the evolutionary scale from 1.4 million to 1.75 milli~n.~ ’ S.Arnott W. Fuller A. Hodgson and I. Prutton Nature 1968 220 561. * G. Milman R. Langridge and M. J. Chamberlin Proc. Nut. Acad. Sci. U.S.A.,1967,57 1804. ’ N. Granboulan and R. M. Franklin J. Virol. 1968 2 129. U. E. Loening J. Mol. Biol.. 1968 38 355. ’ J. J. Pene E. Knight jun. and J. E. Darnell jun. J. Mol. Biol.,1968 33 609. ’ C. Portier and V. Nigon Biochim. Biophys. Acta 1968 169,540. ’ J. R. Rawson and E. Stutz J. Mol. Biol. 1968 33 309; K. E. Schuit and D. E. Buetow Biochim. Biophys. Acta 1968 166. 702. 552 M. J. Waring An unexpectedly close correlation between the (G + C):(A + U) ratios of the two r-RNAs in various organisms has been described.' Sequence analysis however is still at an early stage. Partial sequence analysis has been performed on oligonucleotides in a pancreatic RNAse digest.' Bases modified by methyl- ation seem to occur in clusters; the major methylated oligonucleotides all occur twice in E.coli 23s-r-RNA. This could be the result of gene duplication during evolution or dimerization of precursor half-molecules.'O Alternatively the apparent duplication of sequences might represent the result of 'convergent' evolution" if for example the duplicated sequences were involved in the two t-RNA-binding sites' believed to occur on the ribosome. Infrared difference spectra indicate that the secondary structure of E. coli r-RNA involves 60% of the total bases in pairing; this agrees reasonably well with estimates from other methods. 5s-RNA.-Full details have been published describing the determination of the nucleotide sequences of E.c0liI4 and KB cell15 SS-RNAs. The KB cell SS-RNA exists in two forms; one has 120 nucleotides and the other is identical but for an additional U at the 3'-end (Figure 1). Since it is an axiom of molecular biology that primary sequence determines secondary and tertiary structure which in turn determine biological function it is a challenging (not to say paradoxical) situation for the molecular biologist to be confronted with mole- cules of accurately known sequence in search of a structure and a function. Yet this is still the case with SS-RNA. Various conformations have been pro- posed (see the 1967 Report) to which may be added a pair of structure^'^ which resemble the clover-leaf conformation suggested for t-RNA (see Figures 1 and 2).Optical data18 and sensitivity to monoperphthalic acid o~idation'~ indicate quite a high degree of base pairing and are perhaps most consistent with the model of Cantor,20 but the case is far from proved. Indeed it seems that E. coli 5s-RNA can exist in more than one structural form21*22 and can undergo a partially reversible transition to a 'denatured' state,22 as can some t-RNAs. F. Amaldi Nature 1969 221 95. F. Amaldi and G. Attardi J. Mol. Biol. 1968 33 737. lo P. Fellner and F. Sanger Nature 1968 219 236. C. R. Woese Nature 1968 220 923. M. Bretscher Cold Spring Harbor Symp. Quant. Biol. 1966 31 289. l3 R. 1. Cotter and W. B. Gratzer Nature 1969 221 154. l4 G. G. Brownlee F. Sanger and B.G. Barrell. J. Mol. Biol. 1968 34 379. l5 B. G. Forget and S. M. Weissmann J. Biol. Chem. 1968,243 5709. l6 B. G. Forget aad S. M. Weissmann Science 1967 158 1695. I. D. Raacke Biochem. Biophys. Res. Comm. 1968,31 528. C. R. Cantor Proc. Nat. Acad. Sci. U.S.A. 1968,59,478. l9 F. Cramer and V. A. Erdmann Nature 1968 218 92. 2o C. R. Cantor Nature 1967 216 513. M. E. Geroch E. G. Richards and G. A. Davies European J. Biochem. 1968 6,325. 22 M. Aubert J. F. Scott M. Reynier and R. Monier Proc. Nut. Acad. Sci. U.S.A. 1968,61,292. Nucleic Acids 553 uOH U U pG-C U-G C-G U-A CG c A A “AU c ‘I CCUGA --A,-u C-GG GU C G-C C C-G C-G C-G G-C A-U U-G C-G U-G CA G-C U-G CA U-A G-U AC OGA A GG 5s-RNA (KB cells) RGURE1 Nucleotide sequence of 5s-RNA from KB cells.The sequence is arranged in the clover leaf‘ formsuggestedby Raacke.l7 Another variant exists which lacks the third U at the 3’-end shown here. 554 M. J. Waring 2OH C A PC A G-C C-G G-C G-C G-C G-C *U CGG ccuAA G Ill II A CCGACGA,G GUC GGT4 U 111 I C C 7Me GG~AGCU uG \ U-A AG A C-G G-C G-C G-C 2’0MeC A UA CAU t-RNA Fet (E.Coli) RGURE 2 Nucleotide sequences of t-RNAs arranged in the clouer- leaf form. Bases in t-RNA A adenosine; AMe position of methyl group unknown ; A lMe 1-methyladenosine; A” N(6)-isopentenyl-adenosine ;A* N(6)-acylated adenosine with bulky substituent on the amino-group; At Nature of modification unknown; Y,fluorescent modified adenosine.G guanosine; GMe,position of methyl group unkown; GIMe 1-methylguanosine ;GZMe,2-methylguanosine ; G7Me, 7-methyl-guanosine ; WMe,N(2)-dimethylguanosine; 2’0MeG 2‘-0-guanosine ; G* nature of modification unknown. C cytidine; CMe position of methyl group unknown; C3Me, 3-methylcytidine ;CSMe, 5-methylcytidine;CAc N(4)-acetylcytidine ; 2’0MeC 2’-O-methylcytidine; Ct probably N(4)-acylated cytidine. U uridine ; 2’0MeU 2’-0-methyluridine ; U* 4-thiouridine. @,.pseudouridine;2’OMe@,2’-O-methylpseudouridine. I inosine; IMe,position of methyl group unknown. T thymidine. X,unknown. Nucleic Acids A OH C C A pG-C C-G G-C G-U A-U U-A U-A U GA A II DGACUC G2~e 5MeCU G 111 I cU G7Me G~AG~~ 'GDiMe C-G AG t-RNA (Yeast) A OH C C A pG-C G-C G-U C-G G-C uu G-C U AGGCC u'A MeG 11111 G CG UCCGG T4 cU,D AG C-G u-A C-G C-G C-G u9 u IMe IGC t-RNA :la (Yeast) 556 M.J. Waring gOH C G pG-C G-U C-G A-U A-U G C-G t U-A u CGUCC~~A DGAG AC G 1111I A*G 2'0MeG 5MeC 777 GCAGGTqC G GGC 2'OMe GC DDAA oiMeG u '/'CG A-U GG // u A-U G~ c A-U 'u G-C A-Y -YA U AiP IGA t-RNA 7'' (Yeast) t0" t-RNA2er differs in 3 bases as shown. c A pG-C C-G G-C G-C GA G-C A-U U CGCAC A 11111 DG A c ucG2~e GCGUG n 1111 G DG7Me GG A AG *-A - - C-G A-U G-C A-3, 2'0MeC A UY 2'0MeG A A t-RNA (Wheat germ) NucIeic Acids 557 OH C G pG-C U-A A-U G-C u-A C -G G-C u CGUCC~~A~M~ 11111 G FC CG GCAGGTqC I I I 5MeCm G-C v G-C A-92'OMe 3MeC A U AiP IGA t-RNA Ser (Rat liver) AOH C C A pC-G U-A C-G U-G C-G G-C G-C u CCCGC~~AM~ 11111 G A-U A-U G- C A-9 CA U AiP GWA t-RNA Tyr (Yeast) M.J. Wating OH C A pG-C G-U U-A U-A U-A C-G G-C U-A G-C C-G 3rc UA OH IAC C A PG-C t-RNA Va I (Baker's yeast) G-U-G-G-G-G-*U GCCcu I1II AGGGA G-C-A-G-A-C A u+ A+ GUA t C Species I,Sum as shown.Su+ . G * replaced by c . rn' Species II uc replaced by CA . Nucleic Acids 559 C A pG-C G-C C-G U-A A-U C-G D G-C (I *U UGCCC UAA A II II G G ACAGG TW I C A C-G A-u U-A C-G A-W CA U A* CtA U t-RNA :et (E.Coli) AO" C c A pG-C G-U U-A U- A U-A C-G -c GGG uc AIM^ 11111 G D DGAqC CCCAGTqC G I 5MeC G UGGC C DC A AA 9-A C-G U-A G-C C-G "C UA I AC t-RNA y"' (Torula yeast) 560 M. J. Waring (This may also be true of 28S-r-RNA.23) Clearly a primary requirement for formulating likely structural models for molecules of known sequence is some means of maximising base pairing; a mathematical approach to this problem has been de~cribed.’~ The function of 5S-RNA remains unknown :its location in the ribosome would suggest some role in protein synthesis for which there is limited e~idence,~’ but little more data are available other than studies on its removal from and association with ribosomal particles.’’9 26 Viral RNA.-Table 1 summarises data on terminal sequences of viral RNAs; the great majority of which were published during 1968. In most in- stances the methods involved specific labelling of one end of the RNA e.g. by attaching a [32P]phosphate group at the 5’-end by use of labelled ATP and polynucleotide kina~e,~~. 39 or selective periodate oxidation of the 3’-terminal nucleoside followed by reduction with [3H]borohydride ;29 then cleavage with an endonuclease isolation of the labelled oligonucleotide and sequence analysis.One new technique however identifies the oligonucleotide derived from a 3’-hydroxy-end on the basis of its unchanged electrophoretic mobility after phosphatase treatment.27 It is striking that all the bacteriophage RNAs begin with a G residue bearing a 5‘-triphosphate group; this invites comparison with t-RNAs most of which also begin with a G residue and all of which bear a 5‘-phosphate (see Figures 2 and 3). Perhaps more striking is the 3‘-terminal --CCA, grouping which occurs in all the viral RNAs (including the three plant viruses) and has long been recognised as characteristic of t-RNA. This similarity has prompted some workers to speculate that perhaps the --CCA, serves to protect both t-RNAs and viral RNAs from nuclease attack in the cell.Be that as it may it now appears that the terminal adenosine of viral RNAs is probably added by some host cell enzyme,41 42 perhaps the enzyme involved in the turnover of the --CCAoH end of t-RNAs since infectivity of R17 RNA 23 H. Singh and D. Keller Biochim. Biophys. Acta 1968 169 150. 24 V. G. Tumanyan L. E. Sotnikova and A. V. Kholopov Doklady Akad. Nauk S.S.S.R. 1966 166,1465. 25 D. M. W. Kirtikar and A. Kaji J. Biol. Chem. 1968 243 5345. 26 P. Morel1 and J. Marmur Biochemistry 1968 7,1141 ;M. A. Q.Siddiqui and K. Hosokawa Biochem. Biophys. Res. Comm. 1968,32 1. 27 J. E. Dahlberg Nature 1968 220 548. H. L. Weith and P.T. Gilham J. Amer. Chem. SOC.,1967,89 5473. 29 D. G. Glitz A. Bradley and H. Fraenkel-Conrat Biochim. Biophys. Acta 1968 161 1. 30 R. Roblin J. Mol. Biol. 1968 36 125; 1968,31 51. 31 D. G. Glitz Biochemistry 1968 7,927. 32 R. De Wachter J. P. Verhassel and W. Fiers F.E.B.S. Letters 1968,1,93. 33 R. De Wachter and W; Fiers J. Mol. Biol. 1967 30,507. 34 M. Watanabe and J. T. August Proc. Nut. Acad. Sci. U.S.A. 1968,59 513. 35 R. De Wachter and W. Fiers Nature 1969 221 233. 36 H. L. Weith G. T. Astenadis and P. T. Gilham Science 1968,160 1459. 37 S. Mandeles J. Biol. Chem. 1967 242 3103. 3a J. Suzuki and R. Haselkom J. Mol. Biol.. 1968,36 47. 39 E. Wimmer and M. E. Reichmann Science 1968,160 1452; E. Wimmer A. Y. Chang J. M. Clarke jun. and M. E. Reichmann J.Mol. Biol. 1968 38 59. 40 D. H. L. Bishop D. R. Mills and S. Spiegelman Biochemistry 1968 7 3744. 41 A. Vandenberghe B. Van Styvendaele and W. Fiers European J. Biochem. 1969,7,174. 42 R. Kamen Nature 1969 221,321. TABLE 1 Terminal sequences of viral RNAs Ref. RNA Sequence Y-end 3’-end f2 pppG ________________G U U A C C A C C C AOH 27 27-29 R17 p p p G pu py ____-_______ G U U A C C A C C C A 30 27 MS2 31 29,33 PPPGGU 1 GUUACCACCCA, 32 or p p p G G G U pppGGGGAAC G C C C U C C U C U C U C C C A, 27,34,35 27,36 QP pppGGGGGAAC $ b ‘Little’variant pppGGGGA A 5. 40 % ofQP 29,37 Turnip yellow A py______-_--_-_l-_-.l-l_-l_-__-38 mosaic virus Satellitetobacco p p A G U -39 necrosis virus 562 M.J. Waring Pu 1-3 Ant icodon t-RNA Homologies * indicates G-C pairing PU indicates a purine py indicates a pyrimidine. RGURE 3 Generalised clover-leafstructure for t-RNA. is retained after removal of the 3’-terminal adenylate (but not if the penultimate C is removed as well).42 Qp RNA provides a curiosity in that both the 5’-terminal sequences shown in Table 1 occur in the progeny of an infection initiated by a single parental phage particle suggesting that some sort of equi- librium between the two types is established during growth.35 On the other hand it seems unlikely3’ that a similar explanation can account for the different 5’-terminal sequences reported for MS2 RNA (Table 1).The apparent comple- mentarity of the terminal sequences of MS2 RNA and Qp RNA has prompted the suggestion that the ends of these molecules might associate by hydrogen bonding to form pseudo-circles.3’ 329 t-RNA.-The complete sequences of six more t-RNA species and variants of them are now known valine t-RNA from Torulopsis utiIis;43 serine t-RNA 43 S. Takernura T. Mizutani and M. Miyazaki J. Biochem. (Japan) 1968 63 277. Nucleic Acids 563 from rat liver ;44 phenylalanine t-RNA from wheat germ ;45 tyrosine t-RNA and the amber suppressor su& t-RNA derived from it from E. COIZ;~~ and the methionine-specific t-RNA and t-RNA from E. coli.47,4a Full details of the yeast phenylalanine t-RNA sequence have also been p~blished.~’ The new sequences are shown together with the five reported earlier (Reports for 1965-1967),in the conventional clover leaf form in Figure 2.The considerable degree of homology in the eleven sequences is clear. One can divide the anatomy of the clover leaf into five distinct sections the ‘amino-acid’ helix of seven base pairs with its unpaired --CCAoH end (top); the seven-nucleotide ‘TJIC’ loop with its helical stem of five base pairs (right); the seven-nucleotide ‘anti- codon’ loop with five base pairs in its stem (bottom; the anticodon itself consists of the three bases at the very bottom); the ‘dihydro-U’ loop with three or four base-pairs in its stem (left); and the ‘extra’ loop or ‘lump’ which varies widely in the different molecules and lies between the TJIC and anticodon arms. Much can be learnt from a detailed study of the apparent homologies.To facilitate comparison a generalised clover leaf structure has been compiled in Figure 3 which draws attention to the most striking regularities in the known sequences. Certain irregularities may also be noted for example the ‘dihydro-U’ loop shows substantial variation and indeed in the E. coli tyrosine t-RNAs does not even contain any dihydro-U. In some cases the base-pairing in the ‘amino- acid’ and ‘TJIC’ stems appears to be imperfect. Knowledge of regions shared in common between different t-RNAs is important for formulating detailed ter- tiary structure models (see later) and for identifying regions concerned with particular functions such as interactions with the ribosomea during protein synthesis.Special interest attaches to the t-RNA:“ and t-RNA:“ sequences in view of the peculiar role of the latter in the initiation of polypeptide synthesis (see the 1966 Report). The two sequences are surprisingly dissimilar (only forty- one nucleotides are held in common); this makes it difficult to determine recog- nition sites for aminoacyl t-RNA synthetases the transformylase or initiation factors for protein synthesis. It is hypothesised that the tertiary structures of the molecules must be of importance for these functions.47 In any event the unique ability of t-RNA to recognise both AUG and GUG codons in ribosome- binding experiments is not due to the fact that two forms differing in the 7-methyl-G or A replacement exist. 50N-formylmethionyl t-RNA has been found in mitochondria of yeast and liver but not in their cytoplasmic protein- synthesising systems.s’ 44 M. Staehelin H. Rogg B. C. Baguley T. Ginsberg and W. Wehrli Nature 1968,219 1363. 45 B. S. Dudock G. Katz E. K. Taylor and R. W. Holley Proc. Nut. Acad. Sci. U.S.A. 1969 in the press. 46 H. M. Goodman J. Abelson A. Landy S. Brenner and J. D. Smith Nature 1968 217 1019. 47 S. Cory K. A. Marcker S. K. Dube and B. F. C. Clark Nature 1968,220 1039. 48 S. K. Dube K. A. Marcker B. F. C. Clark and S. Cory Nature 1968 218,232. 49 U. L. Rajbhandary and S. H. Chang J. Bid. Chem. 1968 243 598. S. Cory S. K. Dube B. F. C. Clark and K. A. Marcker F.E.B.S.Letters 1968 1,259. 51 A. E. Smith and K. A. Marcker J. Mol. Bid. 1968 38. 241. 564 M.J. Waring The modified nucleotide Y which occurs next to the anticodon in t-RNAPhe of yeast and wheat germ is strongly fluorescent and occurs in t-RNAPhe of other organisms.52 The base is a rather hydrophobic adenine derivative of as yet unknown structure; it can be selectively removed by mild acid without breaking the chain and the treated t-RNA can still be charged with phenyl- alanine but it is unable to bind to ribosomes in the presence of poly-U or to transfer its Phe to growing peptides.53 Thus bases in the anticodon loop as well as the anticodon itself are needed for codon-anticodon interaction but the integrity of the anticodon loop cannot be necessary for recognition by the aminoacyl synthetase. Similar conclusions may be drawn from other instances where base changes in the presumed anticodon affect the coding properties of the t-RNA but still permit charging with the amino-a~id.~~ Especially eloquent is the production of tyrosine-specific su,’; amber suppressor t-RNA by a mutation which substitutes C for G* in the anticodon of a minor tyrosine t-RNA of E.(Figure 2). Codon-anticodon interaction also seems to require the helical stem of the anticodon loop? There is evidence that the ‘amino-acid’ helical stem and the unpaired --CCAoH end may be directly involved in recognition by the aminoacyl synthetaseS6 and ribosomal peptidyl transferase5’ respectively. Large fragments derived from t-RNA may help to identify regions which interact specifically and non-specifically with ribo- some~.~~ Although the susceptibility of t-RNA sequences to chemical and enzymic attack yields results in broad agreement with the clover leaf model (p.566) the most compelling reasons for believing that it is basically correct are to be found in Figures 2 and 3 in the high degree of homology between different sequences which becomes evident when they are written in the ‘standard’ clover-leaf form. The thermal denaturation pattern of E. coli methionine t-RNA shows its structure to be perceptibly more stable than that of t-RNA (ref 59) which is consistent with an unusually high proportion of GC pairs in helical regions (compare Figure 2) but definitive proof of the clover leaf will probably have to wait for detailed X-ray diffraction examination of t-RNA crystals.At the end of 1968 papers from six laboratories reported that crystallisation of t-RNA had been achieved. Crystals were obtained with formylmethionine 52 D. Yoshikami G. Katz E. B. Keller and B. S. Dudock Biochim. Biophys. Acta 1968 166 714; L. M. Fink T. Goto F. Frankel and I. B. Weinstein Biochem. Biophys. Res. Comm. 1968,32 963; B. S. Dudock G. Katz E. K. Taylor and R. W. Holley Fed. Proc. 1968 27 342. 53 R. Thiebe and H. G. Zachau European J. Biochem. 1968 5 546; Biochem. Biophys. Res Comm. 1968,33,260. 54 J. Carbon and J. B. Curry Proc. Nat. Acad. Sci. U.S.A. 1968 59 467; J. Mol. Biol. 1968 38,201 ;G. Sundharadas J. R. K&ze D. Soll W. Konigsberg and P. Lengyel Proc. Nat. Acad. Sci. U.S.A. 1968 61 693. 55 B. F. C. Clark S. K. Dube and K.A. Marcker Nature 1968 219 484. s6 L. H. Schulman and R. W. Chambers Proc. Nat. Acad. Sci. U.S.A. 1968,61 308. s7 R. E. Monro J. CernB and K. A. Marcker Proc. Nut. Acad. Sci. U.S.A. 1968,61 1042. 58 A. D. Mirzabekov D. Griinberger and A. A. Bayev Biochim. Biophys. Acta 1968,166,68. 59 T. Seno M. Kobayashi and S. Nishimura Biochim. Biophys. Acta. 1968 169,80. Nucleic Acids t-RNAc0 and phenylalanine t-RNA6' from E. coli ;serine t-RNA,62 phenylal- anine t-RNA,63 and formylmethionine t-RNA61 from yeast; and even with unfractionated mixtures of ~-RNAs.~~ Unit cell dimensions were given for formylmethionine t-RNAgO and phenylalanine t-RNA6' of E. coli but the data were clearly rudimentary. While the clover-leaf model provides an acceptable representation of the secondary structure of t-RNA there remains the possibility that the molecule may have an ordered tertiary structure.Low-angle X-ray scattering by t-RNA solutions6s and X-ray diffraction from oriented fibres66 both indicate that the arms of the clover leaf may be folded together possibly in pairs to form an RGURE 4 Two-dimensional projection of a proposed tertiary struc- ture for yeast phenylalanine t-RNA after Cramer et and personal communication from Professor F. Cramer. 6o B. F. C. Clark B. P. Doctor K. C. Holmes A. Klug K. A. Marcker S. J. Morris and H. H. Paradies Nature 1968,219,1222; S. H. Kim and A. Rich Science 1968 162 1381. 61 A. Hampel M. Labanauskas P. G. Connors L. Kirkegard U. L. RajBhandary P. B. Sigler and R.M.Bock Science 1968 162 1384.H. H. Paradies F.E.B.S. Letters 1968 2 112. 63 F. Cramer F. v. d. Haar W. Saenger and E. Schlimme Angew. Chem. 1968,80,969. 64 J. R.Fresco R.D. Blake and R.Langridge Nature 1968,220 1285. 65 J. A. Lake and W. W. Beeman J. Mol. Biol. 1968,31 115. 66 B. P. Doctor W. Fuller and N. L. Webb Nature 1969 221 58. 566 M. J. Waring H shape with the ‘amino-acid’ stem stacked on the T$C-containing arm and the anticodon arm stacked on the arm which bears the dihydro-U loop.66 A more detailed model proposing additional fixation of the arms has been based on measurements of aminoacylation rates and susceptibility of adenosines to monoperphthalic acid oxidation at various temperatures (Figure 4). In this model the helical regions of the dihydro-U-containing arm the amino-acid-bearing arm and the T$C-containing arm are packed to form a trigonal prism with hydrogen bonding between the $C in the T$C loop and the AG in the dihydro-U loop and between the CC at the --CCAoH end and the GG in the dihydro-U loop.These interactions are possible for all the known t-RNA sequences (Figure 3). The anticodon arm extends away from the trigonal prism in the opposite direction. For different species of t-RNA various addi- tional base pairs are feasible which would provide further stabilisation and might help to confer characteristic differences in tertiary structure.67 Such differences doubtless of importance for recognition would be largely deter- mined by the size of the dihydro-U loop and the central part of the molecule at the junction of the arms,including the ‘extra’loop.Weak ‘additional’ interactions like those proposed in this model could account for conformational changes in t-RNAs which might be of functional significance and might also include the reversible denaturation referred to in the 1967 Report. Recent evidence indicates that the native + denatured interconversion involves breakage and re-formation of hydrogen-bonded base pairs as well as changes in base stack- ing,68 and that the nature and concentration of cations (but not necessarily Mg+ +) are 69 Related to the phenomenon of denaturation is the occurrence of aggregates of t-RNA especially dimers frequently observed during purification procedures. It now appears that at least in some instances dimers are formed from monomers by intermolecular interactions resembling normal intramolecular interactions.70p 71 Dimers of yeast alanine t-RNA are interconvertible with monomers by the action of heat ;7 their hypochromism and denaturation spectra are so similar to those of the monomer that they must be presumed to contain the same extent of base pairing; they accept two mole- cules of alanine in the enzymic charging reaction but their anticodon sequences are much less readily accessible for specific cleavage by RNAse T,. These observations7’ suggest the sequence of events in Figure 5. DNA.-An entertaining recent discovery is the finding that certain mutants of E. coIi produce considerable amounts of DNA-less cells which can be separated from the normal (DNA-containing) cells.72* 73 Two such mutants have been 67 F.Cramer H. Doepner F. v. d. Haar E. Schlimme and H. Seidel Proc. Nat. Acad. Sci. U.S.A.,1968 61 1384. ‘* T. Ishida and N. Sueoka J. Biol. Chem. 1968 243 5329. 69 T. Ishida and N. Sueoka J. Mol. Biol. 1968 37 313. 70 A. Adams and H. G. Zachau European J. Biochem. 1968 5 556. 71 J. S. Loehr and E. B. Keller Proc. Nat. Acad. Sci. U.S.A. 1968 61 1115.. 72 H. I. Adler W. D. Fisher A. Cohen and A. A. Hardigree Proc. Nat. Acad. Sci. U.S.A. 1967 57 321. 73 Y.Hirota F. Jacob A. Ryter G. Buttin and T. Nakai J. Mol. Biol. 1968 35 175. unfolded MONOMERS by heat DIMER RGURE 5 Dimers of t-RNA possible structure and mechanism of interconversion with monomers.568 M. J. Waring studied ;in both cases the DNA-less cells arise because of a more or less normal septation process occurring to one side of the nuclear material but while one mutant produces DNA-less cells of relatively normal dimension^,^ in the other mutant septation occurs near one or both poles of the cell giving DNA- less ‘minicells’.72 Apart from their lack of DNA the cells contain protein and RNA and carry out several normal metabolic processes but not DNA-dependent syntheses.72*73 Minicells can be mated with F+ male strains of E. coli; after mating the minicells can be reisolated and extracted to yield the transferred DNA which very probably is the F epis~me.~~ Application of electron microscopy to the study of DNA continues to yield important results especially with circular DNA (see below).High-contrast staining techniques have been de~cribed,~ and a micro-procedure has been developed which requires as little as 0.01 pg of nucleic acid.76 The extreme sensitivity of electron microscope observations enabling single molecules of DNA to be seen has been employed to measure the diffusion coefficient of DNA at essentially ‘zero’ c~ncentration.~~ A by-product of ths work worth noting for its practical implications in work with very dilute DNA solutions is the finding that adsorption of DNA to glass was undetectable with detergent- cleaned glass but did occur with glass cleaned by chromic acid and water.77 An elegant method for studying deletion mutations at the level of the DNA molecule has been described:78 if DNAs from a deletion mutant and the wild- type organism are denatured and annealed together the hybrid molecules have a lowered contour length and a ‘bush’ (formed by the wild-type strand which is in effect looped out) visible at the position of the deletion (Figure 6).Thus the position and length of the deleted region can be mapped directly on the DNA molecule. Partial sequences have been determined for the cohesive ends of phage h DNA by use of the DNA polymerase-catalyzed ‘repair’ reaction in which the 3’-ended strands are lengthened by incorporation of nucleotides comple- mentary to those in the protruding 5’-ended strands7’ In this way thirteen residues of dG thirteen of dC seven of dA and seven of dT were added the complementarity implicit in these values is further evidence that the cohesive ends are indeed complementary and the sum shows that each protruding strand is twenty nucleotides long.79 The partial sequence data are shown in Figure 7.Ofthe other phage DNAs known to possess cohesive ends a h-related group will form mixed hybrid concatemers with h DNA and the phages from which these DNAs are derived will ‘help’ h DNA in infectivity assays; these effects are not shared by a h-unrelated group but within that group a similar relationship 74 A. Cohen W. D. Fisher R. Curtis tert. and H. I. Adler Proc. Nut. Acud. Sci. U.S.A. 1968 61 61. 75 C. N. Gordon and A. K. Kleinschmidt Biochim. Biophys. Actu 1968 155 305. 76 H. D. Mayor and L. E. Jordan Science 1968,161 1246.D. Lang and P. Coates J. Mol. Biol. 1968 36 137. 70 R. W. Davis and N. Davidson Proc. Nat. Acad. Sci. U.S.A. 1968,60,243. 79 R. Wu and A. D. Kaiser J. Mol. Biol. 1968 35 523. 6 FIGURE Electron micrograph of a renatured hybrid DNA molecule formed by annealing denatured wild-type A DNA with denatured DNA from a deletion mutant. A bush marking the position of the deletion mutation is visible about half-way along the length of the molecule. c.s.-~.P.5G8 Nucleic Acids 569 between cohesive end complementarity and helper function can be observed." This is the first evidence of biological function of cohesive ends. The h-related and h-unrelated DNAs also differ with respect to the stability of interaction of their cohesive ends.'l A genetic function Ter has been identified which generates the ends of mature h DNA ;it might control the production of a highly specific nuclease.82 (3') (r-strand) (5') PyAG Il A G GGGGGC GC GC,, AA I1 (5') (1-strand) (3') FIGURE 7 Partial nucleotide sequence of the cohesive ends of phage h DNA The structure of a native linear h DNA molecule is repre- sented by two lines representing the l-strand and r-strand with 3'-and 5'-terminal nucleotides identified.The nucleotides added in the presence of DNA polymerase which have been identified are written beyond the ends of the lines. An additional dG residue is located at one of the three caret marks." For many years it has been a puzzle to explain why denatured DNA always seems to retain a few per cent of the transforming activity of native DNA.Compelling evidence has now been obtained that this residual activity is associated with a small proportion of 'naturally occurring' cross-linked DNA molecules which resist the irreversible strand separation characteristie of normal DNA molecules upon denaturation. 83-'5 The cross-links appear to be located at or close to the ends of the molecules and it is suggested84 that they might arise during shear breakage encountered during DNA extraction and purification perhaps by a mechanism similar to that shown in Figure 8. A novel and rigorous approach to problems of helix-random coil transition in polydeoxyribonucleotides is suggested by results obtained with alternating dAT oligomers of known chain length.At low salt concentrations they form one-chain hairpin helices while at higher salt concentrations two-chain helices form which when heated rearrange into hairpins before 'melting' completely into random coils.86 If more than thirty-two nucleotides long they can be coil- verted into covalently closed circles by polynucleotide ligase ; the circles can 8o M. Mandel and A. Berg Proc. Nat. Acad. Sci. U.S.A.,1968,60 265. 81 M. Mandel and A. Berg J. Mol. Biol. 1968 38 137. 02 S. Mousset and R. Thomas Nature 1969 221 242. 03 R. Rownd D. M. Green R. Sternglanz and P. Doty J. Mol. Biol. 1968,32,369; B. M. Alberts and P. Doty ibid. p. 379; C. Mulder and P. Doty ibid.,p. 423. 84 B. M. Alberts J. Mol. Biol. 1968 32 405.'' M. R. Chevallier and G. Bernardi J. Mol. Biol. 1968 32 437. 86 I. E. Schemer E. L. Elson and R. L. Baldwin J. Mol. Biol. 1968,36,291. 570 M. J. Waring Reaction with solvent FIGURE 8 Possible mechanism for formation of cross-links during breakage by shear ofDNA. The percentages indicate the relative weighting of the alternative pathways as suggested by the degree of cross-linking observed in standard DNA preparations. The creation of ion pairs on breakage is assumed C+ indicating a carbonium ion. Free-radical mechanisms are also possible. It should be noted that breakage could require incipient solvolysis in which case no reactive intermediates of any type need be formed.84 form a base-paired helix (with a loop at each end) which is markedly more resistant to thermal denaturation than the helix formed by the equivalent linear oligomer.87 Circular DNA.-Closed circular duplex DNAs with their twisted super- helical structure or supercoils continue to attract attention.One of the chief sources is mitochondria1 DNA from higher organisms which seems always to occur in the form of 5 p twisted circles.88* 89 In lower organisms such as yeast however the situation appears to be different and as yet rather confused heterogeneous linear molecules are often found and sometimes circles.88 -It is suggested that the complexity of the picture might be due to fragmentation of larger perhaps circular molecules. 88r 89 Twisted circular intracellular forms of two more phage DNAs have been fo~nd,~~.~~ and also ‘minicircles’ (extremely small closed circular duplexes) in uninfected ba~teria.~ Bacterial ” B.M. Olivera I. E. Schemer and I. R. Lehman J. Mol. Biol. 1968,36,275. E. F. J. Van Bruggen C. M. Runner P. Borst G. J. C. M. Ruttenberg A. M. Kroon and F M. A. H. Schuurmans Stekhoven Biochim. Biophys. Acta 1968 161,402. 89 ‘Biochemical Aspects of the Biogenesis of Mitochondria,’ ed. E. C. Slater J. M. Tager S. Papa and E. Quagliariello Adriatica Editrice Bari Italy 1968. 90 L. Shapiro L. I. Grossman J. Marmur and A. K. Kleinschmidt J.Mol. Biol. 1968,33,907 C. J. Avers F. E. Billheimer H. P. Hoffmann and R. M. Pauli Proc. Nut. Acad. Sci. U.S.A. 1968 61 90; Y.Suyama and K. Miura ibid. 1968 60,235; G. E. Sonenshein and C. E. Holt Biochem.Biophys. Res. Comm. 1968 33 361 ; D. R. Wolstenholme and N. J. Gross Proc. Nut. Acad. Sci. U.S.A. 1968 61 245. 91 M. Rhoades and C. A. Thomas jun. J. Mol. Biol. 1968,37,41. 92 C. S. Lee N. Davidson and J. V. Scaletti Biochem. Biophys. Res. Comm. 1968 32 752. 93 N. R. Cozzarelli R. B. Kelly and A. Kornberg Proc. Nut. Acad. Sci. U.S.A. 1968,60,992; C. S. Lee and N. Davidson Biochem. Biophys. Res. Comm. 1968 32 151. Nucleic Acids 57 1 sex factors can be isolated in the form of twisted circles ;94 they have been used to study the rate of production of single-strand breaks in DNA by X-irradiation of cells,95 and the method has been calibrated to permit measurement of molecular weights of the DNA of several F’ element^.'^ Estimates of the number of superhelical turns in naturally occurring closed circular duplexes have been made by several methods The values obtained have on the whole been in very good agreement and lead to the conclusion that the number of turns is proportional to the molecular weight of the DNA (Table 2).The method based on untwisting caused by the intercalating drug ethidium bromide”. lo2(see the 1967 Report) remains the simplest and most straight- forward technique. Confidence in its fundamental assumption that intercala- tion of ethidium uncoils the double helix by 12”may be gained from the agree- ment with values derived by other methods (Table 2). Use of ethidium also enables the number and sense of supercoils to be varied over a continuous range at will a facility which can be expected to prove valuable in investigation of the hydrodynamic behaviour of closed circular DNAs.The alkaline denatura- tion method for estimating superhelical turns depends upon the disruption of a small percentage of the base pairs in the DNA which occurs at pH values just below the pH at which nicked or linear molecules are denatured ~ompletely.~~ In this condition both strands are still intact and topologically bonded but the supercoils have been lost.98 Disruption of more base pairs would be expected to lead to superhelix formation in the opposite (left-handed) sense before the molecule collapses into the double-stranded cyclic coil form. These changes in supercoiling have been observed by electron microscopic examination of p~lyoma~~’ and papillomalo6 DNAs after progressive denaturation by heating to various temperatures in the presence of formaldehyde.The initial stages of denaturation seem to occur preferentially in only a few regions the relative positions of which can be mapped.’” Direct electron microscopy of super- coiled DNA also provides an estimate of supercoiling turns since each turn should give rise to a visible ‘crossover’ of the double helix. In early work little confidence was placed in this type of measurement :it is often difficult to be sure 94 D. Freifelder J. Mol. Biol.. 1968 34 31. 95 D. Freifelder J. Mol. Biol. 1968 35 303. 96 D. Freifelder J Mol. Biol. 1968 35 95. 97 L. V. Crawford and M. J. Waring J. Mol. Biol. 1967 25 23. ’* J. Vinograd J. Lebowitz and R.Watson J. Mol. Biol. 1968,33 173. 99 W. Bauer and J. Vinograd J. Mol. Biol. 1968 33 141. loo M. J. Waring Nature 1968 219 1320. H. Bujard J. Mol. Biol. 1968 33 503. L. V. Crawford and M. J. Waring J. Gen. Virology,1967 1 387. lo’ G. J. C. M. Ruttenberg E. M. Smit P. Borst and E. F. J. Van Bruggen Biochim. Biophys. Acta 1968 157,429. lo4 V. C. Bode and L. A. MacHattie J. Mol. Biol. 1968 32 673. M. F. Bourguignon and P. Bourgaux Biochim. Biophys. Acta 1968 169,476. E. A. C. Follett and L. V. Crawford J. Mol. Biol. 1967 28,455. E. A. C. Follett and L. V. Crawford J. Mol. Biol. 1967 28 461; E. A. C. Follett and L. V. Crawford ibid. 1968 34 565; M. F. Bourguignon Biochim. Biophys. Acta 1968 166 242. T ul -4 N TABLE 2 Number of superhelical turns in closed circular duplex DNAs Mol.wt Number DNA Method Solvent (millions) of turns Polyoma 3-2 Ethidium O.OSM-~I%-HC~ -12 97 Alkaline denaturation Buoyant CsCl -15+1 98 SV-40 3-2 Ethidium 1M-NaC1 -16 f3.5 99 g Buoyant CsCl -12.7f1.5 4 Phage +X 174 3.4 Ethidium O.OSM-tris-HC1 -12 100 P B replicative form Crossover count O-~M-NH, acetate (-113 88 f' @Q Bovine papilloma 4.9 Ethidium O.OSM-tris-HC1 -18+3 101 Human papilloma 5.3 Ethidium -20 102 Shope papilloma 5.3 Et hidium -20 102 Chick liver Ethidium O.OSw-tris-HC1 -40 103 10-1 1 acetate (-)35 +6 88 mitochondria Crossover count O-~M-NH Phage h 31 Crossover count Ionic strength 0.06 (-)117 f11 104 intracellular form Crossover count Ionic strength 2.0 (-112 Nucleic Acids 573 of an accurate count partly because it is not always possible to distinguish left- handed from right-handed crossovers.Some workers have however claimed that the sense of crossovers can be seen,91* lo5and at least in two cases agree- ment with the ethidium intercalation method is good (Table 2). Examples are shown in Figure 9. The ratio of sedimentation coefficients of closed circular DNA and 'nicked' circles provides a crude estimate of supercoiling (crude because the ratio is rather insensitive to the number of turns if it is greater than about threeg9) but it has been used to show that the extent of super- coiling is influenced by temperaturelo8 and ionic strengthlo4* '08* log (com-pare also Figure 9). This leads to the important conclusion that the pitch of the DNA double helix varies with salt concentration and temperature.It also suggests a simple explanation for the origin of supercoils i.e. that they arise because the helix pitch increases in response to differences between the environ- ment in which closure was made and that used to study the DNA in uitro. This would be consistent with the finding that the number of supercoils is apparently a function of the length of the DNA (Table 2). It is not certain however that the ionic strength effect alone could account for the observed extent of super- coiling in natural DNAs.lo4 How a closed circular duplex DNA could replicate in uiuo poses an interesting topological problem but replicating circles have been seen.' ' O Catenated circular DNA molecules have now been reported in sea urchin eggs,'" phage h-infected E.c~li,~' and mitochondria from a number of mam- malian tissues. 'l2 The mammalian mitochondrial DNAs all showed one major difference from the leucocyte mitochondrial DNA of patients with chronic granulocytic leukaemia they contained no circular dimers which accounted for as much as 26 % of the complex mitochondrial DNA from leukatmic cells. '' Sedimentation velocity data have been published for various species of catenanes and circular oligomers of mitochondrial DNA; the most surprising result was the finding that catenated dimers of open (nicked) circles sediment slower than the free monomeric circles themselves. '' Binding of Drugs.-A number of drugs most of which are potent anti- tumour agents appear to bind to DNA both in uiuo and in uitro.It is generally believed that their biological effects and occasional therapeutic value may be accounted for by interference with the structure and function of DNA. This topic has been reviewed.'00* 'I4 The mitomycin antibiotics activated by reduc- tion act as powerful bifunctional alkylating agents and form cross-links be- tween the DNA strands :l ''their action is similar to but probably not identical coiled DNA also provides an estimate of supercoiling turns since each turn lo' J. C. Wang D. Baumgarten and B. M. Olivera Proc. Nut. Acud. Sci. U.S.A. 1967 58 1852. log J. A. Kiger jun. E. T. Young,jun. and R. L. Sinsheimer J. Mol. Biol. 1968,33 395.'lo T. Ogawa J.4.Tomizawa and M. Fuke Proc. Nat. Acad. Sci. U.S.A. 1968 60,861; R.H. Kirschner D. R. Wolstenholme and N. J. Gross ibid. p. 1466. 111 L. Pik6 D. G. Blair A. Tyler and J. Vinograd Proc. Nut. Acud. Sci. U.S.A. 1968,59 838. ''' D. A. Clayton C. A. Smith J. M. Jordan M. Teplitz and J. Vinograd,Nature 1968,220,976. 'I3 B. Hudson and J. Vinograd Nature 1969 221 332. G. Hartmann W. Behr K. A. Beissner K. Honikel and A. Sippel Angew. Chern.Internat. Edn.. 1968 7 693. FIGURE Electron micrographs of closed circular h DNA molecules. Examples of twisted 9 (1 28 crossings) and relatively untwisted (22 crossings) circular h DNA mole-cules the former prepared from low-salt solution and the latter prepared from high-salt solution.c.s.-Jp. 573 574 M. J. Waring with the action of nitrogen rnustards'l6 which are of great importance in cancer chemotherapy. Another group of drugs form reversible complexes with DNA by intercalation of their planar polycyclic ring systems between adja- cent base pairs of the double helix. ''' Among these proflavine is important for its role in the production of frame-shift mutants (see the 1966 Report) while ethidium has found valuable application in the study of closed circular duplex DNA (see previously and the 1967Report). Intercalation has recently become of interest in the interaction of nucleosides with polynucleotides (p. 549).A third mode of interaction with DNA is provided by the peptide-containing antibiotic actinomycin D. This drug (I) is much used by molecular biologists Sarcosine Sarcosine L-N-L-N-I \L-Thr/" \L-Thr/o \ / Actinomycin D.The 2-amino-4,6-dimethylphenoxazinone 3-ring system constitutes the chromophore of the molecule the 7-position is indicated. as a specific inhibitor of DNA-dependent RNA synthesis ;''* it is also a powerful inhibitor of certain tumours probably for the same reason. Its binding to DNA is specific for deoxyguanosine residues in double-helical (not denatured or single-stranded) DNA ;moreover it is the 2-amino-group of guanine which is particularly important since actinomycin will bind to a synthetic dAT copolymer containing occasional 2,6-diaminopurine-thyminepairs but not to ordinary dAT copolymer."g The binding of actinomycin to DNA requires the unsubstituted amino-group and quinoidal oxygen of the chromophore.' ' A molecular model which explains these results proposes that the drug molecule attaches itself in the narrow groove of the DNA and forms hydrogen bonds W. Szybalski and V. N. Iyer in 'Antibiotics 1 Mechanism of Action,' ed. D. Gottlieb and P. D. Shaw Springer-Verlag New York 1967 p. 21 1. I l6 P. D. Lawley Prog. Nucleic Acid. Res. and Mol. Biol. 1966 5 89. 'I' L. S. Lerman J. Mol. Biol. 1961,3 18 ;M. J. Waring in 'Biochemical Studies of Antimicrobial Drugs,' ed. B. A. Newton and P. E. Reynolds Symp. SOC. Gen. Microbiol. Cambridge University Press 1966 16 235; A. Blake and A. R. Peacocke Biopolymers 1968 6 1225. E. Reich and I. H. Goldberg Prog. Nucleic Acid. Res.and Mol. Biol. 1964 3 183. E. Reich A. Cerami and D. C. Ward in ref. 115 p. 714. Nucleic Acidr between the chromophore amino-group and quinoidal oxygen and the 2-amino-group N-3 and sugar ring oxygen of deoxyguanosine.''* This model has recently been challenged. The new data revealed hydrodynamic changes associated with actinomycin binding which have hitherto been regarded as characteristic of intercalation ; moreover it is shown that bulky substituents on the 7-position of the chromophore cause profound changes in the kinetics of association and dissociation,12' yet this side of the chromophore is away from the DNA in the narrow groove-binding model. Accordingly an inter- calation model is proposed with the specificity for deoxyguanosine explained by electronic interactions in the x-complex formed in an intercalated struc- ture.I2' It will take a good deal of work to apply crucial tests to distinguish which model is more correct ; preliminary studies indicate that actinomycin affects the supercoils of closed circular DNA in the same fashion as ethidiurn.lz2 This is more readily explicable by the intercalation model ;yet flow dichroism measurements indicate that the plane of the actinomycin chromophore is inclined at an angle of 23 & 5" to the perpendicular to the helix axis; this is more compatible with the narrow groove-binding model.' 23 Thank's are due to N.L. Webb for collating data on sequences of t-RNA and drawing diagrams of t-RNA molecules. L. D. Hamilton. W. Fuller and E.Reich Nature 1963 198 538. W. Miiller and D. M. Crothers J. Mol. Biol. 1968,35 251. 122 M. J. Waring Biochem. J. 1968,10!3,28P. lZ3 M.Gellert C. E. Smith D. Neville and G. Felsenfeld J. Mol. Biol. 1965 11 445.