VOL. 48, 1962 BIOCHEMISTRY: J0NE.S' .LVD YIRENBERG 2115 QCALITATIVE St-RVEY OF RiVA CODEWORDS* BY OLIVER W. JONES, JR.? urn M~RSITL~LL W. SIRESBERG XYATIONAL HEART INSTITUTE, ?I.ITION.%L INSTITUTES OF HEALTH, BETHESDA, YARYIA.YD Camwunicated by Richard B. Roberts, O&her 29, 1962 h degenerate genetic code was suggested a number of years ago by Cramowt and by Crick.? In such a code, an amino acid may be directed into protein by two or more codewords. Previous work, demonstrated that C%mino acids were directed into protein by synthetic polynucleotides in cell-free E. cnti extracts." and that leucine incorporation was stimulated by either poly c'G,`r I-C'. or L-?l.++ Thus the code was shown to be degenerate with respect to leucine.4a Initially. all of the codewords found contained V. However. assuming a triplet code, the proportion of L compared with other nucleotides in rodewords seemed unusually high, for natural template F&.-L such as viral RX-\, did not contain such a preponderance of I-. To resolve this paradox. a more degenerate code was pro- posed with both non-I: and U cont.aining codewords.5 An alternative hypothesis was advanced by Roberts, who suggested a doublet code, for in such a code the proportions of nucleotides would be within the range found in viral RSA.`. 8 The existence of non-r codewords was suggested when poly AC preparations were found to direct small amounts of proline and threonine into protek9* ?`) Recently, in a careful study, Bretscher and Grunberg-Manago clearly demonstrated coding by non-U words. lo Several poly -K! preparations were reported to code well for proline, threonine, histidine, and glutamine. This work indicated that other 2116 BIOCHEMISTRY: JOh`ES Ah-D NIREh-BERG PROC. x. A. s. non-l' polynucleotidee might have template activities. In this communication, further qualitative analysis of coding by such polynucleotides will be reported. Mate&z& and M-.-The preparation of E. coli extracts (preincubated, DSAase-treawd L&?O fractions). and the techniques used in aashing, plating, and counting protein precipitatee have been rep~ti.~. 6 Polyribonucleotider were synthesized according to the method of Singer and GIIBB." The base-ratio of each polynucleotide preparation was determined by incubation in 0.4 .T KOH at 25" for 18 hours. Under these conditions. little deamination occurred.lz Such mild conditions were not sufficient to hydrolyze certain polymers: however. in such cases inrubation in 0.3 .Y K?)H at 37" for 18 hours resulled in complete bydrolysis.`J Mononucleotidr products u-ere separated either by paper elertrophorek IFVhatman 30. 3 MM paper. O.OS+ ammonium formate. pH 3.7 ' or bv dtscending paper chromatigrapby (Whatman ."io. 3 M%f pawr and s sokent s,wten! conkming 0.1 M sodium phosphate, pH 7.0. and 3 M ammonium sulfate!. Two per rent or greater contamination of pol,vnurleotides bx IY would have been detH%pd. .Vo contamination t)!. U was found. Reartinn mixtures used t.o determine Cl'-amino acid incorporation int,(> protein contained thr following components: 0.1 M Tris (hpdrox!meth~lamincRthane`~ pH i.8: 0.01 M magnesiurr state: 0.05 M h-Cl: 6 X lO-s M mercaptoethanol: 1 X IO-" M ATP; 5 X 10-s M potassium phosphoenolpyruvate; 20 ~g /ml of crystalbne phosphoenolpyruvat kinasr ( California Bicrhem. Corporation); 0.8 X IO-' M C'*(-amino acid (approximately 60.000 rounts /minute ireaction mix- turr); 3.2 X IO-' M each of I!, CL%-amino acids minus the C"-amino acid: 10 M of polvnurlr~+ tide/reartion mixture where swi6ed: and E. &i preincubated p30 extrart,s I 1-2 mg protein reaction mixture). Total volume of each IWLCtitJn mixture was 0.3 ml. &action mixtures were incubated at. 37' for 30 minutes; thus total incorporation rather than rate xvw8p measured. Stimuti oj amino acid inwrp~ration Iry polynueieotides containing jaw Gasc~s: The data of Table 1 demonstrate that synthetic polynucleotides containing four bases stimulate the incorporation of a large number of amino acids into protein. In TABLE 1 .!h-IN~LATION OF kXN0 -4CID k!ORPORATlON BY POLI' UGa4C Pdynucleotidc UGAC 1CGAC UGAC UGAC UGAC IiGAC I- 56 IT.56 P 3 I? 23 r 27 I- 27 2 32 26 2 1: 45 2 4: G 21 G 22 *ratio m&d pi3 can c 2 c c i? 524 A 22 2 `2: cm c 9 Dtai&Bti0Il C'cmnino acid Alanine Arginiqe $J$c=Y;s Glvcine Hikidine laoleucine Leuhne k$%&ine Ph&ahnine Serine Threonine Typtophw Tvrosine &line Total -4p231 110 69 10 ii: 20 68 168 10 1s; SO 1 T!l 15 23 1: I .Oi5 -4~232 A~233 Ap234 JU2&8 JuZ510 Incorporation above control (4,u~molesl* 127 270 Fii 152 31 5 212 99 5; 2 12 9 IO 9 25 14 11 ,. 663 25 12 30: 90 ii 13 0 i 0 54 l!l 78 44 0 F1 16 6E 67 11 70 1: A `0 13 5.@36 867 2.651 5z 10; VOL. 18, 1962 BIOCHEMISTRY: JONES .-LVD .VIRlLVBERG 2117 the last column is given the basal level of C1%mino acid incorporation obtained in the absence of polynucleotide: other figures refer to the net increase above basal incorporation due to addition of polynucleotide. The base-ratios of the poIynucleo- tides vary widely. The hfth polynucleotide (Ju-258) contains approximately equal proportions of U, G, A, and C, whereas the other polynucleotides contain predomi- nant amounts of two or three nucleotides. -411 of the polynucleotides were active in directing amino acid incorporation except polynucleotide Ju-2510. Although 10 erg of polynucleotide were added to each reaction mixture, the total amount of CW amino acid directed into protein by each polynucleotide varied more than N-fold. As we have shown previously, the template activity of polynucieotides is dependent upon factors other than nucleotide sequence. For example, large polymersof-chain length greater than 100 units are considerably more active than shorter ones." Single-stranded polynucleotides are active, whereas double or triple stranded poly- mers are not.J In addition, randomly mixed copolymers which have a high degre of secondary structure are inactive in coding.= In particular, polymers containing much G have little activity, possibly because of G-G interactions. Thus, the relative inactivity of the last poly UGAC preparation (Ju-2510) should not be ascribed necessarily to the presence of a high proportion of nonsense nucleotide sequences. Such considerations make it diflicult to compare with validity the relative abilities of different polynucleotides to code for the same amino acid; thus, `such comparisons shovld be made ,urith caution. The fact that polynucleotides con- taining four bases coded so well for so many amino acids strongly suggests that most nucleotide sequences can be read. In addition, a high proportion of U clearly was not required for messenger Rs-1 activity. St-imdat&m of am&o acid i~wn-poration by pdy ACG: The coding activities of polymers which did not contain I' are given in Table 2. Base-ratio analysis of each poly ACG preparation failed to detect contamination by U. Poly ACG preparations TABLE 2 %&WLATION OF h.VINO kID hCOFIPfXWl'ION BY POLY ACG Polvnucleotide MY3 ACC ACC ACG Designation J251 M 76 M 75 x 74 C~Qmino acid Incorporation shove control (4.rmoIes)* Alanine 123 45 56 85 8 Arginine 1'28 30 40 7.4 Aspartic acids .- .f& 0" 0 34 1: Glutamic acid" i: .x3 21 Glycine 5 Histidine 71 sit `LI 0 9 9: 1:; 5 Lsoleucine 0 0 ii 20 Leucine 10 40 Lysine 84 5 x 23 Metbionine 1 320 : 0 P$;;iaIanine 14; 1% 1 42 9 1: : .Serine is? 256 24 30 .5.j 45 Tbreonine 11 1R 11 ;ww&m V"Sne 0 1 0 4 1: x 4: 18 1 - - 45.i 1 Total 2.;m xii 33; 22 o ~\ur.Molea reprenata the difference between C"-emino acid incorporation intn protein in the presence and sbaence of polynucleotides. &say procedures we described tinder .Katsn& and Methods. 2118 BIOCHEMISTRY: JOSES -4?i-D NREKBERG hoc. K. A S. stimulakd the incorporation of many amino acids t&ed, including alanine. argininc,. glutamic acid, lysine? proline, and threonine. Such high incorporations of glutamic acid, lysine, and ohreonine were not observed previously. Jh -4 number of amino acids did not appear to be coded by any -4CG preparations, wbicb suggested that I- ma! be an absolute requirement in coding for some amino acids. Stim.ulation 0-f amino acid incurporatitm by polynwleotides cont.aininij two bases: The polynucleotides in Table 3 are listed in order of decreasing C content. In accord with the findings of Bretscher and Grunberg-Manago.`" poly -4C s:imulaird incorporatiou of proline. threonine. and histidine. In a&$ion. poly -4C n-as found to direct aspartic arid. glutamir acid, and l+ne int,o pro:ein.3E Poly _4C (J-104 : contained 9; per rent C. yet actively directed proline into protein. Thus. it appear' that one codeword for proline may wntain only C. Relatively large amounts of lysine were directed int,o protein by ,4C J-109 and J-108 whic*h contained 67 and 80 per cent ,4 respectively. Tbus. the lysine codcu-ord may contain only 9. The data of Table 4 demonstrate the e5ects of poly CG and AG preparat,iona in directing amino acids int.o protein. The first three CG poleymers comained high proportions of C and directed alanine, arginine, and proline into prot>ein. The lasr poly CG preparation (F-135) contained 91 per cent G, and xas inactive s template RX;A. Poly ,4G directed incorporation of glutamic acid and Iysine into protein. Quantitahe aspeck oj data: A comparative study of polynucleotides of I-arying base-ratios is helpful in evaluating amino acid incorporation data. for relative amino acid incorporations can be easily correlated with changes in base-ratio. Occasional inexmsistencies and the significance of minor incorporations become apparent. Isotope dilution experiments were performed routinely to detect, the possible presence of radioactive impurities in C1kmino acids. The presence of Cl*-impuri- ties seemed unlikely, for incorporation of a Cl'-amino acid was reduced sharply if the reaction mixture contained both a Cl*-amino acid (0.05 pmole) and the same TABLE 3 ~TIMTTLATIOX OF AMINO ACIII INCORPOFLATIO!~ BT POLT At? Polynuclechide Base-ratio moleA per cent Designation C~+muno acid Alanine Arginine hpartic acid* Glutamic acid- Glpcine Hihiine em;>e Lysine Methiontie Ph'h;;~alanine Swine Threonine p.w.w~~" V'aline Total AC 5104 8 cl 4 5 0 i x 62: ii! 23 14 0 723 AC L!4 6103 1 4 0 19 16 "0 1: 10 0 1,132 i!' x 1.27: AC AC AC AC A 12 A 30 A Si A 80 cf43 c 70 c 33 c 20 5102 JlOl JIOR JlO8 Incorporatmr. above rontroi i ~Irpmolrs~* 0 0 0 0 75 170 176 105 1 f (: :, i 0 0 793 1.64: 1.616 70; Mmw polynuclrot!df eonfroi 11 12 24 15 6 g 3 5 10 I1 9 46 VOL. 48, 1962 BZOCHEMZSTRY: JOSES A-h-D YZREMBERG a119 . Designation kiMI41 41 71 C'camino acld Alanine 30 20 Arginine 39 16 Aspartic acid S 0 Glutamic acid= 8 0 Glvcine - Hiitidine ii 1: Isoleucine Leuciue Lvaine kiethionine Phenylnlanine Proline Seriue Threonine pY!Jw!J;feh~" V&e Total Fl20 F 135 J 106 .I 107 Incorpwation above control i &mmle3~+ t; 0 10 0 s 0 0" 0 ; 44 12 10 - 0 2 ; 0 0 0 39 7 13 1-I 8 42 4?l 14 4 332 * ~&f&s represents the difference berween C'+mino acid incorporation into prutein in the presence and absence of polynuciwtides. Detads of tbe assay procedures are described under Matmds and Methoda. Cl?-amino acid (1.0 pmole). The purity of each Cl*-amino acid also was deter- mined by paper electrophoresis followed byradioautographyas described previouslyy5 Limiting amounts of polynucleotides were added to reaction mixtures and total ammo acid incorporations were measured rather than rates of ammo acid incorpora- tions. E. cola' extracts contain nucleases which rapidly degrade synthetic poly- nucleotides and the nuclease content may vary from one preincubated S-30 prepa- ration to another. Since many difjFerent enzyme extracts were used in this study, the data are not useful for quantitatioe analyses. Comparisons between theoretical fretiuencies of triplets, etc., in polynucleotides and relative ammo acid incorpora- tions have not been presented because the data do not permit such calculations to be made with ~curncy. The data demonstrate only qualitative aspects of the code: that is, nucleotide caompositions of words and the degree of code degeneracy. SlLmmary oj Incorpcrration D&z.-Table .5 summarizes all of the coding data previously pubiishedR-6, S- bn and rhose obtained in this study. Only polynucleo- C'%mllno WI11 Phenvldxune ProI& Lysine Threonine se l-he Vdine Leucine Uycine CpthW Glutamic acid 31 C"-ammo uxd drimulsced by pdy-* Isoleucine Clc 8 ) Trvptoph;m CG(ti! TqToslne Cli 9 j Irginine CG( 15) Msthionine CAC( I ) Hi&dine hC( 101 Alanine CG(l1; Aspartic acids AC!8) * Polymen used for these caieuistions soutain the mitimd number of bases ~leceeawy to direct the Cl'-amino acid8 mm pl0tein. t Xumberg ;a parentheses refer to: ,+mules of amino acid incorporsted X lOO/Sum of ineurporstion of all amino ecd~ af the optimal bree ratio assayed. 2120 BIOCHEMISTRI': JO.?`ES A.VD h-IRE-+-BERG PROC. K. -4. \ tides containing t,he minimum number of bases capable of stimulating an amint, acid into protein are given in Table 5. The roding of proline by poly C and I>-& by poly -4 was suggested by the poly A4C experiment.> presented in Table 2. l'h,. fact that poly C and poly A code so weakly may be due to inhibitory effect> 0; secondary structure. In solut,ion at avid pH. poly A is double-stranded.!Y 3 ant! at neutral pH. both poly A and pofy C have parGaIly ordered strurl ure.?; -4 surprising wnclusion drawn from this summary is that almost every amino a~+i tested could be coded by a polymer cxont,aining only two bases. Mrthioninr c*ouitf be coded only by poly I-G.4 as previously reportrd,$. lc but sinw the amoun: O: methionine dirwwd into protein was small. this c*od&wd remain: qurstionat)lr~. Assuming a triplet (*ode. a summary of codewords estimat,ed t,hus far is prr+nl~~ll in Table 6. Prel-iously. poly I-CG was found t,o direct alanine and arginint in: I; AGU_ cc.4 protein and codewords containing U, C, and G were proposed for these amino aci&.43 5, 1:. 18 The observed frequencies of incorporations5 suggest coding of alanine and arginine by either I-CG or CCG. but not b\- both codeword-, ln addition, the data of Table 4 show that poly CG c&es for alanine and argininc,. thus, codewords corresponding to t,hese amino acids do not appear to contain V. S~IIW it is not possible at this time t,~ d&@&h between triplet and doublet codes. etc., the assignments in Table 6 represent current approximations of code- WONTS. It seenw probable that additional codewords w-ill be found. Di~sitm.-Codeword specijkity in protein synthesis: The term degencrar!! refers to the phenomenon whereby one amino acid is coded by two or man' ~d(`- words. This term is inadequate when applied to the mechanism of coding. for it does not indicate codeword speci$city -4 degenerate code may halve high or 1~ specificity depending upon the fidelity of protein synthesis. In most cases lb(' fidelity of protein s.ynthesie in viva appears to be high, and amino acid rrplacvmc'n:` other than those due to muiat.ion, have not been found. H~ww~. ahho& th(' amino acid sequenrk analyses would reveal mist.aS;es at one sit.t occurring v%tI i1 VOL. 18, 1962 BIOCHEMISTRY: JONES .4AVD YIRE.VBERG 2131 frequency higher than I or 2 per cent, they would not reveal occasional mistakes occurring at different sites. Thus, occasional coding errors of 1 or 2 per cent, dii tributed at random over entire protein molecules, might not be detected. In the in L&O system, coding during protein synthesis displays striking specihc- ity.22 In Table 3, for example, poly AC preparations do not direct the incorporation into protein of alanine, arginine, glycine, isoleucine, leucine, methionine, phenylala- nine, tryptophan, tyrosine, or valine. The specificity of coding by poly CG and AG preparations in Table 4 is equally apparent. Such negative data clearly demon- strate the very high fidelity of codeword recognition during protein synthesis. The codewords corresponding to both leucine and valine contain U and G.&-" -4lthough the nucleot.ide content of these codewords is identical, each word was shown to code only for the appropriate amino arid.?? Thus. nucleotide sequence a~ well as chemical structure confers specificity upon codewords. However, one unclarified example of ambiguity has been found. Poly U directs about 3-S per cent as much leucine into protein as phenylalanine.S Bretscher and Grunberg-Manago also have reported this phenomenoni In the absence of phenylalanine poly G coded for leucine about 50 per cent as well as it would code for phenylalanine.`23 E$iciency of synthetic RLVA4 in coding: In spite of the previously mentioned di&ulties in comparing template activities of RNA preparations with Merent chain lengths and degrees of secondary structure, it seems clear that synthetic poiynucleotides containing 4, 3, or 2 bases code as well in this system as natural template RXA obtained from viruses.3, 24-s The efficiency in coding displayed by synthetic polynucleotides suggests that most nucleotide sequences direct amino acids into protein, and that few nonsense nucleotide sequences are present. Although aiternative explanations of coding efficiency, such as nonrandom polynucleotides or nonsequential reading of template RSA, may be considered, such efficiency cannot be ascribed simply to random error in directing amino acids into protein, for amino acids are coded with marked specificity. Considerations such as these may be used to approximate the coding ratio. In a doublet code only 16 base permutations are possible. Thus, the information content would be insufficient to code specifically for all amino acids. Triplet and quadruplet codes would contain 64 and 2.56 codewords respectively. Since almost every amino acid ixsted was found to be coded by polynucleotides con- taming only 2 bases, specific and efficient coding by quadruplet words would not seem likely. Tue data sugge 0% either coding of all amino acids by tripiers or coding of some by triplets and ers by doublets (mixed doublet-triplet code). Recently Weisblum, Benzer, and Hohey" have established a molecular basis of degeneracy by demonstrating that multiple species of transfer RX-1 recognize different codewords with specificity. Multiple peaks of transfer RX-1 correspond- ing to at least four amino acids have been found independently by Holley et al.,% Sueoka el a1.,?g and Doctor et aLa If a triplet code is assumed, each cell would require almost 6-l transfer RX-\ species. Alternatives which do not require so many transfer RNA species deserve consideration. For example, Donohue and others have described many models other than Watson-Crick pairing. 71 The demonstrated interaction between poly -1 and poly I,"? and the type of base-pairing suggested by Hoogsteen33 also might be 2122 BIOCHEMISTHY: JO.?-ES 9A-D a~?REA-BERG PROC. ?;. -4. 4 cited. Theories which require recognition of either the 2- or &substituents of basesS4 are not supported by the demonstration that hypoxant,hine functions in codeword< like G.15. 35 L The Z&amino group of G does not appear t,o be required. A triplet code ma\- be construct,ed xherein correct hydrogen bonding brtnwn two out of three nucleotide pairs may, in some cases. suffiw for coding. 0xrw.t pairing of a base at one position in the triplet sometimes may be optional. II should be noted that a triplet code of t,hie type in some respects would bear a super- ficial resemblance t'c! a doublet code and would he in aword n-it h all of t hr data -4ny theory concerning t#he physical hasi? of the vode$nust a; t,empt IO explair~ the following esperiment,ally obtained da.ta : ; 1) high coding, efficiency by sync hp; 11' polvnucleotides; (~2) marked codeword specificity : (3, degvnerat,e codewords. (4)`the P-amino group of G is not essential for proper coding: (5) RX-$ with a high degree of secondary struct#ure has little ability t.o code: and ;6,1 almost al! amincb acids t,ested can be poded bg polynucleotides containing only two bases. Summary.-Synthetic polynucleot,ides containing 4. 3, or 2 bases have twn found t,o direct amino acids into prot,ein with high efIiriency and specificity. Milan>. additional RX.4 codewords which do not contain uridylic acid haw been found Almost all amino acids could be coded by polynucleotidee containing on!y 2 haw- These results have been discussed in terms of the general nature of the code. It is a pleasure to thank Lynn Hurwitz, Linda Greenhowe. and Glenn5 Baile? for vahwblr technical assistsme. * Presented at the Symposium on l~form&onul Mac~omol~rules. !Cew Brunanick. Xl`ew J~;w Sept. 1962 (proceedin@ in press:. t The following abbreviations are used: A, adenylir acid; G: guanylir acid; C. cytidylic arid V, uridylic acid. I Gamow, G., Xature, 173, 318 (19541. * Crick, F. H. C., in Stwcture and Function o-f Genetic Elements. Brookhaven S;!-rnwi:i iri Biology, No. 12.35 (1959). * Nirenberg, M. R.. and J. H. Matthaei, these PROCEEDINGS. 47, 1588 t 1961:. ' Martin. R. G., J. H. Matthaei, 0. W. Jones, and M. W. ?&&erg. Bznchem Rzo,vh?/,+ Kr, (lomm., 6, 410 (19621. 6 Matthaei, J. H., 0. a. Jones, R. G. Martin and M. IV. ?iirenbwg. these PROCEE~~~~(:~~ 46. 666 (1962). 6 Speyer. J. F., P. L+ngyel. C. Basilio, and S. Ochna. thee PROCEEDINGS, 48, 63 ( IW `Roberts, R. B.. thew PROCEEDINGS. 48, 8!X (1962:. 8&d., 1245 (1962,. SJones, 0. W., and R. G. Martin, Pad. Proc., 21, 414 (1962). ~Bretscher, M. S., and M. Grunberg-Manago. .\-ature, 195, 283 (1962). I1 Singer, M. F., and J. K. Guss, J. Biol. Chem., 237, 182 ! 19621. I* We thank Dr. M. Grunberg-Manago for this protocol. I* Davidson, J. X., and R. M. S. Sm&rj Rio&m. J.. 52, 594 :`I952 I. I4 Unpublished observations. U Singer, M. F.. 0. W. Jones, and M. VI-. Kirenberg. unpublished nbsen-ationf. I6 Lengvel. P., J. F. Spever, and S. Ochoa. these PROCEEDIKGS, 47, 1936 I 19611. Ii Leng&l, P.. J. F. Sp&er, C. Baeilio. and S. Ochoa, these PROCEEDINGS. 48, 2% ( INil 18 Speyer, J. F., P. Lengyel. C. Basilio. and S. Ochoa, these PROCEEDINGS. 48, 44 1 ( lW;L' I9 Fresro, J. R., and P. Dot,?, J. -4m. rhem. Sot., 79, 3928 [1957:. 9 Rich, -4., D. R. Davies. F. H. C. Crick, and J. D. Watson. J. Mol. Bzoi.. 3, 71 1~' 81 Fresco, J. R.. Trans. .V. J-. dead. SC%., Se&~ II, 21, 653 (1959:. ** ?;irenbwg. M. H.. ,I. H. Matthaei, 0. W. Jones, S. H. Barondek: and R. G %faVt,i'r "`! Proc., in press. VOL. 48 196" , - BfOf`HEMfSTRY: K.lPL-4-V .i.VD Cd H.V 21`23 ?a Nireoberg, M. W , J. H. Mstthaei, and 0. W. Jones, unpublished observations. ?+ Ofengand, J., aand R. Haselkorn, Biocha. Biophys. Res. Corm., 6, 469 ( 19tZ ). a Tsugita, A., H. Fraenkel-Conrat, M. W. Sirenberg, and J. H. .Matthaei, these PROCEEDINGS, 48, 34-16 ( 1962). 1 Nathans. D., G. Yotsni. J. H. Schwartz, and N D. Zinder, these PROCEEDIXGS, 48, l-U4 19623. n Weisblum, B., S. Benaer, and R. W. Halley, these PROCEEDINOS, 48, 1449 (196&Z). a Halley. W. H.. B. P. Doctor, S. H. Merrill, and F. M. Saad, Biochim. Biophys. dctu, 35, 252 (1959). `!g Sueoka, X., and T. Yamane, these PROCEEDINGS, 48, 1154 (196Z2!. m Doctor, B. P.. J. hpgar. and R. W. Halley, b. Bid. C&m., 236, 1117 ( 1961). 31 Donohue, J., these PROCEEDINGS, 42, 60 (1956). J* Rich, A., Suture, 181, $21 I' 1958). 23 Hwgsteen. K.. .1&z C'rysf.. 12, 82`2 i 1959). YJ' Wtwse. C. R.. .Vntrrre. 194, 111-L I 196Y2!. s Basilio. C.. A. Wahha. P. Lengyel, J. F. Bpeyer, and 8. Ochoa, these PROCEEDINGS, 48, 613 1962). a Poly .IC is reported to code for glutamine.*' We have been unable to obtain Cl'-glutamine :hnd C'Qqaragine. Therefore, we cannot ditrerentiate between codewords corresponding to ,+ach frc?t? acid and its respective amide.