Reprinted tram the Proceedings of the NATIONAL ACADEMY OP SCIENCEX Vol. 47, No. 10, BP. 1.588-1602. October, 1961. 9"HE' DEPENDENCE OF CELL- FREE PROTEIN SYNTHESIS IN E. COLI lrPON Nil TURALLY OCCURRING OR SYNTHETIC POLYRIRONUCLEOTIDES BY MARSHALL W. ~IHENBEE~G ANI) J. HEINHICH MATTHAEI* NATIONAL ISSTITUTES OF HEALTH, BETHESDA, MARYLAND Comnunicated by Joseph E. Smadel, August 3, 1961 A stable cell-free system has been obtained from E. coli which incorporates C14-valine int)o protein at, a rapid rate. It was shown that this apparent protein synthesis was energy-dependent, was st'imulated by a mixture of L-amino acids, and was markedly inhibited by RNAane, puromycin, and chloramphenico1.l The present, communication describes a novel characteristic of the system, that is, a requirement for templat'e RNA? needed for amino acid incorporation even in the 1590 BIOCHEMISTRY-: XIRENBERG A.VD MATTHAEI PROC. N. A. S. RNAase (Worthington Biochemical Company) at, 35" for 60 min. RiXAase was destroyed by four phenol extrart#ions performed as given above. After the last phenol extraction, the samples were dialyzed against standard buffer minus mercaptoethanol. RNA samples were treated with trypsin by incubation with 20 rg per ml of twice recrystallized trypsin (Worthington Biochemical Company) at 35' for 60 min. The solution was treated four times with phenol and was dialyzed in the same manner. The radioactive amino acids used, their source, and t,heir respective specific activities are as follow-s: U-C14-glycinr, U-C14-L-isoleucine, U-C14-L-tyrosine, LT-C1l-L-leucine, U-C14-L-proline, L-histidine-2(ring)-C14, U-C'4-L-phenklalanine, U-Cl4-L-threonine, L-methionine (methyl-C'4) , U-C~4-L-nrgininej and U-C14-L-lysine obtained from Nurlear-Chicago Corporation, 5.8, 6.2, 5.95, 6.25, 10.5, 3.96, 10.3, 3.9, 6.5, 5.8, 8.3 mC/mIVI, respectively; C14-L-aspartic acid, C14-Lglu- tamic acid, C14-L-a1anine, obtained from Volk, 1.04, 1.18,0.75 m C/m&l, respectively; C)-L-trypto- phan-3 Cl", obtained from New England Nuclear Corporation, 2.5 mC/mM; SZS-L-cystine obtained from the Ahhott Lahorat,orics . , 2.4 mC/mM; U-C14-L-serine obtained from the Kuclcnr-Chicago Corporation, 0.2 mC/mM. Other materials and methods used in this study- are described in the accompanying paper.' All assays were performed in duplicate. Results.---Stimulation by ribosomal RNA: In bhe previous paper,l it was shown that DWAase markedly decreased amino acid incorporation in this system after 20 min. For t.he purpose of this invest,igat,ion, 30,000 X g supernatant fluid frac- tions previously incubatfed with DXAase and other components of the reaction m&m-es (Incubated-S-30 fractions) were used for many of the experiments. Figure 1 shows that incorporation of C14-L-valine into protein by Incubated- S-30 frac%ion was stimulated by t,he addition of purified E. coli soluble RKA. Maximal stimulation was obt,ained with approximately 1 mg soluble RSA. III some experiment's, increasing t'he concentration S-fold did not further st#imulate t,he q&em. Soluble RXA was added to all reaction mixtures unless ot,herwise specified. Figure 2 demonstrates that E. coli ribosomal RKA preparations markedly stimu- 2oo r- ~-- --T-~---- 7----T- I 1 1 I 0 05 10 1.5 20 2.5 30 MG. SOLUBLE RNA FIG 1:-- Stimulation of amino acid incor- porat,ion int,o protein by E. coli soluble RNA. Composition of reaction mistures is specified in Table 1. Samples were incubated at 35" for 20 min. Reaction mixtures contained 4.4 mg. of Incubated-S-30 protein. FIG. 2.-St,imulation of amino acid incor- poration into protein by E. coli ribosomal RNA in the presence of soluble RNA. Com- position of reaction mixtures is specified in Table 1. Samples were incubated at 35" fo1 20 min. Reaction mixt,ures contained 4.4 mg of Incubated-S-30 protein and 1.0 mg E. coli soluble RKA. 10 20 30 40 50 60 70 80 90 MINUTES FIc:. X.--Uependencr of C'4-I,-valiue inwrpwatioll iuto protriu upon ribosomul 1tSh. `l'ht~ composition of the react,ion mixtures and thr incubation wutlitions :tw prrwntrtl in Tahlr I Reaction misturrs contained 0.98 my of E. co/i soluk)le RNA :tud 4.4 mg of Inc~ut):~tc~ti-S-:~~-l)~~t(~ill. lated incorporation of C14-valine into protein even though maximully stimulating concentrations of soluble RKA were present in t,hc reaction miktures. A linrar relationship between t,he concent&ion of ribonomul K SA and C14-valine incorporu- t,ion into protein was obtained when low ~mcent rations of ribosomul RKA were used. Increasing the soluble RNA concentration rip to :&fold did noi rt~pl:~~ tht> effect observed when ribosonial 1iS.A \vas aticird. The effect of ribusomul KKA ill st~irnulatiug illc,olpolatioli of ("`-valinc illto prot,ein is prc,srlltcd in more detail in Figure 3. In t'hr absence of ribosomal RS.l. incorporation of Cl"-saline into protein by thca inctlbatcad-S-80 fruchu n-as quite low when compared with S-:30 (not, incubated before, storagv at, - 15") a11d stopped almostj complet,ely after 30 min. At. low concent,rations of ribosoniul KS-A, mxsi- mum amino acid incorporation into protein was proportional to the amount, of ribosomal RNA added, suggesting stoichiomctric rather t(han cat,alytk a&on of ribosomal RNA. Totma incorporation of C14-vsline int'o prot,ein was incrcuscd more t,hnn 3-fold by riboaomal RKA in this experimrlit even in thr prescncc of maximally stimulat~ing conc+entrations of soluble RNA. Ribosomal RNA mlq be added at any time during t,he course of the reaction. and, aftrr frlrthrr in(*llh;l.tioll, ali increase in incorporation of CIJ-valinr into prokin will rrsult. Characteristics of arni,rw ucid incorporation stimulated by ribosomal RNA: In Table 1 are prcsmkd the rharactjerist,ics of C14-T,-valinr incorporzat,ion illto protein stimulated by t,hP addit,iotl of rihosomal RNA. Amino acid incorporation was strongly inhibited hy 0.15 pmoles of rhloramphenicol and 0.20 ~moles,/ml reaction mixture of purnmyrin. Flu-thwmorc, the incorporation was complet,cly dependent llpon the addition of AT1 and an ATP-gwerating system and was totally inhibited by IO Kg/ml R?;Aascl. l~~(]~tival~nt amountIs of DYAase had no effect upon the inwrporation stinull:~tcd by the addition of rihosomal RNA. Placing a ribosomal li?;X prcparatiou in a boiling water bath for 10 min did not) destroy its Cl"-valinc iil(~ol,poratiolI acLt,i\,it,y; instead, a slight incwaw in activity was consistrntly ob- ~crwd. HOWYW, wh(an thwc RX.4 prrparat#ions were placed in a boiling water bath, a cwpiolw, whit,cL prwipit,ate rwultrd. Upon cooling t,h(h suspension in an ice bath, the prwipitatc immrdiatply dissolvrd. The data of Table I also demonstrate that, t,he incorporation of amino acids into protein in the presence of ribosomal RNA was further stimulated by the addit,ion of a mixt8ure of 20 I,-amino acids, suggesting cell-free protein synthesis. C- and S-trrminal analyses of the rihosomal RNA-dependent product, of t&he rcac:t,ion mere perfomwd with carboxypeptidase and I-fluoro-Z,&dinit,robenzene rrspwti\~cl~~ (Dr. Frank Tict,zr kindly performed these analyses). Four per cent of thr radioactivity was rclrascd from the C-t,crminal end and 1% was associat,ed \vith the S-trrminal end. The remainder of the Cl"-label was int,rrnal. Similar rcwtlts wcr(' obtaiwd whw rractions were performed using S-30 enzyme fractions which had not, twcw troa,tcd with DKAasc. Protein preripitatrs isolat,ed from re- action mixtlu'cs after incubation were compltt8ely hydrolyzed wit,h I-ICI, and the Cl"-label incmorporatcd into protein was drmon&ated to be valine by paper chro- matography. Many of t'hc experiment,s presented in this paper were performed with enzyme frn.at,ions prr>parcd with I)FAase added tjo reducr their viscosity. Ribosomal RNA also stimulated Cl*-valine incorporation when wzyme extracts prepared in the absence of DNAase were used. To be effective in stimulating amino acid incorporat8ion into protein, the ribo- somal RNA required the presence of washed ribosomes. The data of Table 2 show TABLE 2 THE INEFFECTIVENESS UF RIB~s~MAL RKA IN STIMULATING C1CL-V.4~1~~ I~~Y~IWORATIOX INTO PROTEIN IN THE PRESENCE OF RIBOSOMES OR 105,000 X g SVFERNATANT SOI,~.TI~SS AI.ONE Additions Counts/nlin Complete 51 l` + 2.1 mg Ribosomal RNA 202 I, - Ribosomrs " " `, " - Ribosomes + 2.1 mg Ribosomal RNA - Supernatant solution - Gupernatant solution + 2.1 mg Rihosomal RNA Deproteinised at zero time that both ribosomes and 105,000 X g supernatant solut'ion were necessary fat ribosomal RNA-dependent amino acid incorporation. ?;o incorporation of amine acids into protein occurred when the 105,000 X g suprrnatant solution alone waf: added t,o ribosomal RNA preparations, demonstrat)ing that, ribosomal RKA prepara- tions were not contaminated with intact ribosomes. This conclusion also was substantiated by showing that the acbivities of ribosomal RXA prcxparat,ions were not destroyed by boiling, although t,he activit,ies of t,he ribosonws wrrr drstroyed by such treatment. The effect of ribosomal RXA upon the incorporation of seven differcut amino acids is present#ed in Table 3. The addition of rihosomal R?;A incwascd th(x incorporation of every amino arid test)ed. The effect shown by ribosomal RXA was not observed when ot#hcr polyanions were used, such as polyadenylic acid, highly polymerized salmon sperm I>SA, or a high-molecular-weightj polymer of glucose carboxylic acid (Table 4). Pretreat- ment of ribosomal RNA with trypsin did not affect its biological activity. How- ever, treatment of t'he ribosomal RNA with eit,her RKAase or alkali rcsult,ed in a complete loss of stimulating activity. The act,ive principle, therefore, appears to hr RiSA. The sedimcntat.ion charac:teri&cs of thr ribosomal RNA preparations were exwlitwd in the Spinco Model E ultraccnt,rifuge (Fig. 49). Particles having the charactrristics of S-30, S-50, or S-70 ribosomes were not observed in these prepara- tions. The SE of the first peak was 23, that, of the second peak 16, and that of the third, small peak, 4. Pret'reatment with trypsin did not affect, the S$ values of the peaks appreciably (Fig. 4C) ; however, treatment, with RNAase completely destroyed the peaks (Fig. 4B), confirming the ancillary evidence which had suggest,ed t,hat the major component was high-molecular-weight RNA. Preliminary attempts at, fractionat#ion of the ribosomal RNA were performed by means of densit,y-gradient centrifugation employing a linear sucrose gradient. The result,s of one such experiment' are presented in Figure 5. Amino acid in- corporation activit#y of the RNA did not follow absorbancy at, 260 rnp; instead, the activity seemed to be concentrated around fraction No. 5, which was approxi- mately one-t.hird of the distance from the hott,om of the tube. These result,s again + Ribosomal KKA + Ribosornal RNA + Rihosomxl RNA + Rihosornwl RN.4 + Hihor;omal RNA + Ribosomal RNA Ikproteinizrd at zero time VOL. 47. 1961 HIOCHh'MIS7'RY: XTRENHERG ANI) MA'I'THAEI 1595 TABLE 5 STIMULATION OF AMINO Acm INCORPORATION BY RNA FRaCTIoNa PREPARED FROM I~TERE~-T SPECIE5 L .Additions Cormts/min/mg lxotein Eone + 0.5 mg E. coli ribosomal RNA :i + 0.5 mg Yeast ribosomal RNA 130 + 0.5 mg Tobacco mosaic virus RNA 872 + 0.5 mg Ehrlich ascites t,unlor microsomsl RI\;A 65 The cornl,onents of the reaction mivtrurs and thP incubation ronditihxtrd at .iT," for ti0 Illin. TARI,b; 8 I Phenylalxnine 2 (;lycine, alanine, serine! aspartic arid, glntamlc acid 3 I,eucine, isoleucine, threonine, methlonine, arginine, histidme, lysinc, tyrosine, t,ryptophnn, proline, valine 4 S3Gyst,eine Deprot,einixed at zero t.ime NO& + 10 pg polyuridylic acid Deprot,einized at zero time None + 10 pg polyuridylic acid Deproteinised at zero time SOiF + 10 pg polyuridylic acid Depruteirrizetl at, zero t,imr None + 10 pg pol>wridplic arid 25 68 38 ) a00 17 20 33 73 2Tfi 9; 113 Components of the rettctiun Inixtores arr presented in Table 1. 0.015 pM of each labeled amino acid -`as used. The unlabeled amino acid misturr was olrlitted. in the Melhods and Moferials section. The specific activities of the labeled amino acids are present reaction mixture. 2.3 mg of protein of preinrubatrd S-30 enzyme fraction were added to each All samples were incubated at 35' for 30 min. TABLE 9 COMPARTSOS OF CHARACTERISTICS OF PRODUCT OF REACTIOX AND POI.Y-L-PHENVLALANINE Treatment Product of reaction Poly-L-phenylalaninP 6 A' HCI for 8 hours at 100" 12 .V HCl for 18 hours at 120~1:~O" Partially hydrolyzed Estrwtiou \\ith :3X , ITl3r in glacial Complrtf4y h~drol~wtl I'arti:tll,y hydrol,wcd CompMcly hytirol,wtl arctic :wi(l Soluhlr SOlUl,l~~ l':stractjion* with the following sol- vcnt.s: H?O, lwnzcne, nitrohcnzcne, chloroform, ?;,i\;-dimethylform- amide, ethanol, petroleum ether, concentrated phosphoric arid, gla- cial acetic arid, dioxanr, phenol, acetone, ethyl acetate! pyridinr, acet,ophenone, formic acid Insoluble Insoluble * The product was said to be insoluble if 10.002 gm of product was soluble in 100 ml of solvent at 24O. Eutrar- tions were performed by adding 0.5 mg of authentic jloly-L-phen~lalanine and the ClLprodnct of a reaction rnixtnw (1800 counts/min) to 5.0 ml of solvent. wntrifugui. The suspensions were viyorously shaken for 30 min at 21' and were The precipitntw wew plated and t,heir radioactivity was determinpd. HBr in glacial acetic wid.$ The product, of the reaction had the same apparent solubility as authentic poly-L-phenylalanine. The product of the reaction was purified by means of its unusual solubility behavior. Reaction mixtures were de- proteinized after incubation, and precipitated proteins were washed in t,he usual manner according to the method of 8iekevitz.22 Dried protein pellets containing added carrier poly-L-phenylalanine were then rxtract,ed with 33 per cent HBr in glacial acetic acid, and the large arnouut, of insoluble material was discarded. Polyphetlyla,larliltr was then prrc:ipitat,rd from solut,iull by the addiGon of Hz0 and was washed several times with H20. Sevrnty per wnt of the total amount of C'4-L-phellylalallille incorporated into protein due tv thr additive of polyuridylic acid could be rrwwrrd by this procedure. Cornplet,e hydrolysis of the purified rea&on product, with 12N HCl followed by paper electrophoresis** demonstrated that, the react,ion product contained C14-phenylalanine. No other radioactive spots were found. Discussiott .-- -111 t,his il~vrstigut,ion, WC have dcw~oustrat~etl that template RNA is a requirrmt~r~t for wll-fwr amit wid it ~wrporatiol~. .\dtiit,ioll of solut)lr~ KKA could not replace template RKA iu this s~yystem. In addit,ion, t,he density-gradient cent~rifugation rxprrirnrlltjs showed tJhat t,hr ac:t,ive fra&iwls itI the riboaomal RNA pwpawtious srdirurl~tjrcl much faslt~r tjhail sohlblr KXA. lt should 1~ noted t,hat ribosomal RNA is clualitwtively difl'rrrtlt from soluble RNA, siuw bases such a$ pseudouruc*il, mrt~hylwt~rd gu:~IIilws! rtv., fo1111t1 ill solut~l~~ RNA., :LW lrot present iu ribosonial l< NA." `Thr hulk of the H.NA in our ribosomal K&A fractions may be inactive as tem- plates, for t~obacw' mosaic: virus RNA wad 20 t,imes as at+ive in stimulatiug amino acid iIl~`Orp~Jl'at~hl ;ts equivalent amounts of E. coli ribosomal RXA. In addition, preliminary frac:t,iouation of rihosorual .KNA indicat,rd t#hat, only a portion of the total RKA was active. It should tw rmphasixrd that ribosomal RNA could not subst,itutr for ribosomes, indicat,ing that, l~ihJSoillW were not, assenlhled from thr added RNA i7L toto. The function of ribosornal RNA remains an wigma, although at letlst part of the total RKA is t8hought t#u serve as templates for protein synthesis and has been termed "nwssrugrr" R?4C'A." -I4 Alternatively, a pwt of thp RX?;,4 rr~ay be essential for t,he syllthwis of a&w rihowmes from smaller ribosomal partic:les.15-21 Ribosonial l< SAk m:~y hr an aggregat,e of subunit,s which can dissociat,e after i$Yqw" t.rt:ut llle11t~.6 h I'hPuol C&T&CJil of &. f:Oll' l'it)OsoIiw j&h two types of 1iXA molecules with s'," of %:3 and 16 (Fig. 4j, eyuivulrut. 10 Ill(J~W&lI' \vrights of l,OOO,OOO and 560,000, respectiveIy.Y~ II' Thew RNA specirn can be degraded by boiling to products having sedimwtation aoefIicients of 13.1,8,8, and 4.4, correspond- ing to molecular wcighta of 288,000, 144,WO, and 29,000. Although the Sedimenta- Ii011 riistribut,ions of t,he latter preparat,ions suggest, a high degree of honlogeneit,y ;LIUVU~ the tuolecuk,s of each class, thesr obServati(JnS do not eliminate the possi- bility t'hat the subunits are linkrd to one another viu covalent bonds." Yreliminary ctvidrnce illdiwtcs t,hat the subunits n~ay be act,ive in our system, since t,he super- natantI solution obtained after boiling E. coli ribosomal R?;A for 10 min and centri- fugation at 10+5,000 X y for 60 min was active. Examination of boiled ribosomal RNA with the spinco Model E ult,racentrifuge showed a dispersed peak with a sedimentat'ion corficient of 4-8. This may be the same material found in the sucrow density-gradient experiment (using non-boiled RNA preparations), \vherc? a small pwk of activity somrwhsrt hwviw t,hatl Soluhl~ HN.4 was usually noted (Fig. 5). In our system, at' low concentrations of ribosomal RKA, amino acid incorporation into protein was proport,ional to the amount8 of ribosomal RXA added, suggest'ing a stoichiometric rather than a catalytic action of ribosomal RNA. In contrast. soluble RNA has been shown to act in a cat,alytic fashion." The results indicate that polyuridylic acid contSains the information for tShe syn- thesis of a protein having many of the charact'eristics of poly-L-phenylalanine. This synthesis was very similar to the cell-free protein synthesis obtained when naturally-occurring template RNA was added, i.e., both ribosomw and 100,000 X g supernat,ant solutions were required, and the incorporation was inhibited by puromycin or chloramphenicol. One or more uridylic acid residues therefore appear to be the code for phenylalanine. Whether the code is of t,he singlet, triplet, etc., type has not, yet been determined. Polyuridylic acid seerningly functions as a synthetic template or messenger R,KA, and this stable, cell-free E. coli system may well synthesizr uny protein wrrc3pontlillg tt) mwniligful information con- tained in added RS,Y Summwy.--A stable, cell-fwe system has been obtained from h'. coli in which the arnount of incorporat,ion of amino acids into protein was dependent upon the addition of heat-stable templat,e RNA preparations. Soluble R?CA could not replace template RNA fractions. In addition, t,he amino acid incorporation rc- quired both riboxomrn and 105,000 X g supernatant solution. The correlation between the amount of incorporation and the amount, of added .RXA suggested stoichiomet,ric rat,her t'han cat,alytic actjivity of the t~emplat~e RNA. The t,emplate RNA-dependent amino acid iworporation also required ATI' and an ATP-generat- ing system, was st,imulated by a complete mixtuw of I,-ttmino acids. and was markedly inhibited by puromycin, c:hloraI~lpllellicol, and K Nhasr. A4ddition of a synthetic polynwleot,ide, polyuridylic acid, specifically resulted in thr incorporutioll of L-phenylalanine into a protein resembling poly-L-phcnylalanine. l'olyuridylica acid appenrs to fun&on as a synthetic template or messenger R?;A. The impli- cations of these findings are briefly discussed. Note udded in. prouj. --The ratio between uridylic acid units of the polymer required and mole- rules of I,-phenylalanine incorporated, in recent experiments, has approached the value of 1: 1. Direct evidence for the number of uridylic acid residues forming the code for phenylalanine as well as for the eventual stoichiometric action of the template is not, yet established. 4s polyuridylic acid codes t,he incorporation of I,-pherlylalanine. l)olyq tidylica :rc,iti $ ,q)c~cific~:rlly mecliat t's the ill- (~orporution of I,-proliue int,o :i `r(:A-prwil)il :ibl~ produ(,t (`ornplcte Il:rt:t on thee findings \vill lw included in :t sutwqucnt public2tion. * Supported by :I NATO Postdoctoral Research Fellowship. t Dr. Frankel-Conrat, personal communication. 1: We thank Drs. Leon A. Heppel and Maxine F. Singer For samples of these polyribonucleotides, and Dr. George Rushizky for TMV-RNA. 5 We thank Dr. Michael Sela for this information. ** We thank Drs. William Drryrr and Elwood R~I~IIJII for performing t,hr high-voltage elwtro- phoretic analyses. ' Matthaei, J. H., and M. h11. Nirenberg, these PROCEEDISGS, 47, 1580 (1961). z Matt,haei, J. H., and >I. W. Nirenberg, Biocheva. h Biophys. Res. C'oonsm, 4, 10-l IlSKl). 3 Matt,haei, J. H., and RI. W. Nirenberg, Fed. Proc., 20, 391 (1961). * Crestfield, A. M., I<. C. Smit,h, md F. I\:. Allen, J. Biol. Chew, 216, 1X5 (1955). B Davis, F. F., A. F. Carlucci, ad 1. F. Roubein. ibid., 234, I.525 ~ICJ.59~. fi Hall, 13. I)., and F'. I )(lt#J,, J. .9[06. Niol., 1, 111 ( 195!1). 7 Osawa, 8.. Hiochin~. Wiophys. Acta, 43, 110 (l!WI). y Aronson, A. I., and B. J. McCarthy, Riophqs. .I., 1, 215 (1961). 1602 BIOCHEMISTRY: XIRENBERG AND MATTHAEI hoc. N. A. 8. 9 Kurland, C. C;., J. Mol. Bid., 2, 83 (1960). lo Littauer, U. Z., H. Eisenberg, Biochim. Biophys. dctu, 32, 320 (1959). I* Hoagland, M. B., and L. T. Comly, these PROCEEDIXGS, 46,1554 (1960). 12 Volkin, E., I,. Astrachan, and J. L. Countryman, Virology, 6, 545 (1958). l3 Nomura, hf., B. 1). Hall, and S. Spiegelman, J. Mol. Biol., 2, 306 (1960). I4 Hall, B. D., and S Spiegelman, these PROCEEDINGS, 47, 137 (1961). I5 Bolton, E. T., B. H. Hoyen, and D. B. Ritter, in Microsomal Particles and Protein Synthesis, ed. R. B. Roberts (New York: Pergamon Press, 1958), p. 18. 16 TissiPrrs, A., J. I). Watson, D. Schlessinger, and B. R.. Hollingworth, J. nlol. Biol., 1, 221 (1959). I7 Tissieres, A., 1). Schlessinger, and F. Cros, these hWCEEDINOP, 46, 1450 (IUfiO). I8 McCarthy, B. J., and A. I. Aronson, Biophys. J., 1,227 (1961). Is Hershey, A. D., J. Gen. Physiol., 38, 146 (1954). 20 Siminovitch, L., and A. F. Graham, Cartnd. .J. MicrobLoI., 2, 583 (1956). 21 Davern, C. I., and M. Meselson, J. Mol. Biol., 2, 153 (1960). Z1 Siekevitz, P., J. Biol. C&m., 195,549 (1952). *3 Britten, R. J., and R. B. Roberts, Science, 131, 33 (1960). 24 Martin, R., and B. Ames, J. Biol. Chem., 236, 1372 (1961). 26 Bamford, C. H., A. Elliott, and \V. E. Hanhy, Synthetic Polyp&&s (New York: i\cadrmic Press, 1956), p. 322.