RNA CODEFORDS AA'0 PROTELV SYATTHESIS> II.  NTIJCLEOT_7-DE
     SEQVENCE OF A VALIh-E R-VA CODEWORD

BY PFIILIP LEDER A4m M.~F~FL~LL ?~IREKBERG

NATIONAL FIEABT INSTIXTE, KATIOXAL INSTITPTES OF HEALTB,
             BETaESDA, MABPLAND

Communicded by R. B. Roberts. June 30. 19.64

Recent studies demonstrate that synthetic @lynucleotides. such as poiy I-. in-
duce with specificit,y the binding of C14-amino-ac~l-sR--4 to ribosomes.: -4  Char-
acteristics of binding in the ab,wnce of poly II also have been reported." C14-
Phe-sRX-4 binding induced by poly Y is thought to represent an early step in pro-
tein synthesis before peptide bond format.ion: however. the precise nature of the
interaction has not been ckified.

To establish the sequence of nucleotides in RX-4 codewords. we devised a rapid,
sensitive method for measuring C'"-a~~-~~o-ac~l-sR~~4 binding to ribosomes. and
have investigated both characterist.ics of biding and the minimum oligonucleotidr
chain length required to direct such binding."  pUpI:pC, p-4p.4~~4, and
pCpCpC directed the binding of Phe-. Lys--. and Pro-sRXA, respectively, where&s
dinucleotides had no effect.  Trinucleotides with 5'-t.erminaI phosphate were
more active than those with no terminal phosphate, whereas trinucleotides with
2'-(3')~terminal phosphate were inactive.6

In this report, trinucleotides of knoun sequence were used to direct sR!iA bind-
ing.  GpUpU, but not its sequence isomers, UpGpU and UpUpG, was shown to
induce Va.l-sRKA binding to ribosomes.  These data indicate that the nucleotide
sequence of an RX\`,4 codeword for valine is GpUpU.

Mai&& and Melhods.-Andyses of poZy- and digonuckotidcs:  (a) Paper elertrophoresti was
performed on Whatman 54 or 3 MM paper in 0.05 M NR&OOH buffer, pH 2.i, at 80 vicm for
0.5 hr with authentic reference markers. If a marker wa not available, the expected mobiit!
ws calcultLti.' ib,! Descending paper rhromatoephv wac performed at room temperature
with Whatman 3 MM paper and the following solvents: (-4 I cont. ~H,0H-N-prc~panc~l-H20,
10/55/35, v/v; (B) 40 gm (h&)8& dissolved in 100 ml 0.1 M sodium phosphate, pH 7.0.&
Bands were visualized under C?; light. (cj Ultraviolet absorption measurements were made in a
Zeies spectrophotometer, and spectra were obtained in a cam recording spectrophotomnleter  The
absorbancy of eluates from paper was read against blank paper eluats  (d) The base ra.tio
of the poly UG preparation (D-132j was O.i4/0.26 (T/G).  The base ratio of GpTpG, GpGpr,
and GpUpU preparations was determined bJ- incubating 2.0 -4"" uniti with 3.5 x 10e3 units of
T1 RNaee* in 0.02 ml of 0.5 M ammonium acetate, pH 4.5, for 2.5 hr at 45". The digestion
products were separated by paper electrophoreeis, eluted, and the absorbancy of each at appr<+
Pritlte Lax was determined.
Preparation and separation of oZigom&otidcs: I-pT was chararterized BE described 6 UpVpG

and UPG were obtained by incubating 3 X 10' A*@' units of poly UG with 1.3 X 10" units of T:-
RPuase (Sankpo Co., Ltd., Tokyo) in 4.i ml of 0.1 IV NH+BC& at 37" for 6 hr.`" The reaction
mixture waa lyophilised, dissolved in 0.1 M (KH& CC&. and terminal phoephati were removed
by incubation with E. coEi alkaline phoephataae, free of diesterase, BE described by Heppel d aI I1
Ohgonucleotide fractions were separated by. paper chromatographp with solvent il for 16 hr.
rTltraviolel-at~sc,rbinp ban& with mobiii&sofJUpG and Uplip{; were elated with H,( 1. I~ophilyzrd.
and purified sepamtsly by elertrophoresis.

GpUpV and Gpr were obtained by digesting pal!- U(; with purified pork hver nuc~Ie~se.~ re-
moving terminal pht~phates, and mptrratinp frartmns b> p&per chromatography and electrt*
phoresis a~ described above. This procedure >-ielded B fra(*t,iou containing TWpc.; and GpV and
also one containing GpUpU, UpGpU, and UpUpG. P&per chromatogrtrphy w1t.h solvent B for

420


VOL. 52. 1%     BIOCHEMISTRY: LEDER AND JIRENBERG             421

36 hr resolved each dinucleotide. The trinucleotide fraction was separated into two bands by
chromatography with solvent B.  The band nearest the origin contained GpUpU of high purity.
It wsa eluted with HP, absorbed on acid-washed Norite, and eluted with 4.71~~ ethanol 0.5 M
ZT&GH. GpCpU was purified agaiu by chromatography with solvent B, eluted, and desalted

h derivative of pancreatic RNase A was used by Dr. Merton Bernfield to catalyze the transfer
t)f uridim+2',3'-cyclic phosphate to GpU. I* UpGpC was purified from the reaction mixture by
paper chromatography and electrophoresis as described for UpUpC.
/7haracti~i&ion of triplets: GpUpU, UpGpC, and UpUpG preparations were purified ss de-
wnbed. Only one spot was obeerved after each was subjected to paper chromatography with
solvents ;L or B and to paper electrophoresis.  Base-ratio and sequence analyses established the
purity, chain length, and base sequence of each triplet preparation.  Base-ratio analyses were as
follows: GpUpU, 1.2 U, 1.0 Up, 1.0 Gp; UpGpU, 1.0 U, 1.0 cp, .OQ Gp; CpCpG, 2.1 Up, 1.6 G.
Base sequence was determined as follows: 2.0 Am units of each trinucleotide was digested
separately with T1-RXrse (as described earlier) and in other reactions, with 2.0 pg of chromato-
graphically purified pancreatic RNsse (Sigma) in O.OSS ml of 0.1 Jf (ZIH,)zCO~, at 37' for 6 hr.
Reaction products were separated by paper electrophoresis. Dinucleotide products were eluted,
lyophilized, and digested with the RNaae to which they were susceptible, and the products again
were separated by electrophoreais.  Nucleoeides and nucleotidm were identified by their electrc-
phoretic mobility ss well as by their spectra.

After digestion of GpUpU, UpGpU, and UpUpG with pancreatic RNase, only the following
producte were obtained: GpUp + U; Up + GpG; and Up + G, respectively. After digestion
t,f GpUp and GpC' with T1-R&se, Gp + Up and Gp + U were obtained, respectively. Ti-
RXaae digestion of GpCpU, UpGpU, and UpCpG yielded Gp + UpU, UpGp + U: and UpUpG,
respectively. Digestion of UpC, UpGp, and UpUpG with pancreatic R&se yielded Up + U,
Up + Gp, and Up + G, respectively.  No other L7-absorbing material was found. Since a 2%
contaminant could have been detected, the purity of each trinucleotide was estimated to be
>98rc.  GpUpU and CpCpU have not been purified previously.
Ribommes, sR:V.$ and E. cola' extra&s: E. cdi W 3100 sRN.4 was prepared from cells grown
to late log phase in 0.54 nutrient broth, 1% gluc~~~~ E. cdi B aRNA was obtained from General
Biochemicals, Inc. Bach Cl*-amin+acyl-sRX.4 was prepared in the presence of 19 Cl*-amino
acids as described elsewhere.' In additional experiments C1*-T.m~RNA also was prepared in
the absence of other amino acids.  CII-GVal, Cl'-GPhe, and C*`-GLeu, uniformly labeled with
radioactivities of ;%I, >ZQ, and 435 cpm/wmole (Packard Corp. scintillation counter), respectively,
were obtained from New England Nuclear Corp.  Ribosomee for binding studies and E. cdi ex-
tracts (DNase-treated, preincubatecl. S36 fraction: for amino acid incorporation into protein
studies were prepared as described elsewhere.". `a

dssays:  Each S&l reaction mixture contained 0.1 M u-is-acetate, pH i.2; 0.02 JI mag-
nesium acetate; 0.05 .U KCl: and 1.0 .4m units of nbosomcs unless otherwise specified.  Incuba-
tions were at 24O for 20 mm unless otherwise `indicated.  C'*-amin~acyl-sRN.4 binding to ribo-
somes was determined. M reported elsewhere, fi by washing the ribosomes on Millipore cilters.
Ribosomes with bound sRNh remained on the &her.

C?-amino acid incorporation into protein was determined ss reported previou~ly.~J
Resuhk-EJect oj Gplpl- and PI* L-G upmz the binding of C"-valine &V--l tu
ribosomes:  The data of Figure 1 shoti that the binding of Val-sR,\i-\ to ribosomes is
stimulated by the addition of GpCpU or poly UG and is dependent upon the con-
centration of either template.  JIore Val-sR_\iA is bound to ribosomes in the pres-
ence of poly UC than in the presence of an equivalent amount of GpUpU.  We
have observed similar differences between the 3ctivity of pUpITpI.7 and poly I?."
In Figure 2, the rates of binding at 0" and 24O are compared, and the specificity
of GpUpC: and poly UG for amino-acyl-sRX-4 is shown.  At O", poly UG markedly
stimulated Val- and Phe-sRNA4 binding to ribosomes, whereas GpCpG stimulated
only Val-sRI\3,\ binding.  At 34" the rates of binding directed by poly UC were
higher than st 0". Slightly higher binding in control reactions was observed


BIOCHEMISTRI-: LEDER A-4-D It-IRE-?-BERG     Paoc. X. A. S.

0' '   ' ' ' ' 1 ' I
   a k 8
   M~YOL'~S BAFE RE~DIESM   7.8 32

FIG. l.-The relation between GpTpV
(01 and poly I`G (D) concentration and Cl'-
Val-eR?;A binding to ribosomes. Each re-
action mixture contained 2.0 AM unitE of
riboeomea and 18.9 MFrnoles of Cl*-Val attached
to 0.44 A"O units of sRl"iA. Poly UG and
GpUpU were added as specified.

(especially of Leu-~33-4). Under the
conditions employed, pal- UG had little
activity in directing Leu-sRKA binding
to ribosomes, compared with its abiliq-
to direct amino acid incorporation into
protein.

Codgword specificity arbd the eflect r!i
Of~-~j cwwcentration: The data of
Figure 3 demonstrate that poly UG stim-
ulates binding of T'al- and Phr-sEEi- to
ribosomes optimally at 0.02-0.03 .a1
C.\lg+-;j. whereas GpUpV directs bind-
ing of Val-sRX4 optimally at higher
concentrations.  ilt higher (big++) con-
centrations, a small and apparently non-
specific increase in Yak. Leu-! and Phe-
sRX*4 binding was observed in control
reaction mix-tures (no addition). Tri-
and polynucleotides induce specific
sRKA binding. GpUpU directed only
Val-sRKA binding, whereas UpGpU and

UpUpG had no effect upon either Val-, Leu-, or Phe-sRXA.  The addition of tri- or
polynucleotides did not. appreciably increase Leu-sRXA binding over that of control
reactions within the range of (Mg++) concentrations tested.
The specifxity of di-, tri-, and polynucleotides in directing amino-acyl-sRX4
binding is shown in Table 1. Poly UG directed Val- and Phe-sRK.4 binding.
whereas poly U directed the binding only of Phe-sRKA.  UpU, GpC, and UpG
did not, stimulate Val-, Phe-, or Leu-sRKA binding.
The effect of GpUpU was tested upon the binding of I 7 other sRS-4 preparations,
each with a different C"-amino accepted (CWasparagine and C"-glutamine-sRK-4
not. tested).  GpUpC was found to direct the binding only of T'al-sR?;A.
Amino acid incorporation into protein:  Since poly UG had little effect upon
Leu-sRX-4 binding to ribosomes, w-e investigated the ability of poly UG to direct

TABLE 1

                CODEWORD SPECIFICITY
              Addition.
Expt.       mrmolee base residues      ClcAminc-AcyLsRKA Bound to Ribmomes, ~~molee
                           C"-Bal-nRE;A     CW%HIRIGA     C'LLeu-sRNA
1         ?; one            0.38         0.22         0.3i
        4.; Poiy r          0.23         4 73         0.22
        4 i Poly LrG        2.65         1.93         0.24
2         None             0.40         0.22         0.62
        4 i GpFpl`         1.11          0.2i         0.56
        4 7 UpGpI-         0.40         0.25         0.55
        4 7 UpUpG         0.33         0 25         0 .44
3          None             0.1s          0.22         0.69
        4.i GpT           0.20         0.22         0 .69
        4.7 UpG           0 19         0.21         0.71
        4.i UpI:          0.11         0 20         0.6T

Obgonucleot~dc spwificrt?- m directing sRKA
C-`-\-al attached to G.3 Am unita of aRN.4 17.0 bindma to ribmomes  Where indicated. 0.10 @m&E
                         ,wnoler C'cPbc attached to 0.; -4' unita of sRNA.
or 21.0 wmoias C'CLeu attached to 1.6 A= +ta of r,RK-4 welt added to reactmn nuxtu~-~.


i.OL. x2. I!w      BIOCHEJIISTRY: LEDER d.VD .VIRE.VBERG             423

Fro. %.-Effect of GpcpU and poly c:G u 1 on
sRNA to riboaomea at 0" and 24".  The sym   the rates of binding of Val-, Leu-, and Phe-
mrmoles of base residues; (0) GpUpU,    ols represent the addition of: ( ~1 poly UC, 4.70
                   2.35 mfimolea of base residues; (0) no addition.  Where
indicated, li mpmoles of Cl'-Vai attached to 0.44 A" units of eRN-4, 8.5 gfirnoles C"-Phe at-
tached to 1.5 A?"0 units sRNA, or Sl.O.~gnolea of C*4-Leu attached to 1.5 A' units of sRXA were
added.  Samples were incubated at the temperatures and for the timea indicated.

amino acid incorporation into protein in E. coli extracts at 0' and 24O (also at 137~
to make these data comparable to previously reported studiesI 15).  -1s shown in
Table 2. poly CG directed Phe, Val, and Leu into protein at 37" and 34', and as re-
ported previousiy,`4, l5 Leu incorporation almost equaled that of Val.  Amino acid
incorporation into protein was not detected at 0".
D~c~cussiun:-Previously we have shown that LpUpcI, -4pAp.\, and pCpCpC
specifically induce the binding to ribosomes of Phe-. Lys-, and Pro-sRX.4. respec-
tively  The present study demonstrates that the trinucleotide, GpUpU. can induce
the binding of Val-sR_\;;I to ribosomes. whereas sequence isomers such as UpGpU
and L-pUpG. and dimucleotides have no activity.  When the specificity of poly CG
was compared with that of GpIYpL-, the polynucleotide was found to direct both

Ammo .irid Incorporated into Prorem. rurmoles
C' 4- rvli ne     C'+heoylahn~ne    C'~Ie1mne
<o 03        <0.03       <o 03
<o 03        <I~.03       <o. 03
<I). 03         0 04         0.04
0 io         1.5i         0.6;
0.10         0.09         0.08
  1 04         3.05         0.96

Each 0.1~ml reaction mixture contained 0.3 mq 15. colr protein ~DNasetreated. premcubated J-30
fraction*). 20 agmoles of archer C'CVsl. C'LPhe. or C'cLeu `4.8. 4.0. 4.0 ~cunea; rmole. respectively'.
in addition to components previously described. m  Reactions were incubated for 10. 20. or 60 min at X0.
24O. or O", respectively.


424

BIOCHEMISTRI*: LEDER AA-D h-IRE-+-BERG     Paoc. K A. s.

LEKM

[lap"] aLnAnnY

FIG. 3.-Relationship between magnesium acetate concentration and C~~-amino-a+sRK~4
binding to riboaomes.  The qrnbols represent the,adchtion of: ( L ) poly VG: 4.70 m#moles
of bass; (0) GpCpV, 2.35mpmoles of base; (0) CpGpU, 2.35m~1moles of base; (3'; TpVpG.
2.35 -molea of base; (c)no addition. 9.10 wmoiee Cl'-Valattached toO.3.4m units of sRX.4.
25.0  rmolea C~%eu attached to 1.5 Aam units of .aRS-4, or 10.0 wmoles Cl<-Phe at,tavhed
to 0.i A' unite of BRSA were added where indicated.
reaction mixture is shown on the abscissa.       The (Mg i `i concentration of each

the appropriate concentration of (Mg ++).   Samples were washed with buffer containing

Phe- and Val-sRXz4 binding, but GpITpV induced the binding of only T'al-sRK.4
We conclude that, the nucleotides of a saline codeword in mRK.4 are arranged in
the sequence GpVpV.  Since VpUpG does not replace Gpl-pV, recognkion occurs
with polarity.

As discussed elsewhere,6 the ability of tri- but not. dinucleotides to induce sRX.4
binding demonstrates directly that triplets may code for amino acids, and agrees
with conclusions derived from both geneticI and biochemical studies.2"
Studies of the incorporation of amino acids stimulat.ed by various polynucleotidcs
together with the data OK amino acid replacements in mutants provide a basis EW


voL. .52, 1964      BIOCHEMISTRY: LEDER A.VD .VIRESBERC            -LL'5

assignment of codewords to amino acids.`"-`8  Unfortunnteiy. widespread de-
generacy and the difhculty of assigning words which contain three different bases
prevent unique assignments.  Several self-consistent schemes are possible. The
additional information provided by the definite knowledge of even a single code-
word such as GpUpU provides a distinction among these different possibilities.
The reassignments suggested to make GpUpC the codeword for valine consistent
with the other data are listed in Table 3.  For example, .4pUpU was assigned to
isoleucine on the basis of studies employing randomly ordered polynucleotides to
direct amino acid incorporation into protein, and the HNO?-induced replacement
of isoleucine by valine in TMV coat protein.

T.mLE 3

      PREDICTED J~WLEOTIDE SEQCENCES DF RX-4 CODEW~FDS
Ammo acid             sequence                  Reierence
V&ne            GpPpC
Lsoleucine          AplTpC          Derived experimentally
                               Ii, 1s
              Ap ( PA)
Tyrosine           17~ I AC j
Manine           GPCPC             16
             cp \CA)

Glutamic scid

Glycine



Argiuiue


Threonine

$gi?           16
Gp(AC)
Gp (GCJ            16
$gg,'  Note: GpCpC ia ruled out
hpGp.4            16
CPGPC
(C)Gp(-i)
Apt CA)          li, 18
(XC)

* The as~~qnrnent of Gphp.1 to glutamic acid ia based npon limited -amino sad incorporation
into protem <data md is. therefore. tenratlve.-. 24
Nucleotides witbin parentheses have not been arranged in sequence. .Unino acid replace-
ments used for these
mducrd by KNOI .n rp redictxms were round in E. coli by Yanofsky and co-workers's or were
            MV by Wittmaon and Wittmann-Liebold" and AISO by f~uu\ca.:*

It should be noted t*hat the predicted sequence for isoleucine, ApIYpU, and also
the GpVpU sequence found for valine do not agree with reported sequence deter-
minations21 which, on the basis of experiments with uncharacterized polynucleo-
tides, assign the sequence -1pUpV to tyrosine, tentatively assign GpVpU to cysteine.
and derive the polarity of ntRXA for protein synthesis.

Although our results show only marginal stimulation of Leu+RS-1 binding by
poly CG. we have consistently induced Leu- and also Cys-sRY.1 binding with poly
UC, when reactions were incubated ;ft 0' and at a (llg-`) concentration of 0.03
.ll.  Leu-sR-\;-4 binding wad readil+nduced by poly 1-C which confirms results
reported by Iiaji snd Kaji."  Further studies with Leu- and C'ys+RS.A are in
progress.

We have shown that specific binding of sRX-1 can be induced at 0" and have
described preliminary evidence which indicates that intact anmm-acy!-aR3A4 can be
eluted from ribosomes after incubation."  -Uthough these data suggest that forma-
tion of peptide bonds is not required for binding to ribosornes. a requirement
for the first transfer enzyme and GTP may exist (~cf. refs. l-4, and 6).
Since synonym codewords corresponding to one amino acid often differ in base
composition by only one nucleotide, `?" it is possible that bases in common may


426

BIOCHEMISTRY: LEDER AA-D AVIREKBERG     PROC. x. A. $.

occupy identical positions within each triplet.  -4 triplet code can be constructed
wherein recognition between two out of three nucleotide pairs ma.y. in some cases.
m&ice for coding; or, alternatively, a base at one position in the tripiet may pair
opt,ionally with more than one base.22-P*  Ambiguous and synonymous codewords
may involve such recognitions.  The development of a rapid method for measuring
sRT\`A binding to ribosomes and the use of trinucleotides of known sequence to
direct such binding should provide a general met,hod of great sinlplicit,\- to test
these possibilities and also to study the genetic function of other nucleotide se-
quencez and interactions bet,ween specific codewor& SRI-4. and ribosomes.

Su3nmqj. -The binding of C14-saline-sRS.4 to ribosomes was directled both b>
the trinucleotide, GpTpT. and by poly UG. but not by TpGpc, I`pVpG, or di-
nucleotides.  GpUpT: had no effect upon the binding to ribosomes of sRP\`A cor-
responding to li other amino acids.  The nucleotide sequence GpVpT was pro-
posed for a valine RX-4 codeword.  The implications of these findings and pre-
dictions based on amino acid replacement data are discussed briefly.

We are grateful to Drs. Leon Heppel and George Rushizky for rheil advice on oligonuc*ieotlde
separations. It is also a pleasure to thank Norma Zabrlskie and Theresa Caryk for their ex-
tremely skillful technical assistance.
The following abbreviations are used: Val, valine; Phe, phenylalanine, Leu. leucinr : p oiy Ty,
polyuridylir acid; poly C, polycytidglic acid; ~011 A, pal:-adenylie acid: pal>- UG, copolymer of
uridylic and guanylic acids: TCA, trichloroacetic acid; sRK.4. transfer RI\`A: mRKA, messenger
RK.4.  For mono- and oligonucleotides of specific BtrUCtUre; the "p" to the left of a terminal
nucleoside initial indicates a 5'-terminal phosphate:  the "p" to the right, a 2'-(3' j-terminal
phosphate.  Internal phosphates of oligonucleotides are (3',5',+linkagea.
1 Arlinghaus, R., G. Favelukes, and R. Schweet, Biochem. Bwphys. Res. Pommun., 11, 92
(1963J.

* Kakamoto, T., T. W. Conway, J. E. Allende, G. J. Spyides. and F. Lipmann, in Synthrsu
and Strudurr of Macrem&cuks, Cold Spring Harbor Symposia on Quantitatk Biology, vol. 25
cl963j, p. 227.

3 Kaji, A., and H. Iiaji. Biochem. Res. CowLmun., 13, 186 (19G): Federation Proc. ( .4 bstv.,, 23,
4% j1964;.

4 Sp:+des, G. J., Federation Proc. 1 rl bstr. j, 23, 31s cl964 j.
6 Cannon, M., R. Brug, and W. Gilbert, J. Mol. Biol., 7, 360 !. 1963).
6 Kirenberg, M. W., and P. Leder, Sctirz~, in press.
7 Markham, R., and J. D. Smith, Biochem. J., 52, 5% ;1953 I.
6 Rushizky, G. K., and H. -4. Sober, J. Biol. Cheer., 237, "SK9 1,1962 I.
Pfiid., 238, 371 (1963).
w Ibid.., 237, 834 (1962 j.
11 Heppel, L. A., D. R. Harkneas, and R. J. Hilmoe, J. Biol. Chem., 237, 841 (1962`1.
I* Bemfield, M. R., and M. W. Krenberg, Abstracti, 148th Xational Meeting, American Chemical
Society. Chicago, Illinois, August 1964.
I1 ?jirenberg. M. W., in Methods 211 Enqw~&gy, ed. S. P. Colonick and I\`. 0. Kaplan (Xea
York: Academir Press, IWX;, vol. 6, p. 17.
14 Nirenberg. hl. W.? J. H. Matthaei, 0. K. Jones, R. G. Nartin, and P. H. Barondes. Frdvatzon
Pm.. 22, 55 (1963).

a Sii~ger, >I. F., 0. W. Jones, and M. W. Sl'irenberg, these PROCEEDINGS, 49, 3% i,l%S,.
If Tanofsky, C.. m Synthms and Structure of Macromokc~ks. Cold spring Harbor Symposia on
Quantitative &olog~, vol. 28 (1963 J, p. 551.

Ii Wlttmann, H. G., and B. Wibtmarm-Liebold, in S.yntheszs and Structure qf Manomokcu1ch.
Coid Spring Harbor Symposia on Quantitativr Biolog\-. vol. 2b 1963 J, p. 5R9.
B Tsugita. .4.. personal communiwtion.
I* Crick, F. H C.; L. Bamett. 5. Brenner, and R. J. WatsTobm, -qVaturrj 192, 1227 11961 :.


1o Nirenberg, M., 0. Jones, P. Leder. B. Clark, W. Sly. and 3. Pestka, SynthPszs and Strw-ture
of YacsomoEe~~~lt~, Cold Spring Harbor Symposia on Quantitative Biology, vol. 2% I 1963 8, p. .?A!).
!I Wahba, A. J., C. BaYilio, J. F. Speper, P. Lengyel, R. S. Miller, and 8. Ochoa, these PILOCEEW
INGS, 48, t683 (l%a).
22 Xrenberg, hf. W., and 0. W. Jonps. in Infow~~&wl .lil~omoler&.u, ed. H. I-ngel rt <IL..
t New York: .h:ademic Press, 1963 1.
?a Jones, 0. W., and M. W. Nirenberg, these PROCEEDINGS. 48, 2113 ( 1962 1.
14 Gardner, R. d., A. J. Wahba, C. Basdio, R. P. lIiiIer, P. Lengyel, and J. F Speyer, r,hese
PROCEEDINGS, 48, %87 (1962).
?5 &bay, G., J. Mol. B&L. 4, 347 (1962).