Reprinted from the Proceedings of the NATIONAL Acanmm OF SCIENCES
       Vol. 48, No. 1, pp. 104-109. January, 1962.

AN INTERMEDIATE IN THE BIOSYNTHESIS OF
POLYPHENYLALANINE DIRECTED BY SYNTHETIC
             TEMPLATE RNA

BY MARSHALL W. NIRENBERG, J. HEINRICH MATTHAE,* AND OLIVER W. JONES

NATIONAL INSTITUTES OF HEALTH, BETHESDA, MARYLAND

Communicated by Richard B. Roberts, November 84, 1961

We have recently found that, simple, synthetically prepared polyribonucleotides
such as polyuridylict acid and polycytidylic acid function as RNA templat,es in
a cell-free protein-synthesizing system prepared from E. eoli.*~ 2  In this system,
poly U contains the information for the synthesis of polyphenylalanine; therefore,
the code for phenylalanine is one or more uridylic acid residues.  Poly U was much
more effective in increasing the rate of cell-free protein synthesis than naturally
occurring informational RNA,2 possibly because the synthesis of a protein containing
only one amino acid was faster than the synthesis of a protein containing 20 amino
acids.  We are using this model system currently to study the enzymatic sequence
of protein synthesis.

Although sRNA and amino acid-activating enzymes have been studied exten-
sively, there is controversy concerning their relationship to protein synthesis.3-6


One purpose of this illvrst.ig:ttion was to determine whether sRK,I is an intjermediat,e
in the synthesis of polyphenylalanine directed by a synthvtic Rx21 template.  Our
results demonstrate that. the irworporation of phcnylalanine into sRS.4 and its
subsequent t,ransfcr from sR9,4 are steps in the synthesis of polyphc~~yl:llanii~e.
fllelhotls.-l~ialyzcd extracts of E. coli containing ribosomes 2nd 100,000 X y supernatant solu-
tions were obtained as described perviously.*  These eskacts correspond to the previously
described S-30 fractions.2  Ribosomcs were sedimerkecl by ccnkifuging S-30 extracts at 105,000
X 9 for 2 hr at 3".  The supernstant sohlt,ion was aspirat'ed and will be referred to hereafter as
100,000 X g nupernatant solution.  The ribosomes rverr res:r~spc~r~dt~J in 0.01 J1 Tris (hydroxy-
meth~l):trllirloIueth:nle, pH 7.8, 0.0 1 :I2 magnesium :icetate, 0.06 JI KCI, and were centrifuged
again at 305,000 X c/ for 2 hr.  The supermttant solution ~V:M dec:tnt<ld anti was disrsrded.  The
ribosorws were washed two mow times in the sxn(~ manner by rcwspension und centrifugation.
Both 100,000 X y supernatnnt solutions ttntl washed ribosomrs were stored in mm11 aliquots under
liquid nitrogen.  Th:~w~d preparations rvc~re not refrowr~ and reused.
sR?JA M-:&S purifwl fronr E. coli 100,000 X y supernatant solutions by phenol ext,raction.  s1~JX-A
was ch:wgcbd with (: "-l,tlc,nyl:tlallirie enzymatically, and the C `l"-amirlon~~l-sBSA \vas purified
by thcl met hod described by von IShrenstein and Lipnrann." The specifii radioactivity of the
C14-:lmilio:lc.~l-s~~~.~ varied front 800 to 86,000 counts/mirr/rng sRSA.  The opt,ical density of
1 mg of sl<.NA in II?0 at, X0 rnp \V:LS assulnrd to br 24."
The RNAase-digrsted alrlillo:lc.~l-sllNA drwribed iu T:tt)l(x 2 \v:~s prcywwl by incubating CII-
I'hcnZl:llarlinc-sIiSA with 1 fig c~r~~st:tlline T~t?;Aax~ (\Vorthington Biochrmical Corporation)
per ml at 37" for 1 hr.  Thr 1~SA:we was rrmovctl Ily 3 wnsccutivr phenol extractions using
equal volurncs of phe1101.  The aclucous phaw was t,hen dialyzed overnight against 1,000 volumes
of HrO.
The :llli:tli-tlegr:ltlctl CL"-pherl~l:tlani~lc-sIt~A described in T:tble 2 was prrpared by incubating
arni,,,:,(,yl-i;R.n'rZ in 0.3 d/ KOH at 35" for 18 hr.  The solution was then neutralized and was
dialyzc~d against 1,000 volumes of II,O. Also, Cl'-I)hcn)l:llanirlc-sltNA was incub:tted in 0.4
M glycaiuc, buffers, 1'11 1 1 .O at, 37" for 1 hr.  The solution was then dialyzed against 1,000 volumes
of 1320.
Rxlio:xct,ive prot,ein pwcipit~atcs were lvnshed and counted as before2 by a modification of the
method of Siekcvitz.7  Protein corrrentrations were dct,ermined by a micro modification of t'he
method of Lo~ry.~  All assays reported in this paper were performed in duplicate.
n~aterials.-U-C'l-L-phenyl:tlamine was obtained from the Nuclear-Chicago Company and had
:I specific radioactivit,y of 5-10 mCjmmole.  The purified trnnsfcr enzyme (aft,cr DEAE column
chromatography, purifird about lo-foldg) and some C14-phen~l:tl:tnine-sR~~~ used were the gener-
ous gift of Daniel Nathans and Fritz Lipmann.  l'olynucleotide phosphorylase eras used to
sgnthesizr poly U.  The molecular xveight of the poly U was brtween 30,000 and 50,000.
IZesults.-Compoilc~llt,s of reaction mixtures are given in Table I.  The data pre-
sentcd in Tal~le I, Pkperimcnt 1, demonstrate that, Cl"-pheuylal:tnillc \\-a~ incor-
poratvd into protein only when poly U was added to the rwctioll misture.  Whrn
2 pm&s of uulatwlcd, C1"-phellylalnllillc were addc;d t,o a reaction misturc contain-
ing 0.019 pmoles of Cl"-phenSrlalaninc, the incorporation of Cl"-phcllylalaniile illt'o
prokin was markedly rcduwcl as was expect,ed, due to dilution of the tracer.
C L"phC~lylalnlli~~c-sR~,~ was preparcd hy incul)atitrg sRSA with Cl"-phenyl-
alaninc and nonpllrified phenylalnninc activating enzyme prcscnt in 100,000 X
g suprrnatnnt solutions.  These extracts catalyzed a phellylalnnille-dcpcndeut
cxchangc of lahclcd pyrophosphntc into ATP. Other amino acid-activating en-
zymes, stimulat'ing a similar exchange, were also present).
In l'ixperimcnt 2, the t,racer present in reaction mixtures was CL4-phcnylalanine-
sRNA in place of frw Cl"-phenylalanine.  ATP, CTP, and UTP were omitted.
In t,he absence of poly U, little Cl"-phenylalaninc was incoporated int,o protein.
However, in the presence of poly U, approximately 60 per cent of the C14-phenyl-


106

BZOCHEMZSTRY: NIRENBERG, M4TTHAEI, AND JO:VES PROC. N. A. S.

TABLE 1

EFFECTOFC'~-PHENYLALANINE UPON C14-P~~~~~~~~~~~~ INCORPORATION INTO PROTEIN
Experiment No.         CP-tracer                   Addition           Counts/min
    1     Cl"-L-Phenylalanine        None
                       - Polyuridylic acid           1,430
                     + 2 pmoles W-L-phenylalanine   ii
                      - ATP, PEP, and PEP kinase        62
                       Kane, zero t,imc                  40
    2     C'cL-Phen3;lalanine-sRNA   None                        683
                      - Polyuridylic acid
                     + 2 pmoles C'Z-L-phenylalanine   6";
                      - GTP, PEP, and PEP kinasc        10
                       K-one, zero time                   6
The reaction mixtures for Experiment 1 contained the following in pmoles/ml:  100 Tris (hydroxymethyl)
aminomvthnne, pH 7.8: 10 magnesium acetate; 50 KCl; 6 mercaptoethanol; 1 .ITP; 5 I,hosl,hoenolpyrui-ate K
salt; 2 ~g I,hoslilroenoll)yru\.ate kinase, crystalline; 0.03 each of GTP, CTP, and UTP; 10 fig polyuridylic a&d;
0.019 qroles of CL"-L-r~llenylalanin~,. -75,000 counts/min;
supernntant solution protein, respectively.        and 0.46 and 1.04 mg of rlbosomal and 100,000 X g
The reaction mixtures for Experiment 2 contained the following in pmoles/ml;  100 Tris (hydroxymethyl)-
aminomethane. pH 7.8; 10 magnesium acetate; 50 KCl; 6 mercaptoethanol: 0.3 GTP; 5 yIrospIloenol~ryn,vate,
K salt: 2 PK l)hos1~11ornolpyrlivate kinase, crystalline; 10 ~6 polynridylic acid; 0.45 mg C'4-pllenylalanine-sRNh,
-1100 counts/min; and 0.26 and 1.01 mg of ribosomal and 100,000 X (I supernatant solution protein, respectively,
Total volur,,e was 0.5 ml.
chloroacetic acid.     Samplru were incubated at 3,s' for 10 min, were deproteiniaed with 10 per cent tri-

&mine was tmnsfcrrcd from sRSA to protein.  The transfer required GTP and
a GW-generating system; howercr, with t)his dialyzed butI unpurified system, ATP
and an ATP-generating system was 70 per cent as effective as GTP in facilitating
t,rnnsfrr.  Addit,ion of 2 pm01es of unlabeled CIS-L-phenylalaninc did not, decrease
appreciably the transfer of C14-phenylalanine from sRN\`A t,o ribosomal prot&, ill
contrast to the data of Espcriment, 1. These results demonst,rate that C'4-phenyl-
alanine ran he transferred undiluted from sRXA to protein in t,he presence of a large
pool of unlabeled, free phenylalanine and that poly U directs the transfer.
In Figure 1, the qunnt'ity of phenylalanine transferred from sRSd into rihosomal

                        1-i -r--17

FIG. l.-Stimulation of C*4-pnenylslanine transfer from sRKA to protein by polvuridylic acid
0.5 pg polyuridylic acid added: A wit.hout polruridylic acid. The components of t,he rrnction
mixtures are given in Table 1, Experiment 2.  0.17 mg of Cl&-phenylalanine-sRIiA.  400 counts/
min, 0.46 mg ribosomal prot,ein and 1.04 mg 100,000 X 9 supernatant fraction protein were added
to each reaction mixture.

protein is plot'ted against time in minutes.  In t,he abdence of poly U, lit,tle C14-
phenylalanine was incorporated.  In the presence of poly U, Cl"-phenylalanine was
transferred rapidly from sRr\TA to protein.  Almost all of the Cl"-phenylalaninc was
transferred during the first five minutes of incubation.

Some control `experiments are present,ed in Table 2 demonstrating that C14-


VOL. 48, 1962 BIOCHEMISTRY: NIRENBERG, MATTHAEI, AND JONES       107

TABLE 2

        AMIINOACYL RNA CONTROL EXPERIMENTS
               Additions                      Counts/min
Cl"-phenylalanine-sRXA                           704
C"-phenylalanine-sRNA treated with RNAsse*
C14-phenylaIanine-sRNA treated with 0.3 M KOH +     ?!i
Cl'-phenylalanine incubated at pH 11 t                    18
None, Zero time                                     4

The composition of the reaction mixtures is presented in Table 1, Experiment 2.  1.04
mg protein in 100,000 X g supernatsnt solution, 0.46 mg ribosornal protein, and 0.45 mg
CL'-phenylalanine sRNA containing -1100 counts/min (before digestion) were added
to each reaction mixture.

* sRNA prepamtions were deproteinised by phenol extraction after RNAase treat-
ment a8 specified in the Methods section.

t Alkali treatment of Cl'-phenylalanine-sRNA and incubation at pH 11 are described
in the Methods section.

phenylalanine-sRNA has properties similar to those reported for aminoacyl-RN.4.
Treatment of C14-phenylalanine-sR~A with RNAase or with 0.3 N KOH destroyed
its activity.  Aminoacyl-sRNA is hydrolyzed to free amino acids and sRxA at
pH lO.`O  IncubaGon of Cl"-phenyl-alanine-sRPu'A at pH 11 result,ed in the hydrol-
ysis of phenylalanine from sRNA as expected.  In addition, our preparations of
C14-phenyl-alanine-sRNA had a sedimentation value of 4.6 as determined by
sucrose density-gradient centrifugation.11
The dat,a of Table 3 demonstrate that both 100,000 X g supernatant fractions

TABLE 3

REQUIREMENT FOR 100,000 X g SUPER~ATANT SOLUTION AND RIB~SOMES
             Additions                    Counts/min

None
- 100,000 X g supernatant solution
- Ribosomes
None, zero time

341
14
;

The components of the reaction mixtures are presented in Table 1, Ex-
periment 2. 0.46 and 1.04 mg protein were present in the ribosome and
105,000 X B supernatant fractions. respectively. 1.0 mg of ClLphenyl-
alanine-sRNA. -&JO countsjmin, were added to each sample.

and ribosomes were required for transfer of Cl"-phenylalanine from sRNA t)o protein.
Since the t,ransfer enzyme is found in 100,000 X g supernat'ant, solut>ions, it seemed
likely that the requirement for this fraction could be replaced by purified transfer
enzyme.  The daha of Table 4, Experiment 1, demonstrat,e that transfer enzyme

                         TABLE 4
   REPLACEMENT OF 100,000 X g SUPERNATA~T FRACTION WITH TRANSFER ENZYME
   Experiment No.                    Addition                   Counts/min
       1         None                                   404
             - Transfer enzyme                          10
             - Polyuridylic acid                          11
               None zero time
       2      + Gi`P, + PEP, + PEP Kinase       5%:
            - GTP, + PEP, + PEP Kinase
            - GTP, + ATP, + PEP, + PEP Kinase     &
            + GTP, + ATP, + PEP, + PEP Kinase          271
            f GTP, - PEP, - PEP Kinase                 17
            f GTP, + rZTP, - PEP, - PEP Kinase           13
            + GTP, + PEP, + PEP Kinase, Zero Time          3

   Components of the reaction mixtures are presented in Table 1, Experiment 2.
   natsnt solution w&8 omitted.                             100,000 X B super-
                 2.0 rmoles of ATP/ml of reaction mixture were present where specified.
  0.46 mg of ribosomel protein were present and 0.095 mg transfer ensyme protein were present except
   where specified. 0.30 mg C'cphenyl~lanine-ERNA, -800 counts/m& were added to each sample in
  Experiment 1 and 0.05 mg, -1300 counts/min, were added to each sample in Experiment 2.  In Ex-
  periment 2. for simplicity. the presence or absence of GTP, ATP, PEP, and PEP kinase are noted.


108     BIOCHEMISTRY: NIRENBERG, MATTHAEI, AND JONES hoc. N. A. S.

could replace 100,000 X g supernatant solution.  The data of Experiment 2 show
that GTP was necessary for the transfer, and that, in this purified system, ATP
could not replace GTP effectively.  The data of Tables 1, 3, and 4 show that C14-
phenylalanine transfer from sRNA t,o protein required ribosomes, transfer enzyme,
poly U, and GTP and a GTP-generating system.
Discussion.-The data presented in this communication demonstrate t'hat amino-
acyl-sRNA is an intermediate in phenylalanine incoporation into protein mediated
by poly U.  In a previous communication, we showed that the protein synt,hesized
had unusual characterist'ics similar to autherkic polyphenylalanine.2 The initial
steps in polyphenylalanine synthesis appear to be:

phenylalsnine-
activating

L-phenylalanine + ATP p
        enZyme ' AMP-phenylalanine + P-P.    (1)

                      phenylalanine-
                        activating
AMP-phenylalanine + sRNA p
         enzyme phenylalanine-sR?u'A + AMP.  (2)

                       Poly u
                       rihosomes
                        transfer
              %F

  Phenylalanine-sRNA 1 1 L polyphenylalanine + sRNA.    (3)

The detailed mechanisms of the steps involved in reaction (3) are under investiga-

tion.  It should be noted that our data do not preclude the possibility of alternat,ive
routes of synthesis of polyphenylalanine.

When poly A and poly U are mixed, doubly- and t,riply-stranded RNA is formed
(U-A and U-U-A).12v l3  We have shown previously that poly U in the doubly-
and triply-stranded state was completely inactive as template RNA.2 Further
experiments have corroborated and extended these findings and will be published
in a separate communication.  These data strongly suggest that the portion of the
RNA molecule which functjions as a templat,e for prot,ein synthesis is single-stranded.
Simple predictions may be made concerning the primary and secondary structures
of the hypothetical "template-recognition portion" of phenylalanine-sRKA.
Since a sequence of one or more uridylic acid residues in poly U is the code for
phenylalanine in this system, it is probable that phenylalanine-sRh-A contains
a complementary sequence of one or more adenylic acid residues which base-pair
with t'he template.  It is also probable that, the portion of sRNA recognizing the
template is single-stranded.

The genetic code may not be universal; it may differ from species to species.
Since sRNA may be a cofactor which functions as an "adaptor" carrying an amino
acid to its proper place on t,emplate RPJA, a variant, sRNA base-pairing with a
different code lett,er of template RNA would substitute one amino acid for another
during prot.ein synthesis.14  Although E. coli sRNA can be used for the cell-free
synthesis of rabbit hemoglobin, 3, 4 it is possible that in species other than E. coli,
poly U may be either meaningless or may serve as a t,emplat'e for a different amino
acid.1  Changes in the code could occur at different stages in the translation of
information from DNA to the finished protein, for example, at the level of t'he DNA
or RNA templates, at the level of sRNA, or at the level of amino acid-activating


VOL. 48, 1962 BIOCHEMISTRY: NIRENBERG, MATTHAEI, AND JONES       109

enzymes.  We arc using the poly U system to determine whether t,he code for
phcnylalanine is the same in different, species.

Summary. -I'henylalanine-soluble RXA was shown t,o be a,n intermediate in the
cell-free synt,hesis of polyphenylalanine directed by a synthetic t'emplate Rr\;A,
polyuridylic acid.

We wish to thank illrs. Linda Greenhouse for her excellent help in performing some of the
analyses.

* NATO postdont,#ml fellow.
t The following abbreviations are used: Polyuridylic acid, poly LT; polyq-tidylic arid, poly C:
RNA, rihonucleic :~(*id; I)NA, deouyribonuclric acid; sRNA, soluble rihonuclcir acid; RNAase,
ribonuclcnsc; ATP, adcnosine triphosphat,e; UTP, uridine triphosphxt)r; CTP, cytidine triphos-
phate; and GTP, guanosine triphosphate; PEP, phosphoenolpyruvate; and PEP kinase, phos-
phoenolpgruvate kinase.

: That t,his may be t,he case is suggested by preliminary experiments performed in collaboration
with Harry Gelboin showing that poly U does not stimulnte incorporation of C"-p~`enlal:tnin(:~l~ll:~rline
in a rat liver amino arid-incorporating system.  Similar results have hccn obtained by 8. Ochon
(personal communication).

1 Nirenberg, 11. W., and J. 1%. Mstthaei, V Internxtionnl Congress of Biorhcmistry, bIoscow,
August, 1961.

2 Nircnbcrg, hl. W., and J. H. Matt~hnei, these PRCKXEX~INGS, 47, 1588 (1961).
3 Beljnnski, ill., Biochim. rt Biophys. Acta, 41, 104 (1960).
' Ehrenstein, G. von, nnd I?. Lipmann, these PROCEEDISCS, 47, 941 (1961).
5 Aronson, A. I., E. T. Bolton, Ii. J. Britten, D. 1~. Cowie, J. I). Ducrksen, R. J. RIcCarthy,
I<. McQuillm, and R. 1~. Kobcrts, Carnegie Institute of Kashing:ton, Iwenrbook (1!)59&1%0), p. 220.
6 Nisman, B., and H. E`ukuh:tm, C. R. Acad. Sci. (Paris), 248, 2036 (1959).
7 Siekevitz, I'., J. Biol. Chem., 195, 549 (196'2).
8 Lowry, 0. H., i+i. J. Rosehrough, A. L. Farr, and R. J. Randall, ihid., 193, 265 (1951).
9 Nathans, I)., and F. Lipmann, these PROCEEDINGS, 47, 49i (1961).
`" Hccht, L. I., 11. I,. Strphrnson, and P. C. Zamecnik, ihid., 45, 505 (1959).
`1 We thank Robwt Martin for performing these dcterminntions for us.
I2 Rich, A., and I). It. I)avies, J. Am. Chenr. Sot., 78, R5M (1956).
I3 Felsenfeld, G., xnd Rich, il., Biochim. et Biophys. Acta, 26, 457 (1957).
I4 Benzer, S., and 13. Weisblum, these PRWEEDISGS, 47, 1149 (19Bll.