Approximation of genetic code via cell-free

protein synthesis directed by template RNA

MARSHALL W. NIRENBERG, J. HEINRICH MATTHAEI,!
OLIVER W. JONES, ROBERT G. MARTIN, AND
SAMUEL H. BARONDES

National Heart Institute, National Institutes of Health, Bethesda, Maryland

Arnoucx THE FIELD OF GENETiIcs attained a high
degree of sophistication years ago, the molecular basis of
genetic information storage and retrieval lagged behind.
Only within the last 10 years has the chemical structure
of DNA been established, mainly through the work of
Chargaff et al. (3), Wilkins et al. (29), and Watson
and Crick (28). Some 8 years ago Gamow (5) proposed
a theoretical code and thereby stimulated a great deal of
interest in this problem. The direct biochemical ap-
proach, that is, comparison of the nucleotide sequence
of a gene with the amino acid sequence of its corre-
sponding protein, posed technical problems of great
magnitude. Some of these difficulties have been circum-
vented by the experimental approach summarized here.

DEVELOPMENT OF CELL-FREE ASSAY
FOR MESSENGER RNA

C'*-amino acid incorporation into protein was studied
in a cell-free E. coli system patterned generally after the
systems established by Tissiéres, Schlessinger, and Gros
(25), also by Lamborg and Zamecnik (g) and by Spiegel-
man and his associates (24). A major difficulty in the
study of cell-free protein synthesis in £. coli systems had
been the necessity for preparing fresh enzyme extracts for
each experiment. Techniques were not available for
stabilization and storage of these enzyme extracts com-
parable to those available for mammalian systems. E. coli
extracts were found to be stable in the presence of
mercaptoethanol and could be dialyzed and_ stored
frozen without undue loss of activity (15).

Independently, Tissiéres et al. (25), Kameyama and
Novelli (8), and Nisman and Fukuhara (20) reported
an inhibition of amino acid incorporation by deoxyri-
bonuclease. We also observed such inhibition and studied
its characteristics (15), for here appeared to be a cell-free
system, in which DNA directed the synthesis of protein.
The effect of deoxyribonuclease on C¥-valine incorpora-
tion is shown in Fig. 1. In the absence of deoxyribonu-
clease, C™-valine was rapidly incorporated into protein.

1NATO Postdoctoral Fellow.

55

Deoxyribonuclease completely inhibited incorporation
occurring after 30 min of incubation (15). This type of
experiment suggested that the deoxyribonuclease-in-
hibited system might be dependent on the availability of
messenger RNA. Crude messenger RNA fractions were
found to stimulate amino acid incorporation and proved
to be a requirement for cell-free protein synthesis (17,
18). This technique afforded a highly sensitive, cell-free
assay for messenger RNA and provided the rationale for
all of our subsequent work.

EFFECT OF SYNTHETIC POLYNUCLEOTIDES IN
DIRECTING PROTEIN SYNTHESIS

Discovery of polynucleotide phosphorylase by Grun-
berg-Manago and Ochoa (7) made available to us
chemically defined polyribonucleotides which were
assayed for messenger activity in the cell-free E. cols
system (17, 18). Figure 2 shows the effect of the syn-
thetic polynucleotide, polyuridylic acid,? on incorpora-
tion of C-phenylalanine. In the absence of poly U, very
little phenylalanine was incorporated into protein.
Addition of 10 pg of poly U to a 1-ml reaction mixture
resulted in a 1,000-fold increase in phenylalanine in-
corporation. The synthesized protein was purified by
extraction of precipitated washed protein with 33%
HBr in glacial acetic acid followed by precipitation with
H,O. The synthesized protein was shown to have unusual
solubility and stability properties characteristic of au-
thentic polyphenylalanine (18). The effect of poly U
in directing the incorporation of 18 other C*4-amino
acids into protein was studied. Marked specificity for
phenylalanine was observed; however, a small stimula-
tion of leucine incorporation was observed (3% that of
phenylalanine) (14).

The data of Fig. 3 show that the template activity of
poly U was dependent on its molecular weight (14).

2The following abbreviations are used: poly U, polyuridylic
acid; poly A, polyadenylic acid; octa A, octaadenylic acid; tetra A,
tetraadenylic acid; poly C, polycytidylic acid; TMV, tobacco

mosaic virus.
56 FEDERATION PROCEEDINGS

Volume 22

 

64009, T

2000 T T t T

 

48,000-

 
 

1500;-

T

32,000

T

1000

T

COUNTS / MIN/mg PROTEIN

   

   

a
+ ONAase a

T
COUNTS /MIN /mg PROTEIN

   

¥

500 16,000)

 

 

 

MINUTES
1 L —o—»-

+ POLY

Fic. 1. Cell-free E. coli assay
for messenger RNA: ® Minus
deoxyribonuclease, A 10 yg de-
oxyribonuclease, Ml 500 yg of
1 crude messenger RNA prepared
from yeast ribosomes added at
4 indicated time. For details see
refs. 15 and 18.

FIG. 2. Stimulation of Cl-L-
phenylalanine incorporation by
polyuridylic acid. @ Minus poly-
4 uridylic acid, & [0 yg poly-
uridylic acid added. For details
see ref. 18.

 

MINUTES
PN

 

0 20 40 60 60 9° IS

45 75

FIG. 3. Relationship between

 

100 |

 

600

mumspul | 80

60

()

1

400} CPM. 40

T

my MOLE [pU/FRACTION

AMINO ACID INCORPORATION

40

CPM /my MOLE

200

1

20

T
%

20

 

 

 

0 e

4

 

+19 0

 

molecular weight of polyuridylic
acid and its activity in stimu-
lating C4-phenylalanine incorpo-
ration into protein. Polyuridylic
acid was separated into fractions
of different sizes by sucrose den-
sity gradient centrifugation. Each
fraction was then assayed for its
ability to stimulate C'+-phenyl-
alanine incorporation. For de-
tails see ref. 14.

FIG. 4. Relationship between
guanylic acid content of different
poly UG preparations and their
their abilities to direct C'-valine,
CJeucine, and C'-eryptophan

Phenylatanine
Voline

4, Leucine 4

Tryptophon

 

| 1} J

 

7 9 I 5
= BOTTOM 1S 0 10

FRACTION NUMBER TOP—»

Larger molecules of poly U were separated from smaller
molecules by centrifugation through a sucrose gradient.
The heavier poly U molecules were considerably more
active than the lighter fractions. Poly U of molecular
weight of 50,000-100,000 was very active in this system;
however, the minimum molecular weight necessary
for activity is unknown.

Stoichiometry experiments demonstrated that phen-
ylalanine incorporation was proportional to the con-
centration of poly U in the proper range. The data of
Table 1 demonstrate the quantitative relationship
between phenylalanine incorporated and poly U present
in reaction mixtures. Results of three different experi-
ments are presented. About 1.5 mymoles of uridylic
acid residue in poly U were required to direct incorpora-
tion of 1.0 mumole of phenylalanine (14). These data
alone cannot be used to determine the number of uridylic
acid residues in a phenylalanine coding unit, for addi-
tional results, discussed later, suggest that a molecule of
poly U may function catalytically, by directing the
synthesis of a number of molecules of polyphenylalanine.

Since one or more uridylic acid residues in poly U
appeared to be the RNA code word which corresponded
to phenylalanine, application of this experimental
technique to determine other code words and the general

% GUANYLIC ACID CONTENT OF POLY UG

20 30 40 50 incorporation into protein rela-
tive to Cl-phenylalanine  in-

corporation.

TABLE 1. Relationship between C4-L-phenylalanine incorporated
and polyuridylic acid present

myumole pU

myumole pU (in mumole C!-L-

 

Exp. No, Polyuridytic Acid) phenylalanine mumole cx
phenylalanine

Lf 11.35 g.12 1.24

2 17.7 10.25 1.73

3 25.7 16.9 1.52

In each experiment the amount of polyuridylic acid present
was limiting. Final volume was 0.5 ml. Reaction mixtures were
incubated for 60 min at 37 C as described elsewhere (14).

characteristics of the genetic code appeared reasonable
(17, 18). Randomly ordered polynucleotides containing
different bases were tested by Ochoa and his collabora-
tors (ta, 10, 11, 22, 23) and by ourselves (13, 14) to see
if such polymers would code for different amino acids.
In Table 2 are presented the effects of randomly ordered
polynucleotides on amino acid incorporations. The
composition of the different polynucleotides was ob-
tained by base-ratio analysis. Since cach preparation
of polynucleotide differs in molecular weight, it is difficult
to compare directly the activity of one polynucleotide
with another. Therefore, the figures in the main part of
the table represent the per cent of an amino acid incor-
January-February 1963

TABLE 2. Amino acid incorporation into protein stimulated
by randomly mixed polynucleotides

 

 

 

 

Polynucleotides and Base Ratios
Amino Acid UA ue UG U oes U wn U rs
U _ 0.87 10 = 0.39 U _ 0.76) Ao 0.050] G = 0.152| G = 0.291
A= 0.13 |C = 0.61 |G = 0.24 C = 0.116] C = 0.502] A = 0.034

Phenyl- 100 100 100 100 100 100

alanine
Arginine ° 0 rl 0 49.3 2.9
Alanine 1.9 ° ° 1.0 40.4 0.9
Serine 0.4 | 160 3.2 3.6 | 170 2.3
Proline o 285 oO ° 188 °
Tyrosine 13 ° ° 8 1.0 8.6
Isoleucine 12 1.0 1.0 4.8 5-4 8.4
Valine 0.6 ° 37 0.4 29.8 75
Leucine 4-9 | 79 36 5.1 | 157 44
Cysteine 4-9 o 35 ° 5-4 46
Tryptophan 1.1 ° 14 ° 1.6 23
Glycine 4.7 ° 12 0.5 9.7 15
Methionine 0.6 ° ° 0.6 1.5 8
Glutamic 1.5 ° ° 1.2 0.44 6.2

acid

 

 

 

 

 

 

 

Figures in main part of the table represent incorporation of
any amino acid compared to phenylalanine incorporation ex-
pressed as percentages (mumoles amino acid incorporated/mp-
moles phenylalanine incorporated X 100). Approximately 25
ug of each polynucleotide was added to each 0.5-ml reaction
mixture; these are described elsewhere (14). Samples were in-
cubated at 37 C for 15 min. Incorporation of C1*-phenylalanine
in counts per minute due to the addition of polynucleotides UA,
UC, UG, UAC, UGG, and UGA were 731, 2,900, 714, 804, 2,144,
and 2,744, respectively. The reproducibility of the above per-
centage figures was +3.

porated compared to phenylalanine incorporation
stimulated by the same polynucleotide. Marked poly-
nucleotide specificity in stimulating incorporation is
apparent.

A degenerate code is one in which two or more dif-
ferent coding units can direct the incorporation of the
same amino acid into protein. Since leucine can be
coded by both poly UC and poly UG, the code, with
respect to leucine, is degenerate (13, 14).

Different preparations of poly UG containing in-
creasing percentages of G were synthesized and assayed
(Fig. 4). The relative amounts of valine, leucine, and
tryptophan incorporations, compared to phenylalanine
incorporation, was determined for each polynucleotide.
Leucine and valine incorporation was proportional to
the amount of G in poly UG. However, significant
tryptophan incorporation occurred only when a high
proportion of G was present. These data suggest that
the coding units for both leucine and valine contain one
G, that the coding unit for tryptophan contains two
G’s, and that each preparation of poly UG contained
equal numbers of coding units for leucine and valine.
Similar experiments were performed with other amino
acids and quantitative estimates were obtained of the
number of G, C, or A residues per coding unit.

These results are summarized in Table 3. The proba-
bility of any sequence of three nucleotides occurring in
randomly ordered polynucleotide of analyzed base

GENETIC MECHANISMS 57

ratio can be calculated and may be compared with
the probability of obtaining a sequence of UUU. The
theoretical probabilities of obtaining such triplets,
relative to UUU, are compared in Table 3 with the
experimentally obtained data. Data obtained experi-
mentally agree strikingly well with theoretical predic-
tions. Such quantitative data demonstrate that the
number of G’s, A’s, or C’s in coding units can be de-
termined experimentally.

The sequence of nucleotides in each coding unit is not
known. The current status of the code is similar to that
of an anagram, where the letters of coding units have
been determined, but the sequence is unknown.

Although the uridylic acid content of RNA viruses is
not excessive, a surprisingly high proportion of U has
been found in coding units thus far. This dichotomy
cannot be explained at the present time. However, it is
probable that only coding units containing U have been
selected for by the assay method and that additional
degeneracies without U will be found.

The question may be asked, ‘‘Do the code words agree
with known genetic data?” In Table 4 coding results are
compared with nitrous acid-induced amino acid sub-
stitutions in tobacco mosaic virus protein. Only amino
acid substitutions found more than once have been
cited. TMV protein contains 158 amino acids, and the
entire sequence is known. Treatment of RNA with
nitrous acid results in deamination of nucleotides (12)
and hence mutations (6). Most nitrous acid-induced
amino acid replacements should be due to the replace-

TABLE 3. Summary of code words*

Coding Unit

Amino Acid Composition Theoretical Experimenta}
Phenylalanine UUU... 100 100
Arginine UGG... 67 49
Alanine UGG... 67 40
Serine UUC... 157 160

(UCG)
Proline UCC... 244 285,
Tyrosine UUVA... 13 13
Isoleucine UUA... 13 12
Valine UUG... 32 37
Leucine UUG... 32 36

(UUC) 157 79
Cysteine UUG... 10.6 10.6

(UGG)?
Tryptophan UGG... II 14
Glycine UGG... 1 12
Methionine UGA... 2 8
Glutamic acid UGA... 2 6
Lysine UAA... ? ?

* Sequence of nucleotides in code words is not specified.
Coding ratio is not known definitively. However, assuming a
triplet code, the probability of a triplet occurring in a poly-
nucleotide, relative to UUU, may be calculated from the base-
ratio data presented in Table 2. For example, if poly UG hada
base ratio of 3U/1G, the probability of obtaining the sequence
UUU would be 34 & 34 X 3g = 2%. The probability of ob-
taining the sequence UUG would be 34 XK 34 X 14 = %. Thus,
3 UUU would occur for 1 UUG and, assuming the frequency of
UUU to be 100%, the frequency of UUG would be 33%. Dots
after each ‘‘word’’ are used to indicate the possible presence of
additional uridylic acid residues.
58 FEDERATION PROCEEDINGS

TABLE 4. Comparison of nitrous acid-induced replacements
in tobacco mosaic virus protein with the nucleotide
composition of RNA coding units

 

 

 

iad Mutant Nucleotide C . | ee
trains* Amino Acid Nucleotide Composition ossible
Replacement | of Corresponding | Nysleotide
A B
I 2 Ser UUC... c
J L J
Phe UUU... U
I 2 Glu UAG... A
1 1 1
Gly UGG... G
I 2 Prol UCC... Cc
J L 1
Leu UUC... U
2 Isoleu UDA... A
J L L
Val UUG... G
5 Arg UGG... Cc
4 L L
Gly UGG... G

 

 

 

 

 

* The mutant amino acid replacement data cited and ob-
tained either by A, Tsugita and Fraenkel-Conrat (30), or B,
Wittmann (26).

ment, during virus replication, of either a cytidine by a
uridine or an adenine by a guanine (4).

All of the amino acid replacements with the exception
of the last are the result of a conversion of either a C to a
U or an A toa G. Such results demonstrate confirmation
between amino acid replacement data and proposed
nucleotide compositions of coding units (14). Amino acid
replacements in corticotropins, insulins, and cytochromes
of different species have been examined also. Over 90 %
of these acid substitutions corresponded to a replacement
of only one nucleotide per coding unit. Such data sug-
gest that at least part of the code may be universal.

Nonoverlapping and overlapping codes would be
read as follows:

UUUUUU UUUUUUU

 

 

 

 

Nonoverlapping Overlapping

Brenner on theoretical grounds ruled out overlapping
triplet codes because of the restrictions such codes would
impose on nearest-neighbor frequencies of amino acids
(2). If the code were overlapping and one nucleotide
in a coding unit were replaced with another due to
mutation, amino acid substitutions might be expected
to occur in clusters. The bulk of amino acid substitutions
found in TMV protein do not occur in clusters, and
these data suggest that the code is of the nonoverlapping
type (26, 30, and personal communication from Tsugita
and Fraenkel-Conrat).

A number of polynucleotides which do not contain U
have been tested (Table 5). Poly AC at the indicated
base ratios determined by analysis, were found to direct
small amounts of threonine and proline into protein.
Non-U-containing polynucleotides of different base
ratios, or more sensitive assays, may be necessary to
demonstrate additional degeneracies.

Volume 22

When poly U is mixed with poly A, double- and triple-
stranded helices are formed which are completely in-
active in directing polyphenylalanine synthesis. The
effects of oligo A and poly A on polyphenylalanine
synthesis are presented in Fig. 5. The stability of the
oligo A-poly U helix is influenced strongly by the chain
length of oligo A. Thus octa A forms more stable helices
with poly U than tetra A. It may be seen that octa A
and poly A are very effective inhibitors of polyphenylala-
nine synthesis. In the reaction mixtures used, triple-
stranded helices would be formed (U-A-U). A 99%
inhibition of polyphenylalanine synthesis was observed
at stoichiometric poly A concentrations, that is, 2 U’s
in poly U per 1 A in poly A. In contrast, addition of
poly C had little effect on polyphenylalanine synthesis.
Many mechanisms for enzyme repression can be en-
visioned with strands of RNA base pairing with parts of
DNA or complimentary RNA. It will be of critical
importance to determine whether the type of repression
demonstrated in vitro with poly U and poly A may occur
also in vivo.

FATE OF MESSENGER RNA

H*-poly U was prepared and its fate in reaction mix-
tures was determined (1). Sucrose density gradient
centrifugation experiments revealed that H*-poly U
became associated with a polydisperse peak of ribosomes
which sedimented faster than 70S ribosomes. Almost no
poly U was found on 70S ribosomes. After 3 min of incu-
bation approximately 90% of the poly U disappeared
and was recovered as mononucleotides. The rest of the
poly U still was associated with 100S ribosomes. After
10 min of incubation most of the poly U had been hy-
drolyzed to mononucleotides (90% were 5’-mononu-
cleotides, 10% were 2’, 3’-cyclic mononucleotides or
3’-mononucleotides.) These data show that poly U is
bound to 1008 rather than 70S ribosomes. In addition,
the observed rapid breakdown of poly U suggests that

TABLE 5. Effects of non-U-containing polynucleotides
on amino acid incorporation

Minus
Poly-
C4.1-Amino Acid nucleotide AGz:t AGug:r AC 5:1 AC wos:r
Phenylalanine 10 10 9 10 10
Valine 10 9 8 10 8
Serine 20 20 20 15 18
Methionine 10 10 9 10 Il
Histidine 10 5 IL 10 10
Leucine 30 30 25 28 30
Lysine 10 10 15 13 Lo
Aspartic acid 10 10 10 12 10
Alanine 10 10 It 10 10
Cysteine 20 10 10 20 10
Glycine 10 10 I 10 10
Glutamic acid 10 20 lo 10 20
Proline 10 10 Il 10 70
Isoleucine ro 10 10 10 It
Tyrosine 10 10 8 10 9
Threonine 10 10 9 30 10
Arginine 10 10 9 8 &

Figures represent the incorporation of C'-amino acids in
uumoles. Base ratios were obtained by analysis.
January-February 1963

Fig. 5. Repression of poly-

GENETIC MECHANISMS

on
©

 

 

phenylalanine synthesis by oligo
and poly A. 12.5 mymoles of 140
uridylic acid residues in poly U
and the indicated amount of
oligo A, poly A, or poly C were
incubated at 24C for 15 min be-
fore adding to reaction mixtures.
12.5 mymoles of uridylic acid in
poly U alone stimulated incorpo-
ration of 2,250 counts/min of
C-phenylalanine into protein.
‘Thus 2,250 counts/min repre-
sents 100% incorporation.

FIG. 6. Comparison between
rate of polyphenylalanine syn-
thesis and rate of H’-poly U de-
uradation. Reaction mixtures
used for assay of precipitable
-poly U contained H*-poly U
and G-phenylalanine. Reaction
mixtures used for assay of phenyl-
alanine incorporation contained °

    
    

80

a
oO

+ CLPHENYLALANINE INCORPORATED
>
o

20

Poly U+ Poly A
: =

 

Poly U + Di-A

@ Poly + Pole

Poly U+ Tetra-A

 
 

3
LoS

T

‘

mHMOLES TCA PRECIPITABLE C!4 PHENYLALANINE /ml @—~@

 

 

 

 

CG"-phenylalanine. The 2 reac- o 12.5
tion mixtures, each in a total
volume of 2 ml, were incubated for 40 min and 0.2-ml aliquots
were removed at the indicated time intervals and were transferred
lo either 2 ml of 10% trichloracetic acid at 4C for assay of phenyl-

only the remaining 10-20% directs protein synthesis.
‘These data and the stoichiometry data of Table 1 sug-
gest that poly U functions catalytically; that is, each
poly U molecule directs the synthesis of more than one
polyphenylalanine molecule (14, 1).

Figure 8 illustrates an experiment performed with
unlabeled poly U and C"-phenylalanine. After 3 min of
incubation, phenylalanine incorporation occurred only
un 1008 ribosomes. After 15 min of incubation, insoluble
protein, presumed to be polyphenylalanine, and poly-
phenylalanine associated with both r1ooS and 70S
ribosomes had been formed. These data are in complete
accord with the findings of Risebrough, Tissiéres and
Watson (21), who have reported that rapidly synthesized
RNA in intact cells becomes associated first with 1008
ribosomes.

MECHANISM OF POLYPHENYLALANINE SYNTHESIS

C-phenylalanine-sRNA was prepared enzymatically,
and C!-phenylalanine was found to be transferred to
protein without excessive dilution in the presence of a
large pool of unlabeled phenylalanine. The transfer
required poly U, ribosomes, purified transfer enzyme
(16), GTP, and a GTP-generating system (19). The
function of GTP is unknown, and undoubtedly several
tnzymes are required for the over-all synthesis. Such
txperiments demonstrate that sRNA is an intermediate
in polyphenylalanine synthesis; thus, the initial enzy-
matic reactions are as follows:

phenylatanine-
activating
enzyme

 

L-phenylalanine + ATP

(x)

 

AMP-phenylalanine + P-P

25
my MOLES [pA] or [pC]

 

aE
3
mpMOLES ALCOHOL PRECIPITABLE URIDYLIC ACID RESIDUES/ml 4-4

1 ° 1 | | I | | °
37.5 0 10 20 30 40
. DURATION OF INCUBATION (MINUTES)

alanine incorporation into protcin or into 0.2 ml of magnesium
acetate and 0.4 ml of absolute ethanol at 4C for analysis of pre-
cipitable H®-poly U (1).

2.00 160
Not incubated

 

4 120
+ 80

440

25k
o eee 0

200

 

 

3 Minutes 70S — $00

1.50 300

COUNTS / MINUTE

1.00

o—~ 0.0. 260my

 

a----& TRITIUM,

10 Minutes

1.00 F-

50 CPM

“ b tee

Bottom I 5 10 15 20 25
TUBE NUMBER

 

 

 

Fic. 7. Association of H’-poly U with ribosomes. Reactions were
begun by addition of 20 umoles uridylic acid residues in H*-poly U
to each 0.25 ml of reaction mixture. After incubation at 35C for
indicated times, samples were briefly chilled in an ice bath and
o.2-ml aliquots were layered on top of sucrose gradients at 3C.
Centrifugation and analyses of each fraction were performed as
described elsewhere (1).
60 FEDERATION PROCEEDINGS

 

  

 

 

 

 

q T I T T
2.00 + 800
3 Minutes 708 4
1.50 600
1.00 400
= =
a
3 50 200 &
ow 254 2°
So 0 o °?
> 200 ~| 800 ‘
_ 1
t
1.50 600 *
1.00 —| 400
1
1 _
1
50 + 200
25 ga 4
0 0
| I 5 10 15 20 25
Bottom TUBE NUMBER

Fic. 8. Incorporation of C!‘-phenylalanine into protein on
ribosomes. A regular reaction mixture was prepared except that
C#-phenylalanine (7 % 108 counts/min umole) was used. 20
mumoles of uridylic and residue in H*-poly U were added to 0.25-
ml reaction mixtures, which were incubated at 35C for the indi-
cated times. 0.2-ml aliquots then were layered on sucrose gradients
at 3C. Centrifugation and analyses of each fraction were performed
as described elsewhere (1).

 

 

 

T T t T T T T T T T
3200} 4
2800+ 4
2400F a, 4
5 Fic. g. Stimulation of C-L-
2000+ = valine incorporation into protein
2 by tobacco mosaic virus RNA.
ook 3 4 0.25-ml reaction mixtures were
° incubated at 37C for 90 min be-
00+ 4 fore deproteinization as de-
scribed elsewhere (19).
soo}- 4
400}- 4
ng THY RNA
Pop ii tof it toy fb

 

20 60 100 140 eo

 

Volume 22
phenylalanine-
activating
enzyme
AMP-phenylalanine + sRNA (2)

 

phenylalaninesRNA + AMP
Poly U
ribosomes

transfer enzyme
3

Phenylalanine-sRNA 343 4

(3)

polyphenylalanine + sRNA

MESSENGER ROLE OF VIRAL RNA

The effects of TMV-RNA in directing protein syn-
thesis in this system were studied in collaboration with
Drs. Tsugita and Fraenkel-Conrat of the Virus Labora-
tory, Berkeley, California (27). The effect of TMV-RNA
on C'-valine incorporation into protein is shown in
Fig. 9. Proportionality between valine incorporation
and TMV-RNA added, in the proper range, was demon-
strated. Part of the C'-protein formed a specific pre-
cipitate with TMV antisera. Also, part of the C4
protein formed could be purified from reaction mixtures
with unlabeled carrier TMV protein by repeated iso-
electric precipitations and DEAE column chromatog-
raphy. Virus reconstitution experiments showed that
the C’*-protein product strongly and specifically inhibited
virus reconstitution. However, the small amount of virus
which did reconstitute in the presence of added unlabeled
TMV protein and RNA contained small but significant
amounts of C-protein product.

TMV protein is an excellent one to characterize, for
its entire amino acid sequence is known and digestion
with trypsin converts it into 12 peptides of established
amino acid composition and chromatographic properties.
The C'-protein synthesized under the direction of
TMV-RNA was purified from reaction mixtures with
added unlabeled carrier TMV protein and was treated
with trypsin. The peptides formed after digestion with
trypsin were isolated by Dowex 1 column chromatog-
raphy. Each peptide was subjected to amino acid
analysis and its radioactivity was counted.

The data of Table 6 summarize a small portion of these
results (27). Seventeen peptides were isolated and ana-

TABLE 6. Analysis of peptides obtained from C'-protein synthesized in the presence of
either C'\4-phenylalanine or C¥-tyrosine under direction of TMV-RNA

. . CH-Phenylalanine, C4.Tyrosine,
Peptide No. Amino Acid Composition counts/min counts/min
2 Thro.» Glur.9 Wali.o (Val) (Arg) 8 oO
3 Thro.1 Ser:., Glug. Proe.; Wale; Pher.o (Try, Lys, Arg) 73 5
Ir Aspr.o Thro.» Sero.s Glyz.o Tyri.o (Arg) 10 44
5 Vali.o Tyry.0 (Arg) 3 39
8 ASp2.7 Thrs 9 Glu3.7 Pro,» Ala. Valo.s Tleug.7 Leut.o (leu, Arg) 3
4 Asp2.o Sero.s Pro..o Phei.s (Lys) 96 ("

Reaction mixtures, described in detail elsewhere, contained TMV-RNA and either C-phenylalanine or CH-tyrosine. After
a go-min incubation at 37 C, ribosomes were removed by centrifugation, and the 100,000 X g supernatant solutions were dialyzed
extensively. Unlabeled, carrier TMV protein was added and then was purified from supernatant solutions by repeated isoelectric
precipitations and DEAE column chromatography. Purified TMV protein was digested with trypsin and 17 peptides were sepa
rated by Dowex 1 column chromatography. Amino acid content and radioactivity of each peptide were determined. When yeast
messenger RNA was added to reaction mixtures in place of TMV-RNA, the peptides did not contain significant radioactivity. In
some Cases, peptides were purified further by paper electrophoresis. Data concerning only 6 typical TMV-peptides are presented

here. See ref. 27 for complete details.
lanuary-February 1963

\vzed; however, space permits presentation of only part
of the data. Each peptide in Table 6 is a characteristic
UMV peptide. TMV-RNA directed C'-phenylalanine
Imi not C-tyrosine into phenylalanine-containing
peptides; whereas TMV-RNA directed C'-tyrosine,
but not C'-phenylalanine, into tyrosine-containing
peptides. Although the counts were low, duplicate
experiments gave similar results. In addition, several
peptides were purified further by paper electrophoresis
und still were found to contain radioactivity. The
N-terminal amino acid of TMV protein is acetylated
und the N-terminal peptide did not contain the expected
("-tvrosine, possibly because acetyl CoA was not added
to reaction mixtures. This is the only chemical difference
detected so far. Thus, it appears justifiable to conclude
that TMV-RNA directs the synthesis of a nonacetylated
protein similar to TMV protein in this cell-free E. coli
system. Such results strongly suggest that a large part
of the code may be universal.

SUMMARY

Messenger RNA fractions were found to direct protein
synthesis in cell-free EZ. coli extracts. This system afforded
a sensitive assay for natural and synthetic messenger
RNA and was used both to determine characteristics of
messenger RNA and to decipher part of the genetic code.

Polyuridylic acid specifically directed the synthesis of
polyphenylalanine via an sRNA intermediate. The
activity of poly U varied with its molecular weight;

REFERENCES

1. Baronpes, S. H., anp M. W. Nirensera. Science 138: 810,
1962.
1a. Basitio, C., A. J. Wansa, P. Lencve, J. F. SPEYER, AND
S. Ocuoa. Proc. Natl. Acad. Sei. U.S. 48: 613, 1962.
. Brenner, S. Proc. Natl. Acad. Sci. U.S. 43: 687, 1957-
. Cuarcarr, E., R. Lirsuirz, C. Green, AND M. E. Hopes. J.
Biol. Chem. 192 : 223, 1951.
. Freese, E. Brookhaven Symp. Biol. 12:63, 1959.
. Gamow, G. Nature 173: 318, 1954.
. Gterer, A., AND K. W. Munpry. Nature 182:1457, 1958.
GRuNBERG-Manaoo, M., anv S. Ocuoa. J. Am. Chem. Soc. 77:
3165, 1955.
. Kamevama, T., anp G. D. Novewti. Biochem. Biophys. Res.
Commun. 2: 393, 1960.
g. Lamsore, M. R., anp P. C. Zamecnix. Biochim. Biophys. Acta
42: 206, 1960.
to. Lencyeu, P., J. F. Spever, C. Basitio, anp S. Ocnoa. Pree.
Natl. Acad. Sci. U.S. 48: 282, 1962.
11. Lencye., P., J. F. Speyer, anp S. Ocuoa. Proc. Natl. Acad.
Sa. U.S. 47: 1936, 1961.
12. Levine, P. A., anD R.S. Tipson. J. Biol. Chem. 111: 313, 1935.
13. Martin, R. G., J. H. Martuaer, O. W. Jones, anp M. W.
NIRENBERG. Biochem. Biophys. Res. Commun. 6: 410, 1962.
14. MartHagl, J. H., O. W. Jones, R. G. Martin, AnD M. W.
NIRENBERG. Proc. Natl. Acad. Sci. U. S. 48: 666, 1962.
15. Marruaet, J. H., anp M. W. Nrrenzere. Proc. Natl. Acad. Sci.
U.S. 4.71580, 1961.

won

Oo wap

GENETIC MECHANISMS 61

longer polynucleotide chains were more active than
shorter ones. Evidence was presented which suggested
that each poly U molecule acted catalytically and
directed the synthesis of more than one molecule of
polyphenylalanine. Single-stranded polynucleotides were
shown to have messenger RNA activity but not double-
or triple-stranded polymers. Polyphenylalanine synthesis
was repressed specifically by poly A, which forms hydro-
gen-bonded helices with poly U. Longer oligo A chains
formed more stable helices with poly U than shorter
chains and thus repressed polyphenylalanine synthesis
more effectively.

Addition of poly U to reaction mixtures resulted in the
formation of polydisperse aggregates of ribosomes sedi-
menting faster than 70S. Polyphenylalanine was formed
initially only on 100S ribosomes. Poly U was degraded
rapidly in extracts to mononucleotides, and such deg-
radation was not dependent on protein synthesis.

Randomly ordered polynucleotides were used to direct
many amino acids into protein, and the nucleotide com-
positions of RNA code words corresponding to 15 amino
acids were determined. Two coding units corresponding
to leucine were found; thus, part of the code was shown
to be degenerate.

The number of nucleotides per code word, overlapping
and nonoverlapping codes, and the probable extent of
code degeneracy were discussed. Tobacco mosaic virus
RNA was found to direct the cell-free synthesis of a pro-
tein resembling tobacco mosaic virus protein; thus, at
least part of the genetic code appears to be universal.

16. Naruans, D., anp F. Lipmann. Proc. Natl. Acad. Sci. U. S. 47:
497, 1961.

17. NiRENBERG, M. W., and J. H. Mattuaei. Intern. Congr. Bio-
chem., 5th, Moscow, 1961.

18. Nrrenpere, M. W., ano J. H. Matruaet. Proc. Natl. Acad. Sei.
U.S. 47: 1588, 1961.

19. NIRENBERG, M. W., J. H. Matruaer, anp O. W. Jongs. Proc.
Natl. Acad. Sci. U. S. 48: 104, 1962.

20. Nisman, B., anp H. Fuxunara. Compt. Rend. 249: 2240, 1959.

at. Risesroucu, R. W., A. D. TisstiRes, anp J. D. Watson. Proc.
Natl. Acad. Sci. U.S. 48: 430, 1962.

22. SPEYER, J. F., P. Lencyer, C. Bastuio, ano S. Ocnoa. Proc.
Natl. Acad. Sci. U. S. 48: 63, 1962.

23. SPEYER, J. F., P. Lencvez, C. Basitio, anp S. Ocuoa. Proc.
Natl. Acad. Sci. U.S. 48: 441, 1962.

24. SPIEGELMAN, S. Intern. Congr. Biochem., 3rd, Brussels, 1955, p-
185.

25. TisstiREs, A., D. SCHLESSINGER, AND F. Gros. Proc. Natl. Acad.
Se. U.S. 46: 1450, 1960.

26. Tsucira, A., anD H. Fraenxet-Conrat. J. Mol. Biol. 4:
73, 1962.

27. TsuciTa, A., H. FRAENKEL-ConraT, M. W. NIRENBERG, AND
J.H. Matruaes. Proc. Natl. Acad. Sci. U. S. 48: 846, 1962.

28. Watson, J. D., ano F. H. C. Crick. Nature 171: 737, 1953-

2g. Wiixkins, M. H. F., A. R. Stokes, anp H. R. Witson. Nature
17: 738, 1953.

30. Wirrmann, H. G. Naturwissenschaften 48: 729, 1961