1424 BIOCHEMISTRY: NATHANS ET AL. Proc. N. A. &.

BIOSYNTHESIS OF THE COAT PROTEIN OF COLIPHAGE f2 BY
E. COLI EXTRACTS*

By D. Nartuans,{ G. Norant, J. H. Scuwartz, ano N. D. ZINDER
THE ROCKEFELLER INSTITUTE

Communicated by Fritz Lipmann, June 15, 1962

Nirenberg and Matthaei! have discovered an assay system in which RNA serves
as an activator of protein synthesis in E. col extracts. RNA fractions from cells,'
synthetic polyribonucleotides,?: ? and viral RNA!‘ can all stimulate amino acid
incorporation into acid-insoluble products in £. coli extracts.. Although in each
case the product formed is presumed to be a protein or polypeptide whose structure
is uniquely determined by the RNA, the products have not as yet been completely
VoL. 48, 1962 BIOCHEMISTRY: NATHANS ET AL. 1425

analyzed. A recent paper by Tsugita et al. presents evidence suggesting that the
coat protein of TMV is synthesized by F. coli extracts.’ With the availability of
a bacteriophage attacking Z. coli and containing RNA (coliphage f2),*it became pos-
sible to use the RNA isolated from this phage to stimulate amino acid incorpora-
tion into protein and to identify at least part of the product as the coat protein
of the phage.

Materials and Methods—Phage RNA and protein: Large amounts of the phage f2 were pre-
pared by scaling up the procedure of Loeb and Zinder.t RNA was prepared by deproteinization
of the phage suspension by shaking with phenol and precipitation of the RNA from the aqueous
phase by the addition of two volumes of ethanol. The RNA was kept at — 20° in distilled water
and used as needed. Its concentration was determined by measuring the optical density at 260
my and assuming a specific absorption of 24 per mg. Phage protein was prepared by extraction
of purified f2 in 67 per cent acetic acid followed by dialysis and lyophilization. The following
properties of the protein are pertinent to this paper. On the basis of chromatography on DEAE-
cellulose, starch gel electrophoresis, and ultracentrifugation, the protein appears to be homo-
geneous. Jt has a probable molecular weight of 20,000 and contains 10 leucine, 5 lysine, 3 ar-
ginine, and no histidine residues per molecule.’ Digestion of the oxidized protein by trypsin gives
nine ninhydrin-reacting components which are resolved by two-dimensional electrophoresis; one
of these has the mobility of free lysine. The details of the digestion procedure used in these ex-
periments will be described below.

E. coli extracts: Preparation of the E. coli extracts was based on the procedure described by
Nirenberg and Matthaei.' £. coli B was harvested in the log phase of growth. The washed,
frozen cell paste was ground with 2'/, parts of Alumina (Alcoa A-303) and was extracted with 1'/.
volumes of 0.01 M pH 7.4 Tris buffer that was 0.01 M in magnesium acetate. After the addition
of 5 ug/ml of DNase, the extract was centrifuged at 15,000 X g for 10 min. To the resulting
supernatant, §-mercaptoethanol was added to a final concentration of 0.005 M, and then it: was
centrifuged at 30,000 x g for 30 min. The 30,000 X g supernatant fraction (S-30) was incu-
bated at 35° for 30 min with the components required for amino acid incorporation into protein:
0.003 M ATP, 0.0002 M GTP, 0.01 M phosphoenolpyruvate, 30 ug/ml pyruvate kinase, 0.01 M
glutathione, 0.011 M magnesium acetate, 0.03 M KCI, 0.05 M Tris HCl pH 7.8, 4 x 107-5 M
of each amino acid, 5 ug/ml DNase, 1 mg/ml sRNA, and a volume of S-30 equal to half the total
incubation volume. After overnight dialysis against 0.01 M Tris HCl! pH 7.8, 0.01 M@ magnesium
acetate, 0.03 Af KCI, and 0.005 M f-mercaptoethanol, the preincubated mixture was stored at
~20°.

For the incorporation experiments to be described, the components present during preincuba-
tion (minus additional sRNA and DNase) were incubated at 35° with the appropriate C'‘-amino
acids using an amount of preincubated S-30 (containing 4.6-mg of ribosomal RNA/ml) equal to a
quarter of the total volume. Amino acid incorporation inta proteins was determined by counting
the hot acid-precipitate (5 per cent trichloroacetic acid, 90° for 15 min) after washing with TCA
and ethanol-ether.

Electrophoresis: High-voltage electrophoresis® of tryptic digests of protein (fingerprinting) was
performed on Whatman 3 MM paper in a Servonuclear Company tank under Varsol (Standard
Oil Co., N.J.). The first dimension was run for | br in pH 4.7 buffer containing by volume 25
mil of acetic acid and 25 ml of pyridine/liter of water. The field strength was 50 volts/em. At
the completion of the run, the paper was dried and the strip containing the peptides cut out and
sewn onto another piece of paper. The second dimension was run for 3-4 hr in pH 1.9 buffer
containing 87 ml of acetic acid and 25 ml of 88 per cent formic acid/liter of water. The field
strength was 20 volts/cm.

Radioactive amino acids: C'-leucine, -arginine, and -lysine with specific activities of 144 mC/
mmole and H-leucine with specific activity of 5 C/mmole were obtained from New England Nu-
clear Corporation, Boston, Mass. C'* counting was done with a windowless gas flow counter.
H? and C™ in the same sample were determined by counting with and without a thin window
and then making the appropriate corrections. :

Results —Characteristics of the amino acid incorporating system: The following
. 1426 BIOCHEMISTRY: NATHANS ET AL. Proc. N. A. S.

 

 

 

 

 

A
0.33 mg RNA /mi
15000 -
4000+
0
®& 3000F
2
9
2D
S 0.13 mg RNA/mi 2
9g . 2
& 3°
a
= 5
G 2000 £
=
a.
- Vo
1000 |-
O RNA
1 1 1 1
30 60 90 120 1 1 4 i J
Minutes OF (OR 03 0.4 05
mg RNA

Fic. 1.—-Kinetics of leucine incorporation with .
different concentrations of f2 RNA. The __ Fic. 2.—Effect of increasing concentra-
incubation mixtures were as described under tion of f2 RNA on leucine incorporation.
Methods and contained C-leucine (1.5 X 107 The components given in Methods were
epm/umole) and the concentration of RNA incubated with C™-leucine (1.5 X 107 epm/
indicated. Each point represents the acid- mole) at 35° for 90 min. Total volume was
insoluble counts in 0.20 ml. 0.50 ml.

experiments show the extent of the stimulation of amino acid incorporation by
E. coli extracts when exposed to f2 RNA. Figure 1 shows the kinetics of incorpora-
tion of leucine at different RNA concentrations. Both the rate of incorporation and
the final amount of product vary directly with the RNA concentration. In Figure
2, the effect of increasing RNA concentration on the amount of leucine incorporated
is shown. With the concentration of components used, there is a linear response
up to 200 ug/ml and then a sharp break in the curve. Above this concentration, a
smaller response per unit of RNA added is observed. Saturation is not achieved
even. at our highest RNA concentration, which stimulates some 40-fold. In
the linear portion of the concentration curve, 0.88.mp moles of leucine is incorporated
per 100° yg of RNA. From this value, one can castulate that about 30 nucleotide
residues are required for each molecule of amino acid incorporated. This estima-
tion is based on the specific activity of the leucine, on the percentage composition
of the leucine in the phage protein, and on the assumption that any other proteins
synthesized have the same percentage of leucine as does the f2 protein.

To demonstrate that the reaction has the properties expected of a protein-
synthesizing system, the antibiotics chloramphenicol and puromycin were added
to the system. Both of these inhibitors of protein synthesis inhibited the incorpora-
tion of leucine (Table 1).

Identification of the product: Two kinds of experiments are to be described by
which we characterize the product. The first involves chromatography of the
Vor. 48, 1962 BIOCHEMISTRY: NATHANS ET AL, 1427

TABLE 1
ErrecT oF PUROMYCIN AND CHLORAMPHENICOL ON RNA-DEPENDENT PROTEIN SYNTHESII

C'+leucine

Conditions {cpm incorporated)
Complete system 3650
+ puromycin (4 X 1074 M) 44
+ chloramphenicol (3 x 1074 M) 163
(Complete system minus f2 RNA 235)

The components given in Methods were incubated with C'+leucine (1.5 X 107 cpm/pmole) at 35° for 60 min.
Total volume was 0.25 ml, containing 50 we of f2 RNA.
product, the other analysis of tryptic digests of the partially purified product. Of
the counts incorporated into the acid-insoluble fraction, 45-55 per cent remains
bound to the ribosomes when the ribosomes are removed by centrifugation, and the
remainder is found in the 105,000 X g supernatant. It is this latter fraction only
that we are studying. Analyses were done only when there was about a 40-fold
stimulation over the background incorporation by the extracts so that the unstim-
ulated product may essentially be neglected.

In the first experiment, the product was prepared by incubating C’4-leucine with
the components described in Methods. The ribosomes were removed by centrifu-
gation at 105,000 X g for 2 hr; 3 mg of f2 protein were added to an aliquot of the
supernatant and this material was chromotographed on DEAE-cellulose in 3
urea (in the absence of urea, f2 protein cannot be eluted) with a gradient of 0.1
to 1.0 M potassium phosphate, pH 7.4. Of the 3,000 acid-precipitable counts
applied, 60 per cent was recovered of which 45 per cent was in the f2 protein peak.

The next two experiments take advantage of the reproducibility of the finger-
prints of f2 protein. In the first of these experiments, the product was prepared
by incubating the components described in Methods with 0.5 mi of preincubated
8-30, 0.8 mg of {2 RNA, and C'-arginine and lysine in a total volume of 2 ml at
35° for 90 min. After removal of the ribosomes by centrifugation at 105,000
X g for 2 hr, the supernatant contained 640,000 epm in the hot acid-precipitable
fraction. To an aliquot containing 140,000 cpm was added 3 mg of f2 protein and
a large excess of C}*-arginine and lysine. The f2 protein was precipitated with
ammonium sulfate at 20 per cent saturation, and the precipitate was washed with
ammonium sulfate and suspended in water. By this procedure, 98 per cent of the
acid-insoluble counts was precipitated with the carrier f2 protein. Although only
2 per cent of the product was in the supernatant, 3 mg of carrier {2 protein was
added to this fraction, and it was carried through the procedure described below as
a control for the adsorption of free C'*-amino acids by the carrier protein.

The two fractions were prepared for trypsin digestion by first precipitating and
extracting with 5 per cent TCA at 90° for 15 min, followed by two TCA washes at
room temperature, and then washing with ethanol-ether and ether. The dried
product was subjected to performic acid oxidation by a modification of Hirs’s pro-
cedure.* The oxidized protein was then digested for 3 hr at 25° in 1 ml of 0.05
M ammonium bicarbonate pH 7.9 containing 4 per cent trypsin by weight of sub-
strate. After lyophilization to dryness, the digest was dissolved in 20 ul of water,
an aliquot containing 105,000 cpm was applied to paper, and the peptides were
separated by two-dimensional electrophoresis.

The electropherograms were radioautographed for five days and stained with
ninhydrin to identify the peptides. Figure 3 shows the electropherogram of the
1428 BIOCHEMISTRY: NATHANS ET AL. Proc. N. A. S.

3s

 

i ese
uae peo

¥

Fic. 3.—Electropherogram and radioautogram of the tryptic peptides from C-lysine- and
arginine-labeled product and carrier f2 protein. A is the ninhydrin-stained paper, B is the radio-
autogram. ‘‘Arg’’ and “‘lys’’ represent markers; O indicates the origin.

ammonium sulfate precipitate. As the product was labeled with C!4lysine and
arginine, and since trypsin cleaves specifically at the carbonyl groups of lysine and
arginine, we would expect that all the peptides would be labeled with the exception
of the carboxy-terminal peptide. The nine peptides of {2 are marked 1 through 9.
It may be noted that all peptides specifically identified with f2 except peptide num-
ber l are labeled. The congruence of the ninhydrin spots with the radioactive spots
is immediately apparent even though the film is not superimposed here on the elec-
tropherogram. On the other hand, the electropherogram of the ammonium sulfate
supernatant had the same ninhydrin peptides but the radioautogram was blank.
This result was to be expected since, as already mentioned, 98 per cent of the TCA-
precipitable counts was found in the original ammonium sulfate precipitate. This
negative result does show, however, that the correspondence of ninhydrin and radio-
active spots is not due to adsorption of C'-lysine and arginine.

A similar control experiment was carried out using egg white lysozyme (Worthing-
ton) as carrier in place of {2 protein. In this instance, it was necessary to make the
solution 60 per cent saturated with ammonium sulfate in order to precipitate the
lysozyme. In other respects, the experiment proceeded as described above except
that 50,000 cpm of the lysine and arginine-labeled product were applied to the paper.
Although we obtained a whole new spectrum of ninhydrin spots in the electrophero-
gram, the spots that appeared on the radioautogram (Fig. 4) corresponded with
those of {2 protein and not with the lysozyme.

Note added in proof: Using C'™-lysine and -arginine, we have repeated the
experiments described above with RNA isolated from E. coli K12 W6, starved of
methionine (courtesy of L. Mandel and E. Borek), and with RNA isolated from
Vou. 48, 1962 BIOCHEMISTRY: NATHANS ET AL. 1429

 

 

    

Fic. 4.—Electropherogram and radioautogram of the tryptic peptides from C‘-lysine- and

arginine-labeled product and carrier lysozyme. A is the ninhydrin-stained paper, B is the radio-
autogram. .
TMV (obtained from D. Caspar). Fingerprints of each product with f2 protein
added as carrier showed no correspondence of radioactivity with the ninhydrin
spots. In addition, the product prepared with f2 RNA was fingerprinted with
TMV protein. Again there were no overlapping spots. We conclude that the
product formed in the presence of {2 RNA is peculiar to the nucleic acid of the
bacteriophage.

The next experiment used a second isotopic label to identify those f2 peptides
containing leucine. Carrier protein labeled with H*-leucine was to be mixed with
the product labeled with C'-leucine. If the product was identical to the phage
protein, those peptides with H3, 1.e., containing leucine, would also have C™, and
those peptides without H’, i.e., containing no leucine, would not have C™.

A small amount of H?-leucine-containing phage was made by growing f2 on a leu-
cine-requiring strain of the host bacteria in the presence of high-specific-activity
H*leucine. This phage was added to a large amount of pure, unlabeled phage, and
protein was prepared. Fingerprinting of this material followed by elution and trit-
ium counting showed that only peptides 2 and 4 contain leucine. C'4-leucine-
labeled product was prepared as described above except for the change in C'4-
amino acid. In this instance, the 105,000 X g supernatant contained 810,000
acid-insoluble cpm. An aliquot with 100,000 cpm (C') was added to the mixed
carrier protein containing 200,000 cpm (H?), and this material was precipitated and
digested as previously described. In this experiment, only the ammonium sulfate
precipitate was analyzed since, again, nearly all of the TCA precipitable counts were.
in this fraction. The ninhydrin-stained peptides were eluted from the electro-
pherogram with hot water and an aliquot counted to determine the C' and H3
content. In Table 2, we can see that only peptides 2 and 4 contain substantial
1430 BIOCHEMISTRY: NATHANS ET AL. Proc. N. A. S.

TABLE 2
Raptoactivity or f2 PErripes
cu Ht

Peptide {epm) (cpm) C“/He
1 200 300 a
2 1,100 2,510 0.44
3 355 0 Lee
4 2,960 10,600 0.28
5 0 80 Lee
6 40 0
7 0 24 Lee
8 232 0 Lee
Origin 4,400 7,000 0.63
Total 9,290 20, 500 a
cpm applied 47 ,000 104 ,000 0.45
% recovered 20 20 a

radioactivity and each has both C'4 and H?. The ratio of the two isotopes in each
peptide is approximately that of the original material applied to the paper. Only
a few faint spots were seen on the radioautogram to indicate other C'*-containing
material; these had no corresponding ninhydrin spots nor did they contain H’*.
The product thus resembles f2 protein. It may be noted (Table 2) that only about
20 per cent of the counts applied to the paper were recovered. The balance
of the counts could not be located even when elutions were performed with 6 N HCl,
nor were they found in the solvents used for the electrophoresis. Since both product
and carrier are lost proportionately, these unexplained losses do not alter the con-
clusion that f2 protein is being made.

Discussion.—The experiments presented leave little room for doubt that the syn-
thesis of {2 coat protein can be mediated by f2 RNA when it is added to E. colt
extracts. Since the isotopic labels appear primarily in the f2 peptides, we conclude
that the coat protein forms a major component of the product that is released from
the ribosomes. Preliminary evidence based on the fact that {2 RNA also stimu-
lates the incorporation of histidine, an amino acid which is lacking in the coat pro-
tein, indicates that other protein is also being made. In this instance, more than
half of the radioactive protein remains bound to the ribosomes.

An important question raised by these results is whether viral RNA acts directly
as a template or through some intermediate such as its complementary strand.
Although the hypothesis that RNA acts directly is more appealing, the latter pos-
sibility cannot, as yet, be excluded.

With the demonstration that viral RNA directs the synthesis of the coat protein
in cell extracts, problems relating to viral replication can now be studied in an ex-
perimental system more easily controlled than the infected cell.

* Supported by grants from the National Foundation, the National Science Foundation, and
the National Cancer Institute, National Institutes of Health, U.S. Public Health Service.

{ Post-doctoral fellow of the U.S. Public Health Service. Present address: Department of
Microbiology, The Johns Hopkins University School of Medicine, Baltimore, Md.

1 Nirenberg, M. W., and J. H. Matthaei, these Procrepines, 47, 1588 (1961).

2 Matthaei, J. H., O. W. Jones, R. G. Martin, and M. W. Nirenberg, these ProcrEpiINes, 48,
666 (1962).

2 Lengyel, P., J. F. Speyer, C. Basilio, and S. Ochoa, these ProcEEDINGS, 48, 282 (1962).

‘ Ofengand, J., and R. Haselkorn, Biochem. Biophys. Res. Comm., 6, 469 (1962).

5 Tsugita, A.. H. Fraenkel-Conrat, M. W. Nirenberg, and J. H. Matthaei, these PRocEEDINGs,
48, 846 (1962).
VoL. 48, 1962 BIOCHEMISTRY: NEWTON AND SNELI. 3431

6 Loeb, T., and N. D. Zinder, these Procrrpines, 47, 282 (1961).

7 Notani, G., and N. D. Zinder, unpublished observations.

8 Ingram, V. M., Nature, 178, 792 (1956); Katz, A. M., W. J. Dreyer, and C. B. Anfinsen,
J. Biel. Chem., 234, 2897 (1959).

° Hirs, C. H. W., S. Moore, and W. H. Stein, J. Biol. Chem., 219, 623 (1956). Two ml of formic
acid were mixed with 0.1 ml of 30 per cent hydrogen peroxide (superoxal) and allowed to stand
for 2 hr at 0°. One-tenth ml of this mixture was added to the dried protein and was incubated
for 1 hr at 0°, Oxidation was stopped by the addition of 1 ml of water and the protein was lyo-
philized.