BACTERIOPHAGE-SPECIFIC PROTEINS IN E. COLI INFECTED WITH
AN RNA BACTERIOPHAGE*

By Dantrex Natuans, Max P. Orscuasr, KATHLEEN EGGEN,
AND YOSHIRO SHIMURA

DEPARTMENT OF MICROBIOLOGY, JOHNS HOPKINS UNIVERSITY SCHOOL OF MEDICINE,
BALTIMORE, MARYLAND

Communicated by W. Barry Wood, Jr., October 28, 1966

The f2 group of RNA coliphages ({2, MS2, R17, fr), first isolated by Loeb and
Zinder,' are closely related, small viruses with single-stranded RNA of about one
million molecular weight (for a review of RNA phages, see ref. 2). The number of
proteins encoded in the viral RNA is therefore small. To date, two phage-specified
proteins have been identified chemically or enzymatically in infected cells, the
phage coat subunit and an RNA synthetase,*~> and the presence of a third protein
has been inferred from genetic experiments. 7 By analysis of radioactive proteins
present in phage-infected cells, Haywood and Sinsheimer have provided evidence for
the presence of four phage-specific, noncoat proteins in cells infected with coli-
phage M82.8

We have recently described a method for labeling phage-specific proteins in E.
coli infected with MS2 under conditions where host protein synthesis is virtually
eliminated by actinomycin.* Such cells synthesize predominantly phage coat
protein, the bulk of which is made later in the infectious cycle than is noncoat
protein. The identification of noncoat protein in those studies depended on the
incorporation of radioactive histidine, which is not present in the virus coat. In
this report. we describe the fractionation of phage-specific proteins by polyacryla-
mide gel electrophoresis. The results show that in addition to phage coat, three
discrete proteins are present in infected cells. One of these proteins is missing in
cells infected with an amber mutant with a defect in phage maturation. In con-
firmation of our earlier results,* it was found that coat protein constitutes about 80
per cent of the total phage-specific protein synthesized in the infected cell. In
addition, preliminary findings are presented on the identification of noncoat proteins
made in cell extracts under the direction of MS2 RNA and on the presence of minor
proteins in the phage particle itself.

Materials and Methods.—The source of most of the bacterial strains and of bacteriophage MS2
has been given previously.* Bacteriophage f2 and an amber mutant of f2 (sus-11) were kindly
supplied by N. D. Zinder, as were E. coli K38, a nonpermissive host, and EZ. col: K37, an amber-
suppressing strain.!! . ,

The procedures for sensitization of E. col C3000 to actinomycin and for incorporation of amino
acids into protein after infection with MS2 have been described.*

To prepare the radioactive protein for electrophoresis, the labeled cells were centrifuged at
10,000 g for 5 min, washed once with cold saline containing nonlabeled amino acids, and the cell
pellet was resuspended in a small volume of water. The cells were broken by sonication for 3
min and all labeled protein was solubilized by addition of '/:0 vol of glacial acetic acid, freshly
deionized urea to a concentration of 0.5 M, sodium dodecy] sulfate (SDS) to a concentration of
1.0%, and mercaptoethanol to a concentration of 0.01 M.2 After incubation at 37° for 60 min,
the solution was centrifuged at 30,000 g for 10 min, and the supernate, which contained all the
radioactive protein, was passed through coarse G25 Sephadex equilibrated with 20% sucrose,

0.1% SDS, and mercaptoethanol 0.01 M. For electrophoresis of the protein effluent, concen-
trated buffer was added to give the appropriate concentration (see below).

1844
Vox. 56, 1966 BIOCHEMISTRY: NATHANS ET AL. 1845

A B C D

 

Fie. 1—Radioautogram of dried gels, prepared as described in the text. (A) C!4pro-
teins prepared from growing £. coli in the absence of actinomycin. A 0.2-ml sample con-
taining 50,000 cpm, derived from 0.1 ml of the original culture, was applied to the gel.
(B) C'-proteins prepared from actinomycin-treated, uninfected cells; 1000 cpm, derived
from 0.3 ml of culture, was applied. (C) C*+-proteins prepared from actinomycin-treated
cells infected with MS2 at a multiplicity of 20; 11,000 cpm, derived from 0.3 ml of culture,
was applied. (D) C'-amino acid-labeled phage particles, grown in actinomycin-treated
cells and purified as indicated in the text; 11,000 cpm was applied. Electrophoresis was

carried out at pH 7.2, and the dry gels were exposed to X-ray film for 11 days. The origin
is indicated by the arrows.

 

Polyacrylamide gel electrophoresis was carried out at 25° at two different pH’s: pH 8.9, using
the procedure described by Davis, or pH 7.2, using the procedure described by Summers e al.'*
In all cases 0.1% SDS and 0.5 M urea were in the gels, and 0.1% SDS in the buffers. Electro-
phoresis at pH 8.9 was done in 10-cm-length 71/.% polyacrylamide gels at 5 ma per tube and
run until the tracking dye:(bromphenoPblue) had moved 9 cm (about 3 hr). At pH 7.2, 10%
polyacrylamide gels were used. They were 6 cm in length and were run at 5 ma per tube for 4
hr, at which time the tracking dye had moved about 4 cm.

Radioactive proteins in the gel were detected by two procedures: radioautography of dried
slices of gel, as described by Fairbanks, Levinthal, and Reeder,‘ or by slicing the gel into 1-mm
segments, eluting with 0.1% SDS, and counting the eluted protein in a liquid scintillation counter.
With the latter procedure, about 90% of the applied radioactivity was recovered. The radio-
autograms were traced with a Joyce-Loebl recording microdensitometer.

Protein synthesis in cell extracts was carried out as previously described, except that the S30
fraction was prepared from £. coli strain Q13, kindly supplied by W. Gilbert. The product was
prepared for electrophoresis as described for the infected cell extract, except that acetic acid was
omitted, EDTA (pH 7.6) was present at a concentration of 0.05 M, and mercaptoethanol at a
concentration of 0.05 M.

The procedure followed for the purification of MS2 has been described.”

Chemicals were obtained from sources noted previously.®

Resulis—Number of phage-specific proteins made in infected cells: In the first
experiment to be described, actinomycin-treated bacteria were infected with MS2
in the presence of C4-lysine, C'4-arginine, C!4-isoleucine, and C!*-valine, and at the
end of 50 min. the proteins were solubilized and prepared for electrophoresis as
decribed in Methods. (By 50 min after infection, the synthesis of noncoat proteins
has ceased and the amount of coat protein present is nearly maximal.®) A second
1846 BIOCHEMISTRY: NATHANS ET AL. Proc. N. A.S.

portion of the culture which had not been infected, and a third portion which had
not been treated with actinomycin nor infected, were prepared for electrophoresis
in the same way. Electrophoresis of each sample was performed at pH 8.9 and at
pH 7.2 as described in Methods. In addition, samples of MS2, grown in actinomy-
cin-treated cells in the presence of the C'*-amino acid mixture and purified in CsCl
as previously described,'® were also solubilized and electrophorized. At the end of
the electrophoretic run, the gels were sliced, dried, and applied to X-ray film. The
results are presented in Figures 1, 2, and 3.

Figure 1 shows a typical radioautogram of the dried gels, and Figures 2 and 3 are
microdensitometer tracings of the radioautograms. As shown in the figures,
normally growing FE. coli showed a variety of labeled proteins, whereas actinomycin-
treated E. colt showed a specific pattern of protein synthesis, i.e., the persistent
synthesis of a small number of proteins. (A similar observation was reported by
Haywood and Sinsheimer.’) In actinomycin-treated, infected cells, four new
protein peaks appear (Figs. 1, 2C’, and 3C), one of which is present in great excess
and corresponds to phage coat protein. As seen in the figures, one of the noncoat
proteins is present in small amount, and this component appears to be somewhat
variable in electrophoretic mobility in different preparations. As shown by the

 

 

Fig. 2.—Densitometer tracings of the Fic. 3.—Densitometer tracings of radioauto-
radioautograms shown in Fig. 1. A, B,C,D grams of C'*-proteins electrophorized at pH 8.9.
are the same as noted in Fig.1. The originis (A) Normal E.coli. (B) Actinomycin-treated E.
indicated by the vertical line or first peak on coli. (C) Actinomycin-treated E. coli infected
the left. Phage-specific proteins are num- with MS2. (D) Purified MS2_ grown in actino-
bered 1-4. The 3rd component is more evi- mycin-treated cells. Origin is indicated by ver-
dent in Figs. 4 and 6. tical line or first peak on the left.
Vow. 56, 1966 BIOCHEMISTRY: NATHANS ET AL. 1847

results of other electrophoretic
runs, the amount of radioactive
material at or near the origin is i
quite variable and often entirely

}

absent (Fig. 4). It is presumably \ i
an aggregate of radioactive protein Aon } '
and not a separate protein com- [ Ne tr \sanvemmnnne

ponent. .
oe Fie. 4.—Densitometer tracing of protein from
That no other phage protein 18 actinomycin-treated, infected cells, showing no radio-
present in the coat fraction is active material at the origin. The sample was pre:
pared as described in Methods and electrophorized at

shown by the electrophoretic pat- pH 7.2. The origin is at the left.
tern of C**-histidine-labeled pro-
teins, isolated from actinomycin-treated, infected cells as described above. Asshown
in Figure 5, phage-specific proteins labeled with histidine, an amino acid lacking
in MS2 coat protein, are present in two major peaks and a third minor (and vari-
able) peak, but not in the coat protein region. (The control peaks can be identi-
fied from Figs. 2 and 3.) These results also establish that the first and second
protein peaks are not aggregated coat protein.

In order to detect additional phage-specific proteins which might be present in
very small amounts, gel slices were exposed to film for longer periods. After

=

_Fia. 5.—Densitometer tracings of C« Fic. 6.—Densitometer tracings of radio-
histidine labeled ere ° ee autograms oa bes — Jong. exposure to
trophoresis at pH 7.2; 5000 cpm was applied putogram and tracings out eown ed Figs 1 and 2
and the film exposed for 12 days. (B) Elec- were exposed to X-ray film for 33 days and
trophoresis at PH 8.9; 7000 cpm was spplied = the ‘film was traced. (4) Actinomycin-
and the film exposed for 5 days. The origin treated. coli. (B) Actinomycin-treated
1s at the lett. E. coli infected with MS2. (C’) Purified MS2

grown in actinomycin-treated cells.

   
1848 BIOCHEMISTRY: NATHANS ET AL. Proc. N. A. S.
A
A I".

[a : }

A; '

I !

A Woy iv a fo
~~ etal! \

c .

Fig. 7.—Analysis of proteins from cells infected with sus-11 or its revertant. The figure on the
left shows the proteins labeled with C'-histidine; that on the right shows the proteins labeled with
C-arginine, C*4-lysine, C'isoleucine, and C-valine. (A) Actinomycin-treated cells. (B)
Actinomycin-treated cells infected with the sus-11 revertant. (C) Actinomycin-treated cells

infected with sus-11. The origin corresponds with the first peak at the left of each tracing, and
the arrow indicates the position of coat protein in a companion gel.

exposure for 33 days, the same gels used for Figures 1 and 2 yielded the tracings
presented in Figure 6. As shown in Figure 6A, a new trace component is seen in
uninfected cells, but no new phage-specific proteins are apparent (Fig. 6B). An
interesting observation emerges, however, in the tracing of the protein from purified
phage particles (Fig. 6C). As seen in Figure 6C, two minor peaks are present, one
of which corresponds to phage-specific peak 2 and the other to phage-specific peak 3.
Neither peak 1 protein nor the protein seen in uninfected cells is detectable in the
phage.

Absence of a specific protein in an amber mutant of f2: “One way of identifying the
proteins separated by electrophoresis is to examine the phage-specific proteins
formed in nonpermissive cells infected with amber mutants of the RNA phage,
since the defective gene gives rise to incomplete polypeptide chains."* For this
purpose, actinomycin-treated E. coli C3000, a nonpermissive strain, was infected
with sus-11, an amber mutant of f2 which has nonsense mutations in both the coat
protein cistron and in the cistron which codes for a protein required for the protec-
tion or organization of viral RNA.” The labeled amino acid in one experiment
with this mutant was histidine, so that the mutation in the coat protein could be
ignored, and in a second experiment a mixture of C!4-amino acids was used. For
comparison a companion batch of cells was infected with a revertant of sus-11 which
was selected on the nonpermissive host, FE. coli K38. A third batch of cells was not
infected. The amounts of C-histidine incorporated into total protein (hot TCA-
insoluble radioactivity®) after 50 minutes were as follows: uninfected cells, 150
cpm per ml of culture; sus-11 infected cells, 400 cpm; sus-11 revertant infected
cells, 1200 cpm. The results of electrophoresis of these preparations are shown in
Figure 7. As shown in the figure, cells infected with sus-11 lack the second protein
Vou. 56, 1966 BIOCHEMISTRY: NATHANS ET AL. 1849

 

 

 

 

 

- 25000
O00F 4 fe
10000+-
ooo 12000 +
c
§ €
E 3 8000
~ wo ‘mad
5 4000, #
eo
2000- 4000/-
1 i n I i l A. 1
10 20 30 40 10 20 30 40
FRACTION NO. FRACTION NO.

Fie. 8.-—Electropherogram of proteins made in cell extracts with MS2 RNA as messenger.
Gels were sliced as described in Methods, and protein was eluted with 0.1% sodium dodecyl
sulfate and counted. The first segment includes the origin. (A) Proteins made in in-
fected cells, labeled with C'amino acid mixture. (B) Proteins made in cell extracts,
labeled with C!Camino acid mixture.

component. This result establishes that the first and second proteins are not
aggregates of the same subunit, and identifies peak 2 as the ‘““RNA-protecting pro-
tein.”” Similar studies are now in progress with a series of amber and temperature-
sensitive mutants of MS2 to identify the other protein peaks.

Electrophoretic analysis of proteins synthesized in E. coli extracts with MS2 RNA as
messenger: Bacteriophage RNA has been shown to direct the synthesis in EF. coli
extracts of intact phage coat as well as histidine-containing, noncoat proteins.™: 28. 18
The lack of purified reference proteins of the latter type has precluded more precise .
identification of these proteins. By applying the same techniques employed in the
analysis of phage-specific proteins formed in infected cells, we have identified three
principal protein products in the cell-free system (Fig. 8). In a separate experiment,
the first and second peaks bave been shown to contain histidine and are therefore
not coat protein aggregates. In double-label experiments, in which C'*-cell-free
product was mixed with H?-infected cells and the mixture prepared for electro-
phoresis, the first peak seen in Figure 8B corresponded in mobility to peak 1 from
infected cells, and the third peak seen in figure 8B corresponded to coat protein.
However, the middle peak of the cell-free product did not coelectrophorese with any
of the proteins seen in infected cells, although it was similar in mobility to the vari-
able peak 3. We presume that this peak represents altered RNA-protecting protein.
As shown in Table 1, protein closely related to coat protein in electrophoretic
mobility constituted about 70 per cent of the total protein product, and each of the
other components was about 15 per cent. These relative amounts are similar to -
those found with phage-specific proteins made in infected cells (Table 1). More
detailed studies of the proteins made in cell extracts are in progress in order to es-
tablish their relationship to the proteins found in infected cells.

Discussion.—Three cistrons have been identified in the f2 group of RNA coli-
phages by complementation tests:* 2! one for the coat protein subunit, one for an
RNA-synthesizing enzyme, and a third for an ‘““RNA-protecting protein.” In the
present study, four phage-specific proteins have been -:detected in infected cells.
1850 BIOCHEMISTRY: NATHANS ET AL. Proc. N. A. 8S.

TABLE 1
Perr Cent or Tota Counts in Each Peak
—_——————Total Counts (%)——_________~
Source of proteins lst Peak 2nd Peak Coat protein
Infected cells (a) 8 ll 81
(b) 10 12 78
Cell extract (a) 13 16 71
(b) 14 17 69

These results were obtained from the experiment described in the legend of Fig. 8. (a) and
(b) are duplicate electrophoretic runs.

One of these is coat protein, a second is the ‘‘RNA-protecting protein,”’ and we infer
that a third is the RNA synthetase. In addition to these known proteins, a fourth,
variable protein component is found in infected cells. Whether the latter is the
product of a fourth cistron, whose function is not vital for virus development, or is a
breakdown product or aggregate of one of the other proteins is not as yet clear.

We have already commented on the striking excess of phage coat protein produc-
tion in infected cells, compared with the other proteins.? Even at 50 minutes post-
infection, the amount of coat protein is seven times that of any other protein com-
ponent. Similarly, in cell extracts the amount of phage coat produced is also in
great excess (4'/.-fold). Moreover, both in infected cells and in cell extracts the
two reproducible minor protein peaks are about equal. These similarities suggest
that the same regulatory mechanisms may be operating in each case. The avail-
ability of a simple procedure for quantitating the different phage proteins synthe-
sized offers the opportunity of exploring these mechanisms both in the infected cell
and in cell-free extracts.

The preliminary findings on the presence in phage particles of two minor phage-
specific protein components suggest that the RNA phage is structually more com-
plex than thought heretofore. The possibility that RNA coliphages contain a
protein which protects the RNA from nuclease cleavage and perhaps serves as the
site of attachment to cells is suggested by the properties of certain mutants of the
phage,*:7 and by the effect of 5-fluorouracil ih producing RNA-deficient nonabsorb-
ing particles. 22. A rough estimate of the amount of the minor components found
in phage particles, one of which corresponds to the ‘“RNA-protecting protein,”
indicates that a phage particle may have a single molecule of each. At this level,
however, it is difficult to distinguish structural protein from contaminating protein,
and other experiments are being done to settle this point satisfactorily. Also, it re-
mains to be shown that the two minor components are actually different proteins,
since one might be a breakdown product of the other.

A final comment concerns the selective inhibition of protein synthesis by actino-
mycin. As shown by the electropherograms of protein from actinomycin-treated,
noninfected cells, only two major radioactive peaks and one trace component are
detectable. Of these, the fastest-moving peak is probably not protein, since it is
present in maximal amount in samples of cells taken immediately after addition of
radioactive amino acids. This result therefore suggests that a specific EZ. coli pro-
tein or class of proteins is made on relatively stable messenger RNA.

Summary.—Four phage-specific proteins have been detected by polyacrylamide
gel electrophoresis in lysates of actinomycin-treated E. col infected with an RNA
bacteriophage. One of the proteins, present in excess over any of the others,
corresponded in mobility to the phage coat protein. A second component has been
Vou. 56, 1966 BIOCHEMISTRY: NATHANS ET AL. 1851

identified as the “RNA-protecting protein” previously inferred from genetic studies,
and a third protein is presumed to be the RNA synthetase. The fourth protein is
variable in electrophoretic mobility and is present in much smaller amount than any
of the others. The “RNA-protecting protein,” as well as the minor variable
protein, were also detected in normal phage particles. Fractionation of the proteins
made in E£. coli extracts in the presence of phage RNA as messenger showed two
distinct minor proteins in addition to phage coat protein. The relative amounts of
the minor proteins synthesized in cell extracts and in infected cells were similar. It
is suggested that the same control mechanisms may operate in the infected cell and
in cell-free extracts regulating the translation of polycistronic viral RNA.

Note added in proof: We have recently learned that E. Vifiuela, I. Algranati, and S. Ochoa (per-
sonal communication) have also used polyacrylamide gel electrophoresis to fractionate the phage-
specific proteins made in actinomycin-treated cells infected with MS2.

* This research has been supported by a grant from the National Institutes of Health, U.S.
Public Health Service. M. P. O. is a postdoctoral fellow, and K. E. is a predoctoral trainee of the
National Institutes of Health.

1 Loeb, T., and N. D. Zinder, these ProceEpinas, 47, 282 (1961).

? Zinder, N. D., Ann. Rev. Microbiol., 19, 455 (1965).

* Weissmann, C., L. Simon, and S. Ochoa, these Procerpines, 49, 407 (1963).

4 August, J. T., S. Cooper, L. Shapiro, and N. D. Zinder, in Cold Spring Harbor Symposia on
Quantitative Biology, vol. 28 (1963), p. 95.

6 Haruna, I., K. Nozu, Y. Ohtaka, and 8. Spiegelman, these ProckEpiNGs, 50, 905 (1963).

6 Lodish, H. F., E. Horiuchi, and N. D. Zinder, Virology, 27, 139 (1965).

7 Heisenberg, M., J. Mol. Biol., 17, 136 (1966).

8 Haywood, A., and R. L. Sinsheimer, J. Mol. Biol., 14, 305 (1965).

® Oeschger, M. P., and D. Nathans, J. Mol. Biol., in press.

10 Shimura, Y., R. E. Moses, and D. Nathans, J. Mol. Biol., 12, 266 (1965).

11 Zinder, N. D., and S. Cooper, Virology, 23, 152 (1964).

12 Summers, D. F., J. V. Maizell, Jr., and J. E. Darnell, Jr., these PRocrEpINGs, 54, 505 (1965).

8 Davis, B. J., Ann. N. Y. Acad. Sci., 121, 404 (1964).

4 Fairbanks, G., C. Levinthal, and R. H™ Reeder, Biochem. Biophys. Res. Commun., 20, 393
(1965).

4% Nathans, D., J. Mol. Biol., 13, 521 (1965).

16 Sarabhai, A. 8., A. O. W. Stretton, S. Brenner, and A. Bolle, Nature, 201, 13 (1964).

17 Lodish, H. F., and N. D. Zinder, J. Mol. Biol., 19, 333 (1966).

18 Nathans, D., G. Notani, J. H. Schwartz, and N. D. Zinder, these Procerpinas, 48, 1424
(1962).

1* Ohtaka, Y., and S. Spiegelman, Science, 142, 493 (1963).

* Horiuchi, K., H. F. Lodish, and N. D. Zinder, Virology, 28, 438 (1966).

21 Gussin, G., personal communication.

22 Shimura, Y., and D. Nathans, J. Mol. Biol., in press.