JOURNAL OF VIROLOGY, Sept. 1978, p. 648-658
Copyright © 1978
American Society for Microbiology
Vol. 27, No. 3
Printed in U.S.A.
Specificity of Interferon Action in Protein Synthesis
PETER M. P. YAU, THERESE GODEFROY-COLBURN, CLAIRE H. BIRGE, TRIPRAYAR V.
RAMABHADRAN, AND ROBERT E. THACH*
Departments ofBiology and Biological Chemistry, Washington University, St. Louis, Missouri 63130
Received for publication 28 April 1978
Inhibitors of elongation steps in protein synthesis such as cycloheximide and
anisomycin mimic interferon treatment in that they specifically inhibit the
synthesis of certain viral proteins. These specific effects are seen only at very low
concentrations of the antibiotics, under conditions where host cellular protein
synthesis, as well as cell viability, are not severely reduced. A qualitatively as well
as quantitatively close correlation between the effects of the two types of agents
has been established for encephalomyocarditis virus, vesicular stomatitis virus
and murine leukemia virus protein synthesis. It is concluded that one of the
primary mechanisms of interferon action may be a nonspecific retardation of one
or more elongation steps,.and that this may be sufficient to account for its effects
on the replication of certain viruses such as encephalomyocarditis and vesicular
The mechanisms by which interferon-treated
cells selectively repress virus replication are not
well understood. Although many metabolic al-
terations have been described which are induced
by interferon, it is not clear whether any ofthese
are specific for viral, as opposed to host, func-
tions. An additional difficulty arises from the
fact that different viruses may be affected at
different stages in their replicative cycles, and
there is strong evidence for inhibition of viral
transcription and translation, as well as matu-
ration processes (3, 7, 23).
In spite of these uncertainties it is clear that
viral protein synthesis is a major, ifnot exclusive,
target for interferon. Metz (23) has described a
general theory to account for the specificity of
interferon effects on viral versus host protein
synthesis. He has assumed that most virulent
viral mRNA's have high efficiencies for initia-
tion, as we have found for encephalomyocarditis
(EMC) virus (8, 10, 18) and as Koch and co-
workers have proposed for reovirus (B. Baxt and
G. Koch, personal communication), vesicular
stomatitis virus (VSV) (25), and poliovirus (26).
Metz has further proposed that few if any host
mRNA's fall into this high-efficiency category.
Thus, any agent which is more inhibitory for
translation ofhigh- than low-efficiency mRNA's
will discriminate against viral protein synthesis.
He points out that if interferon treatment pro-
duced such conditions in the cell cytoplasm, this
would account for its selectivity in protein syn-
Metz was only able to speculate as to what
biochemical features might distinguish high-
from low-efficiency mRNA's, and therefore he
was unable to describe in detail the nature ofthe
proposed changes induced by interferon. Some
recent experiments on EMC viral protein syn-
thesis shed light on this problem. First, it has
been found that EMC RNA is a more effective
initiator than host mRNA in vitro (8, 18) and
that this is probably due to its preferential in-
teraction with messenger discriminatory initia-
tion factors. Second, it has recently been shown
that EMC RNA initiates at a much higher rate
than host mRNA in vivo, to the extent that the
elongation rate on viral mRNA is limiting; in
contrast, the rate of initiation is the slow step in
host mRNA translation (10). These two obser-
vations taken together suggest that the interac-
tion of mRNA with initiation factors may ordi-
narily be the rate-determining step for protein
synthesis in the uninfected cell. It also appears
likely that viruses have evolved mechanisms for
speeding up the initiation reaction, possibly by
increasing their affinity for initiation factors.
This would bestow on them the dual advantages
of (i) being able to assure that a single entering
virus mRNA would be rapidly translated, and
(ii) being able to outcompete host mRNA's,
which ordinarily would be present in excess, for
initiation factors late in the course of infection.
potheses as to how interferon might work. Thus,
in general terms, if interferon treatment were to
slow some non-discriminatory step in protein
synthesis which occurs subsequent to the puta-
tive messenger discriminatory step to the point
that the former becomes rate limiting for all
INTERFERON ACTION IN PROTEIN SYNTHESIS
translation, then the selective advantage ofviral
mRNA's would be lost and they would be trans-
lated at the same relative rates as host mRNA's.
How can this hypothesis be tested? One obvious
way is simply to use inhibitors of elongation in
low doses so as to create a new rate-limiting step
subsequent to initiation (10). If the above hy-
pothesis is correct, then this should eliminate
the advantage in translation that viral mRNA's
usually enjoy and thereby reduce virus protein
synthesis, and hence virus yield, without seri-
ously affecting host protein synthesis. It has
already been shown that EMC RNA translation
is greatly reduced relative to host translation by
treatment with elongation inhibitors (10). The
critical experiment, then, is to see if the produc-
tion ofEMC virus is severely inhibited by a dose
of elongation inhibitor low enough to not sub-
stantially affect host functions. (Only ifthis con-
dition is met does the comparison between an-
tibiotics and interferon have any significance,
since inhibition of viral functions by high drug
concentrations is to be expected and is a trivial
result.) The results described in the present com-
munication show that this is true, and that elon-
gation inhibitors and interferon have similar ef-
fects on rates of cellular protein synthesis and
overall cell growth rate, as well as on virus
production. These results have also been ex-
tended and generalized in similar studies with
VSV and murine leukemia virus (MuLV). The
major conclusion to be drawn from this work is
that interferon may well exert its messenger
discriminatory effects in protein synthesis by
slowing nonspecifically one or more elongation
steps. Moreover, for some viruses, such as EMC
and VSV, this effect could well be the primary
means by which interferon inhibits viral repli-
MATERIALS AND METHODS
Growth of cells. A line of mouse L cells (strain
LPA) obtained from E. Knight of E. I. duPont de
Nemours and Co., Wilmington, Del., was grown in
suspension and monolayer cultures. Growth conditions
were identical to procedures previously described by
Knight (15). MOPC 460TC, a mouse plasmacytoma
cell line adapted to grow in tissue culture, was propa-
gated in Leibovitz L-15 medium supplemented with
10% fetal calf serum as previously described by Rob-
ertson et al. (29). SC-1 mouse cells chronically infected
with an AKR-like MuLV (supplied by J. Hartley and
W. Rowe, National Institutes of Health) were main-
tained in McCoy 5a medium containing 50 U of peni-
cillin per ml, 50 gg of streptomycin per ml, and 10%
heat-inactivated fetal calf serum.
Growth and assay of viruses. EMC virus was
grown in Krebs ascites cells as previously described
(18). It was further purified by sucrose gradient sedi-
mentation and banded on a cesium chloride gradient
as described by Kerr and Martin (14). Cesium chloride
gradient-purified virus was then dialyzed overnight
against phosphate-buffered saline (PBS) at 4°C and
diluted with PBS to a concentration of 1012 PFU/ml.
VSV was obtained from S. Schlesinger of Washington
University and grown in L cells.
Virus titer was determined by the hemagglutination
assay (18) or by plaque assay on mouse L cells (22).
Production and assay of interferon. L cells for
interferon production were grown in a T flask with a
75-cm2 growth area until they reached confluency.
DEAE-dextran, purchased from Sigma Chemical Co.,
St. Louis, Mo., was added to a final concentration of
50Ag/ml;a complex of polyinosinic and polycytidylic
acid, purchased from Microbiological
Walkerville, Md., was added to a final concentration
of 5,tg/ml. After 90 min of incubation at 37°C, growth
acid:polycytidylic acid complex was removed by aspi-
ration, cells were washed thoroughly with PBS three
times, fresh growth medium was added, and cultures
were further incubated for 20 h at 37°C. Crude inter-
feron was obtained after centrifugation at 10,000 x g
for 10 min. The clarified supernatant was stored at
Interferon was assayed by a microassay technique
on L cells as described by Knight (15) and Armstrong
(2). EMC virus was used as the challenge virus in our
assay. Mouse standard reference interferon was ob-
tained from the Research Resources Branch, National
Institute ofAllergy and Infectious Diseases. Interferon
preparations used in this study had on the order of 106
U/mg of protein, at a concentration of 600 U/ml.
Infection and pulse-labeling of L cells or
MOPC 460TC cells, and analysis ofthe proteins.
L-cell monolayers were grown to confluency (3 x 105
cells) in 2-cm2 plastic wells. Where indicated, they
were treated with interferon for 24 h prior to infection.
Infection was done in serum-free Eagle minimal essen-
tial medium at a multiplicity of 5. After a 30-min
adsorption period at room temperature, unadsorbed
virus was removed by aspiration and thorough rinsing
with PBS. The cells were refed complete medium and
shifted to 370C (zero time). At the indicated time
[35S]methionine as follows: monolayers were drained,
rinsed three times with PBS, and incubated for 15 min
with 1 ml of Earles balanced salt solution containing
5 ,uCi of [nS]methionine (high-specific-activity grade
from Amersham Corp., Arlington Heights, Ill.). Cyclo-
heximide was added at zero time, as indicated in the
figure legends, and was present until harvest time.
At the end of the labeling period cells were trypsin-
ized and rinsed with PBS, and a sample was counted
in trypan blue. Total incorporation of 5S into proteins
was measured by trichloroacetic acid precipitation of
a constant number of cells. Electrophoretic analysis of
protein was done on samples containing 2 x 105 cells.
After trichloroacetic acid precipitation at a final con-
centration of 5%, the proteins were washed twice with
acetone, redissolved in sample buffer, and analyzed by
polyacrylamide gel electrophoresis (PAGE) in sodium
dodecyl sulfate (SDS) as previously described (18).
Two gel systems were used: linear 7.5 to 20% polyacryl-
amide gradient gels for EMC virus and reovirus pro-
VOL. 27, 1978
YAU ET AL.
teins, and 12.5% gels for VSV proteins. Infection of
MOPC 460TC cells by EMC virus was carried out as
described (18). Pulse-labeling with [35S]methionine
was done as described above, on 2 x 106 cells.
Virus yield determination. At 12 h postinfection,
EMC-infected cells were frozen and thawed twice to
release virus; virus titers were determined by plaque
assay on mouse L cells or by hemagglutination assay.
VSV titers were determined at 8 h postinfection. Assay
samples containing cells and media were frozen and
thawed twice and assayed on L cells.
Quantitation of viable L cells. At specified time
postinfection, medium was removed from cultures by
aspiration, and monolayers were rinsed with normal
saline, stained with crystal violet solution for 15 min,
and rinsed in water after staining. The bound dye was
then eluted with ethylene glycol monoethyl ether, and
optical density was measured at 560 nm.
Labeling of MuLV proteins. SC-1 cells chroni-
cally infected with MuLV were labeled in early log
phase (3 x 104 to 5 x 104 cells per cm2) in 10-cm2
plastic dishes, after 20 h of interferon treatment where
so indicated. They were incubated for 15 min in Hanks
balanced salt solution (containing cycloheximide as
indicated in the figure legend) and then labeled for 1
h in 0.5 ml of the same buffer containing radioactive
amino acids (either 1 mCi of 3H-amino acid mixture
per ml from New England Nuclear, Boston, Mass., or
90t&Ciof [35S]methionine per ml). The cell layer was
then rinsed once with normal growth medium (con-
taining interferon as indicated) and chased in that
medium for 4.5 h.
Cycloheximide-treated cells were lysed in situ by
addition of Triton X-100, sodium deoxycholate, and
SDS to final concentrations of 1, 0.5, and 0.1%, respec-
tively. Interferon-treated cells were drained from the
chase medium, rinsed with PBS, and lysed in PBS
containing the same concentrations of detergents (im-
mune precipitation buffer).
Analysis of MuLV-specific proteins by immu-
noprecipitation and SDS-PAGE. Samples were
spun at 20,000 x g for 20 min. Samples containing
between 25 and 80 ,ug of cell protein, or the correspond-
ing amount of chase medium, were made up to 500,ul
with immune precipitation buffer and incubated with
2piof goat anti-AKR MuLV immune serum (supplied
by J. Gruber, Office of Program Resources and Logis-
tics, National Cancer Institute) for 30 min at room
temperature. A 30-pl amount ofFormol-treated Staph-
ylococcus aureus (supplied by S. Cullen, Washington
University, as a 10% suspension in PBS containing
0.1% gelatin) was then added, and incubation was
resumed for 1 h. After two washes with PBS contain-
ing0.1% gelatin, the immune complex wasresuspended
in electrophoresis sample buffer (5 M urea, 1.4% SDS,
5%,8-mercaptoethanol, 8% glycerol, 0.21 M Tris-chlo-
ride buffer, pH 7.1, and 0.01 M EDTA), boiled for 5
min, and analyzed by SDS-PAGE as described above.
The 7.5 to 20% gradient gel system was used.
Visualization and quantitation ofradioactivity
on polyacrylamide gels. Gels containing a sufficient
amount of 35S were dried and exposed to Kodak XR-
5 film. Otherwise they were treated for fluorography
and exposed to preflashed XR-5 film at -70°C (17).
The amount of radioactivity in a given region ofthe
gel was evaluated, either by scanning the fluorogram
with a Joyce-Loebl densitometer or by dissolving the
gel slice in perchloric acid and hydrogen peroxide
according to Mahin and Lofberg (21); scintillation
counting ofthe dissolved gel was done in 3a70 cocktail
(Research Product International Corp., Elk Grove Vil-
Inhibitors of elongation steps in protein syn-
thesis such as cycloheximide and anisomycin are
ordinarily not thought to be messenger specific
in their effects. Nevertheless, when used at very
low concentrations, these agents can be highly
selective in inhibiting virus as opposed to host
translation. This effect is demonstrated in Fig.
1, where cycloheximide at a concentration of 0.1
,uM virtually eliminates the synthesis of virus
proteins while leaving the pattern ofhost protein
synthesis relatively unaffected. We have argued
elsewhere that this is due to the fact that elon-
gation is rate limiting forEMC RNA translation,
whereas initiation is rate limiting for host trans-
lation (10). In particular, it should be noted that
the selective effects produced when cyclohexi-
mide is added to cells simultaneous with virus
(as in Fig. 1) are also obtained when it is added
[35S]methionine pulse (10). This result shows
that the selective effects of cycloheximide are
due to an instantaneous slowing of elongation
steps per se and not to inhibition ofsome other,
unrelated early event in virus infection.
Results similar to those shown in Fig. 1 have
also been obtained with another elongation step
inhibitor, anisomycin (11). At a concentration of
0.13 LM, viral protein synthesis is undetectable
above the cellular protein synthesis background
(data not shown). In contrast, inhibitors of ini-
tiation steps such as pactamycin (Fig. 2), her-
ringtonine, and hypertonic initiation block do
not produce this result. Indeed, with the latter
the discrimination is reversed, viral translation
being more resistant than host, as originally
reported by Nuss et al. (26) for poliovirus (data
not shown). Results similar to those in Fig. 1
were also obtained using MOPC-460TC cells in
place of L cells.
The selective inhibition of viral protein syn-
thesis shown in Fig. 1 suggests that the overall
production of EMC virus should be inhibitable
by very low concentrations of cycloheximide.
That this is indeed true is shown in Fig. 3A.
Here it is evident that an antibiotic concentra-
tion (0.3 ,M) which only results in a 20% inhi-
bition of the overall rate of protein synthesis in
infected cells inhibits virus production by over
97%. It is interesting to compare the degree of
specificity obtained by cycloheximide to that
shortly before the
INTERFERON ACTION IN PROTEIN SYNTHESIS
FIG. 1. Effect ofcycloheximide on protein synthesis in EMC virus-infected or control L cells. Infected and
control cultures ofL cells were treated with varying concentrations of cycloheximide at the time of infection
with EMC virus. At 4.5 h postinfection, cultures were pulsed withI5S]methionineand analyzed by SDS-
PAGE and autoradiography. Cycloheximide concentrations for infected (i) and control (c) cultures were:
none, lanes 1i and Ic; 0.1 pLM, lanes 2i and 2c; 0.3 M, lanes 3i and 3c; 1.0 uM, lanes 4i and 4c. Virusproteins
are indicated by arrowheads.
produced by interferon treatment. The results
presented in Fig. 3B show that interferon pro-
duces effects which in many respects are similar
to those obtained with cycloheximide. Ofpartic-
ular significance is the fact that interferon pro-
duces a slight but reproducible inhibition of the
overallrate ofproteinsynthesis in EMC-infected
cells. Thus, at a dose of interferon (10 U/ml)
that gives a 94% inhibition of virus yield, the
overall rate of protein synthesis is inhibited by
13%. These numbers are similar to those men-
tioned above for cycloheximide. However, it is
important to note that the analogy between
interferon and cycloheximide is by no means
perfect, and that wide discrepancies occur when
higher concentrations of the two are compared:
increasing concentrations of cycloheximide ulti-
mately block all protein synthesis, whereas a
plateau is reached with interferon beyond which
further addition of this substance produces no
further inhibitory effect.
The experiments shown in Fig. 3 have been
repeated at a higher multiplicity of infection (30
PFU/cell). Similar results were obtained; how-
ever, the doses of interferon and cycloheximide
required to reduce virion production were in-
creased, but by similar amounts (data not
It was of interest to compare the abilities of
interferon and cycloheximide to protect cells
from virus-induced lysis. An experimentshowing
the number of live cells remaining as a function
of time after various treatments is shown in Fig.
4. It is evident that the relatively high dose of
cycloheximide which reduces virus yield by
>99% (1 ILM,cf.Fig. 3) provides virtuallycom-
plete protection from virus-induced cell lysis. A
lower dose (0.1 tiM) provides only partial protec-
tion, comparable to that achieved by 40 U of
interferon per ml. Both agents also retard cell
growth (16, 33), although the significance of this
result is not yet clear (at least part of the inhi-
VOL. 27, 1978
YAU ET AL.
FIG. 2. Effect ofpactamycin on protein synthesis in EMC virus-infected and control L cells. Infected and
control cells were treated with varying concentrations ofpactamycin at the time of infection. Pulsing with
[ S]methionine and subsequent electrophoresis andautoradiography were conductedas in Fig. 1. Pactamycin
concentrations for infected (i) and control (c) cultures were: none, lanes 1i and 1c; 0.1 pLM, lanes 2i and 2c; 0.03
ELM, lanes 3i and 3c; 0.05puM, lanes 4i and 4c; 0.07pM, lanes 5i and 5c. Totalprotein synthesis in infected cells
was inhibited at these concentrations by 0, 14, 32, 39, and 50%, respectively.
bition seen with interferon might be due to
The analogy between elongation inhibitors
and interferon also holds true for viruses other
than EMC. This is demonstrated in Fig. 5 and 6,
where the effects of cycloheximide on VSV pro-
tein synthesis and virion production are shown.
It is evident that VSV translation is selectively
inhibited relative to host translation by low con-
centrations of cycloheximide (Fig. 5), in a man-
ner similar to that observed forEMC translation
(Fig. 1). However, in contrast to the situation
with EMC-infected cells, both cycloheximide
and interferon appear to stimulate the overall
rate of protein synthesis in VSV-infected cells
(Fig. 6), although virion production is again
drastically reduced by both agents. The reason
for the apparent stimulation is not clear, al-
though we suspect that it may be due to an
inhibition by cycloheximide or interferon of the
production of a VSV-specific cytotoxic protein
which would be inhibitory of all protein synthe-
sis in the infected cell. Such an interpretation
would be consistent with the recent report of
Stanners et al. (32), who have described a mu-
tatable virus function which is related to both
cell cytotoxicity and an inhibition ofhost as well
as viral protein synthesis.
Up to this point our results have only shown
that elongation inhibitors can mimic interferon
treatment. They do not prove that interferon is
really acting on virus protein synthesis by the
same mechanism. One way of investigating this
question is to see whether viruses whose protein
synthesis is not affected by interferon are also
relatively resistant to elongation inhibitors. The
best-studied example of such a virus is MuLV,
for which a number of research groups have
INTERFERON ACTION IN PROTEIN SYNTHESIS
FIG. 3. Effectsofcycloheximide and interferon treatmenteon rate oftotalprotein synthesismandEMC virus
yield from L cells. (A) The indicated amount of cycloheximide was added at zero time, simultaneously with
EMC virus. Totalprotein synthesiswas measured at4.5hpost-infection bypulse-labelingwithI'sS]methionine
for 15 mm followed by trichloroacetic acidprecipitation, membrane filtration (Millipore Corp), and scintil-
lation counting. Virusyield was scored by both hemagglutination andplaque assay from infected cultures at
12 hpostinfection; datapoints were superimposabk. (B) The indicated amount of interferon was added 24 h
prior to infection with EMC virus. Totalprotein synthesis was measured at 4.5 hpostinfection. Bothpulse-
labeling and virusyield determinations were the same as described above.f'~S]methionine incorporation is
expressedrelative to the amountincorporated by uninfected cells in the absenceofcycloheximideorinterferon
iu kligb nefrn n
M vrs nefrn
~ OR OT NETO
tocell at zertie iutnosywt
close symblsan rkeiesdnt EM
t th iniae el
VOL. 27, 1978
YAU ET AL.
- Ih d%
FIG. 5. Effect ofcycloheximide onprotein synthesis in VSV-infected or controlL cells. Infected and control
cultures were treated with varying doses ofcycloheximide at the time of virus infection. At 6 h postinfection,
cultures werepulsed with f S]methionine for 15 min. The labeledproteins were thenprocessed and analyzed
by SDS-PAGE and autoradiography. Cycloheximide concentrations for infected (i) and control (c) cultures
were: none, lanes lc and 1i; 0.1 uM, lanes 2c and 2i; 0.3 M, lanes 3c and 3i; 1.0 p.M, lanes 4c and 4i. Virus
proteins are indicated by arrowheads.
reported that interferon treatment has little or
no effect on viral protein synthesis (7, 28, 35).
Thus, if elongation inhibitors and interferon are
acting on protein synthesis by similar mecha-
nisms, we would expect to find thatMuLV trans-
lation is not selectively inhibited by the former.
That this is indeed the case is shown in Fig. 7A,
where it is evident that the synthesis of the
MuLV protein p30 is inhibited by cycloheximide
to exactly the same extent as total cellular trans-
lation. Figure 7B serves as a control experiment,
showing that in our system interferon treatment
does not selectively block p30 synthesis, al-
though the release of virus into the medium (as
indicated by the amount of labeled p30 that can
be chased into the medium) is considerably re-
duced, as reported by others (1, 7, 28, 35). It
should be noted, however, that the similarity of
interferon and cycloheximide effects on p30 syn-
thesis cannot be extended to their effects on
MuLV production. In fact, cycloheximide only
inhibits virus production to the extent that it
inhibits viral protein synthesis. (When cells were
exposed to varying levels of cycloheximide for
prolonged periods oftime, i.e., several days, viral
production was inhibited to the same extent as
protein synthesis, in direct proportion to anti-
biotic concentration [data not shown].) In con-
trast, very low levels of interferon (10 U/ml)
block virus release without affecting p30 synthe-
sis appreciably (Fig. 7B). Thus, even though the
analogy between interferon and elongation in-
hibitors cannot account for all the effects of
interferon onMuLV metabolism, the above data
strongly suggest that their resemblance is more
than superficial, and that they may indeed act
on protein synthesis by similar mechanisms.
Further evidence for this comes from a study
of the effects of cycloheximide on reovirus pro-
tein synthesis (T. Brendler, C. Birge, P. Yau,
Wiebe and Joklik (34) have reported that the
INTERFERON ACTION IN PROTEIN SYNTHESIS
FIG. 6. Effects ofcycloheximide and interferon on rate oftotalprotein synthesis and virusproduction from
VSV-infected and control L cells. (A) The indicated amount of cycloheximide was added at zero time,
simultaneously with VSV. Total protein synthesis was measured at 6 h postinfection by pulse-labeling with
[5S]methioninefor 15 min, followed by trichloroacetic acid precipitation, membrane filtration (Millipore
Corp.), and scintillation counting. Virus yield was scored byplaque assay on L cells from infected cultures at
8 h postinfection. (B) The indicated amount of interferon was added 24 h prior to infection with VSV. Total
protein synthesis and virus production were measured as described above.If5S]methionineincorporation is
expressed relative to the amount incorporated by uninfected cells in the absence ofcycloheximide or interferon
synthesis of the various reovirus proteins is in-
hibited to different degrees by interferon treat-
ment. Some, such as a3, are relatively resistant
to interferon, whereas others are more sensitive,
the most strongly inhibited being A1. Again, if
cycloheximide and interferon are acting via sim-
ilar mechanisms, then treatment of reovirus-in-
fected cells with cycloheximide should produce
a quantitatively sirnilar differential effect. Pre-
liminary experiments have confirmed this pre-
diction and will be described in detail elsewhere.
The experiments described above show that
inhibitors of elongation steps in protein synthe-
sis mimic interferon in their effects on viral as
opposed to host protein synthesis for the three
viruses tested, EMC virus, VSV, and MuLV.
Moreover, the two types of agents showed sim-
ilar effects on the lysis of virus-infected cells.
Whereas most of the data were obtained using
cycloheximide, similar results in several experi-
ments with EMC were seen with anisomycin
(see also reference 10). Inasmuch as these anti-
biotics act via quite different molecular mecha-
nisms (9), it seems likely that other inhibitors of
elongation will have comparable effects. (The
action of the pokeweed antiviral agent, which
has been shown by Obrig et al.  to be an
inhibitor of elongation, supports this view.)
It should be noted that all of the comparative
studies reported here were made using crude
interferon. Recently we have repeated many of
the key observations by using a 50-fold-purified
preparation: qualitatively and quantitatively
similar results were obtained in all cases, except
that the more purified preparations showed less
inhibition of cell growth rate (cf. Fig. 4).
The mechanism by which elongation inhibi-
tors and interferon act on protein synthesis are
likely to be similar, since translation ofmRNA's
which are resistant to the latter are also rela-
tively resistant to the former (as shown by the
studies on MuLV protein synthesis). This con-
clusion is consistent with previous reports of an
inhibitory effect ofinterferon on elongation steps
in protein synthesis (5, 6, 13, 30). However, the
precise mechanism by which elongation
slowed by interferon is not clear. A major obsta-
cle to drawing too close an analogy between
interferon action and the effects of elongation
inhibitors is that whereas the latter tend to
increase the size of host polysomes and leave
unchanged the size of viral polysomes (10), the
former clearly leads to a specific reduction in
viral polysome size (36). This discrepancy is
difficult to explain at present and suggests that
important differences between the modes of ac-
tion of the two types of agents do exist. One
VOL. 27, 1978
p0 by chon, al MuVifce
C1cls A Clswr ricbtdwt ccoeiiea niaeo
15 min,labeled with a dH-amino acid mixture for 1 h in Hanks balanced salt solution containing cyclohexi-
mide, and then returned to normnal medium for a 4.5-h chase. Viralproteins in thepool ofintracellular plus
extracellular material corresponding to 3 cm' of cell layer were analyzed by immunoprecipitation and
electrophoresis as explained in the text. The area ofthe gel corresponding top30 was cut out, dissolved, and
counted. The incorporation of radioactivity in total intra- plus extracellular protein was deternined by
trichloroacetic acidprecipitation. (B) Cells were treated with interferon for20hpriorto labeling,preincubated
for 15 min with Hanks balanced salt solution, labeled with f35S]methionine for 1 h in the same buffer (without
interferon), and chased for4 h in normal medium containing interferon. Immuneprecipitation analysiswas
done on 25jigof cellular protein or on the equivalent amount of chase medium. Radioactivity in p30 was
determined either by direct countingas above (0) or by densitometry (A). Open and closed circles refer to the
sum ofintracellular plus extracellular incorporation oflabel.
ak aacdsl ouin abldwt 3Smtinn
p.30 by chronicallyMuLV-infectedSC-i1 cells. (A) Cells werepreincubatedwith cycloheximide as indicated for
o ntesm buf er(ihu
FIG. 7. Effectofcycloheximide and interferon treatment ontheproduction ofthemajorstructuralprotein
mechanism which would be consistent with all
the data is that interferon might reduce the rate
of only one or a few steps, which are randomly
distributed along the full length of the mRNA.
These sensitive sites would affect a rapidly ini-
tiating message by reducing the number of ri-
bosomes on the mRNA (i.e., polysome size). The
extent of this reduction would be in proportion
to the fraction of the message which lies to the
3' side of the sensitive site proximal to the 3'
end. (There would be no changein the ribosome
density along that portion of the mRNA which
lies to the 5' side of this sensitive site, since
ribosomes should be maximally close packed in
this region in any case .) In contrast, the
average number of ribosomes on a slowly initi-
ating mRNA might decrease or increase, de-
pending upon the locations ofthe sensitive sites.
(A reduction in ribosome density toward the 3'
end of the mRNA might be counterbalanced by
an increase toward the 5' end, where the nor-
mally low ribosome density would be enhanced
by the "queuing up" of ribosomes proxiimal to
the first sensitive site encountered.) Indeed, the
introduction of a few sensitive sites at random
might have little or no discernible effect on the
aggregate size distribution of slowly initiating
host mRNA's. A possible mechanism for the
creation of such sensitive sites might be the
selective inactivation of a few tRNA molecules
in the interferon-treated cells. There is evidence
both for and against such interferon-induced
tRNA inactivation occurring in vivo (4, 5, 31,
37). Thus, this hypothesis must be viewed as
speculative at present. In any event, our results
suggest that some step(s) which occurs subse-
quent to the messenger discriminatory step in
initiation is the primary target of interferon for
certain viruses, such as EMC and VSV, since its
effects are mimicked so closely bythe antibiot-
It is of interest to note that interferon has
been reported to inhibit not only an elongation
step(s) in protein synthesis, but an initiation
step(s) as well (5, 12, 23, 30). However,the latter
typeofinhibition ispredominant onlyin extracts
prepared from interferon-treated, virus-infected
cells (12, 23); incontrast,in the absence of virus
infection the primary targetof interferon seems
to be the elongation step (5, 30). Thus,it seems
YAU ET AL.
INTERFERON ACTION IN PROTEIN SYNTHESIS
possible that the effects of interferon treatment
on protein synthesis may be separable into two
phases. The first phase might involve inhibition
only of an elongation step(s). In this phase the
cell might not be irreversibly injured, although
its growth would be slowed (16, 33), and the
translation of infecting virus mnRNA will be dif-
ferentially inhibited. This inhibition might be
sufficient to block the replication ofsome viruses
under certain conditions. However, in other
cases virus replication might continue through
this phase at a slow rate for a sufficient time to
allow the accumulation of some virus-specific
molecule, such as double-stranded RNA, which
would serve to activate the second phase in the
interferon response. This phase would be char-
acterized by the inhibition of initiation steps, as
well as other steps in virus replication. Some or
all of these second-phase effects may not dis-
criminate between host and virus functions,
thereby leading to the shutoff of all macromo-
lecular metabolism and ultimately to cell death.
As previously noted by others (13), this would
limit the spread ofthe virus infection in an intact
animal. Itwould thus serve as a "backup" system
to the first-phase effects.
The rationale for why general inhibitors of
elongation steps should specifically inhibit the
translation of certain viral mRNA's more than
that of host mRNA's has been described above.
In agreement with the prediction of Metz (23)
and the results of others (10, 25, 26), our findings
suggest that any viral mRNA whose translation
is sensitive to interferon will be found to initiate
unusually rapidly in its normal host cell. It must
be emphasized that certain elements of this ra-
tionale are still hypothetical, however, and re-
main to be proven experimentally. Suffice it to
add here that our results are entirely consistent
with the work of others on protein synthesis
kinetics (19, 20).
Finally, it should be pointed out that the
antiviral effects shown by elongation inhibitors
may prove useful in the treatment of certain
viral diseases in higher organisms, including hu-
mans. The very low concentrations required for
antiviral action in tissue culture suggests that
the ordinarily toxic side effects ofthese antibiot-
ics may be minimized in whole organisms.
This work was supported by Public Health Service grants
from the National Cancer Institute (5R01 CA 13008;5T32 CA
09129; 5P30 CA 16217) and a grantfrom the National Science
Foundation (PCM 76-03079). P.Y. was supported byaSigma
1. Allen,P.T.,H.Schellekens,L J. L.D. vanGriensven,
Rauscher murine leukemia virus (MuLV-R) to interfer-
ons in two interferon-responsive cell lines. J. Gen. Virol.
2. Armstrong, J. A. 1971. Semi-micro, dye-binding assay
for rabbit interferon. Appl. Microbiol. 21:723-725.
3. Billiau, A., V. G. Edy, H. Sobis, and P. DeSomer. 1974.
Influence of interferon on virus particle synthesis in
oncornavirous carrier lines. II. Evidence for a direct
effect on particle release. Int. J. Cancer 14:335-340.
4. Colby, C., E. E. Penhoet, and C. E. Samuels. 1976. A
comparison of transfer RNA from untreated and inter-
feron-treated murine cells. Virology 74:262-264.
5. Content, J., B. Lebleu, U. Nudel, A. Zilberstein, H.
Berissi, and M. Revel. 1975. Blocks in elongation and
initiation of protein synthesis induced by interferon
treatment in mouse L cells. Eur. J. Biochem. 54:1-10.
6. Content, J., N. Lebleu, A. Zilberstein, H. Berissi, and
M. Revel. 1974. Mechanism of the interferon-induced
block ofmRNA translation in mouse L cells: reversal of
the block by transfer RNA. FEBS Lett. 41:125-130.
7. Friedman, R. M., E. H. Chang, J. M. Ramseur, and
M. W. Meyers. 1975. Interferon-directed inhibition of
chronic murine leukemia virus production in cell cul-
tures: lack of effect on intracellular viral markers. J.
8. Golini, F., S. S. Thach, C. H. Birge, B. Safer, W. C.
Merrick, and R. E. Thach. 1976. Competition between
cellular and viral mRNAs in vitro is regulated by a
messenger discriminatory initiation factor. Proc. Natl.
Acad. Sci. U.S.A. 73:3040-3044.
9. Grollman, A. P., and M. T. Huang. 1976. Inhibitors of
protein synthesis in eukaryotes, pp. 125-167. E. H.
McConkey, (ed.), Protein synthesis. Marcel Dekker,
10. Jen, G., C. H. Birge, and R. E. Thach. 1978. Comparison
ofinitiation rates ofencephalomycarditis virus and host
protein synthesis in infected cells. J. Virol. 27:640-647.
11. Jimenez, A., L. Sanchez, and D. Vasquez. 1975. Si-
multaneous ribosomal resistance to trichodermin and
Biochim. Biophys. Acta 383:427-434.
12. Kerr, I. M., R. E. Brown, M. J. Clemens, and C. S.
Gilbert. 1976. Interferon-mediated inhibition of cell-
free protein synthesis in response to double-stranded
RNA. Eur. J. Biochem. 69:551-561.
13. Kerr, I. M., R. M. Friedman, R. E. Brown, L. A. Ball,
and J. C. Brown. 1974. Inhibition of protein synthesis
in cell-free systems from interferon-treated, infected
cells: further characterization and effect of formylme-
thionyl-tRNA. J. Virol. 13:9-21.
14. Kerr, I. M., and E. M. Martin. 1972. Simple method for
isolation ofencephalomyocarditis virus ribonucleic acid.
J. Virol. 9:559-561.
15. Knight, E., Jr. 1975. Heterogeneity of purified mouse
interferons. J. Biol. Chem. 250:4139-4144.
16. Knight, E., Jr. 1976. Antiviral and cell growth inhibitory
activities reside in the same glycoprotein of human
fibroblast interferon. Nature (London) 262:302-303.
17. Laskey, R. A., and A. D. Mills. 1975. Quantitative film
detection of 'H and "4C in polyacrylamide gels by fluo-
rography. Eur. J. Biochem. 56:335-341.
18. Lawrence, C., and R. E. Thach. 1974. Encephalomyo-
carditis virus infection of mouse plasmacytoma cells. I.
19. Lodish, H. F. 1971. Alpha and beta globin messenger
ribonucleic acid: dfferent amounts and rates ofinitiation
of translation. J. Biol. Chem. 246:7131-7138.
20. Lodish, H. R. 1974. Model for the regulation of mRNA
translation applied to haemoglobin synthesis. Nature
21. Mahin, D. T., and R. T. Lofberg. 1966. A simplified
method of sample preparation for determination of
tritium, carbon-14 or sulfur-35 in blood or tissue by
in Saccharomyces cerevisiae mutants.
VOL. 27, 1978
658 Download full-text
YAU ET AL.
22. Martin, E. M., J. Malec, S. Sved, and T. S. Work. 1961.
Studies on protein and nucleic acid metabolism invirus-
infected mammalian cells. Biochem. J. 80:585-597.
23. Metz, D. 1975. The mechanism of action of interferon.
24. Metz, D. H., M. Esteben, and G. Danielescu. 1975. The
effect of interferon on the formation of viral polyribo-
somes in L cells infected with vaccinia virus. J. Gen.
25. Nuss, D. L., and G. Koch. 1976. Differential inhibition
of vesicular stomatitis virus polypeptide synthesis by
hypertonic initiation block. J. Virol. 17:283-286.
26. Nuss, D. L., H. Opperman, and G. Koch. 1975. Selective
blockage of initiation of host protein synthesis in RNA-
27. Obrig, R. G., J. D. Irwin, and B. Hardesty. 1973. The
effect of an antiviral peptide on the ribosomal reactions
of peptide elongation enzymes, EF-I and EF-II. Arch.
Biochem. Biophys. 155:278-289.
28. Pitha, M. P., W. P. Rowe, and M. N. Oxman. 1976.
Effect of interferon on exogenous, endogenous and
chronic murine leukemia virus
29. Robertson, D. L., P. Yau, D. C. Dobbertin, S. K.
Sweeney, S. S. Thach, T. Brendler, and R. E.
Thach. 1976. Relationships between intracisternal type
A and extracellular oncornavirus-like particles pro-
duced in murine MOPC-460 myeloma cells. J. Virol.
30. Samuels, C. E. 1976. Mechanism of interferon action.
Studies on the mechanism of interferon-mediated inhi-
cells. Proc. Natl. Acad.
bition of reovirusmessenger RNA translation in cell-
freeproteinsynthesis systems frommouse ascites tumor
cells. Virology 75:166-176.
31. Sen,G.E.,S. L.Gupta, G. E. Brown, B. Lebleu, M. A.
Rebello, and P.Lengyel. 1976. Interferon treatment
of Ehrlich ascites tumor cells: effect on exogenous
mRNA translation and tRNA inactivation in the cell
extract. J. Virol. 17:191-203.
32. Stanners, C.P., A. M. Francolur, and T. Lam. 1977.
Analysisof a VSV mutant with attenuated cytopatho-
genicity: mutation in viral function, P, for inhibition of
protein synthesis. Cell 11:273-281.
33. Stewart, W. E., L. Gresser, M. G. Tovey, M.-T.Bandu,
and S. I. Goff. 1976. Identification of the cell multipli-
cationinhibitory factors in interferon preparations as
interferons. Nature (London) 262:300-302.
34. Wiebe, M.E., and W. K. Joklik. 1975. The mechanism
of inhibition of reovirus replication by interferon. Vi-
35. Wu,A.M.,A.Schultz, and R. C. Gallo. 1976. Synthesis
oftype C virus particles from murine-cultured cells
induced by iododeoxyuridine. V. Effect of interferon
36. Yakobson, Y.,C.Prives, J. R. Hartman, E.Winocour,
and M. Revel. 1977. Inhibition of viral protein synthe-
sis inmonkey cells treated with interferon late in simian
virus 40lytic cycle. Cell 12:73-81.
37. Zilberstein, A. A., D. Dudock, H. Berissi, and M.
Revel. 1976. Control ofmessenger RNA translation by
minor species of leucyl-transfer RNA in extracts from
interferon-treated L cells. J. Mol. Biol. 108:43-54.
its interaction with dexamethasone.