JouRNAL OF VIRowLy, Dec. 1975, p. 1575-1583
Vol. 16, No. 6 Printed in U.SA.
Copyright © 1975 American Society for Microbiology
Effect of Cordycepin (3'-Deoxyadenosine) on Virus-Specific RNA Species Synthesized in Newcastle Disease VirusInfected Cells SUSAN REICH WEISS AND MICHAEL A. BRAIT* Department of Microbiology, University of Massachusetts Medical School, Worcester, Massachusetts 01605 Received for publication 23 July 1975
Cordycepin (3'-deoxyadenosine) has no effect on the size or relative proportions of Newcastle disease virus-specific 18-22S mRNA species nor on the amount or size of the polyadenylic acid associated with them. Cordycepin does, however, cause an inhibition of incorporation of [3H]uridine into 50S virus-specific RNA relative to 18-22S RNA. This inhibition is probably not a direct effect of the drug on the synthesis of 508 viral RNA. Like cycloheximide, another drug which inhibits 50S RNA accumulation in paramyxovirus-infected cells, cordycepin inhibits protein synthesis as measured by amino acid incorporation. It is likely that the inhibition of 50S RNA accumulation is a secondary effect of protein synthesis inhibition. This is supported by the finding that concentrations of cordycepin and cycloheximide, which inhibit protein synthesis to the same extent, have the same effect on the ratio of 50 to 18-22S virus-specific RNA. During infection of chicken embryo cells by Newcastle disease virus (NDV), 50S singlestranded virion RNA is replicated and singlestranded 18-22S and 35S RNAs complementary to virion RNA are synthesized (4; B. SpanierCollins, C. W. Clinkscales, M. A. Bratt, and T. Morrison, manuscript in preparation). The 1822S and 35S RNAs contain polyadenylic acid [poly(A)I sequences (21; Spanier-Collins et al., in preparation) and serve as messengers for virus-specific proteins in cell-free protein-synthesizing systems (14; T. Morrison, L. E. Hightower, B. Spanier-Collins, S. R. Weiss, and M. A. Bratt, manuscript in preparation; SpanierCollins et al., in preparation). In NDV-infected chicken embryo cells, there are also partially base-paired virus-specific RNA species sedimenting from 35 to 50S which are intermediates in the synthesis of virion RNA and mRNA (B. Spanier-Collins and M. A. Bratt, manuscript in preparation). It would be useful to be able to preferentially inhibit viral genome replication, transcription of mRNA, or the synthesis of the poly(A) associated with this mRNA. This might allow us to independently study the mechanism and molecules and intermediates involved in one of the processes and, if poly(A) synthesis were preferentially inhibited, to learn something about the function of viral poly(A). Cordycepin (3'-deoxyadenosine) appears to have differential effects on the synthesis of various RNA species in eukaryotic cells. It has little effect on the synthesis of heterogeneous
nuclear RNA but inhibits poly(A) synthesis and the appearance of mRNA on polysomes (1, 9, 13, 17, 18). It also seems to cause a reduction in size of ribosomal precursor RNAs and to inhibit mitochondrial RNA synthesis (18). We therefore tested this drug for differential effects on NDV-specific RNA species synthesized in infected chicken embryo cells. We report here that cordycepin preferentially inhibits incorporation of [3H]uridine into 50S virus-specific RNA relative to 18-22S without affecting the size or relative proportions of the 18-22S mRNA's or the amount or size of their associated poly(A). Our results suggest that the effect on 50S RNA is probably not a specific effect of the drug on viral RNA synthesis but rather a secondary effect of the inhibition of protein synthesis by this drug. MATERIALS AND METHODS Solutions, chemicals, and abbreviations. The following buffers were used: standard buffer (0.1 M NaCl; 0.01 M Tris, pH 7.4; and 0.002 M EDTA); sodium dodecyl sulfate (SDS)-standard buffer (standard buffer with 0.5% SDS); solubilizing buffer (0.1 M NaCl; 0.01 M Tris, pH 8.5; 0.002 M EDTA; 1% SDS; and 1% mercaptoethanol); phenol buffer (solubilizing buffer without SDS); Tris-saline (0.15 M NaCl; 0.005 M KCI; 0.025 M Tris, pH 7.3; and 1 g of glucose per liter); SSC (0.15 M NaCl and 0.015 M sodium citrate); column-binding buffer (2x SSC + 0.1% SDS); column elution buffer (0.01 M Tris, pH 7.4; 0.1% SDS). The standard medium used consisted of Eagle minimal essential medium supplemented with 2.5% calf serum, 2.5% tryptose phosphate broth, and 0.07% NaHCO3.
1575
1576
WEISS AND BRArTT
[3H]uridine (28 Ci/mM), [3H]adenosine (15.5 to 27.1 Ci/mM), 32P-labeled inorganic phosphate, [3H]amino acids (1 to 10 Ci/mM), aquasol, and Omnifluor were obtained from New England Nuclear Corp. Bovine pancreatic RNase was obtained from Schwarz Bioresearch, Ti RNase and cycloheximide (Actidione) were from Calbiochem, oligo(dT)-cellulose (T2) was from Collaborative Research, and cordycepin was from Sigma Chemical Co. Actinomycin D (ActD) was generously provided by Merck, Sharp & Dohme, Rahway, N.J. Virus purification and infectivity assays. NDVHP (Israel HP, 1935) (3) was grown in embryonated hens' eggs and then concentrated from the allantoic fluid by centrifugation in a Spinco SW27 rotor for 1 h at 4 C and 24,000 rpm (74,800 x g) in standard buffer through 20% sucrose onto a 65% sucrose-D20 pad. Virus at the interface of the two sucrose layers was then diluted in Tris-saline containing 2% calf serum and stored at - 70 C. Infectivity titrations by plaque formation (PFU per milliliter) were carried out as described previously (5) but using standard medium (see above). Virus growth curve. Growth curves were done as previously described (11) except adsorption was for 45 min and incubation was at 40 C. The end of the virus adsorption period is defined as zero time. Cell culture procedures. Secondary cultures of chicken embryo cells were prepared and grown in a 5% CO2 atmosphere as previously described (3) but in standard medium. Cultures were plated at 5 x 106 cells/100-mm tissue culture plate and used after 48 h of incubation at 40 C. Intracellular RNA. Virus-specific RNA was labeled in the following manner. Cultures were infected at an input multiplicity of 5 PFU/cell with NDV-HP and incubated at 40 C. ActD at 10 ,ug/ml was always added 45 min before labeling. Cordycepin or cycloheximide was added at the concentrations and times indicated. Labeling was with 50 to 100 ,uCi of [3H]uridine or [3H]adenosine per ml for the periods indicated in each experiment. After labeling, the cells were solubilized in solubilizing buffer, phenol extracted three times with buffersaturated phenol, and precipitated twice in ethanol as previously described (4). Host cell RNA to be used as marker was labeled during a 24-h incubation with 500 ,uCi of 32p per ml and then purified as described for virus-specific RNA. Fractionation and analysis of RNA. Velocity sedimentation analysis were performed as follows. RNA dissolved in SDS-standard buffer was layered on 12-ml linear 15 to 30% sucrose gradients containing the same buffer. Centrifugation was carried out at 22 C and 39,000 rpm (180,000 x g) in a Spinco SW41 rotor for 4 h. Fractions (0.4 ml) were collected from the top using an ISCO (Instrumentation Specialties Co.) fractionator, which measured adsorbance at 260 nm, coupled to an ISCO fraction collector. Samples of each fraction were counted directly in aquasol or trichloroacetic acid precipitated and counted using previously described procedures (6). RNase digestion was done by incubating RNA at 37 C for 1 h in 2x SSC, 50 itg of pancreatic RNase per ml, and 25 U of Ti RNase per ml.
J. VIROL.
Oligo(dT)-cellulose binding was done in the following way. RNA, in 1 ml of column-binding buffer, was added to a column of 0.1 g of oligo(dT)-cellulose (previously washed with binding buffer) in a Pasteur pipette. The column was washed with three 1ml aliquots of binding buffer and then four 1-ml aliquots of column elution buffer. These 1-ml fractions were trichloroacetic acid precipitated. Most of the radioactivity is in the first 2 ml of binding buffer and the first 2 ml of elution buffer. Therefore, the RNA in these fractions is designated as binding and nonbinding RNA, respectively. Acrylamide gel electrophoresis was carried out using 10-cm 4 or 10% ethylene diacrylate crosslinked polyacrylamide gels according to the procedures, including gel slicing and counting, of Collins and Bratt (8). Amino acid labeling. Cells were infected as in RNA labeling experiments. Drugs were added after adsorption of the virus. Labeling was for 30 min with 10 ,1Ci of [3H]amino acids per ml at the times indicated. After labeling, cells were washed with Trissaline and removed from the plates with 1 N NaOH. Samples of this were trichloroacetic acid precipitated after boiling for 5 min.
RESULTS Inhibition of virus multiplication. Figure 1 shows that 50 ,ug of cordycepin per ml, in the presence or absence of 10 Ag of ActD per ml, inhibits the appearance of infectious virus almost completely (>95%) if added soon after infection. Cordycepin is significantly less effective when added 4 h after infection. Effects on intracellular virus-specific 1822S RNA and poly(A). Figure 2 shows the sucrose velocity sedimentation patterns of intracellular virus-specific RNA labeled with [3H]uridine (Fig. 2A) or [3H]adenosine (Fig. 2B) in the presence of 10 ,ug of ActD per ml and in the presence or absence of 50 ,ug of cordycepin per ml. The usual 18-22S, 358, and 50S NDV RNA species (4) are present in the cultures that did not receive cordycepin. In drug-treated cultures, the major effect of cordycepin is an apparent decrease in the ratio of 50 to 18-22S RNA. To more precisely analyze any possible effects of cordycepin on the 18-22S RNA, [3H]uridine-labeled 18-22S RNA, synthesized in the absence (Fig. 3A) or presence (Fig. 3B) of cordycepin, was analyzed by gel electrophoresis. The patterns are similar to those obtained previously (8, 21), and there are no major differences in the migration rates or relative proportions of these RNA species when synthesized in the presence or absence of cordycepin. The following experiments suggest that there is little, if any, effect of cordycepin on the synthesis of the poly(A) associated with NDV 1822S mRNA. [3H]uridine and [3H]adenosine-labeled 18-22S RNA were pooled from gradients like those in Fig. 2, and aliquots were trichloro-
EFFECTS OF CORDYCEPIN ON NDV RNA
VOL. 16, 1975
1577
Control I-I
Q,10 :D LL
a0
D at
0
-o0 Cordycepin at 0 Time
U) (n .-_
0.1ka
2
4 8 6 Hours Post Infection
10
FIG. 1. Effects of cordycepin on virus production. Growth curves were done as described in Materials and Methods. Symbols: No drug additions (a); 50 pg of cordycepin per ml added at O time (O); 50 pg of cordycepin and 10 pg of ActD per ml added at 0 time (A); 50 pg of cordycepin per ml added at 4 h postinfection (0).
acetic acid precipitated after incubation with and pancreatic RNases. or without Ti [3Hluridine-labeled RNA had only a small amount of RNase resistance, as expected for single-stranded RNA (Table 1). [3H]adenosinelabeled RNA extracted from control cells was approximately 18% resistant to RNase. We have previously shown that this RNase resistance represents poly(A) (21). [3H]adenosine-labeled RNA extracted from cordycepin-treated cells was about 16% resistant to RNase. This suggests there is about the same amount of poly(A) associated with 18-22S RNA from control or drug-treated cells. The small difference in RNase resistance of [3H]adenosine-labeled RNA from control and cordycepin-treated cells, which is observed in some experiments, is probably due to a greater contribution of radioactivity in ActD-resistant host cell background in untreated cells than in cordycepin-treated cells (see Fig. 2B). This background is highly RNase resistant (approximately 60%) in untreated and drug-treated cells. As another test of poly(A) content, we chro-
matographed [3H]uridine- and [3H]adenosinelabeled 18-22S RNA using oligo(dT)-cellulose columns. Poly(A)-containing RNA should bind to these columns (2). Approximately 70% of the [3H]uridine-labeled RNA and 90% of the [3H]adenosine-labeled RNA from control or drug-treated cells bound to these columns (Table 1). The same percentage of radioactivity from control and drug-treated cells is associated with enough poly(A) to bind to the columns. This suggests that the same percentage of 1822S RNA molecules from control and drugtreated cells contain poly(A). Finally, we determined the size of the RNasefrom fragments resistant [poly(A)] [3H]adenosine-labeled 18-22S RNA extracted from control (Fig. 4A) and drug-treated (Fig. 4B) cells. The RNase-resistant fragments were purified by phenol extraction and then co-electrophoresed In 10% polyacrylamide gels with chick cell RNA from the 4S region of a sucrose gradient. The size distributions of poly(A) in both cases are the same. As we have previously reported (21), NDV poly(A) is heterogeneous in
1578
WEISS AND BRAIT
J. VIROL.
A. 3H- Uridine 50S
35S
41 3
-
I
B. 3H-Adenosine
I
35S
50S
18S
8H 6t
41p .:I
Is
2- - > f A..gsrl
7~~~~~~'i
e E_~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
Bottom
10
20 30 Fraction Number
Top
FIG. 2. Effects of cordycepin on NDV-specific RNA. Intracellular NDV RNA (solid lines) was labeled in the presence ofActD from 4 to 9 h postinfection with [3H]uridine (A) or [3H]adenosine (B) in the absence (-) or presence (0) of 50 .g of cordycepin per ml and purified and fractionated as described in Materials and Methods. RNA from uninfected cells (dotted lines), labeled with [3Hadenosine in the presence of ActD, in the absence (@) or presence (0) of cordycepin, is also shown in (B). Samples (0.01 ml) of each fraction were counted in aquasol. When cordycepin was used, cells were pretreated with the drug for45 min before labeling. (The majority of counts at the top of the gradients [fractions 32 -39] are trichloroacetic acid-soluble counts as they are not detected when samples are trichloroacetic acid precipitated.)
size, with an average size of 4-5S. Thus, cordycepin has no effect on the size of the poly(A) associated with NDV mRNA. Effects of protein synthesis inhibition on the ratio of 50S to 18-22S RNA. The results described thus far suggest that cordycepin has
little or no effect on NDV mRNA. However, the reduction in incorporation of [3H]uridine and [3Hladenosine into 5(S intracellular RNA relative to 18-22S (Fig. 2) is consistently observed when infected cells are labeled in the presence of 50 ,ug of cordycepin per ml. The amount of
EFFECTS OF CORDYCEPIN ON NDV RNA
VOL. 16, 1975
1579
16 14 12 10
6 5 4
3 2 0
0
N
0 0~X
C)X 0~ L
N
:I
8 7 6 5 4
) to
3 2
Origin
20
40 Slice
60 80 Number
100
FIG. 3. Electrophoretic analysis of the intracellular 18-22S RNA species. [3H]uridine-labeled 18-22S RNA (-) labeled in the absence ofcordycepin (A) or in the presence ofcordycepin (B) were co-electrophoresed with 32P-labeled 188 and 28S chick cell RNA (0). Electrophoresis was for 11 h in 4% gels.
inhibition is dependent on drug concentration and the time period of treatment with the drug before labeling (data not shown). This effect of cordycepin on the ratio of 50S to 18-22S RNA is similar to that produced by treating paramyxovirus-infected cultures with the protein synthesis inhibitor cycloheximide (6, 20). Cordycepin has also been reported to inhibit protein synthesis (7). Therefore, it seemed possible that the reduction of 50S RNA accumulation relative to
18-22S might be a secondary effect of the inhibition of protein synthesis and not a direct effect on 50S RNA synthesis. To quantitatively compare the effects of cordycepin with a known inhibitor of protein synthesis, we first determined the extent to which cordycepin inhibits protein synthesis in chicken embryo cells under conditions used in our experiments. This was done by measuring the rate of accumulation of [3H]amino acids into trichloroa-
1580
WEISS AND BRATT
J. VIROL.
TABLE 1. RNase sensitivity and oligo(dT)-cellulose binding of NDV mRNA Counts (%) resistant to Ti and Counts (%) bindpancreatic RNase treatmenta RNA ing to oligo(dT)-
[3H]uridine-labeled 18-22S RNA (no cordycepin) [3H]uridine-labeled 18-22S RNA (+ cordycepin) [3H]adenosine-labeled 18-22S RNA (no cordycepin) [3H]adenosine-labeled 18-22S RNA (+ cordycepin)
Expt 1
Expt 2
Expt 3
Avg
cellulose
3.0 4.4 17.8 16.0
3.6 2.1 18.2 18.6
4.1 2.8 17.8 12.8
3.6 3.1 17.9 15.8
67.3 68.2 88.4 88.7
a Fractions were pooled from gradients similar to those in Fig. 2, and duplicate samples were treated with RNase as described in Materials and Methods. Three separate preparations of RNA were used (experiments 1, 2, 3). Percentage of resistance is defined as the ratio of trichloroacetic acid-precipitable counts before and after treatment, times 100. bFractions were pooled as in a, and the amount of counts binding to oligo(dT)-cellulose columns was determined as described in Materials and Methods. Percentage of binding is defined as the ratio of columnbinding counts to total counts added to the column, times 100.
N
I
6 5 4 3 2
6 5 4 3 2
0
Hours
10 20 30 40 50 60 70 80 Slice Number FIG. 4. Electrophoretic analysis of poly(A). Poly(A) (a) from 18-22S NDV RNA labeled in the absence (A) or presence (B) of cordycepin was coelectrophoresed with 32P-labeled 4, 5, and 7S chick cell RNA. Electrophoresis was for 5 h in 10% gels. Arrows designate the positions of marker cell RNAs.
cetic acid-precipitable material at various times after the addition of cordycepin to uninfected and infected chicken embryo cells. This rate of accumulation in the presence of cordycepin (relative to the untreated control) decays with a half-life of about 2 to 2.5 h (Fig. 5). The
3
5
Post
Infection
7
FIG. 5. Effects of cordycepin and ActD on amino acid incorporation. NDV-infected (solid line) or mock-infected (dashed line) cells were labeled with [3H]amino acids for 30 min at the times shown, and counts incorporated into trichloroacetic acid-precipitable material were determined. Ten micrograms of ActD (a) per ml, 10 pg of ActD and 50 pg of cordycepin (O) per ml, or 50 pg of cordycepin (0) per ml was added at the time of infection.
kinetics of inhibition are independent of the time during infection that the drug is added (data not shown). A small inibition of trichloroacetic acid-soluble radioactivity is also observed in the presence of cordycepin (data not shown), but this is not enough to account for the inhibition of trichloroacetic acid-precipitable radioactivity seen here. We also determined the effect of ActD on amino acid accumulation, because it was used in all the RNA labeling experiments. ActD (10 ,ug/ml) inhibits amino acid incorporation but with slower kinetics than cordycepin (Fig. 5), and ActD and cordycepin together show an effect similar to cordycepin alone.
EFFECTS OF CORDYCEPIN ON NDV RNA
VOL. 16, 1975
Next we determined the concentration of cycloheximide that inhibits amino acid incorporation to the same extent as 50 Ag of cordycepin per ml. We found that 0.1 ,ug of cycloheximide per ml (in the presence of 10 ,ug of ActD per ml) inhibits amino acid incorporation by about 50% in NDV-infected chicken embryo cells (data not shown) and, after a 2.5-h incubation, 50 Ag of cordycepin per ml (in the presence or absence of ActD) causes a similar inhibition. (The amount of inhibition after 2.5 h of cordycepin treatment varies from 50% in many experiments to as much as 65%, as seen in Fig. 5.) Finally, we compared the patterns of [3H]uridine-labeled virus-specific RNA ex-
1581
tracted from cells that were treated with cordycepin or cycloheximide with RNA patterns from untreated control cells (Fig. 6). All of the RNA was labeled in the presence of ActD. When cordycepin or cycloheximide was used, it was added to the cells 2.5 h before labeling and was present during labeling. Figure 6A shows the control pattern of RNA. The usual 18-22S, 35S, and 50S species are present (The relative amounts of virus-specific 50S and 18-22S RNA synthesized in the presence of ActD by NDVHP varies greatly from one experiment to the next [compare the patterns in Fig. 2A and Fig. 6A] but are always constant within one experiment.) Figure 6B shows the pattern of RNA
4
3
2
I
16
12
8 4
sMGM
10
20
30TOP
Fraction
Botom
10
20
30 "P
Number
FIG. 6. Comparison of NDV-specific RNA synthesized in the absence or presence of cordycepin and cycloheximide. Intracellular virus-specific RNA was labeled, in the presence of ActD, with [3H]uridine for 90 min at 6.5 h postinfection. Cordycepin or cycloheximide, as indicated, was added 2.5 h before labeling in (B), (C), and (D) and continued to be present during labeling. RNA was purified and fractionated as described in Materials and Methods. Gradient fractions were trichloroacetic acid precipitated.
1582
WEISS AND BRATT
J. VIROL.
TABLE 2. Effect of cycloheximide and cordycepin on the ratio of 50S to 18-22S RNA Treatment
5OS/18-22Sa
50S/18-22S as % of
control°
Control Cordycepin, 50 ,ug/ml Cycloheximide, 0.1 ,ug/ml Cycloheximide, 50 j,g/ml
2.57 1.28 1.64 0.05
100.0 50.6 63.5 1.9
a Total counts in the 50S and 18-22S regions of the gradients in Fig. 6 were calculated, and the total 50S counts were divided by the total 18-22S counts. b The values in the first column were divided by the ratio for the control and multiplied by 100.
extracted from cells that were treated with 50 ,zg of cordycepin per ml, and Fig. 6D shows the pattern of RNA from cells that were treated with 0.1 ,ug of cycloheximide per ml. In both the cordycepin- and cycloheximide-treated cells, the amount of label in 50S RNA relative to 1822S is reduced. The ratios of 50S to 18-22S are about 51% of the control in RNA labeled in the presence of 50 ,ug of cordycepin per ml and 64% of the control in RNA labeled in the presence of 0.1 ,ug of cycloheximide per ml (Table 2). Therefore, when amino acid incorporation is inhibited by about 50% by cycloheximide or cordycepin, there is approximately the same effect on the ratio of 50S to 18-22S RNA. Figure 6C shows the pattern of RNA from cells that were treated with a much higher concentration of cycloheximide, 50 ,ug/ml. Under these conditions, where incorpbration of [3H]amino acids into trichloroacetic acid-precipitable material is inhibited by 95%, incorporation of [3H]uridine into 50S RNA is almost completely inhibited. The ratio of 50 to 18-22S RNA labeled in the presence of 50 ,g of cycloheximide per ml is 2% of that of the control RNA (Table 2). DISCUSSION Cordycepin has no effect on the size or relative proportions of the six to seven 18-22S NDV mRNA species; nor does it have an effect on the poly(A) associated with this RNA. This is in contrast to the poly(A) associated with the heterogeneous nuclear and mRNA of eukaryotic cells, the synthesis of which is inhibited by cordycepin (1, 9, 13, 17). NDV RNA is synthesized solely in the cytoplasm (4). NDV virions and cores isolated from virions by detergent treatment can synthesize 18-22S RNA (6, 12) and its associated poly(A) (21) in vitro. There is also evidence suggesting that NDV poly(A) is not transcribed from polyuridylic acid sequences in virion RNA (16). This sug-
gests that NDV has its own poly(A)-synthesizing activity, but it is still possible that the enzyme that synthesizes viral poly(A) is a cellular, cytoplasmic, cordycepin-insensitive enzyme that is incorporated into virus particles. It has been shown that the protein synthesis inhibitor cycloheximide alters the ratio of 50S to 18-22S RNA in paramyxovirus-infected cells (6, 20), probably because it inhibits the replication of virion RNA (Spanier-Collins and Bratt, in preparation). We have shown here that cordycepin, another inhibitor of protein synthesis, also alters this ratio. At concentrations ofcordycepin and cycloheximide that inhibit amino acid incorporation to the same extent, there is a similar effect on the ratio of 50 to 18-22S RNA. The inhibition of amino acid incorporation by cordycepin increases with time of drug treatment (Fig. 5); the effect of cycloheximide is faster. By 1 h after addition of the drug, the maximum effect is seen (data not shown). It is noteworthy that cycloheximide and cordycepin, which show different kinetics of inhibition of amino acid accumulation and almost certainly work through different mechanisms, have similar effects on nucleotide incorporation into NDV intracellular 50S RNA relative to 18-22S RNA. Although we have not ruled out all other possibilities, it seems probable that the effect of cordycepin on 5S RNA is a secondary effect of protein synthesis inhibition. Whereas cycloheximide and cordycepin always inhibit the incorporation of [3H]uridine into 5S RNA relative to 18-22S RNA, they have different effects on the absolute number of counts incorporated into 18-22S RNA. Cycloheximide consistently causes an increase in the amount of incorporation into 18-22S RNA. The effect is most dramatic at the high cycloheximide concentration (see Fig. 6). Cordycepin, on the other hand, causes a decrease in the absolute amount of label incorporated into 18-22S RNA, but only when RNA is labeled under certain conditions. In experiments like the one shown in Fig. 2, where labeling is for 5 h after a 45-min drug pretreatment, approximately the same amount of label is incorporated into 18-22S RNA with or without cordycepin. When cells are labeled for 90 min after a 2.5-h pretreatment, as in Fig. 6, less 18-22S RNA is labeled in the presence of cordycepin than in the control. These differences in the absolute amounts of label incorporated do not necessarily indicate changes in rates of RNA synthesis. For example, adenosine has been reported to compete with uridine for transport into the cell and thereby alter the amount of [3H]uridine taken up by cells (19). Cordycepin,
VOL. 16, 1975
EFFECTS OF CORDYCEPIN ON NDV RNA
an adenosine analogue, may also show these effects. The inhibition of total incorporation may be a reflection of less [3H]uridine being available for RNA synthesis in cordycepintreated cells after a long pretreatment with drugs. It has previously been shown that the growth of NDV and Sendai virus (15) is inhibited by cordycepin, whereas the growth of vesicular stomatitis virus (10) and influenza virus (15) is not. We have confirmed that the growth of NDV is drastically inhibited by cordycepin. Our results argue against the interpretation that the different sensitivities of these viruses to cordycepin are due to different sensitivities of the viral transcriptase or poly(A)-synthesizing activities to the drug. The growth inhibition, at least in the case of NDV, does not result from an inhibition of poly(A) synthesis as has been suggested by Mahy et al. (15). However, it is probably at least partially due to an inhibition of 50S viral RNA accumulation, which is probably a secondary effect of the inhibition of protein synthesis. ACKNOWLEDGMENTS We gratefully acknowledge: the help of Sue Anne Lynch; the technical assistance of Eivor Houri and Agnes Blau; the National Science Foundation for research grant GB-30595X and the National Institute of Allergy and Infectious Disease for Public Health Service research grant AI 12467-01 under which this study was conducted; and the National Institute of General Medical Sciences for Public Health Service training grant GM-00177-16 under which S.R.W. was a predoctoral trainee.
LITERATURE CITED 1. Adesnik, M., M. Salditt, W. Thomas, and J. E. Darnell.
2.
3.
4.
5.
1972. Evidence that all messenger RNA molecules (except histone messenger RNA) contain poly (A) sequences and that the poly (A) has a nuclear function. J. Mol. Biol. 71:21-30. Aviv, H., and P. Leder. 1972. Purification of biologically active globin messenger RNA by chromatography on oligothymidylic acid-cellulose. Proc. Natl. Acad. Sci. U.S.A. 69:1408-1412. Bratt, M. A., and W. R. Gallaher. 1969. Preliminary analysis of the requirements for fusion from within and fusion from without by Newcastle disease virus. Proc. Natl. Acad. Sci. U.S.A. 64:636-543. Bratt, M. A., and W. S. Robinson. 1967. Ribonucleic acid synthesis in cells infected with Newcastle disease virus. J. Mol. Biol. 23:1-23. Bratt, M. A., and H. Rubin. 1967. Specific interference among strains of Newcastle disease virus. 1. Demon-
6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16. 17.
18.
19.
20. 21.
1583
stration and measurement of interference. Virology 33:598-608. Clavell, L. A., and M. A. Bratt. 1971. Relationship between ribonucleic acid-synthesizing capacity of ultraviolet-irradiated Newcastle disease virus and its ability to induce interferon. J. Virol. 8:500-508. Colby, D. S., V. Finnerty, and J. Lucas-Lenard. 1974. Fate of mRNA in L-cells infected with mengovims. J. Virol. 13:858-869. Collins, B. S., and M. A. Bratt. 1973. Separation of the messenger RNA of Newcastle disease virus by gel electrophoresis. Proc. Natl. Acad. Sci. U.S.A. 70:2544-2548. Darnell, J., L. Philipson, R. Wall, and M. Adesnik. 1971. Polyadenylic acid sequences: role in conversion of nuclear RNA into messenger RNA. Science 174:507-510. Ehrenfeld, E. 1974. Polyadenylation of vesicular stomatitis virus mRNA. J. Virol. 13:1055-1060. Gallaher, W. R., and M. A. Bratt. 1974. Conditional dependence offusion from within and other cell membrane alterations by Newcastle disease virus. J. Virol. 14:813-820. Huang, A. S., D. Baltimore, and M. A. Bratt. 1971. Ribonucleic acid polymerase in virions of Newcastle disease virus: comparison with the vesicular stomatitis virus polymerase. J. Virol. 7:389-394. Jekinek, W., M. Adesnik, M. Salditt, D. Sheiness, R. Wall, G. Molloy, L. Philipson, and J. E. Darnell. 1973. Further evidence on the nuclear origin and transfer to the cytoplasm of polyadenylic acid sequences in mammalian cell RNA. J. Mol. Biol. 75:515432. Kingsbury, D. W. 1973. Cell-free translation of paramyxovirus messenger RNA. J. Virol. 12:1020-1027. Mahy, B. W. J., N. J. Cox, S. J. Armstrong, and R. D. Barry. 1973. Multiplication of influenza virus in the presence of cordycepin, an inhibitor of cellular RNA synthesis. Nature (London) N. Biol. 243:172-174. Marshall, S., and D. Gillespie. 1972. Poly U tracts absent from viral RNA. Nature (London) N. Biol. 243:172-174. Mendecki, J., S. Y. Lee, and G. Brawerman. 1972. Characteristics of the polyadenylic acid segment associated with messenger ribonucleic acid in mouse sarcoma 180 Ascites cells. Biochemistry 11:792-798. Penman, S., H. Fan, S. Perlman, M. Rosbash, R. Weinberg, and E. Zylber. 1970. Distinct RNA synthesis systems of the HeLa cell. Cold Spring Harbor Symp. Quant. Biol. 35:561475. Plagemann, P. G. W., and M. F. Roth. 1969. Permeation as the rate-limiting step in the phosphorylation of uridine and choline and their incorporation into macromolecules by Novikoff Hepatoma Cells, Competitive inhibition by phenethyl alcohol, persantin, and adenosine. Biochemistry 8:4782-4788. Robinson, W. S. 1971. Sendai virus RNA synthesis and nucleocapsid formation in the presence of cycloheximide. Virology 44:494-02. Weiss, S. R., and M. A. Bratt. 1974. Polyadenylate sequences on Newcastle disease virus mRNA synthesized in vivo and in vitro. J. Virol. 13:1220-1230.
Vol. 16, No. 6 Printed in U.SA.
Copyright © 1975 American Society for Microbiology
Effect of Cordycepin (3'-Deoxyadenosine) on Virus-Specific RNA Species Synthesized in Newcastle Disease VirusInfected Cells SUSAN REICH WEISS AND MICHAEL A. BRAIT* Department of Microbiology, University of Massachusetts Medical School, Worcester, Massachusetts 01605 Received for publication 23 July 1975
Cordycepin (3'-deoxyadenosine) has no effect on the size or relative proportions of Newcastle disease virus-specific 18-22S mRNA species nor on the amount or size of the polyadenylic acid associated with them. Cordycepin does, however, cause an inhibition of incorporation of [3H]uridine into 50S virus-specific RNA relative to 18-22S RNA. This inhibition is probably not a direct effect of the drug on the synthesis of 508 viral RNA. Like cycloheximide, another drug which inhibits 50S RNA accumulation in paramyxovirus-infected cells, cordycepin inhibits protein synthesis as measured by amino acid incorporation. It is likely that the inhibition of 50S RNA accumulation is a secondary effect of protein synthesis inhibition. This is supported by the finding that concentrations of cordycepin and cycloheximide, which inhibit protein synthesis to the same extent, have the same effect on the ratio of 50 to 18-22S virus-specific RNA. During infection of chicken embryo cells by Newcastle disease virus (NDV), 50S singlestranded virion RNA is replicated and singlestranded 18-22S and 35S RNAs complementary to virion RNA are synthesized (4; B. SpanierCollins, C. W. Clinkscales, M. A. Bratt, and T. Morrison, manuscript in preparation). The 1822S and 35S RNAs contain polyadenylic acid [poly(A)I sequences (21; Spanier-Collins et al., in preparation) and serve as messengers for virus-specific proteins in cell-free protein-synthesizing systems (14; T. Morrison, L. E. Hightower, B. Spanier-Collins, S. R. Weiss, and M. A. Bratt, manuscript in preparation; SpanierCollins et al., in preparation). In NDV-infected chicken embryo cells, there are also partially base-paired virus-specific RNA species sedimenting from 35 to 50S which are intermediates in the synthesis of virion RNA and mRNA (B. Spanier-Collins and M. A. Bratt, manuscript in preparation). It would be useful to be able to preferentially inhibit viral genome replication, transcription of mRNA, or the synthesis of the poly(A) associated with this mRNA. This might allow us to independently study the mechanism and molecules and intermediates involved in one of the processes and, if poly(A) synthesis were preferentially inhibited, to learn something about the function of viral poly(A). Cordycepin (3'-deoxyadenosine) appears to have differential effects on the synthesis of various RNA species in eukaryotic cells. It has little effect on the synthesis of heterogeneous
nuclear RNA but inhibits poly(A) synthesis and the appearance of mRNA on polysomes (1, 9, 13, 17, 18). It also seems to cause a reduction in size of ribosomal precursor RNAs and to inhibit mitochondrial RNA synthesis (18). We therefore tested this drug for differential effects on NDV-specific RNA species synthesized in infected chicken embryo cells. We report here that cordycepin preferentially inhibits incorporation of [3H]uridine into 50S virus-specific RNA relative to 18-22S without affecting the size or relative proportions of the 18-22S mRNA's or the amount or size of their associated poly(A). Our results suggest that the effect on 50S RNA is probably not a specific effect of the drug on viral RNA synthesis but rather a secondary effect of the inhibition of protein synthesis by this drug. MATERIALS AND METHODS Solutions, chemicals, and abbreviations. The following buffers were used: standard buffer (0.1 M NaCl; 0.01 M Tris, pH 7.4; and 0.002 M EDTA); sodium dodecyl sulfate (SDS)-standard buffer (standard buffer with 0.5% SDS); solubilizing buffer (0.1 M NaCl; 0.01 M Tris, pH 8.5; 0.002 M EDTA; 1% SDS; and 1% mercaptoethanol); phenol buffer (solubilizing buffer without SDS); Tris-saline (0.15 M NaCl; 0.005 M KCI; 0.025 M Tris, pH 7.3; and 1 g of glucose per liter); SSC (0.15 M NaCl and 0.015 M sodium citrate); column-binding buffer (2x SSC + 0.1% SDS); column elution buffer (0.01 M Tris, pH 7.4; 0.1% SDS). The standard medium used consisted of Eagle minimal essential medium supplemented with 2.5% calf serum, 2.5% tryptose phosphate broth, and 0.07% NaHCO3.
1575
1576
WEISS AND BRArTT
[3H]uridine (28 Ci/mM), [3H]adenosine (15.5 to 27.1 Ci/mM), 32P-labeled inorganic phosphate, [3H]amino acids (1 to 10 Ci/mM), aquasol, and Omnifluor were obtained from New England Nuclear Corp. Bovine pancreatic RNase was obtained from Schwarz Bioresearch, Ti RNase and cycloheximide (Actidione) were from Calbiochem, oligo(dT)-cellulose (T2) was from Collaborative Research, and cordycepin was from Sigma Chemical Co. Actinomycin D (ActD) was generously provided by Merck, Sharp & Dohme, Rahway, N.J. Virus purification and infectivity assays. NDVHP (Israel HP, 1935) (3) was grown in embryonated hens' eggs and then concentrated from the allantoic fluid by centrifugation in a Spinco SW27 rotor for 1 h at 4 C and 24,000 rpm (74,800 x g) in standard buffer through 20% sucrose onto a 65% sucrose-D20 pad. Virus at the interface of the two sucrose layers was then diluted in Tris-saline containing 2% calf serum and stored at - 70 C. Infectivity titrations by plaque formation (PFU per milliliter) were carried out as described previously (5) but using standard medium (see above). Virus growth curve. Growth curves were done as previously described (11) except adsorption was for 45 min and incubation was at 40 C. The end of the virus adsorption period is defined as zero time. Cell culture procedures. Secondary cultures of chicken embryo cells were prepared and grown in a 5% CO2 atmosphere as previously described (3) but in standard medium. Cultures were plated at 5 x 106 cells/100-mm tissue culture plate and used after 48 h of incubation at 40 C. Intracellular RNA. Virus-specific RNA was labeled in the following manner. Cultures were infected at an input multiplicity of 5 PFU/cell with NDV-HP and incubated at 40 C. ActD at 10 ,ug/ml was always added 45 min before labeling. Cordycepin or cycloheximide was added at the concentrations and times indicated. Labeling was with 50 to 100 ,uCi of [3H]uridine or [3H]adenosine per ml for the periods indicated in each experiment. After labeling, the cells were solubilized in solubilizing buffer, phenol extracted three times with buffersaturated phenol, and precipitated twice in ethanol as previously described (4). Host cell RNA to be used as marker was labeled during a 24-h incubation with 500 ,uCi of 32p per ml and then purified as described for virus-specific RNA. Fractionation and analysis of RNA. Velocity sedimentation analysis were performed as follows. RNA dissolved in SDS-standard buffer was layered on 12-ml linear 15 to 30% sucrose gradients containing the same buffer. Centrifugation was carried out at 22 C and 39,000 rpm (180,000 x g) in a Spinco SW41 rotor for 4 h. Fractions (0.4 ml) were collected from the top using an ISCO (Instrumentation Specialties Co.) fractionator, which measured adsorbance at 260 nm, coupled to an ISCO fraction collector. Samples of each fraction were counted directly in aquasol or trichloroacetic acid precipitated and counted using previously described procedures (6). RNase digestion was done by incubating RNA at 37 C for 1 h in 2x SSC, 50 itg of pancreatic RNase per ml, and 25 U of Ti RNase per ml.
J. VIROL.
Oligo(dT)-cellulose binding was done in the following way. RNA, in 1 ml of column-binding buffer, was added to a column of 0.1 g of oligo(dT)-cellulose (previously washed with binding buffer) in a Pasteur pipette. The column was washed with three 1ml aliquots of binding buffer and then four 1-ml aliquots of column elution buffer. These 1-ml fractions were trichloroacetic acid precipitated. Most of the radioactivity is in the first 2 ml of binding buffer and the first 2 ml of elution buffer. Therefore, the RNA in these fractions is designated as binding and nonbinding RNA, respectively. Acrylamide gel electrophoresis was carried out using 10-cm 4 or 10% ethylene diacrylate crosslinked polyacrylamide gels according to the procedures, including gel slicing and counting, of Collins and Bratt (8). Amino acid labeling. Cells were infected as in RNA labeling experiments. Drugs were added after adsorption of the virus. Labeling was for 30 min with 10 ,1Ci of [3H]amino acids per ml at the times indicated. After labeling, cells were washed with Trissaline and removed from the plates with 1 N NaOH. Samples of this were trichloroacetic acid precipitated after boiling for 5 min.
RESULTS Inhibition of virus multiplication. Figure 1 shows that 50 ,ug of cordycepin per ml, in the presence or absence of 10 Ag of ActD per ml, inhibits the appearance of infectious virus almost completely (>95%) if added soon after infection. Cordycepin is significantly less effective when added 4 h after infection. Effects on intracellular virus-specific 1822S RNA and poly(A). Figure 2 shows the sucrose velocity sedimentation patterns of intracellular virus-specific RNA labeled with [3H]uridine (Fig. 2A) or [3H]adenosine (Fig. 2B) in the presence of 10 ,ug of ActD per ml and in the presence or absence of 50 ,ug of cordycepin per ml. The usual 18-22S, 358, and 50S NDV RNA species (4) are present in the cultures that did not receive cordycepin. In drug-treated cultures, the major effect of cordycepin is an apparent decrease in the ratio of 50 to 18-22S RNA. To more precisely analyze any possible effects of cordycepin on the 18-22S RNA, [3H]uridine-labeled 18-22S RNA, synthesized in the absence (Fig. 3A) or presence (Fig. 3B) of cordycepin, was analyzed by gel electrophoresis. The patterns are similar to those obtained previously (8, 21), and there are no major differences in the migration rates or relative proportions of these RNA species when synthesized in the presence or absence of cordycepin. The following experiments suggest that there is little, if any, effect of cordycepin on the synthesis of the poly(A) associated with NDV 1822S mRNA. [3H]uridine and [3H]adenosine-labeled 18-22S RNA were pooled from gradients like those in Fig. 2, and aliquots were trichloro-
EFFECTS OF CORDYCEPIN ON NDV RNA
VOL. 16, 1975
1577
Control I-I
Q,10 :D LL
a0
D at
0
-o0 Cordycepin at 0 Time
U) (n .-_
0.1ka
2
4 8 6 Hours Post Infection
10
FIG. 1. Effects of cordycepin on virus production. Growth curves were done as described in Materials and Methods. Symbols: No drug additions (a); 50 pg of cordycepin per ml added at O time (O); 50 pg of cordycepin and 10 pg of ActD per ml added at 0 time (A); 50 pg of cordycepin per ml added at 4 h postinfection (0).
acetic acid precipitated after incubation with and pancreatic RNases. or without Ti [3Hluridine-labeled RNA had only a small amount of RNase resistance, as expected for single-stranded RNA (Table 1). [3H]adenosinelabeled RNA extracted from control cells was approximately 18% resistant to RNase. We have previously shown that this RNase resistance represents poly(A) (21). [3H]adenosine-labeled RNA extracted from cordycepin-treated cells was about 16% resistant to RNase. This suggests there is about the same amount of poly(A) associated with 18-22S RNA from control or drug-treated cells. The small difference in RNase resistance of [3H]adenosine-labeled RNA from control and cordycepin-treated cells, which is observed in some experiments, is probably due to a greater contribution of radioactivity in ActD-resistant host cell background in untreated cells than in cordycepin-treated cells (see Fig. 2B). This background is highly RNase resistant (approximately 60%) in untreated and drug-treated cells. As another test of poly(A) content, we chro-
matographed [3H]uridine- and [3H]adenosinelabeled 18-22S RNA using oligo(dT)-cellulose columns. Poly(A)-containing RNA should bind to these columns (2). Approximately 70% of the [3H]uridine-labeled RNA and 90% of the [3H]adenosine-labeled RNA from control or drug-treated cells bound to these columns (Table 1). The same percentage of radioactivity from control and drug-treated cells is associated with enough poly(A) to bind to the columns. This suggests that the same percentage of 1822S RNA molecules from control and drugtreated cells contain poly(A). Finally, we determined the size of the RNasefrom fragments resistant [poly(A)] [3H]adenosine-labeled 18-22S RNA extracted from control (Fig. 4A) and drug-treated (Fig. 4B) cells. The RNase-resistant fragments were purified by phenol extraction and then co-electrophoresed In 10% polyacrylamide gels with chick cell RNA from the 4S region of a sucrose gradient. The size distributions of poly(A) in both cases are the same. As we have previously reported (21), NDV poly(A) is heterogeneous in
1578
WEISS AND BRAIT
J. VIROL.
A. 3H- Uridine 50S
35S
41 3
-
I
B. 3H-Adenosine
I
35S
50S
18S
8H 6t
41p .:I
Is
2- - > f A..gsrl
7~~~~~~'i
e E_~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
Bottom
10
20 30 Fraction Number
Top
FIG. 2. Effects of cordycepin on NDV-specific RNA. Intracellular NDV RNA (solid lines) was labeled in the presence ofActD from 4 to 9 h postinfection with [3H]uridine (A) or [3H]adenosine (B) in the absence (-) or presence (0) of 50 .g of cordycepin per ml and purified and fractionated as described in Materials and Methods. RNA from uninfected cells (dotted lines), labeled with [3Hadenosine in the presence of ActD, in the absence (@) or presence (0) of cordycepin, is also shown in (B). Samples (0.01 ml) of each fraction were counted in aquasol. When cordycepin was used, cells were pretreated with the drug for45 min before labeling. (The majority of counts at the top of the gradients [fractions 32 -39] are trichloroacetic acid-soluble counts as they are not detected when samples are trichloroacetic acid precipitated.)
size, with an average size of 4-5S. Thus, cordycepin has no effect on the size of the poly(A) associated with NDV mRNA. Effects of protein synthesis inhibition on the ratio of 50S to 18-22S RNA. The results described thus far suggest that cordycepin has
little or no effect on NDV mRNA. However, the reduction in incorporation of [3H]uridine and [3Hladenosine into 5(S intracellular RNA relative to 18-22S (Fig. 2) is consistently observed when infected cells are labeled in the presence of 50 ,ug of cordycepin per ml. The amount of
EFFECTS OF CORDYCEPIN ON NDV RNA
VOL. 16, 1975
1579
16 14 12 10
6 5 4
3 2 0
0
N
0 0~X
C)X 0~ L
N
:I
8 7 6 5 4
) to
3 2
Origin
20
40 Slice
60 80 Number
100
FIG. 3. Electrophoretic analysis of the intracellular 18-22S RNA species. [3H]uridine-labeled 18-22S RNA (-) labeled in the absence ofcordycepin (A) or in the presence ofcordycepin (B) were co-electrophoresed with 32P-labeled 188 and 28S chick cell RNA (0). Electrophoresis was for 11 h in 4% gels.
inhibition is dependent on drug concentration and the time period of treatment with the drug before labeling (data not shown). This effect of cordycepin on the ratio of 50S to 18-22S RNA is similar to that produced by treating paramyxovirus-infected cultures with the protein synthesis inhibitor cycloheximide (6, 20). Cordycepin has also been reported to inhibit protein synthesis (7). Therefore, it seemed possible that the reduction of 50S RNA accumulation relative to
18-22S might be a secondary effect of the inhibition of protein synthesis and not a direct effect on 50S RNA synthesis. To quantitatively compare the effects of cordycepin with a known inhibitor of protein synthesis, we first determined the extent to which cordycepin inhibits protein synthesis in chicken embryo cells under conditions used in our experiments. This was done by measuring the rate of accumulation of [3H]amino acids into trichloroa-
1580
WEISS AND BRATT
J. VIROL.
TABLE 1. RNase sensitivity and oligo(dT)-cellulose binding of NDV mRNA Counts (%) resistant to Ti and Counts (%) bindpancreatic RNase treatmenta RNA ing to oligo(dT)-
[3H]uridine-labeled 18-22S RNA (no cordycepin) [3H]uridine-labeled 18-22S RNA (+ cordycepin) [3H]adenosine-labeled 18-22S RNA (no cordycepin) [3H]adenosine-labeled 18-22S RNA (+ cordycepin)
Expt 1
Expt 2
Expt 3
Avg
cellulose
3.0 4.4 17.8 16.0
3.6 2.1 18.2 18.6
4.1 2.8 17.8 12.8
3.6 3.1 17.9 15.8
67.3 68.2 88.4 88.7
a Fractions were pooled from gradients similar to those in Fig. 2, and duplicate samples were treated with RNase as described in Materials and Methods. Three separate preparations of RNA were used (experiments 1, 2, 3). Percentage of resistance is defined as the ratio of trichloroacetic acid-precipitable counts before and after treatment, times 100. bFractions were pooled as in a, and the amount of counts binding to oligo(dT)-cellulose columns was determined as described in Materials and Methods. Percentage of binding is defined as the ratio of columnbinding counts to total counts added to the column, times 100.
N
I
6 5 4 3 2
6 5 4 3 2
0
Hours
10 20 30 40 50 60 70 80 Slice Number FIG. 4. Electrophoretic analysis of poly(A). Poly(A) (a) from 18-22S NDV RNA labeled in the absence (A) or presence (B) of cordycepin was coelectrophoresed with 32P-labeled 4, 5, and 7S chick cell RNA. Electrophoresis was for 5 h in 10% gels. Arrows designate the positions of marker cell RNAs.
cetic acid-precipitable material at various times after the addition of cordycepin to uninfected and infected chicken embryo cells. This rate of accumulation in the presence of cordycepin (relative to the untreated control) decays with a half-life of about 2 to 2.5 h (Fig. 5). The
3
5
Post
Infection
7
FIG. 5. Effects of cordycepin and ActD on amino acid incorporation. NDV-infected (solid line) or mock-infected (dashed line) cells were labeled with [3H]amino acids for 30 min at the times shown, and counts incorporated into trichloroacetic acid-precipitable material were determined. Ten micrograms of ActD (a) per ml, 10 pg of ActD and 50 pg of cordycepin (O) per ml, or 50 pg of cordycepin (0) per ml was added at the time of infection.
kinetics of inhibition are independent of the time during infection that the drug is added (data not shown). A small inibition of trichloroacetic acid-soluble radioactivity is also observed in the presence of cordycepin (data not shown), but this is not enough to account for the inhibition of trichloroacetic acid-precipitable radioactivity seen here. We also determined the effect of ActD on amino acid accumulation, because it was used in all the RNA labeling experiments. ActD (10 ,ug/ml) inhibits amino acid incorporation but with slower kinetics than cordycepin (Fig. 5), and ActD and cordycepin together show an effect similar to cordycepin alone.
EFFECTS OF CORDYCEPIN ON NDV RNA
VOL. 16, 1975
Next we determined the concentration of cycloheximide that inhibits amino acid incorporation to the same extent as 50 Ag of cordycepin per ml. We found that 0.1 ,ug of cycloheximide per ml (in the presence of 10 ,ug of ActD per ml) inhibits amino acid incorporation by about 50% in NDV-infected chicken embryo cells (data not shown) and, after a 2.5-h incubation, 50 Ag of cordycepin per ml (in the presence or absence of ActD) causes a similar inhibition. (The amount of inhibition after 2.5 h of cordycepin treatment varies from 50% in many experiments to as much as 65%, as seen in Fig. 5.) Finally, we compared the patterns of [3H]uridine-labeled virus-specific RNA ex-
1581
tracted from cells that were treated with cordycepin or cycloheximide with RNA patterns from untreated control cells (Fig. 6). All of the RNA was labeled in the presence of ActD. When cordycepin or cycloheximide was used, it was added to the cells 2.5 h before labeling and was present during labeling. Figure 6A shows the control pattern of RNA. The usual 18-22S, 35S, and 50S species are present (The relative amounts of virus-specific 50S and 18-22S RNA synthesized in the presence of ActD by NDVHP varies greatly from one experiment to the next [compare the patterns in Fig. 2A and Fig. 6A] but are always constant within one experiment.) Figure 6B shows the pattern of RNA
4
3
2
I
16
12
8 4
sMGM
10
20
30TOP
Fraction
Botom
10
20
30 "P
Number
FIG. 6. Comparison of NDV-specific RNA synthesized in the absence or presence of cordycepin and cycloheximide. Intracellular virus-specific RNA was labeled, in the presence of ActD, with [3H]uridine for 90 min at 6.5 h postinfection. Cordycepin or cycloheximide, as indicated, was added 2.5 h before labeling in (B), (C), and (D) and continued to be present during labeling. RNA was purified and fractionated as described in Materials and Methods. Gradient fractions were trichloroacetic acid precipitated.
1582
WEISS AND BRATT
J. VIROL.
TABLE 2. Effect of cycloheximide and cordycepin on the ratio of 50S to 18-22S RNA Treatment
5OS/18-22Sa
50S/18-22S as % of
control°
Control Cordycepin, 50 ,ug/ml Cycloheximide, 0.1 ,ug/ml Cycloheximide, 50 j,g/ml
2.57 1.28 1.64 0.05
100.0 50.6 63.5 1.9
a Total counts in the 50S and 18-22S regions of the gradients in Fig. 6 were calculated, and the total 50S counts were divided by the total 18-22S counts. b The values in the first column were divided by the ratio for the control and multiplied by 100.
extracted from cells that were treated with 50 ,zg of cordycepin per ml, and Fig. 6D shows the pattern of RNA from cells that were treated with 0.1 ,ug of cycloheximide per ml. In both the cordycepin- and cycloheximide-treated cells, the amount of label in 50S RNA relative to 1822S is reduced. The ratios of 50S to 18-22S are about 51% of the control in RNA labeled in the presence of 50 ,ug of cordycepin per ml and 64% of the control in RNA labeled in the presence of 0.1 ,ug of cycloheximide per ml (Table 2). Therefore, when amino acid incorporation is inhibited by about 50% by cycloheximide or cordycepin, there is approximately the same effect on the ratio of 50S to 18-22S RNA. Figure 6C shows the pattern of RNA from cells that were treated with a much higher concentration of cycloheximide, 50 ,ug/ml. Under these conditions, where incorpbration of [3H]amino acids into trichloroacetic acid-precipitable material is inhibited by 95%, incorporation of [3H]uridine into 50S RNA is almost completely inhibited. The ratio of 50 to 18-22S RNA labeled in the presence of 50 ,g of cycloheximide per ml is 2% of that of the control RNA (Table 2). DISCUSSION Cordycepin has no effect on the size or relative proportions of the six to seven 18-22S NDV mRNA species; nor does it have an effect on the poly(A) associated with this RNA. This is in contrast to the poly(A) associated with the heterogeneous nuclear and mRNA of eukaryotic cells, the synthesis of which is inhibited by cordycepin (1, 9, 13, 17). NDV RNA is synthesized solely in the cytoplasm (4). NDV virions and cores isolated from virions by detergent treatment can synthesize 18-22S RNA (6, 12) and its associated poly(A) (21) in vitro. There is also evidence suggesting that NDV poly(A) is not transcribed from polyuridylic acid sequences in virion RNA (16). This sug-
gests that NDV has its own poly(A)-synthesizing activity, but it is still possible that the enzyme that synthesizes viral poly(A) is a cellular, cytoplasmic, cordycepin-insensitive enzyme that is incorporated into virus particles. It has been shown that the protein synthesis inhibitor cycloheximide alters the ratio of 50S to 18-22S RNA in paramyxovirus-infected cells (6, 20), probably because it inhibits the replication of virion RNA (Spanier-Collins and Bratt, in preparation). We have shown here that cordycepin, another inhibitor of protein synthesis, also alters this ratio. At concentrations ofcordycepin and cycloheximide that inhibit amino acid incorporation to the same extent, there is a similar effect on the ratio of 50 to 18-22S RNA. The inhibition of amino acid incorporation by cordycepin increases with time of drug treatment (Fig. 5); the effect of cycloheximide is faster. By 1 h after addition of the drug, the maximum effect is seen (data not shown). It is noteworthy that cycloheximide and cordycepin, which show different kinetics of inhibition of amino acid accumulation and almost certainly work through different mechanisms, have similar effects on nucleotide incorporation into NDV intracellular 50S RNA relative to 18-22S RNA. Although we have not ruled out all other possibilities, it seems probable that the effect of cordycepin on 5S RNA is a secondary effect of protein synthesis inhibition. Whereas cycloheximide and cordycepin always inhibit the incorporation of [3H]uridine into 5S RNA relative to 18-22S RNA, they have different effects on the absolute number of counts incorporated into 18-22S RNA. Cycloheximide consistently causes an increase in the amount of incorporation into 18-22S RNA. The effect is most dramatic at the high cycloheximide concentration (see Fig. 6). Cordycepin, on the other hand, causes a decrease in the absolute amount of label incorporated into 18-22S RNA, but only when RNA is labeled under certain conditions. In experiments like the one shown in Fig. 2, where labeling is for 5 h after a 45-min drug pretreatment, approximately the same amount of label is incorporated into 18-22S RNA with or without cordycepin. When cells are labeled for 90 min after a 2.5-h pretreatment, as in Fig. 6, less 18-22S RNA is labeled in the presence of cordycepin than in the control. These differences in the absolute amounts of label incorporated do not necessarily indicate changes in rates of RNA synthesis. For example, adenosine has been reported to compete with uridine for transport into the cell and thereby alter the amount of [3H]uridine taken up by cells (19). Cordycepin,
VOL. 16, 1975
EFFECTS OF CORDYCEPIN ON NDV RNA
an adenosine analogue, may also show these effects. The inhibition of total incorporation may be a reflection of less [3H]uridine being available for RNA synthesis in cordycepintreated cells after a long pretreatment with drugs. It has previously been shown that the growth of NDV and Sendai virus (15) is inhibited by cordycepin, whereas the growth of vesicular stomatitis virus (10) and influenza virus (15) is not. We have confirmed that the growth of NDV is drastically inhibited by cordycepin. Our results argue against the interpretation that the different sensitivities of these viruses to cordycepin are due to different sensitivities of the viral transcriptase or poly(A)-synthesizing activities to the drug. The growth inhibition, at least in the case of NDV, does not result from an inhibition of poly(A) synthesis as has been suggested by Mahy et al. (15). However, it is probably at least partially due to an inhibition of 50S viral RNA accumulation, which is probably a secondary effect of the inhibition of protein synthesis. ACKNOWLEDGMENTS We gratefully acknowledge: the help of Sue Anne Lynch; the technical assistance of Eivor Houri and Agnes Blau; the National Science Foundation for research grant GB-30595X and the National Institute of Allergy and Infectious Disease for Public Health Service research grant AI 12467-01 under which this study was conducted; and the National Institute of General Medical Sciences for Public Health Service training grant GM-00177-16 under which S.R.W. was a predoctoral trainee.
LITERATURE CITED 1. Adesnik, M., M. Salditt, W. Thomas, and J. E. Darnell.
2.
3.
4.
5.
1972. Evidence that all messenger RNA molecules (except histone messenger RNA) contain poly (A) sequences and that the poly (A) has a nuclear function. J. Mol. Biol. 71:21-30. Aviv, H., and P. Leder. 1972. Purification of biologically active globin messenger RNA by chromatography on oligothymidylic acid-cellulose. Proc. Natl. Acad. Sci. U.S.A. 69:1408-1412. Bratt, M. A., and W. R. Gallaher. 1969. Preliminary analysis of the requirements for fusion from within and fusion from without by Newcastle disease virus. Proc. Natl. Acad. Sci. U.S.A. 64:636-543. Bratt, M. A., and W. S. Robinson. 1967. Ribonucleic acid synthesis in cells infected with Newcastle disease virus. J. Mol. Biol. 23:1-23. Bratt, M. A., and H. Rubin. 1967. Specific interference among strains of Newcastle disease virus. 1. Demon-
6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16. 17.
18.
19.
20. 21.
1583
stration and measurement of interference. Virology 33:598-608. Clavell, L. A., and M. A. Bratt. 1971. Relationship between ribonucleic acid-synthesizing capacity of ultraviolet-irradiated Newcastle disease virus and its ability to induce interferon. J. Virol. 8:500-508. Colby, D. S., V. Finnerty, and J. Lucas-Lenard. 1974. Fate of mRNA in L-cells infected with mengovims. J. Virol. 13:858-869. Collins, B. S., and M. A. Bratt. 1973. Separation of the messenger RNA of Newcastle disease virus by gel electrophoresis. Proc. Natl. Acad. Sci. U.S.A. 70:2544-2548. Darnell, J., L. Philipson, R. Wall, and M. Adesnik. 1971. Polyadenylic acid sequences: role in conversion of nuclear RNA into messenger RNA. Science 174:507-510. Ehrenfeld, E. 1974. Polyadenylation of vesicular stomatitis virus mRNA. J. Virol. 13:1055-1060. Gallaher, W. R., and M. A. Bratt. 1974. Conditional dependence offusion from within and other cell membrane alterations by Newcastle disease virus. J. Virol. 14:813-820. Huang, A. S., D. Baltimore, and M. A. Bratt. 1971. Ribonucleic acid polymerase in virions of Newcastle disease virus: comparison with the vesicular stomatitis virus polymerase. J. Virol. 7:389-394. Jekinek, W., M. Adesnik, M. Salditt, D. Sheiness, R. Wall, G. Molloy, L. Philipson, and J. E. Darnell. 1973. Further evidence on the nuclear origin and transfer to the cytoplasm of polyadenylic acid sequences in mammalian cell RNA. J. Mol. Biol. 75:515432. Kingsbury, D. W. 1973. Cell-free translation of paramyxovirus messenger RNA. J. Virol. 12:1020-1027. Mahy, B. W. J., N. J. Cox, S. J. Armstrong, and R. D. Barry. 1973. Multiplication of influenza virus in the presence of cordycepin, an inhibitor of cellular RNA synthesis. Nature (London) N. Biol. 243:172-174. Marshall, S., and D. Gillespie. 1972. Poly U tracts absent from viral RNA. Nature (London) N. Biol. 243:172-174. Mendecki, J., S. Y. Lee, and G. Brawerman. 1972. Characteristics of the polyadenylic acid segment associated with messenger ribonucleic acid in mouse sarcoma 180 Ascites cells. Biochemistry 11:792-798. Penman, S., H. Fan, S. Perlman, M. Rosbash, R. Weinberg, and E. Zylber. 1970. Distinct RNA synthesis systems of the HeLa cell. Cold Spring Harbor Symp. Quant. Biol. 35:561475. Plagemann, P. G. W., and M. F. Roth. 1969. Permeation as the rate-limiting step in the phosphorylation of uridine and choline and their incorporation into macromolecules by Novikoff Hepatoma Cells, Competitive inhibition by phenethyl alcohol, persantin, and adenosine. Biochemistry 8:4782-4788. Robinson, W. S. 1971. Sendai virus RNA synthesis and nucleocapsid formation in the presence of cycloheximide. Virology 44:494-02. Weiss, S. R., and M. A. Bratt. 1974. Polyadenylate sequences on Newcastle disease virus mRNA synthesized in vivo and in vitro. J. Virol. 13:1220-1230.
