VIROLOGY
184,768-772
(1991)
Effect of Nucleotide JACQUELINE M. PIZARRO, Jo& Unit of Virology,
lnstituto
de Nutricidn
Analogues
on Rotavirus Transcription
and Replication
L. PIZARRO, JORGE FERNANDEZ, ANA MARIA SANDINO, AND EUGENIO SPENCER y Tecnologla Received
April
de 10s Alimentos, 23,
199 1; accepted
Universidad June
de Chile,
Macul5540,
Santiago
11, Chile
19, 199 1
The effects of several nucleoside and nucleoside triphosphate analogues were studied on rotavirus replication and transcription. Nucleoside triphosphate analogues modified at sugar residues were capable of inhibiting in vitro transcription, including adenosine-9-P-o-arabinofuranoside 5’-triphosphate, 3’-deoxyadenosine 5’-triphosphate, adenosine 5’-triphosphate 2’,3’-dialdehyde, guanosine 5’-triphosphate 2’,3’-dialdehyde, and cytosine-9-fl-o-arabinofuranoside 5’triphosphate. Two dialdehyde derivatives, adenosine 5’-triphosphate 2’3’ dialdehyde and guanosine 5’-triphosphate 2’,3’-dialdehyde, were irreversible inhibitors, forming a stable complex with the viral polypeptide VP3. The effect of the corresponding nucleosides of the inhibitory analogues was studied in SA-11 rotavirus-infected MA-104 cells. Adenosine-9-P-o-arabinofuranoside and 3’-deoxyadenosine were effective inhibitors of RNA synthesis, an effect that could be due to their inhibition of viral transcription. o 1991 Academic Press. tnc.
Rotavirus is the single most important cause of acute gastroenteritis in infants and young children in developed countries (3). The viral particle is made of a double icosahedral capsid surrounding the viral core containing the viral genome (1). The outer capsid is made of two polypeptides named VP4 (88 kDa) and VP7 (38 kDa) which play an important role in cell adsorption and penetration (2). The inner capsid is made up of the most abundant viral polypeptide, VP6 (42 kDa) (9). The viral core is composed of three polypeptides, VP1 (128 kDa), VP2 (95 kDa), and VP3 (88 kDa). The viral genome consists of 1 1 double-stranded RNA segments which code for at least 10 polypeptides, 6 of which are structural (I). The plus strand of each viral RNA is capped at the 5’ end and is identical to viral mRNA (8). Like other members of the Reoviridae family, rotaviruses RNA synthesis is catalyzed by a RNA polymerase activity associated with the polypeptide VP1 , which transcribes mRNA from each of the 11 genomic segments (24). The viral polymerase is activated in vitro after subjecting the particle to a thermal shock or EDTA treatment (22). Rotavirus transcription involves other associated enzymatic activities present in the viral core, in addition to the RNA polymerase activity (15). These include the inner viral capsid which is an absolute requirement for transcription even though it does not have any apparent enzymatic activity associated for RNA synthesis (78). The 5’ cap addition to each of the viral transcripts is catalyzed by a guanylyltransferase activity associated with VP3, present in the viral core ( 15). The formation of the 5’ end cap is important because it commits transcription toward the elon0042.6822/91
$3.00
CopyrIght 0 199 1 by Academic Press. Inc. All rights of reproduction in any form reserved.
gation of the RNA chain causing a decrease in K, values for the ribonucleoside triphosphates (23). As in several eukaryotic and viral systems including rotavirus, ATP seems to play an important role in transcription, coupling the ATP hydrolysis to transcription and as a requirement for the polymerization itself (13, 16, 23). Inhibition of ATP hydrolysis during in vitro transcription abolishes RNA synthesis (23). This may be due to an enzymatic activity which uses ATP as substrate for hydrolysis, an activity which is different from the polymerase activity itself (6, 10). Based on the nucleotide requirement for all the viral enzymes involved in transcription, the effect of several nucleotide analogues including some described as antiviral agents was studied on in vitro and in vivo rotavirus RNA synthesis. We focused our interest on the effect of ATP analogues due to the important role that ATP plays during transcription. To determine the inhibitory effect of a particular analogue on the in vitro transcription, the analogue was added to a standard transcription mixture, in the presence of heat-activated double-shelled rotavirus particles and incubated for 30 min at 45°C. The incorporation of [3H]UMP into acid-insoluble material was determined as described previously (22). Each analogue was tested using concentrations ranging from 1 to 10 m/l/l to overcome any effect derived from the K,,, of the polymerase for the particular analogue. A nucleotide analogue was only considered an inhibitor if, with a concentration of 4 mll/l or less, 50°b inhibition was observed. This concentration was selected because it corresponds to the optimum concentration of each nucleoside triphosphate for transcription in the in 768
SHORT
1..
I 1236.56769
769
COMMUNICATIONS
1234567 imMl analog
FIG. 1. Inhibition of in vitro rotavirus transcription by nucleotide analogues. The results are expressed as percentage from the control reaction done in the absence of the analogue. The analogue was added to a standard mixture at the indicated concentrations and [3H]UMP incorporation was determined as radioactivity associated to acid-insoluble material as described previously (22). (0) Ara ATP, (0) Ara CTP, (X) 3’-dATP, (m) 2’,3’-dialdehyde ATP, (0) 2’,3’-dialdehyde GTP.
vitro assay. All the analogues tested were purchased from Sigma Chemical Co. The nucleotide analogues tested were grouped according to their modifications in phosphate bonds, base, sugar, and base-sugar residues. The most important analogues inhibiting transcription were adenosine 5’-triphosphate 2’,3’-dialdehyde (2’,3’-dialdehyde ATP) and guanosine 5’-triphosphate 2’,3’-dialdehyde (2’,3’-dialdehyde GTP), producing 81.6 and 76.7% inhibition at 4 mM, respectively, and approximately 50% inhibition at 1 mM(Fig. 1 b). They are acyclic analogues which have in the sugar residue the 2’,3’ bond oxidized to aldehyde and do not contain available OH groups for the formation of the phosphodiester bond on the growing RNA chain. 3’-Deoxyadenosine 5’-triphosphate such as 3’-dATP, an analogue which has modified the 3’-OH in the sugar residue, gave 75.9% inhibition at 4 mM (Fig. 1 c). This analogue has been described as an inhibitor of poly(A)synthesis acting as a chain terminator at the 3’ end of the chain (6, 11, 17). In this case it may be inhibiting the enzymatic system associated with viral transcription, because the viral transcripts do not contain poly(A) sequences. Adenosine-9-P-o-arabinofuranoside 5’-triphosphate (Ara ATP), a powerful inhibitor of several steps of the replication and transcription of some DNA and RNA viruses (5, 10, 12, 20) and cytosine-9-@-o-arabinofuranoside 5’-triphosphate (Ara CTP), were able to inhibit transcription 74.4 and 70.0%, respectively, at 4 mM (Fig. 1 a). Other ATP analogues which contain a modification in the purine ring (such as adenosine 5’-triphosphate-N-oxide, xanthosine 5’-triphosphate, 8-bromoadenosine 5’-triphosphate, and 1 ,/V6-ethenoadenosine 5’-triphosphate) or in the sugar moiety (such as 2’-dATP, 2’-dCTP, 2’dGTP, 2’-dlTP, and 2’-3’-ddTTP) or in both sugar and
base (such as 2’-deoxyinosine 5’-triphosphate) were not considered for further study because they inhibited in vitro transcription less than 50% at 4 mM. Some of them at higher concentrations inhibited RNA synthesis but this effect may correspond to what is seen with an excess of the regular nucleotide substrates at similar concentrations. To characterize the nature of the inhibition produced by the analogues defined as inhibitors an experiment was designed as follows: activated virus was incubated for 30 min at 45°C in a standard mixture of transcription including the indicated analogue at 7 mM, a concentration ensuring the maximal inhibitory effect. The virus was then separated by centrifugation and RNA synthesis was measured in the supernatant, as previously described (22). The pelleted virus was washed and resuspended with 1 vol of 50 mM TrisHCI buffer, pH 7.8, and allowed to transcribe by addition of a new transcription mixture but in the absence of the analogue. After incubation for 30 min at 45”C, RNA synthesis was measured again. Table 1 shows the percentage of transcriptional activity detected before and after the above treatment. The results show that for all the analogues tested, with the exception of 2’,3’-dialdehyde ATP and 2’,3’-dialdehyde GTP, there was an inhibitory effect which could be overcome by washing out the analogue (Table 1). The inhibition produced by the two dialdehyde nucleotides could not be reversed by washing, suggesting that they may irreversibly modify a viral polypeptide required for viral transcription. Nucleotide analogues possessing 2’,3’dialdehyde groups have been reported to react through the formation of covalent bonds with protein residues involved in nucleotide binding, by forming Schiff bases between the aldehyde groups and amino acid residues of the protein (13, 16). To identify the rotavirus polypeptide responsible for
TABLE RECOVERY
OF THE TRANSCRIPTIONAL
1 ACTIVITV
OF THE PREINCUBATED
VIRUSINTHEPRESENCEOFTHEINHIBITORANALOGUEAFTERVIRALPURIFICATION
Inhibitory
analogue
Ara CTP (8.64 mM) Ara ATP (7.00 mM) 3’-dATP (7.02 mn/l) 2’,3’-dialdehyde ATP (7.00 2’,3’dialdehyde GTP (7.00
% Inhibition
rnk!) mM)
93.4 92.2 92.6 93.4 82.5
% Recovery” 83.7 87.7 59.7 26.4 26.0
a The percentage of recovery was corrected for a factor involving the percentage of recovery of the transcriptional activity of virus subjected to the same treatment but in the absence of the analogue.
770
FIG. 2. Fluorogram viral polypeptidei.
SHORT
showing
the binding
COMMUNICATIONS
of [(r-32P]-2’,3’-dialdehyde
the binding of 2’,3’-dialdehyde ATP, purified viral particles were incubated in the presence of 50 mM TrisHCI, pH 7.9, 20 mM MgCI,, and increasing amounts of [a-32P]-2’,3’-dialdehyde ATP (0.5 to 2 &i) prepared from [(Y-~*P]ATP (10 $Xml) (4). The reaction mixture was incubated at 37°C for 20 min and then stopped by the addition of 10 mM EDTA. Viral polypeptides were then subjected to SDS-PAGE and processed for autoradiography. Under these conditions onlyoneviral polypeptide was able to form a stable complex with the analogue (Fig. 2). In comparison to the 35S-labeled rotavirus SA-11 polypeptides, it migrates with a mobility corresponding to the viral core polypeptide VP3. Similar results were obtained with 2’,3’-dialdehyde GTP, showing that it binds to VP3 identified as the rotavirus guanylyltransferase (15). This has been shown by the electrophoretic mobility of the polypeptide present in the subviral particles such as the viral core made of VP1 , VP2, and VP3. Furthermore it has been reported that the viral polypeptide VP1 is the rotavirus RNA polymerase (24). Based on the above results it may be concluded that the 2’,3’-dialdehyde ATP and 2’,3’-dialdehyde GTP bind to rotavirus VP3. To perform the in viva studies only the analogues that inhibit in vitro transcription at the levels described above were selected to study their in vivo effect on viral double-stranded RNA replication. Due to the nature of the dialdehyde analogues they were excluded from the in vivo study but others which have been described as inhibitors of viral transcription or replication were included. For this purpose MA-104 cells were infected with 1.5 PFU of trypsin-activated SA-11 rotavirus per cell as described previously (19) and incubated at 37°C with different concentrations of the nucleosidic form of
ATP to viral proteins.
The
markers
indicates
the migration
position
of the
the analogues or other known antiviral agents. No inhibition of the viral RNA replication was observed at millimolar concentrations of 5,6-dichloride-@-o-ribofuranosyl benzymidazole and the antiviral agent ribavirin, both base modified (data not shown). Similar to in vitro results the most potent inhibitors of RNA synthesis were the sugar-modified nucleotide analogues. However, analogues such as 2’,3’-dideoxytimidine and acycloguanosine were not inhibitors of transcription or replication (data not shown). The analogues 3’-deoxyadenosine (cordycepin) and adenosine-9-@-o-ribofuranoside (Ara A) were very efficient inhibitors of the in vivo viral RNA replication, at concentrations of 0.05 and 0.4 mM, respectively (data not shown). Ara CTP showed a strong in vitro inhibitory effect but it failed to inhibit in vivo transcription or replication in a very wide range of concentrations tested. This result is unexpected because this analogue has been described as an effective antineoplasic agent but it must be phosphorylated in order to become involved in nucleic acid synthesis (7, 27). The discrepancy between the in vitro and in vivo effect may be due to the permeability of the MA104 cells to the nucleoside or to the inability of the cellular kinases to phosphorylate the nucleoside sufficiently at the concentration required to exert an inhibitory effect in vivo. In order to determine the critical time point for the inhibition of viral RNA replication by cordycepin and Ara A the analogues were added at different times after infection. Under the infection conditions described above viral transcription is initiated shortly after infection and reaches a maximum after 4 hr. By 6 hr postinfection most of the single-stranded viral RNA has been synthesized and viral RNA replication has actively
SHORT a
12
3
4
56
7
8
91011
COMMUNICATIONS
dmiA -1 c2 r_L3 t5 t6
+--IO .“_ .*
.
tll
b Cl +---2 +3 +--4 t5 t6
4 t-10 tll
FIG. 3. Effect of nucleoside analogues on in viva transcription and replication. The nucleosidic form of the analogue was added to infectedcellsat0,0.5. 1.0, 1.5,2.0,3.0,4.0,and6.0hrpostinfection. and then the cells were further incubated until 7 hr postinfection. At the end of the incubation, infected cells were scrapped and phenol was extracted, processed for PAGE, and stained with silver nitrate. Control experiments were carried where the analogue was added together with the viral inoculum. After washing with fresh media the cells were incubated for 7 hr and processed as above. (a) Gel electrophoresis of a time course inhibition of synthesis of double-stranded RNA by 0.4 mll/l cordycepin and (b) 0.8 mlLl Ara A. Lanes 3 to 10 correspond to the analogue addition at times 0, 0.5, 1 .O. 1.5, 2.0, 3.0, 4.0, and 6.0 hr postinfection, respectively. Lanes 1, 2, and 1 1 correspond to the RNA extracted from the mock-infected cells, the adsorption control, and the viral control done in the absence of the analogue. Positions of the genome segments and transcripts are indicated by arrows.
taken place using viral mRNA as template (14, 19). The analogues were tested at concentrations where viral RNA replication was drastically inhibited without any detectable cytotoxic effect after 12 hr of incubation, i.e., 0.4 mMcordycepin and 0.8 mMAra A. The experiment was designed to determine if the effect of the analogues was during an early event like transcription
771
or during viral replication itself. The effect of the addition of 0.4 mM cordycepin on genomic RNA synthesis at different times postinfection is shown in Fig. 3a. Double-stranded RNA synthesis was not detected at 8 hr postinfection when the analogue was added during the first 3 hr of infection (Fig. 3a, lane 8). However, when the analogue was added 3 to 4 hr postinfection no difference was observed in double-stranded RNA synthesis when compared with the control done in absence of the analogue (Fig. 3a, lane 11). Thus under experimental conditions, inhibition of viral RNA replication is produced at a point where most of the singlestranded RNA is being synthesized and before viral RNA replication takes place (19). Similar results were obtained with 0.8 mMAra A which is an efficient inhibitor of the synthesis of double-stranded RNA when added 1 and 5 hr postinfection (Fig. 3b, lanes 3-6). Also at later times of analogue addition, no effect on RNA replication was observed (Fig. 3b, lanes 7-10). Controls done by adding either analogue in the viral adsorption mixture during the course of an hour are shown in Fig. 3a and 3b, lane 2. These controls were done to demonstrate that the analogues did not have any effect on viral infection, since the genomic RNA synthesis is not altered if compared with the viral control done in absence of the analogue. These results indicate that the inhibition of the double-stranded RNA synthesis may be due to the inhibition of the singlestranded RNA synthesis rather than a direct effect on replication itself. This is supported by the fact that 2 hr postinfection the first round of mRNA synthesis is taking place and little or no double-stranded RNA synthesis is detected (19). This can also be demonstrated by the observation that when the analogue was added at later times, i.e., 3.0, 4.0, and 6.0 hr postinfection, double-stranded RNA was detected (Fig. 3a, lanes 8-10 for 0.4 mM cordycepin and Fig. 3b, lanes 8- 10 for 0.8 mMAra A). At these later times of addition the levels of RNA synthesis were similar to those obtained in experiments performed in absence of the analogue (Figs. 3a and 3b, lane 11). Since viral mRNA plays two roles, one for polypeptide synthesis and the other as a template for RNA replication, the early inhibitory effect may be due to the inhibition of mRNA synthesis necessary for viral polypeptide synthesis which is required for replication or may be due to a direct effect on viral translation which is actively taking place at 3 hr postinfection. This later possibility is highly unlikely due to the nature of the analogues. Rotavirus mRNA is also required as a template for double-stranded RNA synthesis, together with some nonstructural polypeptides (14). These results may explain why the analogues inhibit viral RNA replication only when mRNA synthesis is inhibited. This also
772
’
SHORT
COMMUNICATIONS
suggests that another viral polypeptide, VP3, in addition to VPl, which has been described as the RNA polymerase, is necessary for rotavirus transcription. Polypeptide VP3 may be part of a multienzymatic complex present in the core of the single-shelled virus responsible of viral transcription (15). These results may be useful for future studies of viral replication and transcription in viva and to approach the development of efficient antiviral agents in the treatment of rotavirus acute gastroenteritis. ACKNOWLEDGMENTS The present research was supported by a Grant from Fondo de lnvestigacibn Cientffica y Tecnol6gica (FONDECYT) No. 1017 and from a SAREC (Sweden) grant to E.S. A. M. Sandino is a recipient of a Fundaci6n Andes fellowship. The authors thank Dr. Clinton Chichester for his revision of the manuscript and Mr. Patricia Ortiz for his technical assistance.
REFERENCES 1. ALTENBURG, B., GRAHAM, D., and ESTES, M. /. Gen. viral. 46, 75-85 (1980). 2. CLARK, S., SPENDLOVE, R., and BARNEIT, B. 1. Wrol. 34, 272-276 (1980). 3. CUKOR. G., and BLACKLOW, N. Microbial. Rev. 48, 157-179 (1984). 4. WANG, G., WANG, S., and PAN, F. Biochem. J. 199, 281-287 (1981). 5. DE CLERQ, E. Biochem. fharmacol. 36, 2567-2575 (1987).
6. GUMPORT, R., EDELHEIT, E., UEMATSU, T., and SUHADOLNIK, R. Biochemistry 15, 2804-2809 (1976). 7. HARRIS, A., POTTER, C., BUNCH, C., BOUTAGY, J., HARVEY, D., and GRAHAME-SMITH, D. Br. J. Pharmacol. 8, 219-227 (1979). 8. IMAI, M.. AKATANI, K., IKEGAMI, N., and FURUICHI, Y. J. l&o/. 47, 125-136 (1983). 125,194-205 9. KALIKA, A., FLORES, J., and GREENBERG, H. Wrology (1983). J., and CHIANG, P. 10. KIM, I., ZHANG, C., CANTONI, G., MONTGOMERY, Biochim. Biophys. Acta 829, 150-l 55 (1985). 11. KOCH, S., and NIESSING, J. EBS Lefr. 96, 354-356 (1978). 12. PARKER, WI., and CHENG, Y. Mol. Pharmacol. 31, 146-151 (1978). 13. PATEL-THOMBRE, U., and BORCHARDT, R. Biochemistry 24, 11301136 (1985). 14. PATTON. J. T. virus Res. 6, 217-233 (1986). 15. PIZ~RRO, J. L., SANDINO, A. M., PIZARRO, J. M., FERNANDEZ, J., AND SPENCER, E. J. Gen. Viral. 72, 325-332 (1991). 16. RAYFORD, R., ANTHONY, D., JR., O’NEILL, R., JR., and MERRICK. W. 1. Biol. Chem. 2260, 15708-l 5713 (1985). 17. ROSE, K., BELL, L., and JACOB, S. Nature 267, 178-l 80 (1977). 18. SANDINO, A., JASHES. M., FAUNDEZ, G., and SPENCER, E. 1. Viral. 60,797-802 (1986). 19. SANDINO, A., PIZARRO, J., FERNANDEZ, J., FELLAY, M., and SPENCER, E. Arch. Viol. Med. Exp. 21, 381-392 (1988). 20. SCHANCHE, J., SCHANCHE, T., UELAND, P., and MONTGOMERY, J. Cancer Res. 44,4297-4302 (1984). 21. SOKAL, J., LEONG, S., and GOMEZ, G. Cancer59,197-202 (1987). 22. SPENCER, E., and ARIAS, M. J. Viral. 40, l-10 (1981). 23. SPENCER, E., and GARCIA, B. 1. Viral. 52, 188-197 (1984). 24. VALENZUELA, S., PIZARRO, J., SANDINO, A., VASQUEZ, M., FERNANDEZ, J., HERNANDEZ, O., PATON, J., and SPENCER, E. /. Viral. 65, in press (1991).
184,768-772
(1991)
Effect of Nucleotide JACQUELINE M. PIZARRO, Jo& Unit of Virology,
lnstituto
de Nutricidn
Analogues
on Rotavirus Transcription
and Replication
L. PIZARRO, JORGE FERNANDEZ, ANA MARIA SANDINO, AND EUGENIO SPENCER y Tecnologla Received
April
de 10s Alimentos, 23,
199 1; accepted
Universidad June
de Chile,
Macul5540,
Santiago
11, Chile
19, 199 1
The effects of several nucleoside and nucleoside triphosphate analogues were studied on rotavirus replication and transcription. Nucleoside triphosphate analogues modified at sugar residues were capable of inhibiting in vitro transcription, including adenosine-9-P-o-arabinofuranoside 5’-triphosphate, 3’-deoxyadenosine 5’-triphosphate, adenosine 5’-triphosphate 2’,3’-dialdehyde, guanosine 5’-triphosphate 2’,3’-dialdehyde, and cytosine-9-fl-o-arabinofuranoside 5’triphosphate. Two dialdehyde derivatives, adenosine 5’-triphosphate 2’3’ dialdehyde and guanosine 5’-triphosphate 2’,3’-dialdehyde, were irreversible inhibitors, forming a stable complex with the viral polypeptide VP3. The effect of the corresponding nucleosides of the inhibitory analogues was studied in SA-11 rotavirus-infected MA-104 cells. Adenosine-9-P-o-arabinofuranoside and 3’-deoxyadenosine were effective inhibitors of RNA synthesis, an effect that could be due to their inhibition of viral transcription. o 1991 Academic Press. tnc.
Rotavirus is the single most important cause of acute gastroenteritis in infants and young children in developed countries (3). The viral particle is made of a double icosahedral capsid surrounding the viral core containing the viral genome (1). The outer capsid is made of two polypeptides named VP4 (88 kDa) and VP7 (38 kDa) which play an important role in cell adsorption and penetration (2). The inner capsid is made up of the most abundant viral polypeptide, VP6 (42 kDa) (9). The viral core is composed of three polypeptides, VP1 (128 kDa), VP2 (95 kDa), and VP3 (88 kDa). The viral genome consists of 1 1 double-stranded RNA segments which code for at least 10 polypeptides, 6 of which are structural (I). The plus strand of each viral RNA is capped at the 5’ end and is identical to viral mRNA (8). Like other members of the Reoviridae family, rotaviruses RNA synthesis is catalyzed by a RNA polymerase activity associated with the polypeptide VP1 , which transcribes mRNA from each of the 11 genomic segments (24). The viral polymerase is activated in vitro after subjecting the particle to a thermal shock or EDTA treatment (22). Rotavirus transcription involves other associated enzymatic activities present in the viral core, in addition to the RNA polymerase activity (15). These include the inner viral capsid which is an absolute requirement for transcription even though it does not have any apparent enzymatic activity associated for RNA synthesis (78). The 5’ cap addition to each of the viral transcripts is catalyzed by a guanylyltransferase activity associated with VP3, present in the viral core ( 15). The formation of the 5’ end cap is important because it commits transcription toward the elon0042.6822/91
$3.00
CopyrIght 0 199 1 by Academic Press. Inc. All rights of reproduction in any form reserved.
gation of the RNA chain causing a decrease in K, values for the ribonucleoside triphosphates (23). As in several eukaryotic and viral systems including rotavirus, ATP seems to play an important role in transcription, coupling the ATP hydrolysis to transcription and as a requirement for the polymerization itself (13, 16, 23). Inhibition of ATP hydrolysis during in vitro transcription abolishes RNA synthesis (23). This may be due to an enzymatic activity which uses ATP as substrate for hydrolysis, an activity which is different from the polymerase activity itself (6, 10). Based on the nucleotide requirement for all the viral enzymes involved in transcription, the effect of several nucleotide analogues including some described as antiviral agents was studied on in vitro and in vivo rotavirus RNA synthesis. We focused our interest on the effect of ATP analogues due to the important role that ATP plays during transcription. To determine the inhibitory effect of a particular analogue on the in vitro transcription, the analogue was added to a standard transcription mixture, in the presence of heat-activated double-shelled rotavirus particles and incubated for 30 min at 45°C. The incorporation of [3H]UMP into acid-insoluble material was determined as described previously (22). Each analogue was tested using concentrations ranging from 1 to 10 m/l/l to overcome any effect derived from the K,,, of the polymerase for the particular analogue. A nucleotide analogue was only considered an inhibitor if, with a concentration of 4 mll/l or less, 50°b inhibition was observed. This concentration was selected because it corresponds to the optimum concentration of each nucleoside triphosphate for transcription in the in 768
SHORT
1..
I 1236.56769
769
COMMUNICATIONS
1234567 imMl analog
FIG. 1. Inhibition of in vitro rotavirus transcription by nucleotide analogues. The results are expressed as percentage from the control reaction done in the absence of the analogue. The analogue was added to a standard mixture at the indicated concentrations and [3H]UMP incorporation was determined as radioactivity associated to acid-insoluble material as described previously (22). (0) Ara ATP, (0) Ara CTP, (X) 3’-dATP, (m) 2’,3’-dialdehyde ATP, (0) 2’,3’-dialdehyde GTP.
vitro assay. All the analogues tested were purchased from Sigma Chemical Co. The nucleotide analogues tested were grouped according to their modifications in phosphate bonds, base, sugar, and base-sugar residues. The most important analogues inhibiting transcription were adenosine 5’-triphosphate 2’,3’-dialdehyde (2’,3’-dialdehyde ATP) and guanosine 5’-triphosphate 2’,3’-dialdehyde (2’,3’-dialdehyde GTP), producing 81.6 and 76.7% inhibition at 4 mM, respectively, and approximately 50% inhibition at 1 mM(Fig. 1 b). They are acyclic analogues which have in the sugar residue the 2’,3’ bond oxidized to aldehyde and do not contain available OH groups for the formation of the phosphodiester bond on the growing RNA chain. 3’-Deoxyadenosine 5’-triphosphate such as 3’-dATP, an analogue which has modified the 3’-OH in the sugar residue, gave 75.9% inhibition at 4 mM (Fig. 1 c). This analogue has been described as an inhibitor of poly(A)synthesis acting as a chain terminator at the 3’ end of the chain (6, 11, 17). In this case it may be inhibiting the enzymatic system associated with viral transcription, because the viral transcripts do not contain poly(A) sequences. Adenosine-9-P-o-arabinofuranoside 5’-triphosphate (Ara ATP), a powerful inhibitor of several steps of the replication and transcription of some DNA and RNA viruses (5, 10, 12, 20) and cytosine-9-@-o-arabinofuranoside 5’-triphosphate (Ara CTP), were able to inhibit transcription 74.4 and 70.0%, respectively, at 4 mM (Fig. 1 a). Other ATP analogues which contain a modification in the purine ring (such as adenosine 5’-triphosphate-N-oxide, xanthosine 5’-triphosphate, 8-bromoadenosine 5’-triphosphate, and 1 ,/V6-ethenoadenosine 5’-triphosphate) or in the sugar moiety (such as 2’-dATP, 2’-dCTP, 2’dGTP, 2’-dlTP, and 2’-3’-ddTTP) or in both sugar and
base (such as 2’-deoxyinosine 5’-triphosphate) were not considered for further study because they inhibited in vitro transcription less than 50% at 4 mM. Some of them at higher concentrations inhibited RNA synthesis but this effect may correspond to what is seen with an excess of the regular nucleotide substrates at similar concentrations. To characterize the nature of the inhibition produced by the analogues defined as inhibitors an experiment was designed as follows: activated virus was incubated for 30 min at 45°C in a standard mixture of transcription including the indicated analogue at 7 mM, a concentration ensuring the maximal inhibitory effect. The virus was then separated by centrifugation and RNA synthesis was measured in the supernatant, as previously described (22). The pelleted virus was washed and resuspended with 1 vol of 50 mM TrisHCI buffer, pH 7.8, and allowed to transcribe by addition of a new transcription mixture but in the absence of the analogue. After incubation for 30 min at 45”C, RNA synthesis was measured again. Table 1 shows the percentage of transcriptional activity detected before and after the above treatment. The results show that for all the analogues tested, with the exception of 2’,3’-dialdehyde ATP and 2’,3’-dialdehyde GTP, there was an inhibitory effect which could be overcome by washing out the analogue (Table 1). The inhibition produced by the two dialdehyde nucleotides could not be reversed by washing, suggesting that they may irreversibly modify a viral polypeptide required for viral transcription. Nucleotide analogues possessing 2’,3’dialdehyde groups have been reported to react through the formation of covalent bonds with protein residues involved in nucleotide binding, by forming Schiff bases between the aldehyde groups and amino acid residues of the protein (13, 16). To identify the rotavirus polypeptide responsible for
TABLE RECOVERY
OF THE TRANSCRIPTIONAL
1 ACTIVITV
OF THE PREINCUBATED
VIRUSINTHEPRESENCEOFTHEINHIBITORANALOGUEAFTERVIRALPURIFICATION
Inhibitory
analogue
Ara CTP (8.64 mM) Ara ATP (7.00 mM) 3’-dATP (7.02 mn/l) 2’,3’-dialdehyde ATP (7.00 2’,3’dialdehyde GTP (7.00
% Inhibition
rnk!) mM)
93.4 92.2 92.6 93.4 82.5
% Recovery” 83.7 87.7 59.7 26.4 26.0
a The percentage of recovery was corrected for a factor involving the percentage of recovery of the transcriptional activity of virus subjected to the same treatment but in the absence of the analogue.
770
FIG. 2. Fluorogram viral polypeptidei.
SHORT
showing
the binding
COMMUNICATIONS
of [(r-32P]-2’,3’-dialdehyde
the binding of 2’,3’-dialdehyde ATP, purified viral particles were incubated in the presence of 50 mM TrisHCI, pH 7.9, 20 mM MgCI,, and increasing amounts of [a-32P]-2’,3’-dialdehyde ATP (0.5 to 2 &i) prepared from [(Y-~*P]ATP (10 $Xml) (4). The reaction mixture was incubated at 37°C for 20 min and then stopped by the addition of 10 mM EDTA. Viral polypeptides were then subjected to SDS-PAGE and processed for autoradiography. Under these conditions onlyoneviral polypeptide was able to form a stable complex with the analogue (Fig. 2). In comparison to the 35S-labeled rotavirus SA-11 polypeptides, it migrates with a mobility corresponding to the viral core polypeptide VP3. Similar results were obtained with 2’,3’-dialdehyde GTP, showing that it binds to VP3 identified as the rotavirus guanylyltransferase (15). This has been shown by the electrophoretic mobility of the polypeptide present in the subviral particles such as the viral core made of VP1 , VP2, and VP3. Furthermore it has been reported that the viral polypeptide VP1 is the rotavirus RNA polymerase (24). Based on the above results it may be concluded that the 2’,3’-dialdehyde ATP and 2’,3’-dialdehyde GTP bind to rotavirus VP3. To perform the in viva studies only the analogues that inhibit in vitro transcription at the levels described above were selected to study their in vivo effect on viral double-stranded RNA replication. Due to the nature of the dialdehyde analogues they were excluded from the in vivo study but others which have been described as inhibitors of viral transcription or replication were included. For this purpose MA-104 cells were infected with 1.5 PFU of trypsin-activated SA-11 rotavirus per cell as described previously (19) and incubated at 37°C with different concentrations of the nucleosidic form of
ATP to viral proteins.
The
markers
indicates
the migration
position
of the
the analogues or other known antiviral agents. No inhibition of the viral RNA replication was observed at millimolar concentrations of 5,6-dichloride-@-o-ribofuranosyl benzymidazole and the antiviral agent ribavirin, both base modified (data not shown). Similar to in vitro results the most potent inhibitors of RNA synthesis were the sugar-modified nucleotide analogues. However, analogues such as 2’,3’-dideoxytimidine and acycloguanosine were not inhibitors of transcription or replication (data not shown). The analogues 3’-deoxyadenosine (cordycepin) and adenosine-9-@-o-ribofuranoside (Ara A) were very efficient inhibitors of the in vivo viral RNA replication, at concentrations of 0.05 and 0.4 mM, respectively (data not shown). Ara CTP showed a strong in vitro inhibitory effect but it failed to inhibit in vivo transcription or replication in a very wide range of concentrations tested. This result is unexpected because this analogue has been described as an effective antineoplasic agent but it must be phosphorylated in order to become involved in nucleic acid synthesis (7, 27). The discrepancy between the in vitro and in vivo effect may be due to the permeability of the MA104 cells to the nucleoside or to the inability of the cellular kinases to phosphorylate the nucleoside sufficiently at the concentration required to exert an inhibitory effect in vivo. In order to determine the critical time point for the inhibition of viral RNA replication by cordycepin and Ara A the analogues were added at different times after infection. Under the infection conditions described above viral transcription is initiated shortly after infection and reaches a maximum after 4 hr. By 6 hr postinfection most of the single-stranded viral RNA has been synthesized and viral RNA replication has actively
SHORT a
12
3
4
56
7
8
91011
COMMUNICATIONS
dmiA -1 c2 r_L3 t5 t6
+--IO .“_ .*
.
tll
b Cl +---2 +3 +--4 t5 t6
4 t-10 tll
FIG. 3. Effect of nucleoside analogues on in viva transcription and replication. The nucleosidic form of the analogue was added to infectedcellsat0,0.5. 1.0, 1.5,2.0,3.0,4.0,and6.0hrpostinfection. and then the cells were further incubated until 7 hr postinfection. At the end of the incubation, infected cells were scrapped and phenol was extracted, processed for PAGE, and stained with silver nitrate. Control experiments were carried where the analogue was added together with the viral inoculum. After washing with fresh media the cells were incubated for 7 hr and processed as above. (a) Gel electrophoresis of a time course inhibition of synthesis of double-stranded RNA by 0.4 mll/l cordycepin and (b) 0.8 mlLl Ara A. Lanes 3 to 10 correspond to the analogue addition at times 0, 0.5, 1 .O. 1.5, 2.0, 3.0, 4.0, and 6.0 hr postinfection, respectively. Lanes 1, 2, and 1 1 correspond to the RNA extracted from the mock-infected cells, the adsorption control, and the viral control done in the absence of the analogue. Positions of the genome segments and transcripts are indicated by arrows.
taken place using viral mRNA as template (14, 19). The analogues were tested at concentrations where viral RNA replication was drastically inhibited without any detectable cytotoxic effect after 12 hr of incubation, i.e., 0.4 mMcordycepin and 0.8 mMAra A. The experiment was designed to determine if the effect of the analogues was during an early event like transcription
771
or during viral replication itself. The effect of the addition of 0.4 mM cordycepin on genomic RNA synthesis at different times postinfection is shown in Fig. 3a. Double-stranded RNA synthesis was not detected at 8 hr postinfection when the analogue was added during the first 3 hr of infection (Fig. 3a, lane 8). However, when the analogue was added 3 to 4 hr postinfection no difference was observed in double-stranded RNA synthesis when compared with the control done in absence of the analogue (Fig. 3a, lane 11). Thus under experimental conditions, inhibition of viral RNA replication is produced at a point where most of the singlestranded RNA is being synthesized and before viral RNA replication takes place (19). Similar results were obtained with 0.8 mMAra A which is an efficient inhibitor of the synthesis of double-stranded RNA when added 1 and 5 hr postinfection (Fig. 3b, lanes 3-6). Also at later times of analogue addition, no effect on RNA replication was observed (Fig. 3b, lanes 7-10). Controls done by adding either analogue in the viral adsorption mixture during the course of an hour are shown in Fig. 3a and 3b, lane 2. These controls were done to demonstrate that the analogues did not have any effect on viral infection, since the genomic RNA synthesis is not altered if compared with the viral control done in absence of the analogue. These results indicate that the inhibition of the double-stranded RNA synthesis may be due to the inhibition of the singlestranded RNA synthesis rather than a direct effect on replication itself. This is supported by the fact that 2 hr postinfection the first round of mRNA synthesis is taking place and little or no double-stranded RNA synthesis is detected (19). This can also be demonstrated by the observation that when the analogue was added at later times, i.e., 3.0, 4.0, and 6.0 hr postinfection, double-stranded RNA was detected (Fig. 3a, lanes 8-10 for 0.4 mM cordycepin and Fig. 3b, lanes 8- 10 for 0.8 mMAra A). At these later times of addition the levels of RNA synthesis were similar to those obtained in experiments performed in absence of the analogue (Figs. 3a and 3b, lane 11). Since viral mRNA plays two roles, one for polypeptide synthesis and the other as a template for RNA replication, the early inhibitory effect may be due to the inhibition of mRNA synthesis necessary for viral polypeptide synthesis which is required for replication or may be due to a direct effect on viral translation which is actively taking place at 3 hr postinfection. This later possibility is highly unlikely due to the nature of the analogues. Rotavirus mRNA is also required as a template for double-stranded RNA synthesis, together with some nonstructural polypeptides (14). These results may explain why the analogues inhibit viral RNA replication only when mRNA synthesis is inhibited. This also
772
’
SHORT
COMMUNICATIONS
suggests that another viral polypeptide, VP3, in addition to VPl, which has been described as the RNA polymerase, is necessary for rotavirus transcription. Polypeptide VP3 may be part of a multienzymatic complex present in the core of the single-shelled virus responsible of viral transcription (15). These results may be useful for future studies of viral replication and transcription in viva and to approach the development of efficient antiviral agents in the treatment of rotavirus acute gastroenteritis. ACKNOWLEDGMENTS The present research was supported by a Grant from Fondo de lnvestigacibn Cientffica y Tecnol6gica (FONDECYT) No. 1017 and from a SAREC (Sweden) grant to E.S. A. M. Sandino is a recipient of a Fundaci6n Andes fellowship. The authors thank Dr. Clinton Chichester for his revision of the manuscript and Mr. Patricia Ortiz for his technical assistance.
REFERENCES 1. ALTENBURG, B., GRAHAM, D., and ESTES, M. /. Gen. viral. 46, 75-85 (1980). 2. CLARK, S., SPENDLOVE, R., and BARNEIT, B. 1. Wrol. 34, 272-276 (1980). 3. CUKOR. G., and BLACKLOW, N. Microbial. Rev. 48, 157-179 (1984). 4. WANG, G., WANG, S., and PAN, F. Biochem. J. 199, 281-287 (1981). 5. DE CLERQ, E. Biochem. fharmacol. 36, 2567-2575 (1987).
6. GUMPORT, R., EDELHEIT, E., UEMATSU, T., and SUHADOLNIK, R. Biochemistry 15, 2804-2809 (1976). 7. HARRIS, A., POTTER, C., BUNCH, C., BOUTAGY, J., HARVEY, D., and GRAHAME-SMITH, D. Br. J. Pharmacol. 8, 219-227 (1979). 8. IMAI, M.. AKATANI, K., IKEGAMI, N., and FURUICHI, Y. J. l&o/. 47, 125-136 (1983). 125,194-205 9. KALIKA, A., FLORES, J., and GREENBERG, H. Wrology (1983). J., and CHIANG, P. 10. KIM, I., ZHANG, C., CANTONI, G., MONTGOMERY, Biochim. Biophys. Acta 829, 150-l 55 (1985). 11. KOCH, S., and NIESSING, J. EBS Lefr. 96, 354-356 (1978). 12. PARKER, WI., and CHENG, Y. Mol. Pharmacol. 31, 146-151 (1978). 13. PATEL-THOMBRE, U., and BORCHARDT, R. Biochemistry 24, 11301136 (1985). 14. PATTON. J. T. virus Res. 6, 217-233 (1986). 15. PIZ~RRO, J. L., SANDINO, A. M., PIZARRO, J. M., FERNANDEZ, J., AND SPENCER, E. J. Gen. Viral. 72, 325-332 (1991). 16. RAYFORD, R., ANTHONY, D., JR., O’NEILL, R., JR., and MERRICK. W. 1. Biol. Chem. 2260, 15708-l 5713 (1985). 17. ROSE, K., BELL, L., and JACOB, S. Nature 267, 178-l 80 (1977). 18. SANDINO, A., JASHES. M., FAUNDEZ, G., and SPENCER, E. 1. Viral. 60,797-802 (1986). 19. SANDINO, A., PIZARRO, J., FERNANDEZ, J., FELLAY, M., and SPENCER, E. Arch. Viol. Med. Exp. 21, 381-392 (1988). 20. SCHANCHE, J., SCHANCHE, T., UELAND, P., and MONTGOMERY, J. Cancer Res. 44,4297-4302 (1984). 21. SOKAL, J., LEONG, S., and GOMEZ, G. Cancer59,197-202 (1987). 22. SPENCER, E., and ARIAS, M. J. Viral. 40, l-10 (1981). 23. SPENCER, E., and GARCIA, B. 1. Viral. 52, 188-197 (1984). 24. VALENZUELA, S., PIZARRO, J., SANDINO, A., VASQUEZ, M., FERNANDEZ, J., HERNANDEZ, O., PATON, J., and SPENCER, E. /. Viral. 65, in press (1991).
