key: cord-321265-il9vbbgk
authors: DEN BOON, JOHAN A.; SPAAN, WILLY J.M.; SNIJDER, ERIC J.
title: Equine Arteritis Virus Subgenomic RNA Transcription: UV Inactivation and Translation Inhibition Studies
date: 1995-11-30
journal: Virology
DOI: 10.1006/viro.1995.0009
sha: 
doc_id: 321265
cord_uid: il9vbbgk

Abstract The expression of the genetic information of equine arteritis virus (EAV), an arterivirus, involves the synthesis of six subgenomic (sg) mRNAs. These are 5′ and 3′ coterminal since they are composed of a leader and a body sequence, which are identical to the 5′ and 3′ ends of the genome, respectively. Previously, it has been suggested thatcis-splicing of a genome-length precursor RNA is involved in their synthesis. This was reevaluated in a comparative analysis of the sg RNA synthesis of EAV, the coronavirus mouse hepatitis virus (MHV), and the alphavirus Sindbis virus. UV transcription mapping showed that the majority of the EAV sg RNAs made at later stages of infection is not derived from a genome-length precursor. However, complete independence of sg RNA synthesis from that of genomic RNA was never observed during the course of infection. The possibility that this resulted from UV irradiation-induced effects on the synthesis of the viral replicase was investigated by inhibiting translation using cycloheximide. For EAV, ongoing protein synthesis was found to be more important for the synthesis of sg RNA than for that of genomic RNA. In general, MHV transcription was extremely sensitive to translation inhibition, whereas EAV genomic RNA synthesis became independent ofde novoprotein synthesis late in infection.

As in the case of coronaviruses, the arterivirus sg RNAs are composed of sequences which are not contiguous Equine arteritis virus (EAV) is the type member of a in the genomic RNA (de Vries et al., 1990) . A common 5 recently reclassified group of enveloped positiveleader sequence of 206 nt, derived from the 5 end of stranded RNA viruses, the arteriviruses (for a review: the genome, is fused to mRNA body sequences which Plagemann and Moennig, 1992) . Other arteriviruses are are colinear with the genomic 3 end. For EAV sg RNAs lactate dehydrogenase-elevating virus of mice (Godeny 6 and 7 the so-called leader-body junction site is a penet al., 1993) , porcine reproductive and respiratory syntanucleotide sequence (5 UCAAC 3), present at the 3 drome virus (Meulenberg et al., 1993a) , and simian hemend of the leader sequence and at the 5 end of the orrhagic fever virus (Godeny et al., 1995) . mRNA body (de Vries et al., 1990) . The sequence analysis of the EAV genome previously Despite the structural similarities between arterirevealed that arteriviruses are evolutionarily related to and coronavirus sg RNAs, their modes of synthesis corona-and toroviruses (den Boon et al., 1991) . Their were reported to be different. Stern and Sefton (1982) , common ancestry is illustrated by the presence of a num-Jacobs et al. (1981) , and Yokomori et al. (1992) have ber of homologous replicase domains and striking simiused UV transcription mapping analyses to investigate larities in genome organization and expression. The EAV coronavirus sg RNA synthesis. They showed that late genome is a 12.7-kb RNA molecule which is structurally in infection the UV sensitivity of the synthesis of the polycistronic (Fig. 1) . From this RNA open reading frames sg mRNAs is not equal to that of the genomic RNA. (ORFs) 1a and 1b are translated into two large replicase Instead, UV target sizes of subgenomic transcripts polyproteins, 1a and 1ab, which are N-terminally identical were concluded to be proportional to their physical but, due to ribosomal frameshifting, C-terminally different size. In contrast, van Berlo et al. (1982) reported ap-(den Boon et al., 1991) . The more distally located ORFs proximately identical UV target sizes for the genomic 2-7, which encode four structural proteins (de Vries et RNA and all sg RNAs of the arterivirus EAV. Although al., 1992) and two proteins of unknown function, are exa slight deviation was observed for the smallest sg pressed from a 3 coterminal nested set of subgenomic RNA, currently known as RNA 7, the data suggested a (sg) RNAs (van Berlo et al., 1982; de Vries et al., 1990) .

posttranscriptional processing mechanism that generates all sg RNAs from a genome-length precursor.

nisms would be essentially different. Upon reexamination by medium containing 10 mg/ml of dactinomycin to block host RNA synthesis. After 30 -60 min, cultures of the experimental details of the previous EAV UV transcription mapping analysis (van Berlo et al., 1982) , we were incubated for 60 -90 min in medium containing 100 -500 mCi of [ 3 H]uridine and 10 mg/ml of dactino-found the applied UV doses to be remarkably low: a maximum dose of 300 ergrmm 02 was used. For compari-mycin. RNA lysates were prepared as described previously (Spaan et al., 1981) . son, the inactivation of the synthesis of sg RNA 2 of the coronavirus mouse hepatitis virus (MHV), which is only 3 kb smaller than the EAV genome, required UV doses RNA electrophoresis of up to 3000 ergrmm 02 (Jacobs et al., 1981) . In view of Denaturing RNA electrophoresis was carried out using these observations, we decided to repeat the experi-1-1.5% agarose gels containing 10 mM MOPS (morpholiments. A comparative analysis on the basis of new UV nepropanesulfonic acid) and 2.2 M formaldehyde. For transcription mapping data for EAV, MHV, and the analysis of 3 H-labeled RNA samples, agarose gels were alphavirus Sindbis virus (SIN) is described in this paper. fixed in methanol, impregnated with 3% PPO (2,5-diphe-The latter was chosen as a reference because the genyloxazole) in methanol, rinsed with water, and dried at nome sizes of EAV and SIN are similar and because the 60Њ. After fluorography, incorporation of label into individsynthesis of the sg 26S RNA of SIN is known to be fully ual RNA species was quantified by excision of bands independent from that of the genome (Brzeski and Kenfrom dried gels and scintillation counting. nedy, 1978). In the context of these experiments, we have also studied the effect of translation inhibition by cyclo-UV transcription mapping heximide on the synthesis of viral RNA.

were infected with an m.o.i. of 50. At 6 1 2 hr p.i., the me-Cells and viruses dium was removed, and cells were UV-irradiated for various intervals with a dose rate of 40 ergrs 01 rmm 02 Equine arteritis virus (Bucyrus strain) and SIN (HR strain) were grown in baby hamster kidney cells (BHK-(the total dose was between 0 and 3600 ergrmm 02 ). of RNA synthesis after t sec of irradiation, N 0 that of the dose of 2400 ergrmm 02 at 4, 5, or 6 hr p.i., and metabolically labeled RNA was subsequently analyzed as de-unirradiated control, and K is a constant (Sauerbier and Hercules, 1978) . Curves were fitted using linear regres-scribed. sion analysis.

Inhibition of protein synthesis In a separate experiment, the effect of UV irradiation at different time points in infection was investigated. EAV-Protein synthesis was inhibited by replacing the medium with medium containing 100 mg/ml of cyclohexi-infected BHK-21 cells were UV-irradiated with a single Spaan et al. (1988) , and Strauss et al. (1984) .

such as the sg RNAs found in EAV-infected cells, are derived from the same precursor molecule. If so, the sensitivity of their synthesis to UV irradiation (UV target the well-characterized alphavirus SIN, using the same size) will for all sg RNAs be identical to the target size irradiation doses and the same BHK-21 cells used for of that precursor. Alternatively, when the transcription the EAV analysis. The length of the 49S SIN genomic of sg RNAs is fully independent from that of the ge-RNA (11.7 kb) is comparable to that of EAV (12.7 kb) (Fig. nome, a UV target size which is proportional to its physical size should be measured for each sg RNA. 1). The replication of SIN involves the synthesis of a EAV-infected BHK-21 cells were UV-irradiated at 6 1 2 hr single 4.1-kb sg RNA (26S) from a well-defined internal promoter on the genome-sized minus-strand template p.i., when RNA synthesis approaches its maximum (van RNA (Ou et al., 1982; Levis et al., 1990) . As shown in Fig. Berlo et al., 1982 , and data not shown). Subsequently, viral RNA synthesis was monitored by [ 3 H]uridine incor-2 and Table 1 , the relative physical and UV target sizes poration. In parallel, 3 H-labeled RNA was isolated from of the genomic and sg SIN transcripts were in good MHV-infected cells which were UV-irradiated at 6 hr p.i. agreement, as was previously shown by Brzeski and Ken-To determine the UV inactivation kinetics for individual nedy (1978) . In addition to this difference with the corona-RNA species, 3 H-labeled RNAs were separated in denaand arterivirus systems, SIN transcription in general was turing agarose gels. Representative fluorographs for both less sensitive to UV irradiation than EAV and MHV tranviruses are shown in Fig. 2A . For both EAV and MHV, scription. we restricted our analysis to the genomic RNA and three sg RNAs (2, 6, and 7). Bands were cut from four agarose UV irradiation has differential effects early and late in gels, derived from two independent experiments. In Fig. EAV infection 2B the average percentages of remaining RNA synthesis have been plotted against the UV dose. These graphs Yokomori et al. (1992) showed that during the early clearly show that for both viruses the sg RNA synthesis stages of MHV infection the UV target sizes of sg RNAs is not directly dependent on the synthesis of a genomeapproach that of the genomic RNA. This observation length precursor RNA. On the other hand, the UV target prompted us to compare the UV target sizes of the EAV sizes of the RNAs, which are reflected by the slopes of transcripts at three different time points in infection. In the curves, are not proportional to their physical sizes Fig. 3 the reduction in the synthesis of RNAs 1, 2, 6, and either (Table 1) . Compared to the genomic RNA, the syn-7 is compared after UV irradiation with a single dose thesis of sg RNAs of both EAV and MHV is more sensitive (2400 ergrmm 02 ) at 4, 5, or 6 hr p.i. and subsequent to UV irradiation than would be expected on the basis metabolic labeling for 1 hr. Quantitative analysis on the of their lengths.

basis of two experiments showed that, similar to the situation in coronavirus-infected cells, UV sensitivities Sindbis virus UV transcription mapping during early and late transcription were different. Upon irradiation at 4 or 5 hr p.i. the relative decrease in the To investigate whether the discrepancy between relasynthesis of the genomic and sg RNAs was similar, tive physical sizes and UV target sizes of genomic and whereas after irradiation at 6 hr p.i. sg RNA synthesis sg RNAs is typical of arteri-and coronavirus transcription, a UV transcription mapping analysis was carried out for was less affected than that of RNA 1 (Fig. 3B) . 

sis. This was irrespective of the stage of infection. As previously reported by Sawicki and Sawicki (1986) , this transcription differential dependence was not observed for MHV. The UV transcription mapping data showed that the Third, the overall MHV transcription was significantly EAV sg RNAs were not produced by processing of a more dependent on de novo protein synthesis than that genome-length precursor RNA. However, unlike the case of EAV and SIN. Even during the late phase of the MHV of the SIN sg RNA, their UV target sizes were larger than infection cycle, when all infected cells were fused into expected on the basis of their physical sizes, even when one large syncytium, both genomic and sg RNA synthesis sg RNA synthesis had reached its maximum. The genowere severely impaired by the addition of cycloheximide. mic RNA encodes the viral replicase, which is involved In contrast, EAV genomic RNA synthesis was nearly in both genomic and sg RNA synthesis. Thus, its inactivatranslation-independent late in infection. In the case of tion as a transcription unit by UV irradiation could have SIN, the transcription of 49S genomic RNA was less afan important side effect: reduction of the amount of fected than that of 26S sg RNA during the earlier stages mRNA for replicase synthesis. Depending on the requireof infection, but the synthesis of both became largely ment of de novo replicase synthesis for sg RNA transcriptranslation-independent as infection progressed. Late in tion, it was feasible that the UV inactivation of genomic infection, 26S RNA transcription was even less affected RNA synthesis indirectly influenced the production of sg by translation inhibition than 49S transcription. RNAs, which would therefore never become completely independent.

DISCUSSION We investigated to what extent EAV RNA synthesis depended on continuous protein synthesis and whether

The EAV UV inactivation experiments described in this paper show that at the peak of sg RNA synthesis conven-the UV transcription mapping results could be explained by a combined effect on transcription and replicase tional cis-splicing, if at all, is not a major mechanism in the production of the EAV sg mRNAs. However, in a translation. Protein synthesis was inhibited by the addition of cycloheximide to EAV-infected BHK-21 cells at fully independent transcription system the slopes of the curves in Fig. 2B , which reflect the UV target sizes, should different time points after infection. One hour later, the effect of translation inhibition on RNA synthesis was be directly proportional to the physical sizes of the transcription units. In that case, the transcription of e.g., the measured by metabolic labeling. RNAs were isolated and analyzed by denaturing gel electrophoresis (Fig. 4) . For EAV genome (12.7 kb) would be expected to be 18 times more sensitive to UV irradiation than the synthesis of the comparison, similar analyses with MHV and SIN were carried out.

smallest sg RNA, RNA 7 (0.7 kb). It is clear that the relative RNA sizes and relative UV sensitivities (Table 1 ) On the basis of these experiments a number of conclusions could be drawn. First, ongoing protein synthesis do not support the conclusion that the transcription of the EAV genomic and sg RNAs are fully independent. was most important during the early phase of infection. Second, compared to genomic RNA synthesis, EAV sg A similar analysis of our MHV data (Table 1 ) also revealed that the transcription of the sg RNAs of coronavi-RNA synthesis was more dependent on protein synthe- ruses is not fully independent from that of genomic RNA.

been published by Jacobs et al. (1981) and Yokomori et al. (1992) . Both groups reported differences between the As for EAV, there are clear differences in UV sensitivity between larger and smaller RNAs, but the quantitative UV target sizes of the various MHV RNAs which are in the same range as our observations. For example, the MHV analysis shows differences between relative physical sizes and relative UV sensitivities. Our results are in UV target size of MHV RNA 7 calculated by Jacobs et al. (1981) is only about eight times smaller than that of geno-general agreement with those which have previously mic RNA, whereas the physical size ratio is about 1:18. Due to the fact that at that time the length of the MHV genome was assumed to be only 16 kb, instead of the currently known 31 kb, Jacobs et al. (1981) concluded that their data supported fully independent transcription of MHV genomic and sg RNAs. Yokomori et al. (1992) , who recently repeated the MHV UV transcription mapping experiments, did not describe a detailed quantitative analysis of the data they obtained late in infection, but concluded that UV sensitivities roughly paralleled mRNA sizes. However, a calculation using the graphs published by Yokomori et al. (1992) revealed only a sevenfold difference between the genomic RNA and RNA 7. Similar observations were made when we reevaluated the data obtained with another coronavirus, infectious bronchitis FIG. 5 . The synthesis of sg RNAs 2, 6, and 7, relative to that of the genomic RNA, during different time intervals throughout infection. Data virus (Stern and Sefton, 1982) . sulted in an almost perfect correlation between physical RNA sizes and UV target sizes ( Fig. 2; Table 1 ). These results demonstrate that the experimental approach was models have been put forward, which are not necessarily mutually exclusive (Jeong and Makino, 1994; Sawicki and appropriate and that our observations for EAV cannot be explained by technical imperfections. Apparently, the Sawicki, 1990; van der Most et al., 1994) . Leader-primed transcription could, for example, generate a first genera-intrinsic properties of coronavirus-like RNA transcription are different from those of alphaviruses.

tion of sg mRNAs, which could subsequently function as templates for sg minus-strand production. These could Although cis-splicing events cannot be ruled out as an early-stage mechanism to produce sg RNAs, another in turn be used to transcribe a second generation of positive-strand sg mRNAs. The recent analysis of the transcription mechanism must be operating in EAV-infected cells and is most likely similar to that of coronavi-transcriptional activity of a number of MHV ts-mutants indicates that sg minus strands are likely to be used ruses. The so-called leader-primed transcription model has been proposed to explain the synthesis of coronavi-as transcription templates late in infection (Schaad and Baric, 1994) . Like the UV transcription mapping results rus sg RNAs (Baric et al., 1983; Spaan et al., 1983 ) and a substantial amount of data in its favor has been re-described for MHV (Yokomori et al., 1992) , our data on EAV indicate that different mechanisms of sg RNA tran-ported (for a review: Spaan et al., 1988; Lai, 1990) . According to this model, sg RNAs are synthesized either scription might be operating simultaneously. The differences in UV sensitivity of sg RNA synthesis early and by an intrinsic ability of the viral leader/polymerase complex to dissociate from its template and subsequently late in infection (Fig. 3) can be interpreted to reflect the levels at which each of these mechanisms participate in resume synthesis at a more distal position or by priming by free leader transcripts at subgenomic promoter se-transcription, at a given stage in infection.

We have considered the possibility that replicase syn-quences (the complements of the so-called intergenic sequences on the positive strand). The conserved thesis was affected by UV irradiation and that genomic and sg RNA synthesis were inhibited as a result. Differen-leader-body junction sequences in EAV sg RNAs (de Vries et al., 1990; den Boon et al., manuscript in prepara- tial effects of translation inhibition early vs late in infection (Fig. 4 ) may be based on the properties as well as tion) possibly represent such complementary promoter sequences. Similar sequence elements have been de-the absolute amount of the replicase. Our results and previously published data from Sawicki and Sawicki scribed for other arteriviruses (Godeny et al., 1993; Conzelmann et al., 1993; Chen et al., 1993; Meulenberg et (1986) on MHV RNA transcription indicated that ongoing protein synthesis is an absolute requirement during the al., 1993b; Zeng et al., 1995) . A single genome-length minus-strand RNA could be the template used for the early stages of infection and that both MHV genomic and sg RNA synthesis at later stages were still significantly synthesis of genomic as well as sg mRNAs. Nevertheless, anti-leader-containing subgenomic minus strands reduced when translation was blocked. In contrast, the EAV data indicated that the synthesis of sg RNAs was in coronavirus-infected cells have been demonstrated (Hofmann et al., 1990; Sethna et al., 1989 Sethna et al., , 1991 . Their more impaired by inhibition of translation than was the case for the production of genomic RNA. In addition, the presence in double-stranded replicative intermediates suggests that these might function in mRNA transcription data presented in Figs. 3 and 4 revealed that sg RNA synthesis is down-regulated relative to the synthesis of (Sawicki and Sawicki, 1990 ). On the basis of these and other observations several coronavirus transcription genomic RNA late in infection (Fig. 5) .

Coronavirus minus-strand synthesis and effect of cycloheximide on coronavirus RNA synthesis

of replicative intermediate RNA of mouse hepatitis virus: Presence of leader RNA sequences on nascent chains

Coronavirus transcription: Subgenomic mouse hepatitis virus replicative intermediates function ruses: Structure and genome expression

Genetics of mouse hepatitis Kinetics of inactivation of infectious bronchitis virus RNA synthesis virus transcription: Evidence that subgenomic negative strands are by UV light

92-copies of replicating coronavirus mRNAs contain anti-leaders

491-562. mic minus-strand RNAs and the potential for mRNA replicons

5755-5764. mic RNA synthesis directed by a synthetic defective interfering RNA Spaan

Coronavirus mRNA 424-434. transcription: UV light transcriptional mapping studies suggest an

early requirement for a genomic-length template

sis of simian hemorrhagic fever virus (SHFV) subgenomic RNAs, junction sequences, and 5 leader

 Brzeski, H., and Kennedy, S. I. T. (1978) . Synthesis of alphavirus-speci-Whether our findings are indicative of a functional polyfied RNA. J. Virol. 25, [630] [631] [632] [633] [634] [635] [636] [637] [638] [639] [640] merase switch, from synthesizing sg RNA to producing Chen, Z., Kuo, L., Rowland, R. R. R., Even, C., Faaberg, K. S., and Plagegenomic RNA, remains to be investigated. Similar to the mann, P. G. W. (1993) . Sequence of 3 end of genome and of 5 regulatory function of nonstructural polyprotein processing end of ORF 1a of lactate dehydrogenase-elevating virus (LDV) and common junction motifs between 5 leader and bodies of seven in alphavirus replication (for a review: Strauss and Strauss, subgenomic mRNAs. J. Gen. Virol. 74, [643] [644] [645] [646] [647] [648] [649] [650] [651] [652] [653] [654] [655] [656] [657] [658] [659] [660] 1994), proteolytic processing may influence the transcrip- Conzelmann, K. K., Visser, N., Woensel, P. V., and Thiel, H. J. (1993) . An alternative explanation for the observed difference in 3339-3346. sensitivity to translation inhibition between genomic and Lai, M. M. C. (1990) . Coronavirus-Organization, replication and expression of genome. Annu. Rev. Microbiol. 44, sg RNA synthesis could be a demand for one or more Levis, R., Schlesinger, S., and Huang, H. V. (1990) . Promoter for Sindbis relatively instable host protein factors involved in sg RNA virus RNA-dependent subgenomic RNA transcription. J. Virol. 64, transcription, the synthesis of which might be affected