- Email: [email protected]
dsRNA-ended genomes in orthobunyavirus particles and infected cells
Virology 489 (2016) 192–193
Contents lists available at ScienceDirect
Virology journal homepage: www.elsevier.com/locate/yviro
Letter to the Editor
dsRNA-ended genomes in orthobunyavirus particles and infected cells
art ic l e i nf o Keywords: Bunyavirus Genome RNPs DsRNA
a b s t r a c t dsRNA-ended genome RNPs accumulate during LaCrosse bunyavirus infection. The possible significance of these dsRNA structures for orthobunyavirus replication and survival are discussed. & 2016 Published by Elsevier Inc.
The RNP genome segments of segmented negative strand RNA viruses (sNSV; Arenaviridae, Bunyaviridae, and Orthomyxoviridae), are pseudo-circular in structure when viewed stretched out in the EM, their genome termini containing RNA sequences that are more or less complementary. In the case of influenza A virus (flu), where the terminal RNA complementarity is insufficient for the RNA ends to be held together stably by base-pairing alone, the ends are held together by the viral RdRp (Baudin et al., 1994). In the case of bunyaviruses and arenaviruses, the various segments contain RNA termini that are highly complementary for 15–20nt, sufficient for the ends to be stably held together by base-pairing alone. In the case of LaCrosse orthobunyavirus (LACV), a fraction of their RNA termini within RNPs, both within infected mammalian and mosquito cells and as purified RNPs, indeed appears to be base-paired as they can be cross-linked with psoralen (Raju and Kolakofsky, 1989). Moreover, cooperation between at least a portion of the 30 and 50 -terminal sequences of Bunyamwera orthobunyavirus (nucleotides 10–15) is required for BUNV RNA synthesis (Barr and Wertz, 2004; Kohl et al., 2004), suggesting that this cooperation is due to nucleotide complementarity that allows base-pairing. Thus, this dsRNA LACV structure containing the genomic termini was thought to be part of the promoter for RNA synthesis. Gerlach et al. (2015) have recently reported the structure of the LACV genomic promoter (i.e., short RNAs representing the 50 and 30 genomic termini) bound to its RdRp, which bears important similarities to the structure of the analogous flu complex (Pflug et al., 2014; Reich et al., 2014). The LACV complex was assembled somewhat differently than that of the flu complex, in that the 30 genomic end and pol were first co-crystallized, and the 50 genomic end was then added to the complex by soaking it into the crystal. This was necessary because the LACV genome ends are so complementary that they form dsRNA when added together, and dsRNA binds poorly to the polymerase (Gerlach et al., 2015). In this structure, the entire 50 and 30 LACV RNA ends bind as ssRNA to separate regions of L. However, LACV L also binds partially annealed 30 promoter sequences with high affinity, and Gerlach et al. were able to resolve a structure in which nucleotides 9–16 were annealed to their complement. Gerlach et al. propose that, similar to flu, the distal portion of the 50 and 30 promoter http://dx.doi.org/10.1016/j.virol.2015.12.016 0042-6822/& 2016 Published by Elsevier Inc.
sequences are base-paired in the context of a LACV pre-initiation complex, in agreement with the requirements for a functional Bunyamwera virus promoter determined by Barr and Wertz (2004) and Kohl et al. (2004). Interestingly, both the flu and LACV structures are described as pre-initiation complexes, as in neither case is the 30 end bound to the RNA synthesis cavity. As mentioned above, some of the LACV genome ends have in fact also formed dsRNA in intact viruses and infected cells. Given other unusual properties of the NP/RNA interactions within orthobunyavirus RNPs, it was possible to devise a scheme where, at least theoretically, these dsRNA-ended genome segments could act as templates for viral RNA synthesis (Reguera et al., 2014). Other properties of these dsRNA-ended genomes during infection, however, were inconsistent with their having an essential role in LACV replication. For example, even though the infecting LACV stock contained S genomes of which 50% could be psoralen cross-linked, cross-linkable S genomes could not be detected early in infection when RNA synthesis was robust. The fraction of S segments that was cross-linkable, including antigenomes, increased steadily during infection, reaching 50% when CPE was severe. In addition, the M and L segments could not be cross-linked at any time to any extent during the infection; the presence of dsRNA-ended genomes was apparently restricted to the S segment. Gerlach et al. (2015) have suggested that the extensive complementarity of the LACV genome ends, due to their propensity to form stable dsRNA, was a major obstacle to reconstituting the putative LACV preinitiation complex in vitro and thereby demonstrating its capacity to synthesis RNA (in contrast to the robust RNA synthesis activity of the flu preinitiation complex (Reich et al., 2014)). In this view, bunyaviruses must avoid stable base-pairing of their highly complementary genomic promoter ends during infection; the free RNA ends would normally be prevented from meeting each other by the sequential mode of RNP assembly, starting at their 50 ends as they exit the replicating polymerase (cf Figure 7, Gerlach et al., 2015). If this view is correct, the dsRNA-ended LACV S segments we see in virus particles and infected cells would presumably represent dead-ends, incapable of acting as templates for viral RNA synthesis. These dsRNA-ended genomes would presumably form during infection when some of
Letter to the Editor / Virology 489 (2016) 192–193
the RNA ends lose contact with the polymerase and anneal to each other. A possible reason for why this annealing is restricted to S segments is that only they are small enough so that the polymerase-free genomic ends can find each other with any frequency during infection. It should be noted that the partially complementary flu genome RNA ends in flu RNPs can also be cross-linked with psoralen to 50% (or more) during infection (Hsu et al., 1987). However, in this case all the genome segments were cross-linkable, and, more importantly, this property of the flu RNPs was present throughout the infection. Thus, there is no indication that cross-linkable flu genome segments are defective. Given that free flu genomic RNAs are not cross-linkable (Hsu et al., 1987), unlike those of LACV (Patterson et al., 1983), it is presumably the presence of the flu polymerase that allows the limited base-pairing in the distal region of the promoter that confers this property. Flu and LACV are proposed to have similar pre-initiation complexes, in which the first 10 nt of the 50 and 30 ends of the genomic promoters are bound as ssRNA to different regions of the polymerase, whereas the more distal nucleotides are base-paired to each other. Nevertheless, the difference in the degree of complementarity of the genome ends of these two sNSV is striking. The flu genome ends are considerably less complementary than those of the bunyaviruses, such that experimentally they bind to their separate regions of the polymerase as ssRNA when added together during co-crystallization, rather than anneal to each other like those of LACV. Terminal genome complementarity is needed, at least in part, to simultaneously maintain genomic (or vRNA) and antigenomic (or cRNA) promoter activity, but this requirement applies to both sNSVs. The extensive bunyavirus terminal complementarity would then appear to have been selected for during evolution not simply to maintain promoter activity, but for a property that is unique to the Bunyaviridae. One such property is that while these infections can be highly cytopathic in humans, they are non-cytopathic and persistent in their natural hosts, e.g., mosquitos (Borucki et al., 2002). These species-specific differences of infection are presumably due, at least in part, to the different ways that their cells respond to the infection, and whether this response is linked to cell death. As the S segment codes for the nucleoprotein that is required in stoichiometric amounts for genome replication, it is possible that the steady accumulation of
193
defective dsRNA-ended LACV S genome segments during infection helps to attenuate the infection and to establish the persistent infection that ensures virus survival.
References Barr, J.N., Wertz, G.W., 2004. Bunyamwera bunyavirus RNA synthesis requires cooperation of 30 - and 50 -terminal sequences. J. Virol. 78, 1129–1138. Baudin, F., Bach, C., Cusack, S., Ruigrok, R.W., 1994. Structure of influenza virus RNP. I. Influenza virus nucleoprotein melts secondary structure in panhandle RNA and exposes the bases to the solvent. EMBO J. 13, 3158–3165. Borucki, M.K., Kempf, B.J., Blitvich, B.J., Blair, C.D., Beaty, B.J., 2002. La Crosse virus: replication in vertebrate and invertebrate hosts. Microbes Infect. 4 (3), 341–350. Gerlach, P., Malet, H., Cusack, S., Reguera, J., 2015. Structural insights into bunyavirus replication and its regulation by the vRNA promoter. Cell 161 (6), 1267–1279. Hsu, M.T.1, Parvin, J.D., Gupta, S., Krystal, M., Palese, P., 1987. Genomic RNAs of influenza viruses are held in a circular conformation in virions and in infected cells by a terminal panhandle. Proc. Natl. Acad. Sci. USA 84 (22), 8140–8144. Kohl, A., Dunn, E.F., Lowen, A.C., Elliott, R.M., 2004. Complementarity, sequence and structural elements within the 3' and 5' non-coding regions of the Bunyamwera orthobunyavirus S segment determine promoter strength. J. Gen. Virol. 85 (Pt 11), 3269–3278. Patterson, J.L., Kolakofsky, D., Holloway, B.P., Obijeski, J.F., 1983. Isolation of the ends of La Crosse virus small RNA as a double-stranded structure. J. Virol. 45 (2), 882–884. Pflug, A., Guilligay, D., Reich, S., Cusack, S., 2014. Structure of influenza A polymerase bound to the viral RNA promoter. Nature 516 (7531), 355–360. Raju, R., Kolakofsky, D., 1989. The ends of La Crosse virus genome and antigenome RNAs within nucleocapsids are base paired. J. Virol. 63, 122–128. Reguera, J., Cusack, S., Kolakofsky, D., 2014. Segmented negative strand RNA virus nucleoprotein structure. Curr. Opin. Virol. 5, 7–15. Reich, S., Guilligay, D., Pflug, A., Malet, H., Berger, I., Crépin, T., Hart, D., Lunardi, T., Nanao, M., Ruigrok, R.W., Cusack, S., 2014. Structural insight into cap-snatching and RNA synthesis by influenza polymerase. Nature 516 (7531), 361–366.
Daniel Kolakofsky Department of Microbiology and Molecular Medicine, University of Geneva School of Medicine, CMU, 9 Avenue de Champel, 1211 Geneva, Switzerland E-mail address: [email protected] Received 5 November 2015 19 December 2015 23 December 2015
Contents lists available at ScienceDirect
Virology journal homepage: www.elsevier.com/locate/yviro
Letter to the Editor
dsRNA-ended genomes in orthobunyavirus particles and infected cells
art ic l e i nf o Keywords: Bunyavirus Genome RNPs DsRNA
a b s t r a c t dsRNA-ended genome RNPs accumulate during LaCrosse bunyavirus infection. The possible significance of these dsRNA structures for orthobunyavirus replication and survival are discussed. & 2016 Published by Elsevier Inc.
The RNP genome segments of segmented negative strand RNA viruses (sNSV; Arenaviridae, Bunyaviridae, and Orthomyxoviridae), are pseudo-circular in structure when viewed stretched out in the EM, their genome termini containing RNA sequences that are more or less complementary. In the case of influenza A virus (flu), where the terminal RNA complementarity is insufficient for the RNA ends to be held together stably by base-pairing alone, the ends are held together by the viral RdRp (Baudin et al., 1994). In the case of bunyaviruses and arenaviruses, the various segments contain RNA termini that are highly complementary for 15–20nt, sufficient for the ends to be stably held together by base-pairing alone. In the case of LaCrosse orthobunyavirus (LACV), a fraction of their RNA termini within RNPs, both within infected mammalian and mosquito cells and as purified RNPs, indeed appears to be base-paired as they can be cross-linked with psoralen (Raju and Kolakofsky, 1989). Moreover, cooperation between at least a portion of the 30 and 50 -terminal sequences of Bunyamwera orthobunyavirus (nucleotides 10–15) is required for BUNV RNA synthesis (Barr and Wertz, 2004; Kohl et al., 2004), suggesting that this cooperation is due to nucleotide complementarity that allows base-pairing. Thus, this dsRNA LACV structure containing the genomic termini was thought to be part of the promoter for RNA synthesis. Gerlach et al. (2015) have recently reported the structure of the LACV genomic promoter (i.e., short RNAs representing the 50 and 30 genomic termini) bound to its RdRp, which bears important similarities to the structure of the analogous flu complex (Pflug et al., 2014; Reich et al., 2014). The LACV complex was assembled somewhat differently than that of the flu complex, in that the 30 genomic end and pol were first co-crystallized, and the 50 genomic end was then added to the complex by soaking it into the crystal. This was necessary because the LACV genome ends are so complementary that they form dsRNA when added together, and dsRNA binds poorly to the polymerase (Gerlach et al., 2015). In this structure, the entire 50 and 30 LACV RNA ends bind as ssRNA to separate regions of L. However, LACV L also binds partially annealed 30 promoter sequences with high affinity, and Gerlach et al. were able to resolve a structure in which nucleotides 9–16 were annealed to their complement. Gerlach et al. propose that, similar to flu, the distal portion of the 50 and 30 promoter http://dx.doi.org/10.1016/j.virol.2015.12.016 0042-6822/& 2016 Published by Elsevier Inc.
sequences are base-paired in the context of a LACV pre-initiation complex, in agreement with the requirements for a functional Bunyamwera virus promoter determined by Barr and Wertz (2004) and Kohl et al. (2004). Interestingly, both the flu and LACV structures are described as pre-initiation complexes, as in neither case is the 30 end bound to the RNA synthesis cavity. As mentioned above, some of the LACV genome ends have in fact also formed dsRNA in intact viruses and infected cells. Given other unusual properties of the NP/RNA interactions within orthobunyavirus RNPs, it was possible to devise a scheme where, at least theoretically, these dsRNA-ended genome segments could act as templates for viral RNA synthesis (Reguera et al., 2014). Other properties of these dsRNA-ended genomes during infection, however, were inconsistent with their having an essential role in LACV replication. For example, even though the infecting LACV stock contained S genomes of which 50% could be psoralen cross-linked, cross-linkable S genomes could not be detected early in infection when RNA synthesis was robust. The fraction of S segments that was cross-linkable, including antigenomes, increased steadily during infection, reaching 50% when CPE was severe. In addition, the M and L segments could not be cross-linked at any time to any extent during the infection; the presence of dsRNA-ended genomes was apparently restricted to the S segment. Gerlach et al. (2015) have suggested that the extensive complementarity of the LACV genome ends, due to their propensity to form stable dsRNA, was a major obstacle to reconstituting the putative LACV preinitiation complex in vitro and thereby demonstrating its capacity to synthesis RNA (in contrast to the robust RNA synthesis activity of the flu preinitiation complex (Reich et al., 2014)). In this view, bunyaviruses must avoid stable base-pairing of their highly complementary genomic promoter ends during infection; the free RNA ends would normally be prevented from meeting each other by the sequential mode of RNP assembly, starting at their 50 ends as they exit the replicating polymerase (cf Figure 7, Gerlach et al., 2015). If this view is correct, the dsRNA-ended LACV S segments we see in virus particles and infected cells would presumably represent dead-ends, incapable of acting as templates for viral RNA synthesis. These dsRNA-ended genomes would presumably form during infection when some of
Letter to the Editor / Virology 489 (2016) 192–193
the RNA ends lose contact with the polymerase and anneal to each other. A possible reason for why this annealing is restricted to S segments is that only they are small enough so that the polymerase-free genomic ends can find each other with any frequency during infection. It should be noted that the partially complementary flu genome RNA ends in flu RNPs can also be cross-linked with psoralen to 50% (or more) during infection (Hsu et al., 1987). However, in this case all the genome segments were cross-linkable, and, more importantly, this property of the flu RNPs was present throughout the infection. Thus, there is no indication that cross-linkable flu genome segments are defective. Given that free flu genomic RNAs are not cross-linkable (Hsu et al., 1987), unlike those of LACV (Patterson et al., 1983), it is presumably the presence of the flu polymerase that allows the limited base-pairing in the distal region of the promoter that confers this property. Flu and LACV are proposed to have similar pre-initiation complexes, in which the first 10 nt of the 50 and 30 ends of the genomic promoters are bound as ssRNA to different regions of the polymerase, whereas the more distal nucleotides are base-paired to each other. Nevertheless, the difference in the degree of complementarity of the genome ends of these two sNSV is striking. The flu genome ends are considerably less complementary than those of the bunyaviruses, such that experimentally they bind to their separate regions of the polymerase as ssRNA when added together during co-crystallization, rather than anneal to each other like those of LACV. Terminal genome complementarity is needed, at least in part, to simultaneously maintain genomic (or vRNA) and antigenomic (or cRNA) promoter activity, but this requirement applies to both sNSVs. The extensive bunyavirus terminal complementarity would then appear to have been selected for during evolution not simply to maintain promoter activity, but for a property that is unique to the Bunyaviridae. One such property is that while these infections can be highly cytopathic in humans, they are non-cytopathic and persistent in their natural hosts, e.g., mosquitos (Borucki et al., 2002). These species-specific differences of infection are presumably due, at least in part, to the different ways that their cells respond to the infection, and whether this response is linked to cell death. As the S segment codes for the nucleoprotein that is required in stoichiometric amounts for genome replication, it is possible that the steady accumulation of
193
defective dsRNA-ended LACV S genome segments during infection helps to attenuate the infection and to establish the persistent infection that ensures virus survival.
References Barr, J.N., Wertz, G.W., 2004. Bunyamwera bunyavirus RNA synthesis requires cooperation of 30 - and 50 -terminal sequences. J. Virol. 78, 1129–1138. Baudin, F., Bach, C., Cusack, S., Ruigrok, R.W., 1994. Structure of influenza virus RNP. I. Influenza virus nucleoprotein melts secondary structure in panhandle RNA and exposes the bases to the solvent. EMBO J. 13, 3158–3165. Borucki, M.K., Kempf, B.J., Blitvich, B.J., Blair, C.D., Beaty, B.J., 2002. La Crosse virus: replication in vertebrate and invertebrate hosts. Microbes Infect. 4 (3), 341–350. Gerlach, P., Malet, H., Cusack, S., Reguera, J., 2015. Structural insights into bunyavirus replication and its regulation by the vRNA promoter. Cell 161 (6), 1267–1279. Hsu, M.T.1, Parvin, J.D., Gupta, S., Krystal, M., Palese, P., 1987. Genomic RNAs of influenza viruses are held in a circular conformation in virions and in infected cells by a terminal panhandle. Proc. Natl. Acad. Sci. USA 84 (22), 8140–8144. Kohl, A., Dunn, E.F., Lowen, A.C., Elliott, R.M., 2004. Complementarity, sequence and structural elements within the 3' and 5' non-coding regions of the Bunyamwera orthobunyavirus S segment determine promoter strength. J. Gen. Virol. 85 (Pt 11), 3269–3278. Patterson, J.L., Kolakofsky, D., Holloway, B.P., Obijeski, J.F., 1983. Isolation of the ends of La Crosse virus small RNA as a double-stranded structure. J. Virol. 45 (2), 882–884. Pflug, A., Guilligay, D., Reich, S., Cusack, S., 2014. Structure of influenza A polymerase bound to the viral RNA promoter. Nature 516 (7531), 355–360. Raju, R., Kolakofsky, D., 1989. The ends of La Crosse virus genome and antigenome RNAs within nucleocapsids are base paired. J. Virol. 63, 122–128. Reguera, J., Cusack, S., Kolakofsky, D., 2014. Segmented negative strand RNA virus nucleoprotein structure. Curr. Opin. Virol. 5, 7–15. Reich, S., Guilligay, D., Pflug, A., Malet, H., Berger, I., Crépin, T., Hart, D., Lunardi, T., Nanao, M., Ruigrok, R.W., Cusack, S., 2014. Structural insight into cap-snatching and RNA synthesis by influenza polymerase. Nature 516 (7531), 361–366.
Daniel Kolakofsky Department of Microbiology and Molecular Medicine, University of Geneva School of Medicine, CMU, 9 Avenue de Champel, 1211 Geneva, Switzerland E-mail address: [email protected] Received 5 November 2015 19 December 2015 23 December 2015