Viruses in the Orthobunyavirus genus have smaller S segments than those of viruses in other genera (Table 42.1). Some orthobunyaviruses encode two polypeptides, the nucleocapsid protein (N) and a nonstructural protein (NSs), in ORFs in cRNA⁴⁸ (Fig. 42.2), whereas others only encode N protein.³⁵¹ The presence of NSs in virus-infected cells was demonstrated for several members of the genus.¹⁴⁰ Only one S segment mRNA species can be found in orthobunyavirus-infected cells⁸⁴; therefore, N and NSs are generated by alternative start codon recognition by ribosomes.¹²⁶,¹⁴⁴

Hantaviruses and nairoviruses encode larger N proteins than viruses in other genera, (Table 42.2, Fig. 42.2).³³²,⁴⁷⁰ Some hantaviruses (e.g., SNV, PUUV, Tula [TULV] and Prospect Hill [PHV] viruses), have ORFs within the N ORF, and an NSs protein has been detected in PUUV- and TULV-infected cells.²²⁷ No NSs protein has been detected in nairovirus infected cells. Only one S-segment mRNA, similar in size to the coding region for N, was identified in hantavirus- or nairovirus-infected cells, indicating that transcription termination occurs shortly after the translation stop codon.⁴⁶⁹,⁵⁶³ Certain hantaviruses (e.g., SNV) have long 3′ noncoding regions (of greater than 700 nucleotides) containing numerous repeated sequences. These repeats may result from polymerase slippage on the vRNA template.⁵⁰⁵

The ambisense coding strategy of the S segments of phleboviruses and tospoviruses produces N from a subgenomic mRNA that is complementary to vRNA, and NSs from a subgenomic mRNA of the same polarity as vRNA (Fig. 42.2).¹¹⁸,²²⁰ Evidence that the mRNA for NSs is copied from cRNA (after genome replication) comes from time-course studies. For the phlebovirus UUKV, N was detected at 4 to 6 hours after infection, whereas NSs did not appear until 8 hours after infection.⁴⁹²,⁵⁴² Likewise, the mRNA for NSs of the tospovirus TSWV was detected in infected plant cells 15 hours later than for N.⁵⁰⁷ In addition, studies with the phlebovirus Punta Toro virus (PTV) demonstrated that protein synthesis inhibitors arrest production of NSs mRNA, but not N mRNA.²²² These results suggest that protein synthesis must occur before the NSs mRNA can be made. In contrast, studies with the phlebovirus RVFV demonstrated the presence of cRNAs to all three RNA segments in purified virions. Moreover, mRNA for the NSs protein was detected as early as 20 minutes after infection, concomitant with the appearance of mRNA for N, suggesting that the ambisense-encoded NSs mRNA are transcribed from the incoming antiviral sense RNA.²²⁵

Sizes of M segments range from ∼3,600 nucleotides to ∼4,900 nucleotides (Table 42.1). All bunyavirus M segments encode the two envelope GPs in a single ORF of cRNA (Fig. 42.2). Previous designations of G1 and G2, which were based on relative migration of the proteins in polyacrylamide gels, have been replaced by designations of Gn and Gc, referring to the amino-terminal or carboxy-terminal coding of the proteins.²⁹⁴ As described below, it is becoming increasingly clearer that the functions of the Gn and Gc proteins are conserved among the five genera.

Some viruses encode NSm proteins, and others do not. Except for the tospoviruses, which use an ambisense strategy to generate NSm from a subgenomic mRNA, a single mRNA, nearly equivalent in size to the cRNA ORF has been detected in virus-infected cells. For tospoviruses, separate, subgenomic messages for the Gn-Gc precursor and for NSm were identified.²⁷³,²⁹⁵ NSm is readily detected in infected plants and is the only M segment nonstructural protein in the Bunyaviridae family to have a clearly defined role (i.e., it is a movement protein [see “M Segment Products,” below]). Another possible role for NSm for some viruses may be as a virulence factor. For example, a genetically engineered RVFV lacking NSm induced more extensive apoptosis than did one with NSm and the expression of NSm significantly inhibited the cleavage of caspase 8 and 9 induced by staurosporine, indicating that the NSm protein suppresses apoptosis.⁵⁸²

The L segments of hantaviruses, orthobunyaviruses, and phleboviruses are of similar size, (∼6,500 nucleotides), whereas those of tospoviruses and nairoviruses are considerably larger (∼9,000 and 12,000 nucleotides, respectively; Table 42.1). All L segments of viruses in the family use conventional negative-sense coding strategies (Fig. 42.2). For each L segment described thus far, there are fewer than 200 nucleotides of total noncoding information, and there is no evidence for additional coding regions in either the cRNA or vRNA.¹¹⁷,¹⁴¹,¹⁴³,²⁰⁹,³⁶⁰,⁴⁶⁵

The mechanisms by which members of the family Bunyaviridae gain access to the host cell’s cytoplasm appear similar to those reported for many other enveloped viruses. The first step involves an interaction between cell surface receptors and viral attachment proteins, Gn and/or Gc. The presence of neutralizing and hemagglutination-inhibiting sites on both the Gn and Gc proteins of phleboviruses and hantaviruses²⁷,²⁵² suggests that both proteins may be involved in attachment. Although it is possible that both are directly involved, it is more likely that both are required due to conformational requirements that depend on dimerization of Gn and Gc.

In general, Gc appears to be the primary attachment protein for orthobunyaviruses in mammalian cells, mosquito cells, and mosquitoes.¹⁹³,²⁶¹,⁵¹⁵ Consistent with this view, expression of M-segment products of three orthobunyaviruses demonstrated that Gc, but not Gn, could effect attachment and entry when tested in a cell-to-cell fusion assay and a pseudotype transduction assay.⁴¹⁹ However, the amino terminal half of BUNV Gc ectodomain is not required for infection of cultured mammalian cells,⁴⁸³ and this region can be replaced by green fluorescent protein (GFP) allowing the creation of viable viruses expressing chimeric Gc-GFP in their virions.⁴⁸⁶

For tospoviruses, the envelope GPs are required only for infection of their arthropod vectors, as evidenced by the ability of envelope-deficient mutants to replicate in plants after
mechanical transmission, but their inability to infect thrips.⁴⁴⁵ In addition, analysis of reassortant viruses demonstrated that mutations that disrupted the Gn-Gc ORF of TSWV ablated transmissibility by thrips.⁴⁹⁵ These studies provide indirect evidence that the L-polymerase protein remains associated with ribonucleocapsids and is active despite the absence of intact virions.

Although there are no specific reports of Gc being the attachment protein of tospoviruses, functional homology to the Gc proteins of other bunyaviruses was proposed based on amino acid sequence homologies detected for iris yellow spot virus.¹⁰⁶ In contrast, however, a study with a soluble, truncated form of TSWV Gn expressed in baculoviruses showed binding of this protein to epithelial cells in the midguts of thrips, suggesting that Gn could mediate attachment in thrips.⁵⁷³ Similar results were reported earlier with the orthobunyavirus LACV, in that virions treated with a protease that cleaved Gc but not Gn exhibited increased binding to the insect vector midgut, but reduced binding to cultured mammalian or mosquito cells.³²⁴ Given that both of these systems use a Gn protein in the absence of Gc, it remains to be determined whether Gn functions for attachment when in its native conformation on virions.

Host cell receptors have not been identified for most viruses in the family; however, pathogenic and nonpathogenic hantaviruses were shown to use β3 and β1 integrins, respectively, to enter endothelial cells.¹⁷⁰ This finding was suggested to relate to pathogenesis, in that binding of hantaviruses to integrins might disrupt their ability to regulate cell-to-cell adhesion and result in the vascular permeability characteristic of hantaviral diseases.¹⁷¹ In addition, decay accelerating factor (DAF)/CD55 a glycosylphosphatidylinositol (GPI)-anchored protein, has been shown to mediate HTNV and PUUV entry across the apical membrane of polarized epithelial cells,²⁷⁷ and gC1qR/p32, a 32-kD glycoprotein that interacts with complement protein C1q, binds HTNV and mediates infection of cultured A549 lung cells.⁹⁸

Shortly after attachment, viruses in the Phlebovirus and Nairovirus genera were observed in phagocytic vacuoles.¹⁴⁵,⁴⁵⁵ This suggested a mode of viral entry similar to that first described for alphaviruses in which the virus is endocytosed in coated vesicles, and inhibitor studies have confirmed this to be so for the animal-infecting bunyaviruses. Orthobunyaviruses, hantaviruses, and nairoviruses enter by endocytosis into clathrin-coated pits,²³⁹,⁴⁶³,⁴⁹⁰ and HTNV proteins were shown to colocalize with clathrin in confocal immunofluorescence studies.²³⁹ In addition, CCHFV requires functional microtubules for infection.⁴⁸⁹ By contrast, entry of UUKV phlebovirus is predominately in a clathrin-independent manner, and viruses enter Rab5a+ early endosomes and subsequently Rab7+ and LAMP-1+ late endosomes.³²³ After endocytosis, acidification of the endocytic vesicles is thought to promote a conformational change in Gn and/or Gc that facilitates fusion of the viral and cellular membranes, thereby allowing the viral genome and polymerase access to the cytoplasm. Infection is blocked if cells were treated with ammonium chloride, which prevents acidification.²²⁹,⁴¹⁹,⁴⁶³,⁴⁸⁴ Additional indirect support for membrane fusion as a mode of entry came from experiments demonstrating the ability of viruses in the family to mediate syncytia formation at low pH.¹⁵³,¹⁸³,²²⁹,³⁸³,⁴¹⁹,⁴⁸⁴,⁵⁷⁴

Computational studies revealed that the Gc proteins have characteristics similar to those of class II fusion proteins identified for viruses in the Flaviviridae and Togaviridae families.¹⁶⁹ Such proteins have internal fusion domains, as opposed to terminal domains found for the class I fusion proteins. Experimental support for this finding was obtained by demonstrating interaction of the fusion peptide postulated for the Gc protein of the hantavirus, Andes virus (ANDV), with artificial membranes.⁵³⁷ In addition, several lines of evidence indicate that the Gc of orthobunyaviruses is involved in membrane fusion. Protease sensitivity assays, detergent partitioning experiments, and antibody-binding studies suggest that Gc undergoes a conformational change at pH conditions where fusion is observed.¹⁸² Mutational analyses of a hydrophobic region (residues 1,066–1,087) of LACV Gc that is well conserved among different orthobunyaviruses¹⁰⁶ supported this notion,²²⁹,⁴²⁰ and the residues flanking the predicted fusion peptide in BUNV Gc (residues 1,058–1,079) were shown to be structurally critical for the conformational change in Gc that occurs during the fusion process.⁴⁸³ LACV Gc alone, however, when expressed from recombinant vaccinia viruses, could not cause cell fusion, suggesting that an association of the two GPs may be needed for membrane fusion,²²⁹ and mutations in the cytoplasmic tail of BUNV Gn severely affected membrane fusion indicating that Gn must also play an important role in the fusion process.⁴⁸⁴ Likewise for HTNV, both Gn and Gc were required to achieve surface expression and cell-to-cell fusion activity.³⁸³ Recombinant LACV carrying mutations in the Gc fusion peptide was impaired for growth in both mammalian and insect cells but were still neurotoxic to neuronal cells, implying that the fusion peptide is a determinate of neuroinvasiveness but not necessarily neurovirulence.⁵⁰¹

To mediate replication, the RdRp must affect numerous enzymatic functions. For primed synthesis of mRNA from genomic templates and nonprimed synthesis of vRNA from cRNA templates (as will be described), the RdRp performs endonuclease, transcriptase, replicase, and probably RNA helicase activities. Motifs common to polymerases in general are conserved in the bunyaviral RdRps, most notably a catalytic core motif.²⁶,²³⁷,²⁸²,³⁶⁰ The N-terminal domain of orthobunyavirus L proteins contains a conserved PD-(D/E)xK amino acid nuclease motif and the isolated domain of LACV L exhibited nuclease activity in vitro; furthermore, bioinformatic analyses indicated that viruses in all the other Bunyaviridae genera contain a similar domain.⁴⁴² The nuclease domain of LACV bears structural similarities to that of the PA subunit of the influenza A virus polymerase, suggesting a common origin of the cap-snatching process of segmented negative sense viruses. None of the other functions has been definitively localized to a region of the RdRp gene or gene product; however, comparison of two nairovirus L proteins with functionally defined regions of other polymerases revealed potential helicase, topoisomerase, and gyrase coding regions as well as an N-terminal cysteine-protease motif typical of the ovarian tumor (OTU) protein superfamily and protease.²⁶² Biochemical and structural analyses demonstrate that in addition to the deubiquitinase activity as shown by cellular OTU proteases, the nairovirus OTU also targets ISG15 modification and thus enhances its role in overcoming host innate immune defenses.⁵,⁷⁷,¹⁶⁵,²³⁰ The OTU domain is not required for the RNA polymerase activity by CCHFV in a minigenome system.⁴³

The template for bunyavirus transcription and replication is not naked RNA but RNA in the form of ribonucleocapsid.¹²⁷,³¹⁹ The 3′ and 5′ nontranslated complementary nucleotides contain signals for both mRNA synthesis and antigenome synthesis, and thus are the genome promoters. For the phleboviruses RVFV and UUKV, the first 13 or the first 10 nucleotides, respectively, of the 3′ end of the genome are sufficient for RNA synthesis activity.¹⁵⁶,⁴²⁴

For the orthobunyavirus BUNV, a minigenome reporter system was used to demonstrate that optimal transcription required exact complementarity at the 3′ and 5′ termini; however, one to three nucleotide deletions could be tolerated at the 3′ terminus, but not the 5′ terminus.²⁶⁸ Although sequence changes at several positions could be tolerated as long as complementarity was maintained, some changes did reduce transcription. Interestingly, mutagenesis experiments indicated that the single exception to complementarity (a U residue at 3′ position 9) was critical for transcription from the genomic promoter, whereas the corresponding 5′ position 9 could be changed without influencing transcription.³³ Therefore, the mismatch itself is not important, but the nucleotide is. This nucleotide, however, does not have to be maintained for transcription activity of the antigenomic promoter.³³ Together, these data suggest that the structure of the panhandles is more important than the sequence, but that the sequence also plays a role in promoter strength and activity.

To further investigate promoter regions in an authentic replication system, S segments with shortened complementary termini were introduced into a BUNV reverse genetics system, and rescued viruses were screened to determine the minimal terminal complementary region needed for virus production.³²² The minimal 3′ and 5′ deletion mutant viable in this system maintained 29 of the 85 untranslated nucleotides on the 3′ end and 112 of the 174 untranslated nucleotides on the 5′ end, whereas those with shorter untranslated regions were not viable. Therefore, although only the complementary regions are needed for promoter activity in vitro, actual virus production also requires some of the unique sequences in the untranslated regions at the 3′ and 5′ termini of BUNV.³²²

The in vivo importance of intact termini was also suggested by a study in which terminally deleted RNAs were shown to accumulate during the establishment of persistent infections with the hantavirus, Seoul virus (SEOV). It was hypothesized that as deletions accrued, fewer replication-competent genomes were present, leading to a downregulation of the replication processes, and possibly to persistence.³⁴² Similar deletions were found on the termini of ANDV M and L, but not S genes, leading to a speculation that differences observed in the relative abundance of Gn and Gc compared to N could reflect down regulation of M-segment expression from genes without intact termini.³⁸⁹

Like influenza viruses, viruses in the family Bunyaviridae prime mRNA synthesis with capped oligonucleotides that are scavenged from host mRNAs (Fig. 42.5). Cleavage of the capped primers is accomplished by endonuclease activity found in virions and associated with the L protein.³⁹⁸,⁴⁴² Unlike influenza viruses, which take primers from newly synthesized mRNAs in the host cell’s nucleus, members of the Bunyaviridae family use primers cleaved from cytoplasmic host cell mRNAs. Evidence has been presented that the hantavirus N protein binds to the 5′ cap of cellular mRNAs to protect them from cell-mediated degradation, and that N accumulates in cytoplasmic processing bodies (P bodies) where protected 5′ caps are sequestered, and hence can serve as a pool of primers to initiate mRNA synthesis.³⁴⁶ A result of this mode of mRNA transcription is the presence of 5′ terminal extensions of approximately 10 to 20 heterogeneous nucleotides that are not found in vRNA.⁴⁷ Studies using anticap antibodies to immunoselect mRNAs¹⁹² have provided direct proof for the presence of caps on the scavenged primers. Further evidence for capped extensions of mRNAs was provided by a study of tospoviruses, in which plants were co-infected with TSWV and with a positive-strand RNA virus (alfalfa mosaic virus). The tospovirus acquired 5′ mRNA extensions with nucleotide sequences that matched those of the other virus with a preference for cleaving at an A residue.¹²⁵ It was suggested that this
preference is due to a need for base-pairing at the ultimate U residue of the gene segments. Consistent with this, later studies indicated that double and triple base complementarity to nucleotides at the ends of the tospovirus gene segments were preferred, even more than the single complementary residue.⁵⁴⁵ Interestingly, it was also found that newly synthesized mRNAs can serve as cap donors in vitro for tospoviruses, suggesting a “resnatching” mechanism.⁵⁴⁴

Other viruses in the family also have nucleotide or nucleotide motif preferences for endonuclease cleavage of capped primers. These preferences vary among the genera, and sometimes among viruses within a genus. A preferred primer sequence, or a favored nucleotide at the site of cleavage due to a need for limited base pairing with the viral genome, appears to be a common feature of primed transcription for bunyaviruses. In one study, the 3′-terminal nucleotides of the scavenged host primers often were similar to the 5′-terminal viral nucleotides.²³⁸ It was proposed that after transcription of two or three nucleotides of the nascent mRNA, the viral polymerase might slip backward on the template before further elongation, resulting in a partial reiteration of the 5′-terminal sequence. An extension of this concept, termed “prime and realign,” was proposed for mRNA transcription of hantaviruses.¹⁶⁸ According to this model (e-Fig. 42.2), priming by host oligonucleotides with a terminal G residue would initiate transcription by aligning at the third nucleotide of the viral RNA template (C residue). After synthesis of a few oligonucleotides, the nascent RNA could realign by slipping backward two nucleotides on the repeated terminal sequences (AUCAUCAUC) (Table 42.3), such that the G becomes the first nucleotide of the nontemplated 5′ extensions (e-Fig. 42.2). The frequent deletion of one or two of the triplet repeats in hantaviral mRNA supports this sort of slippage mechanism and suggests that sometimes the initial priming might start at the C residue of the third triplet in the conserved sequence rather than at the C of the second triplet.¹⁶⁸

Bunyavirus transcription is unique among negative-sense RNA viruses in that functional viral mRNA synthesis requires ongoing protein synthesis,³¹,⁴⁰⁰,⁴³⁰,⁴³²,⁴³³,⁵⁵³ a finding that at face value appears incompatible with the presence of a virion transcriptase. In the absence of protein synthesis, only short transcripts are produced in vivo and in vitro. If the in vitro reaction is supplemented with rabbit reticulocyte lysate, however, full-length RNAs are synthesized. The translational requirement is not at the level of mRNA initiation, but rather during elongation or, more precisely, to prevent the transcriptase from terminating prematurely. A model to account for these observations proposed that in the absence of ribosome binding and protein translation, the nascent mRNA chain and its template can base-pair, thereby preventing progression of the transcriptase enzyme.³⁷ Recently this model (e-Fig. 42.3) has been tested using BUNV model templates containing translational stop codons,³¹ and the results showed that translation of nascent mRNAs prevented transcription termination. Thus orthobunyaviruses couple transcription and translation, a feature commonly found in prokaryotes, but rare in eukaryotic cells.

For gene segments with simple negative-sense coding, synthesis of M- and S-segment mRNAs terminates about 40 to 100 nucleotides before the end of the genome RNA template.¹⁰⁹,¹⁴⁸,²³⁵,³⁹⁹ Although most other negative-strand RNA viruses use a stretch of nucleotides rich in U residues to signal transcription termination and polyadenylation, no such tract has been consistently identified for members of the family Bunyaviridae. This is supported by experimental evidence showing that bunyavirus mRNAs are not 3′ polyadenylated,¹,⁴⁰⁰,⁵⁴² but many have the potential to form stem-loop structures that are probably involved in enhancing translation of viral mRNAs.⁵³,⁵⁵¹

By using a reporter system that involved mutated S segments, the transcription termination signal for the orthobunyavirus BUNV was mapped to a 33-nucleotide region within the 5′ nontranslated region of the S segment.³² Within this region, a 6-nucleotide motif, 3′-GUCGAC-5′ was critical to transcription termination. In addition, changing the nucleotides in this transcription termination signal revealed a second downstream transcription termination signal, which had a 5-nucleotide motif, 3′-UGUCG′-5′, that was also found within the 33-nucleotide region of the upstream site, and that partially overlapped the critical 6-nucleotide motif. The finding that there are no U-rich regions in the transcription termination signal of BUNV is consistent with the absence of poly-A tails on the mRNAs.

Comparison of these S-segment sequences with those of other orthobunyaviruses revealed a high degree of conservation, suggesting that similar motifs also function for transcription termination throughout the genus, at least for S segments. A motif similar to that on the BUNV S segment was observed within the BUNV L, but not the M segment.³² Experimental mapping of BUNV L and M mRNA termination sites has not been reported.

Analysis of the L mRNAs of SNV hantavirus²¹⁸ and RVFV and Toscana (TOSV) phleboviruses⁷ showed they were co-terminal with the L vRNA template, suggesting L mRNA synthesis terminates by run-off of the RNA template.

A U-rich motif was proposed as the site of termination for the M mRNA and a C-rich motif (CCCACCC) as termination site for the S mRNA of SNV hantavirus.²¹⁸ Fine mapping
of M-segment mRNA termination sites for three phleboviruses (RVFV, TOSV, and SFSV) showed that termination occurred immediately following a C-rich domain in the template at a conserved motif 3′-C_1-3 GUCG/A-5′.⁷ Although previously it was thought the C-rich region itself was involved in transcript termination¹⁷⁸ deletion of this region in template RNA did not affect specific termination at the identified motif.⁷

The mechanism of transcription termination of the ambisense genes of phleboviruses and tospoviruses was initially thought to involve RNA secondary structure in the intergenic region.¹⁰³,¹¹⁸,¹⁴⁶,¹⁸⁷,¹⁸⁸,⁴⁹³ However, reanalysis of computer predictions of hairpin formation in these intergenic regions suggest that they are unlikely to form because of their complexity or low energy.⁷ Detailed mapping of mRNA termination sites for the N and NSs mRNA of RVFV, TOSV, and SFSV phleboviruses showed that the 3′ ends of the mRNAs contained most of the intergenic sequence and indeed overlapped each other.⁷ Furthermore, termination occurred for both messages at the same motif, 3′-C_1-3GUCG/A-5′, as in the M segment. Two copies of this motif are thus present, one in vRNA (genome) and one in the cRNA (antigenome). The conservation of this motif in viral genomes that display considerable sequence diversity suggests a similar mechanism for transcription–termination in the M and S segments. Deletion of either copy of the motif in RVFV S segment did not prevent correct termination of unaffected N or NSs mRNA, but the mutant viruses were attenuated in growth compared to wild-type.⁷

The change from primary transcription to replication requires that the RdRp, either acting alone or in concert with undefined viral or cellular factors, must at some point, switch from primed mRNA synthesis to independently initiating transcription at the precise 3′ end of the template to produce a full-length transcript. The processes involved in making that switch from primary transcription to genome replication have not been defined completely for any member of the family. Presumably, some viral or host factor is required to signal a suppression of the transcription termination signal responsible for generation of truncated mRNA and also to prevent the addition of the capped and methylated structures to the 5′ termini of the cRNAs. There is no question that genome replication and subsequent secondary transcription are prevented by translational inhibitors such as cycloheximide. These results indicate that continuous protein synthesis is required for replication of the genome. Although not proven, it is likely that synthesis of N is required for genome replication, as described for other negative-strand RNA viruses such as the rhabdovirus, vesicular stomatitis virus, and the paramyxovirus Sendai virus. For these viruses, encapsidation by N seems to serve as an antitermination signal, thus allowing full-length genome synthesis.

A prime-and-realign model was also postulated for the nonprimed transcription of hantaviral vRNA and cRNA, except that transcription would initiate with pppG alignment at the third nucleotide (C residue) of the template RNA. After synthesis of several nucleotides, polymerase slipping would realign the nascent RNA such that the initial priming G residue would overhang the template. It was further theorized that nucleolytic activity of the L protein might remove the overhanging G, leaving a monophosphorylated U residue at the nascent 5′ end. The presence of the monophosphorylated U on HTNV RNA was experimentally demonstrated¹⁶⁸ (e-Fig. 42.2).

Indirect evidence suggesting that a prime-and-realign method of initiation is used by phleboviruses was obtained by using a reconstituted transcription system to study the polymerase recognition sequence at the 3′ termini of the ambisense S segment RNA of RVFV. In those studies, mutational analysis of the terminal nucleotides revealed that the first 13 nucleotides are required for polymerase recognition, but that one of the two terminal dinucleotides (UGUG) could be removed without deleterious effects on transcription.⁴²⁴ These data also suggested that realignment is not a prerequisite for transcription initiation. In contrast, a recent study using the BUNV minigenome system showed that the viral polymerase was able to repair both insertions and deletions in model template RNAs, although this did not appear to involve the prime-and-realign mechanism.⁵⁵⁹

Schematics of transcription and replication based on information above are presented in Figures 42.5 and 42.6.

Signals for the encapsidation of the vRNAs and cRNAs of orthobunyaviruses, phleboviruses, nairoviruses, and hantaviruses were found entirely within the noncoding 3′ and 5′ regions of minireplicon reporters.⁵²,¹⁵⁴,¹⁵⁵,¹⁵⁷,³²¹,³⁴⁵,³⁴⁸,³⁸² Mutational
analysis of these terminal regions for BUNV revealed a conserved sequence within nucleotides 20–33 of the 5′ end of vRNA that was critical for packaging.²⁷⁰ Variable packaging efficiencies were noted for the L, M, and S segment–derived minireplicons, and packaging appeared to be independent for each.

Beyond the need for host-derived primers for initiating mRNA synthesis, little is known about the role that host factors might play in viral replication. One putative host transcription factor is an ∼40-kD protein isolated from the main insect vector of TSWV, and it was shown to bind to the viral RdRp at its C-terminus and to the viral RNA at its N-terminus.¹¹⁹ Both this protein and reticulocyte lysates were needed for RNA synthesis in vitro, indicating that other cellular factors are also required. Expression of this protein in mammalian cells rendered them permissive to TSWV replication.¹¹⁹ As is the case for a number of other negative-strand RNA viruses, heat-shock protein 90 (Hsp90) appeared essential for La Crosse virus replication in that virus yields were diminished in cells treated with Hsp90 inhibitors like geldanamycin due to destabilization of the viral L protein.¹⁰⁵

The N protein is the most abundant viral product in virions and infected cells. N plays several important roles in viral replication. In addition to protecting the RNA from degradation, N also interacts with L and Gn and Gc, although the exact interactions have not been defined. A critical step in N interaction with RNA appears to be the ability of the proteins to oligomerize. Oligomerization characteristics of the N proteins are likely different among the genera, which is probably not surprising, given the size differences of the N proteins, which range from approximately 19kD for orthobunyaviruses to 54kD for hantaviruses and nairoviruses (Table 42.2).

Chemical cross-linking studies indicated that the ribonucleocapsids of the phlebovirus RVFV, either from authentic virions or reconstituted from bacterially expressed protein, contain dimers of N as the basic oligomer.²⁹⁷,⁴⁴⁰ The crystal structure of RVFV N has been determined to 1.93Å resolution, which revealed that it has a novel all-helical fold without a positively charged surface RNA binding cleft as seen for other negative-sense RNA virus nucleocapsid proteins.⁴⁴⁰ Extensive ribonuclease digestion of authentic or reconstituted RNAs released heterogenous N-RNA multimers, consistent with the lack of observed helical symmetry. In contrast, hantaviral N proteins form stable trimers, which are postulated to assemble on the viral RNA, followed by further protein–protein interactions that encapsidate the entire RNA.²⁴⁹ The N–N interactions are likely to be electrostatic, with both amino- and carboxy-terminal residues participating.²⁵⁰ The N-terminal residues of hantaviral N proteins are predicted to form a coiled coil domain, providing a trigger for trimerization, and containing charged residues important for intermolecular interactions.⁹,¹⁰,¹³,⁵⁶² For TULV, carboxy-terminal amino acids 393 to 398 were shown to be crucial for trimerization, and amino acids 1 to 43 provided a secondary interaction region.²⁴⁹ These data are in agreement with those obtained with other hantaviruses (e.g., SNV and HTNV, for which carboxy-terminal and amino-terminal regions were shown to be involved in the homotypic interactions of N).⁹,⁵⁸⁹

Orthobunyaviruses and tospoviruses, in contrast to hantaviruses, do not appear to assemble preformed trimers, but instead multimerize by adding one N protein at a time. This hypothesis comes from chemical cross-linking experiments, which showed that the N protein of the orthobunyavirus BUNV was able to form a range of multimers from dimers to high–molecular-weight structures, with a preponderance of tetramers.³⁰⁸ Deletion of N- or C-terminal portions of N prevented multimerization and also resulted in loss of functionality in a minireplicon system. The data were interpreted to indicate a head-to-head and tail-to-tail multimerization model of individual N proteins. Likewise, for the tospovirus TSWV, analysis of deletion mutants identified binding regions both at the amino- and carboxy-termini of N, and although the type of multimers formed was not examined closely, it appears to be a continuous addition process, like that proposed for the orthobunyaviruses.²⁴²,⁵⁴⁰

In addition to protein–protein interaction, the amino- and carboxy-terminal regions of the tospovirus N were also implicated in RNA binding.⁴⁴⁷ In contrast, the RNA-binding site of hantaviral N was localized to a central, conserved region of the protein.⁴⁷⁴ For hantaviruses and bunyaviruses, RNA-binding studies with bacterially expressed, hexahistidine-tagged N proteins indicated that N preferentially interacts with the 5′ end of the vRNA.³⁸⁵,⁴⁷⁵ In another study, high-affinity binding of hantaviral N to panhandle RNA was demonstrated, and it was suggested that such binding is related specifically to the trimers, which were able to discriminate viral from nonviral RNA, whereas monomer or dimer subunits bound nonspecifically to RNA.³⁴⁸ These data suggest a requirement for multimerization of N for effective RNA binding and encapsidation, and implicate the panhandle structure as the trigger for encapsidation. Hantavirus N is also able to bind 5′ capped mRNAs (to generate a source of primers for transcription initiation; see above), and the cap and vRNA binding sites have been mapped to separate domains within the protein.³⁴⁹ Furthermore, binding to the mRNA cap induces a conformational change in N, and the structurally altered N loaded with the capped primer binds specifically to the conserved 3′ terminus of vRNA.

The S segments of viruses in the Phlebovirus and Tospovirus genera and some viruses in Orthobunyavirus and Hantavirus genera produce NSs proteins as well as N proteins. Sizes of NSs range from 10 kD for orthobunyaviruses and hantaviruses to more than 50 kD for tospoviruses (Fig. 42.2). The NSs protein of the phlebovirus RVFV is phosphorylated and accumulates in the nuclei of infected cells, where it forms fibrillar structures (e-Fig. 42.4).⁵¹³ The major phosphorylation sites of the RVFV NSs were mapped to serine residues located near the carboxy-terminus of the protein.²⁶⁷ Expression studies revealed that fibrils could form in the absence of other viral proteins; however, if the carboxy-terminal region of NSs was removed, NSs were still transported to the nuclei, but did not coalesce into fibrils.⁵⁸⁴ Similar fibrillar structures of NSs were observed for some strains of TSWV, but the structures appeared only in the cytoplasm.

Comparing the deduced amino acid sequences of NSs for five different phleboviruses revealed homologies of only 17% to 30%.¹⁷⁸ However, comparison of the NSs gene sequences for a number of strains of a single phlebovirus RVFV showed that certain areas were highly conserved.⁴⁵⁹ These data suggest that there may be a strong evolutionary pressure to maintain distinct portions of the NSs for individual viruses, but that the remainder of the protein can diverge without affecting the function of NSs.

Earlier work suggested that the NSs of phleboviruses and tospoviruses were only synthesized after viral replication; that is, after the ambisense vRNA has been copied to cRNA and mRNA was transcribed from cRNA. Recent work with RVFV, however, indicates that at least for this phlebovirus, virions can encapsidate cRNA and that the incoming cRNA ribonucleocapsids can serve as templates for NSs mRNA.²²⁵ The outcome of this is that NSs are present early in infection and might have a role in early replication. The NSs of three orthobunyaviruses all inhibited BUNV RNA synthesis in a dose-dependent manner in a minireplicon system.⁵⁶⁷ The results were suggested to indicate a function for NSs in regulating the activity of the RdRp. In addition to a possible role for NSs in early replication processes, NSs proteins have other important roles in viral infection, such as interferon antagonism, host-range determination, and gene silencing. These effects will be described in the context of effects on host cells.

The viral envelope GPs Gn and Gc are translated from a single mRNA complementary to vRNA. The polyprotein precursor of Gn and Gc is not seen in infected cells, and has been observed only by in vitro translation of RNA transcripts in the absence of microsomal membranes.⁵⁴² For most viruses, both Gn and Gc are preceded by signal sequences; therefore, cleavage of the polyprotein precursor is likely mediated by host signalase.¹⁵⁰ A pentapeptide motif, WAASA, which is highly conserved among hantaviruses, was shown to be the polyprotein cleavage site for HTNV, by mutational analysis of the signal sequence preceding Gc.³¹⁸

M-segment gene products have a cysteine content of approximately 4% to 7%. Positions of these residues are highly conserved in the M-segment products of related viruses, suggesting that extensive disulfide-bridge formation may occur and that the positions may be crucial for determining correct polypeptide folding. The secondary structure of the proteins is also involved in immunogenicity, as indicated by the finding that neutralizing or protective epitopes are often nonlinear.

All M-segment polyproteins display variable numbers of predicted transmembrane regions, and a hydrophobic sequence at the carboxy-terminus, indicative of a membrane anchor region. Therefore, the M-segment translation products of viruses in the family Bunyaviridae are typical class 1 membrane proteins, with the amino terminus exposed on the surface of the virion and the carboxy-terminus anchored in the membrane.

The Gn and Gc proteins of all viruses in the family possess asparagine-linked oligosaccharides. Examination of the oligosaccharides attached to the Gn and Gc proteins of orthobunyaviruses revealed that Gc has mostly high-mannose glycans, whereas Gn contains both complex and a novel intermediate-type oligosaccharide.³²⁶,³⁷⁷,⁴⁷⁹ In contrast, the GPs of hantaviruses were found to be mostly of the high-mannose type.⁴⁶⁷ These findings indicate that the proteins are incompletely processed through the Golgi. This is likely related to retention of Gn and Gc in the Golgi, where assembly occurs, and which will be discussed more fully.

Nairoviruses appear unique among viruses in the family in that they have both N-linked and a region of heavily O-linked carbohydrates (a mucin-like domain, as will be described).⁴⁶¹

The M-segment gene-product processing events differ among the genera. Hantaviruses produce only Gn and Gc,⁴⁷¹ whereas some viruses in the Phlebovirus and Orthobunyavirus genera produce NSm from the same M-segment mRNA as Gn and Gc. The NSm protein of orthobunyaviruses is encoded between the Gn and Gc proteins.¹⁴⁹ Its hydropathy profile suggests that it is a membrane-bound protein, and expression studies demonstrated that it localizes to the Golgi along with Gn and Gc.³⁰¹ Although a function for this protein has not been identified, it was suggested that it might be involved in facilitating virion assembly in the Golgi.³⁰¹

For the phlebovirus RVFV, a NSm of approximately 14 kD is cleaved from the amino-terminus of the polyprotein translation product (e-Fig. 42.5). Sequence studies revealed five in-frame translation initiation codons (AUG) upstream of the amino-terminus of Gn. Mutational analysis of those codons indicated the fourth and the fifth AUGs to translate Gn and Gc. Whilst NSm was found to originate from the second AUG and a 78-kD polypeptide, representing uncleaved NSm and Gn, was translated from the first AUG.¹⁹⁸ Pulse-chase experiments revealed no precursor–product relationship between the 78- and 14-kD proteins; therefore, it appears that use of the first and second AUGs, respectively, is what dictates generation of these proteins.⁷⁷ Both the 14-kD NSm and the 78-kD polypeptide are found in abundance in RVFV-infected cells,⁷⁸ suggesting that they might play a role in replication or morphogenesis.

An even larger NSm (∼30 kD) is cleaved from the amino terminus of the M-segment polyprotein of the phlebovirus, PTV (e-Fig. 42.5). In vitro expression studies indicated that both envelope GPs could be produced in the absence of the NSm coding region,³³⁵ but other studies to evaluate the use of the 13 potential translation initiation codons present in the NSm coding information have not been reported. No homology between the NSm proteins of PTV and RVFV was apparent.²²¹

In contrast to the mosquito-borne phleboviruses PTV and RVFV, the tick-borne phlebovirus UUKV does not produce an NSm (e-Fig. 42.5). The first (and only) initiation codon preceding sequences of the envelope GPs is located 17 amino acids upstream of the amino terminus of Gn⁴⁴⁹; hence, the Gn protein of UUKV appears to be analogous to the 78-kD NSm–Gn fusion product produced by RVFV (although there is no obvious sequence homology of these predicted products). Until a function can be assigned to NSm, it is impossible to determine whether UUKV replicates in the absence of such a function or accomplishes whatever function is required without removal of a portion of the amino terminus of Gn.

M-segment polyprotein processing events for nairoviruses are more complicated and appear to differ from those of other viruses in the family (e-Fig. 42.6). Nucleotide sequences of DUGV, CCHFV, and NSDV revealed a coding strategy similar to other viruses in the family, that is, a single ORF.³³¹ Studies on CCHFV demonstrated the polyprotein is co-translationally
cleaved into 140-kD PreGn and 85-kD PreGc precursors and a 15-kD NSm protein.¹⁴,⁴⁴,⁴⁶¹,⁴⁶²,⁵⁵⁵ PreGn is processed to the mature 37-kD Gn by cleavage with a cellular secretory pathway protease, SKI-1, following a consensus motif, RRLL, early in the secretory pathway.⁴⁶¹ SKI-1 cleavage generates a subprecursor of the other two polypeptides that is cleaved by a furin-like protease late in the secretory pathway.⁴⁶¹ This cleavage site is completely conserved among several CCHFV strains, suggesting that it has importance for viral replication or pathogenicity.⁴⁶¹ The amino-terminal portion of this subprecursor is a highly variable (among CCHFV strains) polypeptide with mucin-like characteristics, including numerous serines, threonines, and prolines and predicted extensive O-linked glycosylation. The carboxy-terminal portion of the subprecursor is a 38-kD nonstructural polypeptide. After furin cleavage, secreted proteins GP38, GP85, GP160, and the mucin-like protein are produced; presumably these products reflect extensive O-linked glycosylation differences. Although the PreGc precursor also possesses a SKI-1–like cleavage motif, SKI-1 cleavage could not be demonstrated for this protein, so it is likely that a related enzyme is responsible for the processing to yield mature Gc.⁴⁶¹