A comprehensive list of Bunyavirales replication promoters reveals a unique promoter structure in Nairoviridae that differs from other virus families

In this study, we compiled the replication promoter structures of all known virus species in order Bunyaviruses. Our analysis focused on five large families of tri-segmented viruses, and the results indicated that the genomes can be divided into two categories: those whose promoter begins with A (families Peribunyaviridae, Phenuiviridae, and Tospoviridae) and those whose promoter begins with U (families Hantaviridae and Nairobiviridae; figure 2). Viral RNA polymerases have been shown to have the ability to initiate RNA synthesis with a purine (G or A), but not with a pyrimidine (U or C)¹⁴. Therefore, the 5′-U genomic end of Hantaviridae and Nairobiviridae is unconventional. The genomes of the LACV and Rift Valley fever virus (RVFV) families Peribunyaviridae and Phenuiviridaerespectively contain a 5′-triphosphate end that starts with A (5′-pppA)^15.16. 5′-pppA is generated by viral RdRp which recognizes the opposite U as a template. As seen in several segmented and unsegmented viral RNA polymerases^17,18,19, the bunyaviral RdRp synthesizes RNA from an internal nt and not from the template end. In LACV, RNA synthesis is initiated with A using the U at position +4 of the antigenome (3′-UCAyouCA) as a template during genome replication⁴. The elongated product, 5′-pppAGU, is realigned to the +1 to +3 position of the antigenomic template (3′-UCAUCA), and is further elongated to generate 5′-pppAGUAGU. Accordingly, the position of U responsible for the initiation of RNA synthesis is likely position +3 in the Phenuiviridae (3′-UGyouG) and Tospoviridae (3′-UCyouC) antigenomes. This indicates that the 5′-pppTHAT and 5′-pppHER the products realign on the 3′-U GUG and 3′-CPUUC of the antigenomes, respectively, and are further elongated to generate 5′-pppACAC and 5′-pppAGAG, respectively, which are precise complementary chains of the antigenome templates. In contrast, the Hantaan virus (HTNV) genomes of the family Hantaviridae, and CCHFV in the family Nairobiviridae contain a 5′-monophosphate end^15.20, suggesting an unconventional RNA processing event during replication. In HTNV, RNA synthesis is initiated with an internal G at position +3 using a C from 3′-AUVSAUC of the antigenome as a model²⁰. Subsequently, the elongated 5′-pppGAU the product realigns on the 3′-AUCAUC to produce more 5′-pppGUAGUA. Then, the extreme 5′-pppG is removed by an endoribonuclease activity of the viral RdRp to produce 5′-pUAGUA (5′-monophosphate end)²⁰. The endoribonuclease activity of RdRp is responsible for the pull-off of the cap which cleaves the 5′ end of the host mRNA for use as a transcription primer²¹. Position -1 of viral HTNV mRNA is G, indicating that viral RdRp can cleave host mRNA after G nt (cleave GpN to produce G/pN) during transcription. In Nairobiviridaethe −1 position of the viral mRNA is C^22.23, and it is also generated via the cap-snatching mechanism. Similar to RNA synthesis in Hantaviridaenairoviral RNA synthesis is assumed to be initiated internally with 5′-pppC at position +2 using the G of 3′-AgHas antigenome as template. Subsequently, 5′-pppCyou the product would realign to the 3′-AGA, and be further elongated to generate 5′-pppCUCU. The 5′-pppC would then be eliminated, resulting in the production of a 5′-monophosphate end. Therefore, although hantaviral RdRp is a conventional enzyme that initiates RNA synthesis with a purine (G), nairoviral RdRp is considered an unconventional enzyme that can initiate with a pyrimidine (C). Such a difference may be important for the targeting of new antivirals specific for nairoviral diseases.

Our analysis further confirmed that most bunyavirus genomes begin with a di- or tri-nt repeat (Fig. 2), which has been previously suggested.^20.24. Repeats can determine the RdRp initiation site (e.g., +2 in Nairobiviridae+ 3 in Hantaviridae, Phenuiviridae and Tospoviridaeand + 4 in Peribunyaviridae), which is important for the mechanism of prime-realign RNA synthesis. The biological significance of the internal position of the initiation of RNA synthesis is unclear. It is likely that the di-nt repeat is restricted to virus families with an ambisense genome, such as Phenuiviridae and Tospoviridae, as good as Nairobiviridae (for which only CCHFV was reported)²⁵. This suggests that the ambisense coding property may be related to the di-nt repeat in the genomic ends. If true, analysis of genomic end repeat patterns may allow the elucidation of novel transcripts in diverse bunyaviral genomes.

The Nairobiviridae The promoter appears to have high G:C complementarity in the 17-21 nt region (Fig. 2A and Supplementary Fig. 2), which likely reflects the high level of G:C complementarity in the promoter region (Fig. 3A ). Interestingly, this GC-rich dsRNA region is located after a spacer region composed of non-complementary bases around the 14th position in all three segments, as previously reported in HAZV and CCHFV.^8.26. We have previously suggested the possibility that the HAZV polymerase may recognize this GC-rich dsRNA as an essential promoter element for the initiation of RNA synthesis via an unidentified domain of the L protein.⁸. This type of specific protein-RNA interaction has been proposed as a suitable target for antivirals against CCHFV, which is closely related to HAZV. Our comprehensive analysis of the list of promoters also suggested that this type of strategy could be applicable to all viruses belonging to the family Nairobiviridae.

In bunyaviruses, genome replication in each segment is regulated by segment-specific promoter strength, but variations of nts (A, U, G, and C) in each promoter region do not differ significantly between L, M segments. and S in all virus families except Nairobiviridae (Fig. 3B). It is possible that promoter strength among segments is determined by slight differences in promoter structure that do not affect the total number of complementarities or nt variations. Viruses in the family Phenuiviridae and family CCHFV Nairobiviridae have an ambisense S-segment, but there is no pattern of nt variation in the promoter that is unique to the S-segment (Fig. 2B). This suggests that the nt variation of the promoter was not affected by the presence of the ambisense segment during the process of viral evolution. In contrast, the nt variation of the nairoviral L segment promoter was different from that of the M and S segments, that’s to say, it was observed that there were fewer Gs and Cs at the 5′ and 3′ ends, respectively (Fig. 2B). The nairoviral L segment is remarkably long compared to other nairoviral segments and the genomes of other virus families (Fig. 4B). This large genome size may be associated with the structure of the promoter.

It is still unclear why the family genome Nairobiviridae is so big. Nairobiviridae is the only tri-segmented virus family that includes hemorrhagic fever viruses classified as BSL-4 pathogens, such as CCHFV. We hypothesized that the large genome size of the family Nairobiviridae may be related to its strong pathogenesis in mammals. Although the length of the segment L in Nairobiviridae is the longest among all bunyaviruses, it is not particularly long among the highly pathogenic viruses of this family (Fig. 5A). On the contrary, our analysis confirmed that among the viruses of the family Nairobiviridae, the M-segment is the largest segment of two highly pathogenic mammalian viruses, CCHFV and NSDV, suggesting that the M-segment contains factors involved in viral pathogenesis. The M segment encodes GPC which is first translated as a polyprotein from mRNA and then cleaved into Gn, Gc and other accessory or uncharacterized proteins. A schematic diagram of several representative nairovirus GPCs is shown in Figure 5B. GPC contains an N-terminal signal peptide and several membrane-spanning domains, and is processed by signal peptidases to generate an N-terminal pre-Gn protein, a C-terminal pre-Gc protein, and a double-membrane NSm protein. The pre-Gn and pre-Gc are then transformed by furin or subtilisin-kexin isozyme-1-like proteases to generate a mucin-like protein containing a large number of O-glycosylation sites, a protein called GP38 (-like), Gn virion envelope glycoprotein and Gc virion envelope glycoprotein²⁷. We have shown that although the sizes of Gn and Gc are similar among virus species, those of O-the glycosylation sites and the GP38-like protein are different; in particular, they are larger in CCHFV and NSDV (Fig. 5B). This suggests that these regions could be determinants of the pathogenicity of Nairobiviridae. GP38 has been proposed to be involved in CCHFV particle formation and viral infectivity²⁷. Analysis of sera from convalescent patients showed high titers of CCHFV GP38 antibodies, indicating the immunogenicity of this protein in humans during natural CCHFV infection. In a mouse model, an antibody against GP38 could protect animals from heterologous CCHFV challenge, indicating an association between GP38 and the strong pathogenesis of CCHFV²⁸. Our current analysis indicates that there is an association between the N-terminal GPC region and viral pathogenesis not only in CCHFV, but also in other highly pathogenic nairoviruses, including NSDV.

In conclusion, we have built an exhaustive list of promoters in Bunyaviruses which included all virus families in that order. Studies on the mechanism of RNA synthesis from Bunyaviruses were restricted to only a few virus species. Analysis of conservation in all promoter structures is useful for prediction of RNA synthesis mechanisms in uncharacterized and newly identified bunyaviruses. The automatic promoter characterization system (Supplementary Table 1) is applicable to all bunyaviruses for which the precise genomic terminal sequences are known.