Structural basis of poxvirus A16/G9 binding for sub-complex formation

Structural basis of poxvirus A16/G9 binding for sub-complex formation Fanli Yang#, Sheng Lin#, Zimin Chen, Dan Yue, Ming Yang, Bin He, Yu Cao, Haohao Dong, Jian Li, Qi Zhao and Guangwen Lu West China Hospital Emergency Department (WCHED), State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, Chengdu, People’s Republic of China; Disaster Medicine Center, West China Hospital, Sichuan University, Chengdu, People’s Republic of China; Laboratory of Aging Research and Cancer Drug Target, State Key Laboratory of Biotherapy and Cancer Center, National Clinical Research Center for Geriatrics, West China Hospital, Sichuan University, Chengdu, People’s Republic of China; School of Basic Medical Sciences, Chengdu University, Chengdu, People’s Republic of China; College of Food and Biological Engineering, Chengdu University, Chengdu, People’s Republic of China

Dear Editor, On July 23rd, 2022, the sudden monkeypox outbreak was declared as a Public Health Emergency of International Concern by the World Health Organization [1]. The latest epidemiological data has revealed a total of 84,733 laboratory-confirmed monkeypox cases with 80 deaths in 110 countries, areas and territories [1], highlighting an unexpected poxvirus-associated pandemic globally. While the COVID-19 pandemic is still surging, the emergence of monkeypox virus infection has unavoidably posed a greater threat to social economy and public health. Poxviruses and the related diseases once again draw worldwide attention after the eradication of smallpox in the late twentieth century [2]. Entry, within the viral life cycle, is the first step for a virus to set up infection and also the most important target for the development of antiviral therapy. It is therefore an urgent issue to characterize the structural features of key viral proteins involved in poxvirus entry.
Poxviruses are a group of enveloped viruses, the entry of which require the fusion between viral envelop and cell membrane. For most of other enveloped viruses, their attachment and entry just depend on one or a few proteins. In poxviruses, however, its fusion process is mediated by a large proteinaceous machinery, which is designated as the entry-fusion complex (EFC) [3,4]. It is believed that poxvirus EFC is composed of at least eleven virus-encoded protein subunits, among which the A16 and G9 subunits (nomenclatures of A16, G9 and other proteins refer to vaccinia virus, a prototype membrane of Orthopoxvirus genus in Poxviridae family [5]) form a sub-complex and play a key role in EFC assembly and function. In addition, the A16/G9 sub-complex is also reported to interact with the viral A26 protein and A56/K2 complex, which could in turn moderate the fusion process [6,7]. In recognition of the important functions of the A16/G9 sub-complex, the atomic structure of this sub-complex, however, remains uncharacterized.
In this study, we reported the crystal structure of the A16/G9 sub-complex from the vaccinia virus. As expected, the two proteins and their homologs of known orthopoxviruses individually exhibit high sequence similarities (Supplementary Figure S1). To verify the sub-complex formation, A16-ectodomain engineered with a C-terminal His-tag and G9-ectodomain with a C-terminal Strep-tag were co-expressed in insect cells (Figure 1(A,B)). Expectedly, the two proteins were easily co-purified during Ni-NTA affinity chromatography and remained bound as a stable complex in solution during gel-filtration chromatography ( Figure 1(B,C)). The structure of the A16/G9 subcomplex were subsequently solved via X-ray crystallography. The final structure, with a resolution of 2.7-Å, was refined to R work = 0.243 and R free = 0.273, respectively (Supplementary Table S1).
Within the crystallographic asymmetric unit of the structure, a single A16 protein and one G9 protein are present in a 1:1 binding mode. Traceable electron densities can be observed for A16 residues L7-C291 and G9 amino acids E8-N268 (Supplementary Figure S2 For A16, its NTD is composed of ten β-strands (β1-β10), three α-helices (α1-α3) and one 3 10 -helix (η1). The strands assemble into three antiparallel β-sheets, which are further interspersed by the helical components to form a compact structure. The CTD of A16 consists of eight α-helices (α4-α11) and three 3 10 -helices (η2-η4). The helical elements of α5-α9 are arranged one-by-one in an anti-parallel mode, stacking into a two-layered helical array. This array is further decorated on one side by helices α4 and η2 and on the other by the intertwined helices α10-α11 and η3-η4, together assembling into a pure helical structure with an extended conformation. A bunch of intra-domain disulfide linkages are observed to form to further stabilize the individual domain structure, including two in NTD (C60/C90 and C70/C128) and seven in CTD (C146/C155, C147/C168, C176/ C185, C204/C213, C236/C245, C247/C270 and C265/C291) (Figure 1(D); Supplementary Figure S3). Sterically, the NTD and CTD of A16 are linked via a relatively long loop, which might be of certain flexibility.
For G9, its NTD also features with an α/β structure. Within this domain, six β-strands (β1'-β6') assemble into two anti-parallel β-sheets, which are further interspersed with four α-helices (α1'-α4') and one additional 3 10 -helix (η1'). In addition, a long N-terminal loop is observed in the NTD of G9, extending out into A16-NTD. The CTD of G9 is composed of ten α-helices (α5'-α14') and two 3 10 -helices (η2'-η3'). These helical components are intertwined together, once again assembling into a pure helical structure as observed for A16-CTD. The individual domain structure of G9 also features with intra-domain disulfides for stabilization, including one in NTD (C88/C117) and four in CTD (C135/C145, C177/C186, C223/C248 and C243/ C268) (Figure 1(D); Supplementary Figure S3). Unlike that of A16, however, the inter-domain linker in G9 is of short length and an inter-domain disulfide-bond (C89/C127) is observed to form, dragging the two domains into close proximity (Supplementary Figure S3). We believe this would result in a limited domain-hinge plasticity between NTD and CTD of G9. It is notable that except for the angle of the inter-domain connections, the A16 and G9 structures predicted by AlphaFold2 seem to match the crystal structures well.
We further performed mutagenesis study to validate the observed A16/G9 interaction. Because large number of amino acids and diverse types of interchain contacts (vdw contacts, hydrophobic interactions and H-bonds) are observed to locate along the A16-G9 interface, it is expected that single mutations or a small number of mutations in A16 and G9 proteins will not disrupt the tight binding of the two entities. Thus, a bunch of important residues involved in the interactions were simultaneously mutated into alanine. The subsequent A16 mutant (A16-mut) contains substitutions: E42A, I67A, S71A, L84A, H86A, F95A, R96A, E166A, D170A, T226A, D257A, P262A, R263A and W266A, and the G9 mutant (G9-mut) contains: L9A, P10A, K11A, R12A, E22A, M23A, K38A, N67A, P70A, T95A, N154A, H192A, Y261A and L264A. The A16 and G9 proteins were then coexpressed in pairs for co-precipitation analyses. Both the wild type and the mutant proteins were well expressed and extracted (Figure 1(K)). While A16-wt and G9-wt were readily co-purified, highly attenuated co-precipitation was observed for the A16-wt/G9-mut pair and no obvious co-precipitation was recorded for the A16-mut/G9-wt and A16-mut/G9-mut pairs (Figure 1(L)). The results demonstrated that those interface residues identified in our structure indeed play important roles in the A16/G9 sub-complex formation.
In conclusion, we report, to our knowledge, the first atomic structure of the poxvirus A16/G9 complex, which we believe would facilitate future studies on the mechanism of poxvirus EFC assembly and guide the antiviral drug design.