Structural basis of the monkeypox virus mRNA cap N7 methyltransferase complex

Abstract

The global outbreak of Mpox, caused by the monkeypox virus (MPXV), has attracted international attention and become another major infectious disease event after COVID-19. The mRNA cap N7 methyltransferase (RNMT) of MPXV methylates the N7 position of the added guanosine to the 5'-cap structure of mRNAs and plays a vital role in evading host antiviral immunity. MPXV RNMT is composed of the large subunit E1 and the small subunit E12. How E1 and E12 of MPXV assembly remains unclear. Here, we report the crystal structures of E12, the MTase domain of E1 with E12 (E1_CTD-E12) complex, and the E1_CTD-E12-SAM ternary complex, revealing the detailed conformations of critical residues and the structural changes upon E12 binding to E1. Functional studies suggest that E1_CTD N-terminal extension (Asp545-Arg562) and the small subunit E12 play an essential role in the binding process of SAM. Structural comparison of the AlphaFold2-predicted E1, E1_CTD-E12 complex, and the homologous D1-D12 complex of vaccinia virus (VACV) indicates an allosteric activating effect of E1 in MPXV. Our findings provide the structural basis for the MTase activity stimulation of the E1-E12 complex and suggest a potential interface for screening the anti-poxvirus inhibitors.

Disclaimer

As a service to authors and researchers we are providing this version of an accepted manuscript (AM). Copyediting, typesetting, and review of the resulting proofs will be undertaken on this manuscript before final publication of the Version of Record (VoR). During production and pre-press, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal relate to these versions also.

Introduction

Monkeypox virus (MPXV) is classified as an orthopoxvirus within the poxviridae family, exhibiting the capacity to induce zoonotic viral diseases. Human cases of MPXV were initially documented sporadically in Africa during the 1970s [1]. Since 2022, there has been a dramatic increase in the number of MPXV cases globally, with the number of detected MPXV infections soaring to more than 48,000 and a total of 13 deaths recorded in less than four months [2]. This epidemiological upswing prompted the World Health Organization (WHO) to officially declare the monkeypox outbreak a global health emergency [3]. MPXV is one of the largest and most complex double-stranded DNA viruses, showing a brick-shaped structure with a length of 220 to 450 nm and a width range of 140 to 260 nm. It has four components: the core, the lateral bodies, the outer membrane, and the outer lipoprotein envelope, of which the core is the central part surrounded by core fibrils and double-stranded viral DNA [4].

During poxvirus-host coevolution, viral mRNA has the same cap structure as host mRNA, which allows the virus to evade detection by the innate immune system and produce viral proteins efficiently [5]. The synthetic cap structure consists of three typical capping reactions [6]. First, RNA triphosphatase (RTPase) hydrolyses γ-phosphate (pppNp-RNA) of nascent RNA to generate diphosphate RNA (ppNp-RNA) and inorganic phosphate (Pi). The guanyltransferase (GTase) then reacts with the α-phosphate of GTP, releasing pyrophosphate (PPi) and forming a covalent enzyme-guanylate intermediate (Gp-GTase). GTase then transfers the GMP molecule (Gp) into the 5 ‘-diphosphate RNA to produce GpppNp-RNA [7]. In the final step, guanine-N7-methyltransferase (N7MTase) transfers methyl groups from S-adenosine-L-methionine (SAM or AdoMet, which can provide methyl groups to methylate the substrate) to coronoguanine to form the cap-0 structure, 7-methyl-GPPPNP (m7GpppNp), and releases S-adenosine-L-homocysteine (SAH or AdoHcy) as a by-product [7].

MPXV belongs to the poxvirus family and encodes a heterodimeric RNA capping enzyme consisting of a large subunit E1 and a small subunit E12 [8]. The RNA capping enzymes encoded by poxviruses share a high degree of homology among the poxvirus family. The E1 subunit of the MPXV contains catalytic modules required to produce cap 0, as does the D1 subunit of the vaccinia virus [8]. The catalytic modules are neatly arranged in the order of the N-terminal RTPase, the middle GTase, and the C-terminal domain (CTD) composed of guanine-N7 MTase [9]. The guanine-N7 MT enzyme activity of the vaccinia virus requires an association with D12 to function efficiently [10]. Considering the high homology between E12 and D12, it is postulated that the CTD of E1 (denoted as E1_CTD) forms a stable complex with E12. Reports showed that the interaction of D12 with the guanine-N7 MTase domain of vaccinia virus capping enzymes induces the conformational changes required for efficient catalysis [11]. However, how the conformational change occurs and the transition from GTase to MTase active site in MPXV remains unclear.

In this study, we solved the MPXV E12, the E1_CTD-E12 complex, and the E1_CTD-E12-SAM ternary complex structures using X-ray crystallography. E1_CTD and E12 form a 1:1 heterodimeric complex in solution and crystal structure. Mutations in the E1_CTD -E12 interface dramatically decreased the binding affinity. Structural comparison of E12 and E1_CTD-E12 complex showed significant conformation changes upon E12 binding to E1. In addition, we found the E1_CTD N-terminal extension (Asp545-Arg562) and the small subunit E12 play an essential role in SAM binding. Superimposition of the AlphaFold2-predicted structure of full-length E1 with different homologs revealed that bound E12 could stimulate E1 activity and form a compact complex. This may provide the structural basis for the MTase activity stimulation of the E1-E12 complex and suggest a potential interface for screening the anti-poxvirus inhibitors.

Results

The overall structure of the MPXV E1_CTD-E12 MTase complex

To reveal the molecular basis of the MPXV MTase complex in evading host antiviral immunity, we investigated the crystal structure of the E1-E12 complex. The full-length MPXV E1 contains three domains, named RTPase domain (residues 1-225), GPase domain (residues 226-502), and N7 MTase domain (CTD, residues 545-844) (Figure 1A). E1 interacts with E12 to form the MPXV MTase complex via the N7 MTase domain (E1_CTD-E12). As obtaining the full-length E1-E12 complex is challenging, we successfully co-expressed and purified the E1_CTD-E12 complex in E. coli. The E1_CTD-E12 complex came out at the elution peak of 79 mL on a Superdex200 16/600 GL column, corresponding to a molecular mass of 68 kDa (Figure 1B).

Figure 1. Crystal structure of the MPXV E1_CTD-E12 complex. (A) The domain architectures of E1 and E12 protein. The RTPase, GPase, and N7 MTase domains (denoted as E1_CTD) of MPXV E1 are colored in gray, blue, and orange, respectively. E12 is colored in green. (B) Size-exclusion chromatographic profile and SDS-PAGE result of the E1_CTD-E12 complex on a Superdex200 16/300 column. M: protein marker. (C) A ribbon representation of the E1_CTD-E12 complex structure, showing 14 α-helices and 23 β-sheets. The E1_CTD and E12 are colored in orange and green, respectively. (D) The electrostatic potential surfaces of the E1_CTD-E12 complex. Red indicates negative potential, and blue indicates positive potential.

After several rounds’ optimization of crystallization conditions, we successfully solved the E1_CTD-E12 crystals structure, which belonged to the space group P6₅22 at a resolution of 3.94  Å (Figure 1C and Table 1). There is one molecule of E1_CTD-E12 within the asymmetric unit, composed of 14 α-helices and 23 β-sheets (Figure 1C). The electrostatic potential surfaces of the E1_CTD-E12 complex are evenly distributed with positive and negative charges. However, the RNA-binding pocket of E1, adjacent to E12, was full of positive charge (Figure 1D).

Table 1. Data collection and refinement statistics

	E12	E1CTD-E12	E1CTD-E12 (H122D) +SAM
PDB codes	8Y2Y	8Y2Z	8ZE4
Data collection
Space group	P3121	P6522	P6522
Unit cell a, b, c (Å)	77.24 77.24 96.94	138.55 138.55 178.36	136.56 136.56 177.94
α, β, γ [°]	90 90 120	90 90 120	90 90 120
Resolution range (Å)^a	55.06-2.2 (2.27-2.2)	45.12-3.94 (4.08-3.94)	44.7-3.33 (3.44-3.33)
Total reflections	335079 (33709)	18870 (1831)	29844 (2902)
Unique reflections	17452 (1688)	9436 (916)	14922 (1451)
Completeness (%)	99.96 (99.94)	99.76 (100.00)	99.78 (99.86)
I/σ(I)	19.7 (2.1)	7.49 (1.67)	13.25 (1.61)
Wilson B-factor	43.48	145.03	100.47
R-merge^b	0.079 (0.4905)	0.053 (0.399)	0.041 (0.412)
R-meas	0.082(0.5032)	0.074 (0.565)	0.058 (0.583)
R-pim	0.019 (0.1119)	0.053 (0.399)	0.041 (0.412)
CC1/2	0.999 (0.953)	0.995 (0.675)	0.999 (0.782)
Refinement statistics
Resolution (Å)	55.06-2.2 (2.27-2.2)	45.12-3.94 (4.08-3.94)	44.7-3.33 (3.44-3.33)
Reflections used in refinement	17446 (1687)	9425 (916)	14905 (1449)
Rwork^c	0.2119 (0.2770)	0.294 (0.389)	0.2793 (0.3801)
Rfree^d	0.2664 (0.3050)	0.327 (0.418)	0.3199 (0.4678)
No. residues	277	580	565
RMSD
Bond lengths (Å)	0.015	0.005	0.003
Bond angles (Å)	1.61	0.83	0.63
Average B-factor	50.55	160.21	104.66
Ramachandran favored (%)	93.75	88.29	89.66
Ramachandran allowed (%)	5.51	10.84	10.34
Ramachandran outliers (%)	0.74	0.87	0.00
No. atoms
Protein or RNA	2325	4730	4477
Ligands or ions	0	0	0
Water	72	0	0

^a Values in parentheses are for the highest resolution shell.

^b R_merge= ∑|I_i − |/∑|I|, where Ii is the intensity of an individual reflection and is the average intensity of that reflection.

^c R_work= ∑||F_o| − |F_c||/∑|F_o|, where F_o and F_c are the observed and calculated structure factors for reflections, respectively.

^d R_free was calculated as R_work using the 5% of reflections that were selected randomly and omitted from refinement.

The interface between E1_CTD and E12

Previous studies showed that the MTase domain of VACV D1 has a feeble intrinsic biochemical activity and requires a stimulating subunit (D12) to exert full MTase activity. The MPXV capping enzyme has a C-terminal MTase domain of E1 subunit and a stimulatory subunit E12. The structure of E1_CTD comprised 11 β-strands and 10 α-helices. The overall view indicates that E1_CTD has a classic SAM-MT fold [12], incorporated seven alternating β-strands, namely β1 (Lys593-Ile597), β2 (Leu615-Thr619), β3 (Lys646-Ile650), β4 (Ile674-Gln678), β5 (Lys706-Met712), β6 (Phe780-Asp787), and β7 (Tyr836-Lys843), surrounded by six α-helices, namely, α1 (Gly563-Tyr580), α2 (Leu605-Gly611), α3 (Asp624-Leu636), α4 (Asp657-Glu665), α5 (Tyr690-Ser698), and α6 (Lys768-Glu777). Different from the core SAM-MT fold, E1_CTD has one α-helix (αa) and four alternating β-strands (βa, βb, βc, βd) between β5 and α5, and three α-helices (αb, αc, αd) between β6 and β7 (Figures S1 and 2A).

The E12 subunit exhibits a high affinity to the E1_CTD domain. The interface-involved residues of E12 were located in the α2 helix (Leu34-Arg49), while the related residues of E1_CTD were mainly located in regions Tyr571-Asp587 (α1) and Glu806-Phe814 (αb-αc) (Figure 2A). Electrostatic and hydrophobic analysis showed that the inner part of the interface is mainly hydrophobic, while the ends are electrostatic interaction. Briefly, residues Lys31 and Arg49 on the E12 formed salt bridges with Asp587 and Glu814 on the E1_CTD, respectively. Also, there is extensive hydrophobic interaction at the E1_CTD-E12 interface. Residues Leu37 and Phe44 on the E12 formed hydrophobic interaction with Leu575 and Phe814 on the E1_CTD, and the side chain of E12 Leu34 formed hydrophobic interaction with the side chain of E1_CTD Leu586 residue (Figure 2B). Next, we investigated the interaction between E1_CTD and E12 using the isothermal titration calorimetry (ITC) and surface plasmon resonance (SPR) methods. SPR result showed a K_D value of 8.15 nM and ITC result showed a K_D value of 35.5 nM (Figures 2C and 2D), indicating that the E12 subunit exhibits a high affinity to the E1_CTD domain.

Figure 2. MPXV E12 exhibits a high binding affinity with E1_CTD. (A) Cartoon representation of MPXV E1_CTD-E12 complex structure, the binding interface was labeled with a red rectangle for key residues depicted with stick models. (B) Ribbon representations of crucial residues in the interface between E12 (green) and E1_CTD (orange). The residues were labeled and shown as sticks. The hydrogen bonds among interacted residues are shown in yellow dashed lines. (C) Measurements of the binding affinity between E1_CTD and E12 by ITC method. The upper panel represents the raw data, revealing the calorimetric response peaks during successive injections of E12 to the E1_CTD plot. The lower panel is the best fit of the raw data after subtracting the heat of dilution from the appropriate buffer blank. The solid line in the lower panel shows the best fit of the raw data to the one-set of sites binding model. (D) SPR assay showing the binding of E1_CTD and E12. (E) The gel-filtration profiles of the E1_CTD-E12 complex, E12, and different mutants. (F) The SDS-PAGE results of the E1_CTD-E12 complex, E12, and different mutants, corresponding to (E). M: protein marker.

To identify the critical residues for the interaction of the E1_CTD-E12 complex, we constructed six E1_CTD-E12 mutants, including F44A and R49A in E12, D587N, L575A, L586A, and double mutant F44A/L586A in E1_CTD-E12. The size exclusion chromatography results of E1_CTD-E12 mutants clearly showed that F44A, D587N, L575A, L586A, and double mutant F44A/L586A significantly decreased the binding of the E1_CTD and E12. However, these mutants did not destroy the E1_CTD-E12 formation, indicating that the interaction of E1_CTD and E12 is coordinated by a variety of amino acids (Figures 2E and 2F).

The solution structures of E12 and E1_CTD-E12 complex

To reveal how E12 bound and stimulated E1 activity, we expressed and purified E12 protein in E. Coli. The E12 was eluted out at the peaks of 83.5 mL on the gel-filtration column, corresponding to a mass of 33 kDa (Figure S2A). Next, the E12 structure was solved by the X-ray crystallography method with a resolution of 2.2  Å in the P3₁21 space group (Table 1). The asymmetric unit of the crystal structure consisted of one E12 molecule. A central feature of the E12 is a strip of parallel and anti-parallel β-strands (β1, β4, β5, β3, β2), which is stabilized by surrounding nine α-helices (Figure 3A). In order to explore whether there will be structural changes in E1 upon binding with E12, we superimposed the E12 to the E1_CTD-E12 structures and found that the overall folding of E12 is very similar in the structures; the root mean square deviation (RMSD) value is 0.544  Å, based on 253 pairs of Cα atoms (Figure 3B). The conformations of the α-helices and β-strands are well conserved, but E12 in the E1_CTD-E12 does show some conformational changes, especially between the Lys118 and Asn126. The density of Asn120 to Glu125 was not observed in any reported E1_CTD-E12 structures, which might be due to the side chain of His122 being too close to the side chain of Lys583 in E1_CTD (Figure 3B). We tried to construct mutants E1_CTD-E12 (H122D) and E1_CTD-E12 (K583A) to stabilize the E1_CTD-E12 complex structure and carried out crystallization trials. Fortunately, the E1_CTD-E12 (H122D)-SAM structure was solved by the X-ray crystallography method with a resolution of 3.3  Å in the P6₅22 space group (Table 1) and the small molecule SAM is clearly observed in the structure.

Figure 3. The solution structures of E12 and E1_CTD-E12 complex. (A) Cartoon presentation of the E12 structure, showing nine α-helices and five β-sheets. (B) Structural superimposition of the E1_CTD-E12 (orange and green) and E12 (purple). Residues His122 of E12 and Lys583 of E1_CTD are shown in stick models. (C) The experimental scattering curve of the MPXV E1_CTD-E12 complex. (D) The distance distribution function curve of the E1_CTD-E12 complex. (E) The crystal structure of the E1_CTD-E12 complex was fitted into the ab initio envelope obtained from SAXS. (F) The experimental scattering curve of MPXV E12. (G) The distance distribution function curve of E12. (H) The crystal structure of E12 was fitted into the ab initio envelope obtained from SAXS.

In addition, we investigated the solution status of the E12 and E1_CTD-E12 complex by the small-angle X-ray scattering (SAXS) and DLS methods (Figures 3C-3H and Table S1). All the proteins behaved well in the solution. The maximum dimension (Dmax) from distance distribution function p(r) for the E12 and E1_CTD-E12 complex were 67.5  Å and 100.4  Å, respectively (Figures 3C and 3F). When superimposed the E12 and E1_CTD-E12 complex crystal structures with the envelopes generated from SAXS data, the E12 and E1_CTD-E12 complex showed high similarities (Figures 3E and 3H).

Analysis of the E1_CTD-SAM binding model

To elucidate the molecular basis of MPXV mRNA cap N7 methyltransferase, we successfully obtained the E1_CTD-E12 (H122D)-SAM ternary complex structure. The SAM molecule is bound in the pocket consisting of four loops (NTD-α1, β1-α1, β2-α3, and α4-α5), forming a closed substrate binding site (Figure 4A). The directly interacting amino acids surrounding the SAM-binding region include Arg548, Leu549, Asn550, Lys573, Asp598, Asp620, Arg655, Gln678, and Tyr683. The adenosine moiety is stacked between conserved Tyr683 and Arg548. In addition, the main chain of Leu549 and Asn550 forms hydrogen bond (H-bond) interaction with the adenine N6 and N1, respectively, and the side chain of Asn550 forms H-bond interaction with the adenine N7. The O atom of the Gln678 and Asp598 forms an H-bond with the homocysteine of the AdoMet. Moreover, the -COOH of Asp620 and the -NH3 of Lys573 form polar contacts with -OH of the adenosine ribose and -COOH in the SAM, respectively (Figure 4A).

Figure 4. Comparison of the SAM binding pocket in MPXV mRNA capping MTase and other homologs. (A) The overall structure of the E1_CTD-E12 (H122D)-SAM ternary complex in cartoon presentation. The zoomed-in area shows the detailed contact analysis between SAM (cyan sticks) and MPXV E1_CTD (orange cartoon). The substrate SAM is shown in a stick model with an electron density map (2Fo-Fc, level = 1.0). (B) Structure comparison of MPXV mRNA capping MT E1_CTD (orange, PDB: 8ZE4) with VACV D1_CTD (pink, PDB: 2VDW). The adenosine nucleoside of SAM in E1 and SAH in D1 adopts a syn-conformation, respectively. (C) Structure comparison of RNA capping MTase members from MPXV (orange, PDB: 8ZE4), African Swine Fever Virus (ASFV, blue, PDB: 7D8U), and Encephalitozoon cuniculi (Ecm1, gray, PDB: 1RI3). The adenosine nucleoside of SAM in E1 (orange) of MPXV adopts a syn-conformation, whereas those in ASFV (blue) and Ecm1 (gray) adopt an anti-conformation. (D) Multiple sequence alignment of MPXV E1 with VACV D1 (PDB: 2VDW), pNP868RMT of ASFV (PDB: 7D8U), RMNT of Homo sapiens (PDB: 3BGV), and ABD1 of Ecm1 (PDB: 1RI3). Identical and similar residues among groups are shown in white font on a red background and in red font on a white background, respectively.

Superimposition of the MPXV E1_CTD-E12 (H122D)-SAM complex structure with the VACV D1_CTD-D12 complex (PDB ID: 2VDW) [11], we found poxvirus mRNA cap N7 methyltransferase are highly conserved in overall structures. The RMSD values between the E1_CTD-E12 and D1_CTD-D12 complexes were 0.430  Å among 276 Cα atoms, only showing some differences in the α linker and β linker regions (Figure 4B). Structure superimposing the MPXV E1_CTD-SAM complex with that of Encephalitozoon (PDB ID: 1RI3) and African swine fever virus ASFV (PDB ID: 7D8U) revealed similar features of core SAM-MT fold, whereas the additions to the core fold are quite differences (Figure 4C). The significant difference was that the MPXV E1_CTD had the N-terminal extension (Asp545-Arg562) before the α1 helix. The N-terminal peptide of VACV D1_CTD involved in SAM binding (closed conformation) and protected D1_CTD from trypsinization [11]. Another minor difference was that the MPXV E1_CTD had three helices αb (Phe788-Met805), αc (Pro809- Lys823), αd (Val829- Tyr835) between β6 and β7. The corresponding regions are involved in residues interacting with E12 to help form a stable E1_CTD-E12 complex structure (Figure 4C).

In addition, we compared the catalytic sites of E1_CTD-E12 with the catalytic sites of MTases from unrelated eukaryotes, prokaryotes, and viruses, namely Homo sapiens, Encephalitozoon, ASFV, and Mimivirus [13-15]. The amino acid sequences alignment of capping enzyme MTase domain found MTases from Homo sapiens, Encephalitozoon, ASFV and Mimivirus have a conserved SAM-binding motif (Gln/Asp-X-Gly-X-Gly-X-Gly). In contrast, the sequence of the poxvirus's SAM-binding domain (Ala-X-Asp-X-Gly-X-Gly) is slightly different (Figure 4D) [12]. Previous structural studies of VACV MTase showed that Asp598 is substituted for the usually highly conserved glycine, making the adenosine adopt the usual syn-conformation in the active site [11]. Also, the essential amino acids that interacted with SAM in poxviruses are not exactly conserved in other MTases (Figure 4D).

Structural basis for the allosteric effect of capping enzyme in poxviruses

We then performed the interaction assay between E1_CTD-E12 and SAM by ITC. The K_D value of 3.87 μM was determined using the one-site binding site model (Figure 5A). The mutated complex Δ(545-562) E1_CTD-E12 (deletion of Asp545-Arg562) shows no detectable binding affinity to SAM (Figure 5B), suggesting that the N-terminal extension (Asp545-Arg562) plays a vital role in the substrate binding process, consisting with the observation in the complex structure (Figure 4A). In addition, the E1_CTD also shows no detectable binding ability to SAM (Figure 5C), indicating that the binding of the small subunit E12 may lead to structure changes in the large subunit E1_CTD.

Figure 5. Binding characteristics of MPXV mRNA capping MTase with its substrate SAM. (A-C) The binding affinity of SAM with E1_CTD-E12 complex (A), Δ(545-562) E1_CTD-E12 complex (B), and E1_CTD (C), respectively. Baseline-subtracted raw ITC data for injections of SAM are indicated in the upper panels of each ITC profile shown. The peaks normalized to the ligand/protein molar ratio were integrated, as shown in the bottom panels. The solid line in the lower panel shows the best fit of the raw data to the one-set of sites binding model. (D) Structural comparison of AlphaFold2-predicted E1 (wheat) with solved E1_CTD-E12 complex (green, PDB: 8Y2Z), shown in wheat and green, respectively. (E) Structural comparison of AlphaFold2-predicted E1 (wheat) with VACV D1-D12 complex (cyan, PDB: 4CKC), shown in wheat and cyan, respectively. The red arrow indicates the rotation of the MTase domain upon binding with E12.

Next, we compared the structure of E1_CTD-E12 with the structure of E1 predicted by AlphaFold2. Structural superposition showed that the AlphaFold2-predicted MTase domain structure in full-length E1 is very similar to the resolved E1_CTD-E12 structure, with an RMSD value of 0.451  Å among 269 Cα atoms (Figure 5D). Interestingly, when superimposition the AlphaFold2-predicted E1 with the VACV D1-D12 complex (PDB ID: 4CKC), the N7 MTase domain is found to be close to the GTase domain with an RMSD value of 0.667  Å among 514 Cα atoms (Figure 5E). The amino acid residues (Ile528-Asn544) at the junction of MTase and GTase were not observed in the D1-D12 complex structure, which might be due to this region being more flexible. Previous studies indicated that poxviruses encode a small subunit that stimulates MTase activity. In the Vaccinia virus, the small subunit D12 is essential; it lacks any MT activity but can stimulate the weak intrinsic MT activity of D1 by 30-50 folds [16]. The previous study and our alignment results confirmed that the binding of the small subunit E12 of the MPXV capping enzyme may lead to an allosteric reaction of the MTase in the large subunit E1 [11]. This reaction makes the overall structure more compact, and the substrate binding sites in the GTase and MTase domains are closer, which may be more conducive to the enzymatic reaction and the transfer of substrate in different subunits.

Discussion

The mRNA cap structure was discovered in the 1970s, and numerous studies have led to a deeper understanding of the biological role of mRNA cap structures. The mRNA cap structure plays an important role in protein translation and innate immunity [17]. Therefore, the virus can produce its own mRNA cap structure, which can effectively synthesize viral proteins and evade the innate immune response of host cells [18]. Due to the wide variety of viruses, mRNA capping has various mechanisms. The GTase and RTPase activities of Chlorella virus (dsDNA) are located in different proteins [19]. Vaccinia virus (dsDNA) and Bluetongue virus (dsRNA) encode multifunctional proteins that produce cap 0 and cap 1 RNA, respectively [20]. Sars-cov-2 ((+) ssRNA) can use the single-stranded nucleic acid binding protein (nsp9) in its transcription and replication complex as a medium to mediate the novel capping process [21]. Flaviviruses ((+) ssRNA) can tether the activities of GTase and MTase to their RNA-dependent RNA polymerase (RdRp) [22]. Alphaviruses ((+) ssRNA) reorder the enzymatic steps of RNA capping. GTP is first modified to become m7GMP and then transferred to processed ppRNA to form m7G-cap [23]. Currently, only the full-length protein structure of vaccinia virus capping enzymes has been fully resolved. In the structure, GTP and SAH identify the GTase and MTase active sites [9]. However, how the substrate moves from one active site to the next and whether conformational changes occur during these transitions have not been observed in any crystalline structure.

We solved the MPXV MT subcomplex E1_CTD-E12 crystal structures and the small subunit E12. By analyzing the interaction interface between E1 and E12 in E1_CTD-E12, it was found that the large and small subunits interacted with each other mainly through hydrophobic interaction (Figures 1 and 2). In the binding process between E12 and E1, the loop of small subunit E12 conformationally changed due to the steric hindrance between the side chain of His122 and Lys583 (Figure 2B). The substrate binding sites in the domain of the MPXV MTase were found to be remarkably conserved by structural comparison (Figures 4A and 4B). However, there were still significant differences between the poxvirus MTases and other MTases. The first is amino acid residues change in the conserved SAM binding site (Figure 4D). Second is the presence of the small subunit E12 that can stimulate E1-MTase activity, which is also unique to poxviruses. The ITC result of Δ(545-562) E1_CTD-E12 and E1_CTD with SAM suggest the essential role of the N-terminal extension (Asp545-Arg562) and the small subunit E12 in substrate binding process (Figures 5A-5C). Despite the high homology between VACV D1_CTD-D12 and MPXV E1_CTD-E12, the mutation study showed that the residue Tyr202 of MPXV (corresponding to residue Cys202 in VACV) can dramatically change the complex homogeneity in gel-filtration profiles (Figure S2C). Moreover, we predicted the full-length E1 protein with the help of AlphaFold2 and found that the structure of the full-length E1 protein was significantly different from that of the full-length protein bound to the small subunit. The differences mainly occurred in the MTase domain with a large allosteric reaction (Figures 5D and 5E). We speculate that the binding of small subunits may make the structure of large subunits more compact, bringing the active sites of different domains closer together, which is more conducive to the movement of substrates between active sites and the occurrence of enzymatic active reactions (Figures 5D and 5E). Overall, the structure of poxvirus capping enzymes is highly conserved, suggesting an essential role in the viral life cycle. Therefore, it is promising to use MPXV capping enzymes as targets for anti-poxvirus drug discovery.

Methods and Materials

Plasmid construction

The full-length E12 (UniProt ID: Q8V4X6) gene and the full-length E1 (UniProt ID: Q8V4Y7) were synthesized by Tsingke, the target genes were amplified by the polymerase chain reaction and subcloned into the pSMT3 vector with a His-SUMO tag. The E1_CTD (residues 545-844) were subcloned into the pSMT3 vector with a His-SUMO tag. The MTase subcomplex E1_CTD-E12 and Δ(545-562) E1_CTD-E12 were subcloned into the pETDuet vector with a 6xHis tag. The E1_CTD-E12 mutants were generated by whole-plasmid PCR in an 18-cycle reaction with steps at 95 °C for 15 s, 65 °C for 30 s, and 72 °C for 3 min per cycle. After digestion with the enzyme DpnI, the PCR products were transformed into E. coli DH5a cells. The positive constructs were determined by DNA sequencing. Mutants F44A, R49A, H122D, Y202C and Y202A in E12, K583A, D587N, L575A, L586A, and double mutant F44A/L586A in E1_CTD-E12 were constructed according to the standard QuikChange Site-Directed Mutagenesis protocol (Stratagene, United States), confirmed by DNA sequencing finally.

Protein expression and purification

The wild-type and mutants of E12, E1_CTD, E1_CTD-E12, and Δ(545-562) E1_CTD-E12 were transformed into E. coli BL21 (DE3) cells for protein expression. E. coli cells were cultured in LB medium at 37 °C. When the OD₆₀₀ reached 0.6-0.8, isopropyl β-D-1-thiogalactopyranoside (IPTG, 0.2 mM) was added to cells to induce protein expression at 18 °C for 18-20 h. Then, the cells were harvested by centrifugation at 6000 rpm for 15 min at 4 °C. The cell pellets were resuspended in a lysis buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10 mM imidazole pH 8.0) and disrupted using a high-pressure homogenizer (JNBIO, China). The cell debris was removed by centrifugation at 17,000 rpm for 60 min at 4 °C. Next, the supernatant was purified using a His Trap^TM HP column (GE Healthcare) and was eluted in a buffer (50 mM Tris-HCl, 500 mM NaCl, 500 mM imidazole pH 8.0). The His-sumo tag was removed by the ULP1 enzyme cleavage, followed by additional Ni-NTA affinity chromatography as previously described. The protein was further purified by gel filtration using a Superdex200 16/600 column (GE Healthcare) in a buffer (20 mM Tris-HCl pH 8.0, 100 mM NaCl).

Crystallization and data collection

The E12 and E1_CTD-E12 complexes were crystallized using the sitting-drop vapor-diffusion method by mixing 0.2 μL protein and 0.2 μL reservoir solution at 18 °C. E12 crystals were grown in a reservoir solution containing 2.0 M ammonium sulfate, 0.1 M CAPS/sodium hydroxide pH 10.5, 0.2 M lithium sulfate with a protein concentration of 15 mg/mL. For the E1_CTD-E12 complex, crystals were grown in a reservoir solution containing 1600 mM sodium phosphate monobasic/400 mM potassium phosphate dibasic, 100 mM sodium phosphate dibasic/citric acid pH 4.2 with a protein concentration of 20 mg/mL. For the E1_CTD-E12 (H122D)-SAM complex, a mixture of E1_CTD-E12 (H122D) and SAM (Aladdin) at a molar ratio of 1:5 was prepared and incubated on ice for 1 h before performance of crystallization experiments. The E1_CTD-E12 (H122D)-SAM complex crystals were grown in a reservoir solution containing 0.1 M sodium acetate trihydrate pH 4.6, 2.0 M sodium formate with a protein concentration of 20 mg/mL. All crystals were briefly soaked in a cryoprotectant solution consisting of 25% (v/v) glycerol dissolved in their corresponding mother liquors before being flash-cooled directly in a liquid-nitrogen stream at 100 K. The X-ray diffraction data were collected at the beamlines of the Shanghai Synchrotron Radiation Facility. The process of data integration, scaling, and merging was performed using the HKL3000 package and the XDS program [24].

Structure determination and refinement

The crystal structure of MPXV E12 was determined by molecular replacement using AlphaFold2-predicted E12 structure (https://alphafold.ebi.ac.uk) as a search model. E1_CTD-E12 and E1_CTD-E12 (H122D)-SAM complex structures were determined by molecular replacement using the crystal structure of the VACV mRNA MTase structure (PDB: 2VDW) as the search model. Cycles of refinement and model building were carried out by using Phenix and COOT programs [25, 26]. All of the structures were displayed and analyzed using the PyMOL program. The collected data and refinement statistics are summarized in Table 1.

Dynamic light scattering (DLS) measurement

The DLS data were collected using the DYNAMICS software from DynaPro NanoStar (Wyatt Technology), operating at a light source wavelength of 658 nm and a fixed scattering angle of 90°. E12 and E1_CTD-E12 were diluted to 1 mg/mL with a buffer containing 20 mM Tris-HCl (pH 8.0) and 100 mM NaCl at 25 °C.

Small angels X-ray scattering (SAXS)

The SAXS data of E12 and E1_CTD-E12 were collected at beamline BL19U2 of the Shanghai Synchrotron Radiation Facility, with a radiation wavelength of 1.03  Å. The protein samples were prepared at 0.1-2 mg/mL concentrations in the buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl). The samples were measured in triplicate, and the sample measurements were adjusted by subtracting the scattering from the buffer alone. Data were analyzed using the software package BioATSAS (https://www.embl-hamburg.de/biosaxs/). The scattering images were averaged and subtracted from the buffer scattering images. Then, using the indirect Fourier transform method, the Rg was estimated. The distribution function p(r) was calculated from the parameter Dmax. The SAXS envelopes of the E12 and E1_CTD-E12 complex were built by GASBOR, as previously reported [25].

Isothermal titration calorimetry (ITC) assay

The ITC measurement was performed using MICROCAL PEAQ-ITC (Malvern) at 25 °C. 20 μM E1_CTD was titrated with 100 μM E12. 20 μM E1_CTD, 20 μM E1_CTD-E12, and 20 μM Δ(545-562) E1_CTD-E12 were titrated with 200 μM SAM, respectively. The fitting curve with a single-binding-site model was performed by the ITC data analysis module provided by the manufacturer.

Surface plasmon resonance (SPR) assay

The SPR binding experiments were carried out using a Biacore 8 K (Cytiva) at 25 °C in the single-cycle mode. All proteins for SPR assays were exchanged to PBST buffer (10 mM Na₂HPO₄, 2 mM KH₂PO₄, pH 7.4; 137 mM NaCl; 2.7 mM KCl, and 0.05% Tween-20). The E1_CTD was immobilized on the CM5 chip and gradient concentrations of E12 (from 781.25 to 25000 nM) was then used to flow over the chip surface at 30 μl min^-1 and the real-time response was recorded. After each cycle, the sensor surface was regenerated using 10 mM glycine-HCl (pH 2.5). KD values for SPR assays were calculated using a 1:1 binding fit model with the software BIAevaluation version 4.0.8 (Cytiva). The SPR binding assays were conducted three times, and the proteins from different batches were used in the replication experiments.

Author Contribution Statement

J.L. and J.D. conceived and designed this study. A.C., N.F., Z.Z., Y.W., Y.S., Y.Z., L.Z., and G. Z. performed research and analyzed data; J.L. and A.C. wrote the manuscript. All authors discussed the results and commented on the manuscript.

Competing Financial Interests

The authors declare no competing interests.

Acknowledgments

We thank the staff of the Shanghai Synchrotron Radiation Facility beamlines BL02U1, BL10U2, BL18U1, BL19U1, and BL19U2 for help with X-ray data and SAXS data collection. This work was supported by grants from the National Key Research and Development Project of China (2021YFC2301500) and the National Natural Science Foundation of China (32161160323) to J.L.

Data Availability

The coordinates and structural factors generated in this study have been deposited in the Protein Data Bank with accession codes 8Y2Y (E12) [https://www.rcsb.org/structure/8Y2Y], 8Y2Z (E1_CTD-E12 complex) [https://www.rcsb.org/structure/8Y2Z], and 8ZE4 (E1_CTD-E12 (H122D)-SAM complex) [https://www.rcsb.org/structure/8ZE4]. The SAXS data and model have been deposited in the SASBDB database with accession codes SASDUL2 for E12 and SASDUM2 for the E1_CTD-E12 complex. Source data are provided in this paper.