Oomycete Nudix effectors display WY‐Nudix conformation and mRNA decapping activity

Phytopathogens secrete effectors to promote colonization of their host plants. These effectors modulate host immune responses by interfering with various pathways, but little is known about their biochemical activities. Several recent reports indicate that some phytopathogen effectors have Nudix hydrolase activity, including Avr3b from Phytophthora sojae (Dong et al., 2011), AvrM14 from Melampsora lini (McCombe et al., 2023), and RipN from Ralstonia solanacearum (Sun et al., 2019). Nudix hydrolases contain the conserved Nudix motif GxxxxxExxxxxxxREUxEExGU (where x is any amino acid and U is an aliphatic, hydrophobic residue), catalyzing the hydrolysis of nucleotide derivatives, including nucleoside diphosphates and triphosphates, nucleotide sugars, nicotinamide adenine dinucleotide + hydrogen (NADH), coenzyme A, and messenger RNA (mRNA) caps (McLennan, 2006). Although effectors with Nudix hydrolase activity have been identified in different phytopathogens, the phylogenetic distribution, structural characteristics, and substrate preference of most such effectors are currently unclear. To investigate the phylogenetic distribution of Nudix effectors in various pathogens, we conducted a phylogenetic analysis of 229 Nudix proteins from 8 oomycetes, 10 fungi, and 2 bacteria (Figure S1). Notably, all oomycete Nudix effectors clustered on a single branch (Figures 1A, S2), while the fungal Nudix effectors were located on distant branches of the phylogenetic tree. Two bacterial effectors clustered together with some Nudix proteins without a signal peptide. Based on these independent evolutionary characteristics, we speculated that Nudix effectors in oomycetes may exhibit specific structural features and functions. Nudix effectors from oomycetes take on a WY-Nudix conformation and have messenger RNA (mRNA) decapping activity (A) Phylogenetic tree of 229 Nudix proteins from selected species. The oomycete, fungal, and bacterial Nudix effectors are highlighted in red, green, and brown, respectively. (B) Overall structure of Avr3b57–303 (coordinates and structure factors have been deposited in the Research Collaboratory for Structural Bioinformatics Protein Data Bank (PDB) with the accession code PDB: 8ZJB). The WY-like domain (deep salmon) consists of three α-helices (α1, α2, and α3). The Nudix (purple) domain consists of three α-helices (α4, α5, and α6) and five β-strands (β1, β2, β3, β4, and β5). The loop region is shown in green. (C) Structure of the Avr3b WY motif. The key residues 70F and 124F are shown as sticks. (D) Structure superposition of the Avr3b N-terminal domain (deep salmon) and Avr3a11 (gray). The key residues in the WY motifs are shown as sticks. (E) Structure of the Avr3b Nudix domain. The key Nudix motif is indicated in light orange. (F) Structure superposition of the Avr3b C-terminal domain (purple) with the Saccharomyces cerevisiae Nudix hydrolase ScDDP1 (diadenosine and diphosphoinositol polyphosphate phosphohydrolase 1) (yellow) (with a root mean square deviation (RMSD) of 1.5 Å). (G) m7G-RNA hydrolysis assay of Avr3b and various mutants. The peaks in absorbance (260 nm) for the hydrolyzed products of m7G-RNA are indicated by red arrows. The data are representative from three independent experiments. (H) Structure of Pp_07435 from Phytophthora parasitica as predicted by AlphaFold2. The RxLR motif, WY-like motif, loop, Nudix domain, and key Nudix motif are colored in blue, cyan, green, gray, and deep salmon, respectively. (Pp_07435 displayed here is a representative oomycete Nudix effector. The structures of other oomycete Nudix effectors are shown in Figure S4.) (I) Oomycete Nudix effectors show a conserved WY-Nudix structural fold. The diagrams of effectors are drawn to scale. The different structural regions of Nudix effectors are indicated with different colors. (J) Hydrolysis of m7G-RNA by oomycete Nudix effectors (Pi_05846, Pp_07435, Pc_024430, Pr_86166, and Hpa_811157). The peaks in absorbance (260 nm) for the hydrolyzed products of m7G-RNA are indicated by red arrows. The data are representative from three independent experiments. (K) Phytophthora capsici infection of Nicotiana benthamiana leaves expressing green fluorescent protein (GFP), Avr3b, Pi_05846, Pp_07435, Pc_020430, Pr_86166, or Hpa_811157. Infection lesions were observed under UV light illumination 48 h after inoculation with P. capsici strain LT263. The diameters of disease lesions were measured and are shown as means ± SE from three independent experiments. Different lowercase letters indicate significant differences (P < 0.01, Duncan's multiple range test). We used the P. sojae effector Avr3b as an example to explore the structural features of oomycete Nudix effectors. Avr3b is a typical RxLR (Arg–any amino acid–Leu–Arg) effector with a conserved RxLR motif at its N terminus (Dong et al., 2011). To determine the crystal structure of Avr3b, we produced Avr3b fragments of different lengths in Escherichia coli BL21 (DE3) and purified the proteins. We tested various conditions for crystallization, yielding a crystal structure for Avr3b57–303, the effector domain of Avr3b, at a resolution of 3.19 Å. The phase was determined by single-wavelength anomalous diffraction, and the final model was refined to Rwork and Rfree values of 23.65% and 29.51%, respectively (Table S1). The overall structure of Avr3b57–303 contained two regions, with a long random coil connecting the N- and C-terminal regions (Figure 1B; Protein Data Bank (PDB) ID: 8ZJB). The first three α-helices from the Avr3b57-303 N-terminal region formed a hydrophobic core, resulting from an interaction between the Phe-70 and Phe-124 residues, exhibiting a structure similar to that of a Trp-Tyr (WY) motif fold (Figure 1C). The WY motif fold is a ubiquitous structural unit in oomycete RxLR effectors that comprises three α-helices and forms a hydrophobic core via two hydrophobic residues, usually Trp and Tyr (Win et al., 2012). When we aligned the N-terminal region structure of Avr3b57–303 to some typical WY motif structures, including Avr3a11 (PDB ID: 3ZR8) and PexRD2 (PDB ID: 3ZRG), we observed they all contained a hydrophobic core composed of two aromatic amino acids, although they showed slight differences in their structural folds (Figures 1D, S3). These results highlight the diversity of WY motif folds in RxLR effectors. The C-terminal domain of Avr3b57–303 showed a common αβα Nudix fold. Five β-strands (β1, β2, β3, β4, and β5) were sandwiched between two antiparallel α-helices (α5 and α6) on one side and helix α4 (oriented perpendicular to the α5 and α6 helices) on the other side (Figure 1E). The Nudix hydrolase motif (amino acids 206–228) corresponded to helix α4 and the linker in this structure and is colored light orange in Figure 1E. We searched for structural homologs of Avr3b57–303 in the PDB using the Dali server, which revealed several structures with high similarity to the Avr3b57–303 C-terminal region, corresponding to the Nudix hydrolase domain of Avr3b. However, we obtained no homologous structures for the WY motif fold at the N-terminal region. The structure most similar to Avr3b is that of budding yeast (Saccharomyces cerevisiae) Nudix hydrolase ScDDP1 (diadenosine and diphosphoinositol polyphosphate phosphohydrolase 1, PDB ID: 7AUK) (Figure 1F). ScDDP1 is reported to hydrolyze multiple substrates, including diphosphoinositol-polyphosphates, diadenosine polyphosphates (ApnA), and the 5′ m7G mRNA cap (Song et al., 2013; Márquez-Moñino et al., 2021). Avr3b produced in E. coli was previously reported to have no hydrolytic activity toward many substrates, such as NAD, NADH, NADPH, adenine diphosphate ribose (ADPR), Ap4A, flavin adenine dinucleotide (FAD), or coenzyme A, while Avr3b produced in plants can hydrolyze NADH and ADPR (Dong et al., 2011). Recently, Avr3b has been found to hydrolyze 2′,3′-cAMP, a newly characterized second messenger that participates in plant immune responses in vitro (Yu et al., 2022). We therefore hypothesized that Avr3b might be similar to ScDDP1 and hydrolyze multiple substrates. Since ScDDP1 can hydrolyze the 5′ mRNA cap, we investigated whether Avr3b could also have this activity. We synthesized m7G-RNA in vitro and incubated it with recombinant purified Avr3b. Indeed, recombinant ScDDP1 and Avr3b each displayed mRNA decapping activity, as evidenced by a clear peak for absorbance at 260 nm of fractions separated by high-performance liquid chromatography (HPLC) (Figure 1G). Importantly, the variant Avr3b-4Q, harboring mutations in the key catalytic sites of the Nudix domain (Dong et al., 2011), could not hydrolyze the m7G-RNA cap, suggesting that the observed mRNA decapping activity of Avr3b is dependent on its Nudix hydrolase activity (Figure 1G). We generated another variant, Avr3b∆WY, lacking the WY motif; Avr3b∆WY possessed an mRNA decapping activity similar to that of intact Avr3b (Figure 1G). These results indicate that the WY motif is not essential for the mRNA decapping activity of Avr3b. Based on our structural analysis, we concluded that Avr3b57–303 shows a distinct WY-Nudix fold not observed in previously reported RxLR effectors (Win et al., 2012). Since Avr3b and other oomycete Nudix effectors clustered along one branch of the phylogenetic tree, we speculated that their structures might be similar. We thus predicted the structures of these Nudix effectors via AlphaFold2. Indeed, all oomycete Nudix effectors exhibited a typical Nudix motif fold in their C-terminal region (Figures 1H, S4). The N-terminal region of these oomycete Nudix effectors consisted of three α-helices that resemble the WY motif fold, although they also showed diversity. Moreover, we identified a long loop between the WY-like motif and the Nudix domain as a common feature of these Nudix effectors (Figures 1H, S4). These results suggest that oomycete Nudix effectors have conserved WY-Nudix structural folds (Figure 1I). To explore whether these predicted effectors possess Nudix hydrolase activity, we successfully cloned one effector from each of five different pathogens: Phytophthora infestans effector Pi_05846, Phytophthora parasitica effector Pp_07435, Phytophthora capsici effector Pc_020430, Phytophthora ramorum effector Pr_86166, and Hyaloperonospora arabidopsidis effector Hpa_811157. We produced these effectors in E. coli and purified them for enzymatic activity analysis. With the exception of Pr_86166, which had hydrolytic activity toward FAD and NADH, the other four effectors failed to hydrolyze NAD, NADH, ADPR, Ap4A, or FAD in vitro (Figure S5). However, two effectors (Pi_05846 and Pp_07435) exhibited mRNA decapping activity similar to Avr3b, while the other three effectors (Pc_020430, Pr_86166, and Hpa_811157) did not (Figure 1J). We also infiltrated constructs encoding each of these oomycete Nudix effectors fused to a green fluorescent protein (GFP) tag into Nicotiana benthamiana leaves and purified each protein with GFP Trap-A beads for mRNA decapping analysis. Consistent with the results obtained with the recombinant purified proteins above, the GFP-tagged effectors Avr3b, Pi_05846, and Pp_07435 displayed mRNA decapping activity when produced in plant cells, whereas Pc_020430, Pr_86166, and Hpa_811157 did not (Figure S6). These results suggest differences in substrate preference among oomycete Nudix effectors, despite their structural similarity. To assess their contribution to pathogen resistance, we inoculated N. benthamiana leaves that had been infiltrated with constructs encoding each Nudix effector with P. capsici for a pathogenicity test. The disease regions formed on leaves expressing each Nudix effector construct were significantly larger than those on leaves harboring the empty vector control, indicating that these effectors promote pathogen infection (Figure 1K). In this study, we not only revealed the independent evolutionary characteristics of oomycete Nudix effectors, but also discovered they adopt a conserved WY-Nudix conformation. The M. lini effector AvrM14 appears to only contain a typical Nudix fold (McCombe et al., 2023). A structural prediction for other fungal Nudix effectors by AlphaFold2 revealed they harbor a typical Nudix fold similar to that of AvrM14 (Figure S7). We performed the same structural prediction for two bacterial Nudix effectors: RipN consisted of three domains with a Nudix domain between two domains of unknown function, (Figure S8) and XCV0537 from Xanthomonas euvesicatoria pv. vesicatoria contained two domains in addition to a type III secretion system (T3SS) peptide region (Figure S8). Thus, these oomycete, fungal, and bacterial Nudix effectors display different structural characteristics (Figures 1I, S9). During preparation of this manuscript, a structure for the Magnaporthe oryzae Nudix effector MoNudix (named MGG_14344/MGG_14156 in this study) was reported (PDB ID: 8SXS, https://doi.org/10.1101/2023.11.14.566975). The MoNudix structure contains a typical Nudix fold that is similar to the structure predicted by AlphaFold2 (Figure S10), supporting the accuracy of the prediction presented in this study for fungal Nudix effectors. We showed here that multiple oomycete Nudix effectors, including Avr3b, exhibit mRNA decapping activity in vitro and in vivo. As an important modification of mRNA, the m7G cap plays a key role in eukaryote gene expression. Recently, the fungal Nudix effector AvrM14 was reported to specifically hydrolyze the m7G mRNA cap in vitro and disturb plant immune pathways (McCombe et al., 2023). Based on the differences in structure and substrate diversity between AvrM14 and Avr3b, we speculate that these two effectors might regulate plant immunity in different ways. Given that Avr3b may mimic a plant Nudix hydrolase to perform virulence function (Dong et al., 2011), we compared the structure of Avr3b to that of Arabidopsis thaliana Nudix hydrolase NUDIX HYDROLASE HOMOLOG 7 (NUDT7; PDB ID: 4ZBP), which is a negative regulator of plant immunity; Avr3b and AtNUDT7 have similar structures in their Nudix domains, with a root mean square deviation (RMSD) of 2.84 Å (Figure S11). We produced and purified AtNUDT7 from E. coli for mRNA decapping assays. Recombinant AtNUDT7 could not hydrolyze the m7G-RNA cap (Figure S12). Since Avr3b and AtNUDT7 can both hydrolyze NADH, ADPR, and 2′,3′-cAMP, Avr3b might imitate AtNUDT7 in hydrolyzing these substrates to promote pathogen infection (Dong et al., 2011; Yu et al., 2022). Avr3b may also mimic other plant Nudix hydrolases that display mRNA decapping activity, such as AtNUDT27, to cause virulence (Mititelu et al., 2023). However, the role of AtNUDT27 in plant immunity is currently unclear. Further research is needed to determine which plant Nudix hydrolase(s) with mRNA decapping activity participates in plant immunity. How these oomycete Nudix effectors regulate host immunity through mRNA decapping activity to promote pathogen infection will be the focus of future research. A previous study showed that Avr3b expressed in plant displays hydrolysis activity on NADH and ADPR, while Avr3b produced by E. coli does not (Dong et al., 2011). Further research determined that Avr3b needs to be processed by plant prolyl-peptidyl cis-trans isomerases (PPIase) to hydrolyze NADH and ADPR, suggesting that Avr3b may undergo conformation changes in plants (Kong et al., 2015). In this study, we determined the structure of Avr3b produced in E. coli and demonstrated that recombinant Avr3b proteins produced in E. coli and N. benthamiana display mRNA decapping activity. However, due to the lack of the structure of Avr3b produced in plants, we are currently unable to compare the structural differences of Avr3b produced in E. coli and plants. Whether there are multiple conformations of Avr3b within plant cells should be investigated and determined in the future. Moreover, we established that the oomycete effectors Avr3b, Pi_05846, and Pp_07435 display mRNA decapping activity, while three other oomycete Nudix effectors (Pc_020430, Pr_86166, and Hpa_811157) do not. The chains and motifs next to Nudix motif usually determine the substrate specificity of Nudix hydrolases (McLennan, 2006). Although oomycete Nudix effectors display similar structures in their Nudix domains, there may be substantial changes in their binding sites to the substrate, the position of their catalytic residues, and/or the number of divalent ions involved. We have demonstrated the importance of four glutamate residues in the Nudix motif of Avr3b for mRNA decapping activity in this study. Further research is necessary to determine the structure of Avr3b in complex with its substrate and metal ions to elucidate its catalytic mechanism. In addition, the substrate preference and catalytic mechanisms of other oomycete Nudix effectors also need to be investigated. We thank the staff of beamlines BL17U1 and BL19U1 at the Shanghai Synchrotron Radiation Facility for assistance during data collection. This research was supported by grants from the National Key Research and Development Program of China (2022YFF1001500), National Natural Science Foundation of China (31721004, 32202261) and Natural Science Foundation of Jiangsu Province (BK20210156). The authors declare no conflicts of interest. Y.W., W.X., F.L., and B.G. designed the research. B.G., Q.H., B.W., and H.W. performed the research. B.G., W.X., Q.H., D.Y., S.D., G.K., and C.H. analyzed the data. B.G. and Q.H. wrote the manuscript. All authors read and approved the contents of this paper. Additional Supporting Information may be found online in the supporting information tab for this article: http://onlinelibrary.wiley.com/doi/10.1111/jipb.13712/suppinfo Figure S1. The number of Nudix proteins from different pathogens Figure S2. Phylogeny of 229 Nudix proteins from oomycetes, fungi, and bacteria Figure S3. Structural alignment of the Avr3b N-terminal domain with PexRD2 Figure S4. Oomycetes Nudix effector structures predicted by AlphaFold2 Figure S5. Enzymatic analysis of five oomycete Nudix effectors on different substrates Figure S6. Hydrolysis of m7G-RNA by oomycete Nudix effectors expressed in Nicotiana benthamiana Figure S7. Fungal Nudix effector structures predicted by AlphaFold2 Figure S8. Bacterial Nudix effector structures predicted by AlphaFold2 Figure S9. Diagram of fungal and bacterial Nudix effector structures Figure S10. Structural alignment of MGG_14344/MGG_14156 with MoNudix Figure S11. Structure alignment of the Avr3b Nudix domain with AtNUDT7 Nudix domain Figure S12. Analysis of m7G-RNA decapping by Arabidopsis Nudix hydrolase AtNUDT7 Table S1. Avr3b57-303 X-ray data collection and refinement statistics Table S2. Sequences and accession IDs of all Nudix hydrolases in this study Table S3. Primers used in this study Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.