A Colletotrichum-unique effector with the Cx11NC motif enhances plant NDPK2 kinase activity to suppress plant immunity

Abstract

Colletotrichum fungi cause destructive diseases among a wide range of hosts worldwide. We found that effector CfEC92 from C. fructicola specifically binds ATP through an unidentified ATP-binding domain, leading to changes in the protein secondary structure. The residues Cys²⁶, Asn³⁸, and Cys³⁹ were critical for ATP binding with CfEC92, and mutations at these sites impaired the ability to suppress host immunity. CfEC92 interacted with MdNDPK2, a negative immune regulator in apple. The CfEC92-ATP complex altered the conformation of MdNDPK2, enhancing its affinity for ATP, and further increasing its autophosphorylation and kinase activity. The activated MdNDPK2 phosphorylated MdMPK3 to suppress host immunity. Homology and functional tests showed that the Cx₁₁NC motif was highly conserved among Colletotrichum species, suggesting that CNC effectors represent a class of broad-spectrum virulence factors. Our findings revealed a mechanism by which Colletotrichum effectors cooperate with helper ATP to promote target protein phosphorylation and suppress host immunity.

Fig. 1.. CfEC92 interacted with the NDK domain of MdNDPK2.

(A) Y2H confirmed the interaction between CfEC92 and MdNDPK2. DDO: SD/-Leu/-Trp; QDO/X/A: SD/-Ade/-His/-Leu/-Trp + X-α-gal + AbA. (B) BiFC assay showed the interaction of CfEC92 with MdNDPK2 in N. benthamiana leaves. Scale bars, 50 μm. (C) Pull-down assay demonstrated that CfEC92 interacted with MdNDPK2 in vitro. (D) Co-IP assay showed that CfEC92 interacted with MdNDPK2 in vivo. Actin as an internal control indicates equal loading. (E) Y2H assay showed that the CfEC92 interacted with all tested mutants (MdNDPK-t1, MdNDPK-t2, and MdNDPK-t3) except for MdNDPK-t4, which lacked the NDK domain. (F) BiFC assay confirmed that CfEC92 interacted with mutants of MdNDPK2 containing the NDK domain in N. benthamiana. Scale bars, 20 μm.

Fig. 2.. MdNDPK2 negatively regulates plant resistance.

(A) RT-qPCR assay of MdNDPK2 expression patterns during the C. fructicola infection. The MdUBQ gene was used as an internal reference. Values represent the mean ± SD of three independent biological replicates. Asterisks represent significant differences (*P < 0.05, Student’s t test). (B) Symptoms on N. benthamiana leaves with silenced NbNDPK2 or overexpressed MdNDPK2. (C) Lesion diameters caused by P. capsici infection in N. benthamiana leaves with silenced NbNDPK2 and overexpressed MdNDPK2. (D) Stable overexpression of MdNDPK2 weakens the resistance of (D) calli (n = 4) or (E) apple leaves (n = 6) to C. fructicola. (F) Lesion diameters were measured at 3 dpi. RT-qPCR analysis confirmed MdNDPK2 expression levels in transgenic plants. (G) Symptoms on apple fruit with overexpressed and silenced MdNDPK2 after inoculation with WT or ΔCfEC92-27. (H) Lesion diameters caused by WT or ΔCfEC92-27 infection in apple fruit overexpressing or silencing MdNDPK2. (I) Symptoms on N. benthamiana leaves with silenced NbNDPK2 and overexpressed CfEC92 after inoculation with P. capsici. Lesion diameters caused by P. capsici infection of N. benthamiana leaves overexpressing CfEC92. tGFP, TRV2-GFP. Values are mean ± SD of three independent biological replicates. Asterisks represent significant differences (*P < 0.05, Student’s t test). In (C), (D), (F), and (H), values are mean ± SD of three independent biological replicates, and different letters represent significant differences [one-way analysis of variance (ANOVA), Tukey test].

Fig. 3.. MdNDPK2 enhances phosphorylation of MdMPK3 relying on its autophosphorylation.

(A) In vitro kinase assay showed that the mutant MdNDPK2^H201A could not undergo autophosphorylation. Coomassie brilliant blue (CBB) staining shows protein loading. (B) Symptoms observed in apple calli overexpressing MdNDPK2 and MdNDPK2^H201A after inoculation with C. fructicola, n = 3. (C) Y2H assays to detect MdNDPK2 interaction with MdMPK3. (D) Symptoms in calli overexpressing MdMPK3 after inoculation with C. fructicola. (E) Lesion diameters in calli overexpressing MdNDPK2 and MdNDPK2^H201A after C. fructicola inoculation. RT-qPCR confirmed the relative expression of MdNDPK2 and MdNDPK2^H201A in the respective calli. Values represent the mean ± SD of three independent biological replicates, and different letters represent significant differences (one-way ANOVA, Tukey test). (F) RT-qPCR analysis of MdMPK3 relative expression in calli overexpressing MdNDPK2 or GFP (control). (G) RT-qPCR analysis of the expression levels of immunity-related genes (MdNPR1, MdPR1, MdPR5, and MdWRKY17) in calli overexpressing MdNDPK2. (H) SA concentrations were detected by enzyme-linked immunosorbent assay kit in calli overexpressing MdNDPK2. In (F) to (H), values represent the mean ± SD of three independent biological replicates, and asterisks represent significant differences (*P < 0.05 and **P < 0.01, Student’s t test). (I) In vitro kinase assays detected the phosphorylation of MdMPK3 in kinase buffer containing MdNDPK2 or MdNDPK2^H201A. GST-MdMPK6 protein was used as a control.

Fig. 4.. CfEC92 protein binds ATP in vitro and in vivo.

(A) ITC assay showed that CfEC92 bound ATP in vitro. Heat flow as a function of time is depicted in red, and the brown dotted line corresponds to the theoretical independent model. ATP was titrated with ddH₂O as a negative control. (B) MST assay showed that CfEC92 bound ATP in vivo. The binding curve and K_d of CfEC92 (red spots) or GFP (green spots) after adding ATP were generated by MO.Affinity Analysis. N/A, no affinity. (C) CD depicts CfEC92 secondary structure variation after incubation with ATP. (D) Second-derivative infrared spectra peak variations show CfEC92 secondary structure alternation after binding with ATP. a.u., absorbance units.

Fig. 5.. The conserved residues for CfEC92 binding ATP are Cys²⁶, Asn³⁸, and Cys³⁹.

(A) The simulation of molecular docking between CfEC92 and ATP is displayed in 2D and 3D using the Discovery Studio Visualizer. (B) Four CfEC92 truncated mutants were generated. (C) MST and (D) ITC were used to verify the key region of CfEC92 binding ATP. (E) MST shows that only the CfEC92^{GFCT30/31/32/33AAAA} mutant containing Cys²⁶, Asn³⁸, and Cys³⁹ residues binds ATP. (F) ITC detection confirms that CfEC92^{GFCT30/31/32/33AAAA} binds ATP.

Fig. 6.. CfEC92 binding ATP is essential for the virulence of C. fructicola.

(A) Transient overexpression of CfEC92 and CfEC92-M in apple leaves weakens resistance to C. fructicola. (B) Lesion diameters were measured at 3 dpi. (C) Pathogenicity assays were conducted using mutant ΔCfEC92-27 and complementation strains (ΔCfEC92-27/CfEC92, ΔCfEC92-27/CfEC92-A, and ΔCfEC92-27/CfEC92-M) on apple leaves. The relative fungal biomass was detected by qPCR. CfEC92-M, mutated Cys²⁶, Asn³⁸, and Cys³⁹ to Ala, not binding ATP; CfEC92-A, mutated Gly³⁰, Phe³¹, Cys³², and Thr³³ to Ala, binding ATP. In (B) and (C), values are mean ± SD of three independent biological replicates, and different letters represent significant differences (one-way ANOVA, Tukey test). (D) Heatmap showing the expression levels of immunity-related genes in apple leaves in response to pathogen inoculation. Asterisks indicate significant differences (false discovery rate < 0.05). ROS, reactive oxygen species; FPKM, fragments per kilobase of exon model per million mapped fragments. (E) Principal component analysis (PCA) plots of phosphoproteomic samples, separating them into three groups. GFP, GFP overexpressed apple samples; OE92, CfEC92 overexpressed apple samples; OEm92, mutant CfEC92-M (not binding ATP) overexpressed apple samples. (F) Expression levels of protein kinases involved in immunity in response to overexpression of CfEC92 and mutant CfEC92-M in apple leaves. LRRKs, Leucine-rich repeat kinases.

Fig. 7.. CfEC92 protein binds ATP to enhance autophosphorylation of MdNDPK2 and then promote phosphorylation of MdMPK3.

(A) In vitro kinase assay of MdNDPK2 autophosphorylation in kinase buffer incubated with CfEC92. MBP protein was used as a negative control. (B) In vitro kinase assay of mutant MdNDPK2^{K/F/R/T/R/N-A} testing phosphorylation in kinase buffer incubated with CfEC92. His-MdNDPK2 was used as a positive control. CBB staining shows protein loading. (C) In vitro kinase assay showing MdNDPK2 autophosphorylation in kinase buffer incubated with CfEC92 and CfEC92-M (not binding ATP). (D) In vitro kinase assay showed that MdNDPK2 autophosphorylation affected by CfEC92-ATP complex in non-ATP kinase buffer. CBB staining shows protein loading. (E) Phosphorylation level of MdMPK3 was detected in GST-MdMPK3 coincubated with MBP-CfEC92 and His-MdNDPK2, MBP-CfEC92 and His-MdNDPK2^H201A, or MBP and His-MdNDPK2 in vitro. (F) MdMPK3 phosphorylation was detected using phospho-P44/P42 MAPK antibody in apple calli overexpressing MdNDPK2 and inoculated with ΔCfEC92-27/CfEC92 or ΔCfEC92-27/CfEC92-M strains or no inoculation. Actin as an internal control indicates equal loading.

Fig. 8.. The CfEC92-ATP complex promotes affinity of MdNDPK2 binding ATP by altering its conformation.

(A) Autophosphorylation of MdNDPK2 in vitro inhibited by 0.15 μM ATP competitive inhibitor STS and restored by CfEC92. (B) ITC assay showed that STS did not interact with CfEC92. (C) Second-derivative infrared spectra of MdNDPK2 interacted with CfEC92. His-MdNDPK2 + MBP-CfEC92 shown in black and His-MdNDPK2 in red. (D) Second-derivative infrared spectra of MdNDPK2 interacted with CfEC92-ATP complex. His-MdNDPK2 + MBP-CfEC92-ATP shown in black and His-MdNDPK2 in red. (E) The spectra of MdNDPK2 upon interaction with CfEC92. His-MdNDPK2 + MBP-CfEC92 shown in black and His-MdNDPK2 in red. (F) The spectra of MdNDPK2 upon interaction with CfEC92-ATP. His-MdNDPK2 + MBP-CfEC92-ATP shown in black and His-MdNDPK2 in red. (G) The distance between γ-P of ATP and N1 and N3 of imidazole structure in His²⁰¹. The interaction was predicted by AlphaFold 3.0 and is displayed in 3D using the Discovery Studio Visualizer. (H) The distances between γ-P and N1 and N3 were slightly different from those of MdNDPK2-bound ATP of (E) above. (I) For the CfEC92-ATP complex, the distances between γ-P and N1 and N3 were reduced compared to those of MdNDPK2-bound ATP of (E) above. (J) A 3D schematic representation of multimolecular docking of CfEC92-ATP, MdNDPK2, and ATP. (K) The affinity between MdNDPK2 and ATP was measured by ITC assay and expressed by K_a values. MdNDPK2^{K/F/R/T/R/N-A} and CfEC92-M mutants with loss of ATP binding. Values are mean ± SD of three independent biological replicates, and different letters represent significant differences (one-way ANOVA, Tukey test).

Fig. 9.. The motif Cx₁₁NC of effector protein in Colletotrichum has ATP-binding activity.

(A) The maximum likelihood phylogenetic tree was constructed using CfEC92 conserved sequences (20 to 44 amino acids) and homologous animal protein sequences. Protein sequences from Lachnospiraceae bacteria were used as outgroups. The conserved sequence (20 to 44 amino acids) of CfEC92 and homologous sequences were aligned using Clustal W, and the alignment was visualized using Jalview. Blue shading intensity reflects the level of amino acid identity at each position. Red shading represents the conserved residue Cys (C), Asn (N), and Cys (C). (B) MST analysis confirmed that the CfEC92 homologous proteins Ch592 (from C. higginsianum) and Cg242 (from C. gloeosporioides) bound to ATP. (C) MST analysis of the proteins Cn022 and Pm566 bound to ATP. (D) Symptoms in N. benthamiana leaves overexpressing Cg242 or Cg242-M (nonbinding ATP) and Ch592 or Ch595-M (nonbinding ATP). (E) Lesion diameters caused by P. capsici infection in N. benthamiana overexpressing Cg242, Cg242-M, Ch592, or Ch592-M. Values represent the mean ± SD of three independent biological replicates, and asterisks represent significant differences (*P < 0.05, Student’s t test). (F) A working model of CfEC92-mediated pathogenicity. During C. fructicola infection, effector CfEC92 is secreted into plant cells. CfEC92 binds ATP, forming a CfEC92-ATP complex. The complex interacts with and alters the conformation of MdNDPK2, enhancing its autophosphorylation and kinase activity. Activated MdNDPK2 phosphorylates MdMPK3 and further suppresses plant immunity.