HpaP modulates type III effector secretion in Ralstonia solanacearum and harbours a substrate specificity switch domain essential for virulence

Abstract

Many pathogenic bacteria have evolved a type III secretion system (T3SS) to successfully invade their host. This extracellular apparatus allows the translocation of proteins, called type III effectors (T3Es), directly into the host cells. T3Es are virulence factors that have been shown to interfere with the host's immunity or to provide nutrients from the host to the bacteria. The Gram-negative bacterium Ralstonia solanacearum is a worldwide major crop pest whose virulence strongly relies on the T3SS. In R. solanacearum, transcriptional regulation has been extensively studied. However, very few data are available concerning the role played by type III-associated regulators, such as type III chaperones and T3SS control proteins. Here, we characterized HpaP, a putative type III secretion substrate specificity switch (T3S4) protein of R. solanacearum which is not secreted by the bacterium or translocated in the plant cells. HpaP self-interacts and interacts with the PopP1 T3E. HpaP modulates the secretion of early (HrpY pilin) and late (AvrA and PopP1 T3Es) type III substrates. HpaP is dispensable for the translocation of T3Es into the host cells. Finally, we identified two regions of five amino acids in the T3S4 domain that are essential for efficient PopP1 secretion and for HpaP's role in virulence on tomato and Arabidopsis thaliana, but not required for HpaP-HpaP and HpaP-PopP1 interactions. Taken together, our results indicate that HpaP is a putative R. solanacearum T3S4 protein important for full pathogenicity on several hosts, acting as a helper for PopP1 secretion, and repressing AvrA and HrpY secretion.

Keywords: T3Es (type III effectors); T3S4 (type III secretion substrate specificity switch); T3SS (type III secretion system).; bacterial wilt; pathogenicity; secretion; translocation.

Figure 1

HpaP is a putative type III secretion substrate specificity switch (T3S4) protein conserved in the Ralstonia solanacearum species complex. (a) Protein identity between HpaP, HrcV and HrcU from R. solanacearum  GMI1000 and homologues from strains representative of the R. solanacearum species complex diversity or from Xanthomonas campestris pv. vesicatoria (Xcv) 85‐10. Sequences were compared using SIAS (http://imed.med.ucm.es/Tools/sias.html). Arrows represent the hrcU‐hrcV‐hpaP operon. (b) Unrooted phylogenetic tree of hpaP based on sequence similarities using the maximum likelihood principle. Ten sequenced strains belonging to the four phylotypes of the R. solanacearum species complex were included. Branch support values were calculated using the approximate likelihood ratio test (aLRT) and are shown only if >70%. Branch lengths indicate sequence divergence. (c) Amino acid sequence alignment of HpaP from R. solanacearum  GMI1000 (GenBank accession: CAB58249) and its homologue HpaC_Xcv 85‐10 (GenBank accession: CAJ22055). Sequences were aligned using MAFFT version 7 (http://mafft.cbrc.jp/alignment/server). Conserved residues are shaded in black. Numbers refer to the HpaP sequence. The red line indicates the localization of the T3S4 domain as described for HpaC_Xcv. Green lines underline the two deletions made in this study, corresponding to hpaP::hpaPΔ_149–153 and hpaP::hpaP Δ_156–160 mutants.

Figure 2

The hpaP mutant is strongly reduced in virulence on Arabidopsis thaliana. (a) Kaplan–Meier survival analysis of A. thaliana plants inoculated with Ralstonia solanacearum wild‐type strain (red squares), hpaP mutant (green stars) and the complemented strain hpaP::hpaP ⁺ (blue circles). Each strain was root inoculated on at least 16 A. thaliana plants. P values from Gehan–Breslow–Wilcoxon tests are shown below the graph. Red boxes indicate P < 0.01. Data were pooled from three independent experiments. (b) Representative photographs taken 6 days post‐inoculation. WT, wild‐type.

Figure 3

HpaP is not secreted in vitro or translocated in planta. (a) After 8, 12 and 18 h of culture in secretion medium, Ralstonia solanacearum wild‐type strain cell pellet (CP) and culture supernatant (SN) were analysed by immunoblotting. HpaP protein (22 kDa) was detected using anti‐haemagglutinin (anti‐HA) antibody. PopP2 was used as a positive control of the functional secretion of the strain used. This Western blot is representative of three independent replicates. (b) Translocation assay of HpaP and PopP2 proteins in Nicotiana tabacum leaves using HpaP‐CyaA′ and PopP2‐CyaA′ (positive control) fusion proteins. Cyclic adenosine monophosphate (cAMP) levels were detected using cAMP Biotrak competitive enzyme immunoassay. Four independent biological replicates were made. WT, wild‐type.

Figure 4

HpaP is able to self‐interact in vitro. Glutathione S‐transferase (GST) and GST‐HpaP were immobilized on glutathione sepharose and incubated with an Escherichia coli lysate containing 6His‐HpaP (His, histidine). Total cell lysates (TE) and eluted proteins (eluates) were analysed using antibodies directed against GST and the 6His epitope. Bands corresponding to GST and GST fusion proteins are marked by asterisks; lower bands represent degradation products. The experiment was repeated three times with similar results.

Figure 5

HpaP modulates the secretion of AvrA, PopP1 and HrpY. Ralstonia solanacearum  GMI1000 wild‐type strain (triangles) and hpaP mutant (squares) were cultivated in secretion medium. Culture supernatants were harvested after 8, 12 and 18 h of culture and analysed by immunoblotting. PopP1 (a), AvrA (b), GALA7 (c) and PopP2 (d) type III effectors, and HrpY pilin (e), were detected using their respective antibodies for three independent biological replicates. Band intensities for all conditions were quantified using GeneTools (Syngene, Cambridge, UK). For each protein, the band intensity corresponding to the wild‐type strain after 8 h of culture was used as an internal reference. For other time points of the wild‐type and of the hpaP mutant, ratios of band intensities compared with this reference were calculated. Means and standard errors at each time point were plotted on the graphs and correspond to the measurement of three independent biological replicates.

Figure 6

HpaP interacts specifically with the type III effector (T3E) PopP1. (a) Interaction studies between HpaP and several T3Es. Yeast cells were co‐transformed by AD‐HpaP and BD‐PopP1, BD‐PopP2, BD‐AvrA or BD‐GALA7. Double transformation and interaction were tested by plating yeasts on synthetic dropout medium lacking leucine and tryptophan (SD/–Leu/–Trp) and synthetic dropout medium lacking leucine, tryptophan and histidine (SD/–Leu/–Trp/–His), respectively. 3‐Aminotriazole (3‐AT) was added to suppress autoactivation when necessary. Three biological replicates were performed giving the same results. (b) In vitro validation of the interaction between HpaP and PopP1. Glutathione S‐transferase (GST) and GST‐HpaP were immobilized on glutathione sepharose and incubated with an Escherichia coli lysate containing 6His‐PopP1. Total cell lysates (TE) and eluted proteins (eluates) were analysed using antibodies directed against GST and the 6His epitope. Bands corresponding to GST and GST fusion proteins are marked by asterisks; lower bands represent degradation products. Experiments were repeated twice with similar results. AD, activation domain; BD, binding domain; His, histidine.

Figure 7

Interaction studies with HpaP deletion derivatives mutated in the type III secretion substrate specificity switch (T3S4) domain. (a) Deletions in the T3S4 domain do not abolish the self‐interaction of HpaP. Immobilized glutathione S‐transferase (GST), GST‐HpaP, GST‐HpaPΔ_149–153 and GST‐HpaPΔ_156–160 were incubated with 6His‐HpaP. Total cell lysates (TE) and eluted proteins (eluates) were analysed using antibodies directed against GST and the 6His epitope. Bands corresponding to GST and GST fusion proteins are marked by asterisks; lower bands represent degradation products. (b) Deletions in the T3S4 domain do not abolish interaction between HpaP and the type III effector PopP1. Yeast cells were co‐transformed by BD‐PopP1 and either AD‐HpaPΔ_149–153 or AD‐HpaPΔ_156–160. Double transformation and interaction were tested by plating yeasts on synthetic dropout medium lacking leucine and tryptophan (SD/–Leu/–Trp) and synthetic dropout medium lacking leucine, tryptophan and histidine (SD/–Leu/–Trp/–His), respectively. 3‐Aminotriazole (3‐AT) was added to suppress autoactivation. Experiments in (a) and (b) were repeated twice with similar results. AD, activation domain; BD, binding domain; His, histidine.

Figure 8

hpaPΔ_149–153 and hpaPΔ_156–160 mutants behaved similarly to the hpaP mutant. (a) PopP1 is specifically less secreted in both hpaP deletion derivative mutants (hpaP::hpaPΔ_149–153 and hpaP::hpaPΔ_156–160) compared with the wild‐type version (hpaP::hpaP ⁺). Ralstonia solanacearum strains were cultivated in secretion medium, cell pellets (CP) and culture supernatants (SN) were harvested after 8, 12 and 18 h of culture and analysed by immunoblotting. Effector proteins PopP1 (43 kDa) and PopP2 (53 kDa) were detected using their respective antibodies. Two independent biological replicates were made with similar results. Both hpaP deletion mutants are strongly reduced in virulence on tomato (b) and Arabidopsis thaliana (c). Kaplan–Meier survival analysis of plants inoculated with R solanacearum wild‐type strain (WT) (red squares), hpaP mutant (green stars) and the deletion mutants hpaP::hpaPΔ_149–153 (yellow diamonds) and hpaP::hpaPΔ_156–160 (black triangles). Each strain was root inoculated on 24 tomato and at least 16 A. thaliana plants. P values from Gehan–Breslow–Wilcoxon tests are shown below each graph. Red boxes indicate P < 0.01. Data were pooled from three independent experiments.