Noncanonical calcium binding motif controls folding of HopQ1, a Pseudomonas syringae type III secretion effector, in a pH-dependent manner

Abstract

Virulence of many gram-negative bacteria relies upon delivery of type three effectors into host cells. To pass through the conduit of secretion machinery the effectors need to acquire an extended conformation, and in many bacterial species specific chaperones assist in this process. In plant pathogenic bacterium Pseudomonas syringae, secretion of only few effectors requires the function of chaperones. This raises a question how chaperone-independent effectors achieve an appropriate conformation for the secretion. One such mechanism was previously described for AvrPto. It contains a pH-sensitive switch, which is involved in unfolding of the effector at the mildly acidic pH corresponding to the pH value of the bacterial cytosol, and refolding at the neutral pH. Therefore, it was proposed that the switch facilitates first translocation of AvrPto and then its maturation once the effector reaches the cytoplasm of host cell. Here we show that an atypical motif of HopQ1, another effector of P. syringae, reversibly binds calcium in pH-dependent manner, regulating the effector thermal stability. Therefore, we propose a model that HopQ1 traversing through the type three secretion system encounters conditions that maintain its extended conformation, while upon delivery into host cell the effector undergoes refolding.

Fig. 1

Disulfide-linked HopQ1 oligomers. Recombinant HopQ1 with a C-terminal 6 × His epitope was subjected to gel filtration coupled to MALS analysis under native (a) or reducing (b) conditions. The cysteine single and double mutants C70A (c), C230A (d), and C70A_C230A (e) were analyzed under native conditions. Blue and red traces correspond to absorbance at 280 nm and 254 nm, respectively; green trace indicates static light scattering at 90° (LS90), and black indicates molecular weights. The derived molar masses for the monomer, dimer, and trimer are 52 kDa ± 2%, 94 kDa ± 1%, and 138 kDa ± 1%, respectively (the theoretical masses for the monomer, dimer, and trimer are 49.8, 99.6, and 149.4 kDa, respectively).

Fig. 2

Depletion of calcium ions induces HopQ1 dimer formation in the presence of the reducing agent. Representative gel filtration runs on a Superdex 200 column are shown. (a) Comparison of canonical and HopQ1-like Ca²⁺ binding motifs (b) Blue trace, recombinant HopQ1 with a C-terminal 6 × His epitope was eluted under reducing conditions (5 mM DTT) as a monomer (elution volume of 14.65 ml). Red trace, on application of chelating agent (1 mM EDTA), HopQ1 was eluted as two peaks corresponding to monomer (elution volume of 14.56 ml) and dimer (elution volume of 12.90 ml). Green trace, on addition of chelating agent and calcium ions (1mM EDTA, 2 mM CaCl₂), HopQ1 eluted only as monomers (elution volumes of 14.71 ml). (c) Green trace, the HopQ1 mutant in the calcium-binding site was eluted only as dimers (peak 13.14 ml). As a control, HopQ1 was used under reducing conditions (blue trace, monomer, elution volume of 14.71 ml) or upon application of chelating agent (red trace; monomer and dimer with elution volumes of 14.75 and 13.01 ml, respectively). Prior to gel filtration the samples were analyzed by immunodetection (Fig. S1a, b). The experiment was performed twice with similar results.

Fig. 3

Thermal stability of 6 × HIS-tagged proteins: HopQ1-WT (red), HopQ1-D107A_D108A (green), HopQ1-N103D_D105G (purple) and RihA (blue). RihA is a pyrimidine nucleoside hydrolase, containing a canonical calcium binding motif, and was used as a control. (a) Relationships between protein melting temperature, pH, and free Ca²⁺ concentration (0 mM—presence of 200 µM EGTA; 8.58 µM—retroactively-estimated calcium concentration in the absence of EGTA, see Figs. S3–S4). For pH adjustment MES (pH ≤ 6.75) and HEPES (pH ≥ 6.75) buffers were used. Lines correspond to local regression (LOESS). (b) The data obtained after at increasing EGTA concentrations. The experiment was performed twice with similar results.

Fig. 4

HopQ1 forms dimers in planta. Confocal images show representative N. benthamiana leaf epidermal cells transiently co-expressing pairs of HopQ1 (a) or HopQ1-D107-D108A (b) fused to ECFP along with the same variants fused to EYFP and nuclear localization signal (NLS). The photographs were taken 72 h after agroinfiltration. For each variant, approximately 20 transformed cells were examined and fluorescence intensity of ECFP was measured in each nucleus (c, d). Bars = 10 μm. Asterisks indicate significantly higher ECFP intensity, when compared to expression of single HopQ1 (a) or HopQ1-D107-D108A (b) fused to ECFP, p-value < 0.001 assessed by Wilcoxon test. The expression levels of proteins used in the experiment was analyzed by immunodetection (Fig. S5).