Computational and cellular exploration of the protein-protein interaction between Vibrio fischeri STAS domain protein SypA and serine kinase SypE

Abstract

Anti-sigma factor antagonists SpoIIAA and RsbV from Bacillus subtilis are the archetypes for single-domain STAS proteins in bacteria. The structures and mechanisms of these proteins along with their cognate anti-sigma factors have been well studied. SpoIIAA and RsbV utilize a partner-switching mechanism to regulate gene expression through protein-protein interactions to control the activity of their downstream anti-sigma factor partners. The Vibrio fischeri STAS domain protein SypA is also proposed to employ a partner-switching mechanism with its partner SypE, a serine kinase/phosphatase that controls SypA's phosphorylation state. However, this regulation appears opposite to the canonical pathway, with SypA being the more downstream component rather than SypE. Here we explore the commonalities and differences between SypA and the canonical single-domain STAS proteins SpoIIAA and RsbV. We use a combination of AlphaFold 2 structure predictions and computational modeling to investigate the SypA-SypE binding interface. We then test a subset of our predictions in V.fischeri by generating and expressing SypA variants. Our findings suggest that, while SypA shares many sequence and structural traits with anti-sigma factor antagonist STAS domain proteins, there are significant differences that may account for SypA's distinct regulatory output.

Keywords: Anti-sigma factor antagonists; STAS; Vibrio; biofilm; phosphorylation; protein-protein interaction.

Figure 1.

Comparison of regulatory pathways. a) the canonical pathway model for single-domain STAS proteins as anti-sigma factor antagonists. With no environmental trigger (grey region), the single-domain STAS protein (blue) is inactive. The anti-sigma factor (red) binds and sequesters its cognate sigma factor (green) preventing gene expression. An environmental signal leads to the dephosphorylation and activation of the STAS protein (white region). The STAS protein binds the anti-sigma factor, releasing the sigma factor and turning on gene expression. b) Model of the syp-dependent biofilm induction pathway in V. fischeri. Sensor kinase RscS (orange) indirectly activates SypG (green) which turns on syp transcription. RscS also regulates phosphorylation of SypE (red). Unphosphorylated SypE acts as a kinase that phosphorylates and inactivates SypA (blue), preventing biofilm formation.

Figure 2.

Computational model of SypA and structural alignment to canonical single-domain STAS proteins. a) AlphaFold prediction of V. fischeri SypA (Uniprot ID Q5DYQ6). Residues are colored based on the AlphaFold prediction confidence scores (pLDDT) which range from 0 to 100. b) Structural alignment of AlphaFold SypA prediction (blue) with SpoIIAA and RsbV structures (grey; PDB ID 1AUZ, 1THN, 1H4X, and 6M36). The black arrow denotes the structural variability in the α1-β3 loop. The box highlights the C-termini.

Figure 3.

Sequence alignment of SypA, RsbV, and SpoIIAA. Alignments are based on protein secondary structures. The conserved phosphorylation site is denoted with a star. Residues labeled with a closed circle are sites of interest from studies of B. subtilis SpoIIAA and RsbV. Shading denotes sequence similarity across all species. Open boxes denote residues that are similar in SpoIIAA and RsbV but from which SypA deviates. Alpha helices and beta sheet strands are indicated above the alignments by red bars and blue arrows, respectively. VfSypA = V. fischeri SypA (Uniprot ID Q5DYQ6), BsRsbV = B. subtilis RsbV (Uniprot ID P17903), LsSpoIIAA = L. sphaericus SpoIIAA (Uniprot ID O32723), BsSpoIIAA = B. subtilis SpoIIAA (Uniprot ID P10727), GsSpoIIAA = G. stearothermophilus SpoIIAA (Uniprot ID O32726).

Figure 4.

Computational model of SypA-SypE complex. a) Optimized ColabFold prediction of SypA (blue) bound to a SypE dimer (green). The serine kinase REC, and PP2C domains of SypE are labeled. b) Closeup view of SypA bound to the serine kinase domain of SypE. The SypA-SypE prediction is overlaid with the structure of SpoIIAA-SpoIIAB (grey, PDB ID 1TH8) and RsbV-RsbW (pink, PDB ID 6M36). The phosphorylation site of the STAS protein and ATP lid of the serine kinases are labeled.

Figure 5.

SypA-Sype binding interaction and simulated mutagenesis. a) Binding interface of the optimized ColabFold prediction with SypA (blue) bound to the SypE serine kinase domain (green). SypA residues are labeled in black and SypE residues are labeled in green. Yellow dash lines denote potential hydrogen bonds. b) Simulated mutagenesis data for ΔΔG_binding and ΔΔStability for the SypA-SypE binding interaction with data represented in graphical form below. Data are the mean ± standard error in kcal/mol each calculated from 10 simulations.

Figure 6.

Impact of sypA mutations on biofilm formation. Colony morphology at 48 h of strains that carry pEAH73 (sypG overexpression) (a) or pKG11 (rscS overexpression) (b). The base strains are as described in Table 1: wild type (WT), ES114; ΔsypE, KV10149sypE; ΔsypA, KV10032; ΔsypA + sypA, KV10163; and ΔsypA derivatives that carry the following sypA alleles: S56A, KV10247; S56D, KV10331; D20A, KV10248; D22A, KV10334; F53A, KV10249; S57A, KV10332; D20A/D22A, KV10401. Cultures were spotted on LBS plates containing chloramphenicol. Representative images are shown.