H-NOX Regulates Biofilm Formation in Agrobacterium Vitis in Response to NO

Abstract

Transitions between motile and biofilm lifestyles are highly regulated and fundamental to microbial pathogenesis. H-NOX (heme-nitric oxide/oxygen-binding domain) is a key regulator of bacterial communal behaviors, such as biofilm formation. A predicted bifunctional cyclic di-GMP metabolizing enzyme, composed of diguanylate cyclase and phosphodiesterase (PDE) domains (avi_3097), is annotated downstream of an hnoX gene in Agrobacterium vitis S4. Here, we demonstrate that avH-NOX is a nitric oxide (NO)-binding hemoprotein that binds to and regulates the activity of avi_3097 (avHaCE; H-NOX-associated cyclic di-GMP processing enzyme). Kinetic analysis of avHaCE indicates a ∼four-fold increase in PDE activity in the presence of NO-bound avH-NOX. Biofilm analysis with crystal violet staining reveals that low concentrations of NO reduce biofilm growth in the wild-type A. vitis S4 strain, but the mutant ΔhnoX strain has no NO phenotype, suggesting that H-NOX is responsible for the NO biofilm phenotype in A. vitis. Together, these data indicate that avH-NOX enhances cyclic di-GMP degradation to reduce biofilm formation in response to NO in A. vitis.

FIGURE 1.

Electronic absorption spectra of avH-NOX as the unligated Fe(II) complex (solid black line), the Fe(II)-CO complex (solid gray line), and the Fe(II)-NO complex (dashed black line) in 50 mM HEPES pH 7.5 buffer and 200 mM NaCl at room temperature. Insert shows the $\alpha /\beta$ region at wavelength range 500 – 600 nm.

FIGURE 2.

Steady-state kinetics of cyclase and phosphodiesterase activities of the AAL and GGAAF variants of avHaCE, respectively. Initial velocity measurements of the diguanylate cyclase activity of the AAL variant (250 nM) as a function of GTP concentration (0 μM - 100 μM) in 50 mM Tris-HCl pH 7.5 buffer containing 5 mM MgCl₂ at 25$°$C (A) and phosphodiesterase activity of the GGAAF variant (15 nM) as a function of cyclic di-GMP concentration (0 μM – 20 μM) in 50 mM Tris-HCl pH 7.5 buffer containing 5 mM MgCl₂ at 25$°$C (B). The data were fit to the Hill equation (plot A) or Michaelis−Menten equation (plot B), respectively. Error bars represent standard deviation from the mean determined from at least three independent trials.

FIGURE 3.

Initial velocities (ΔAbs_360nm min⁻¹) determined using a modified Invitrogen EnzChek^™ assay were plotted as a function of enzyme activity without and with avH-NOX as the Fe(II)-unligated (Fe(II)-avH-NOX) or Fe(II)-NO (NO-avH-NOX) complex. The initial velocity of the cyclase activity of the AAL variant (250 nM) with and without 2.5 μM Fe(II)-avH-NOX or NO-avH-NOX is reported relative to the initial velocity of the cyclase activity of the AAL variant in the absence of avH- NOX (white bars). The initial velocity of the phosphodiesterase activity of the GGAAF variant (50 nM) with and without 0.50 μM Fe(II)-avH-NOX or NO-avH-NOX, is reported relative to the initial velocity of the phosphodiesterase activity of the GGAAF variant in the absence of avH- NOX (shaded bars). Errors bars represent the standard deviation from the mean of at least three independent trials. NS = not significant and * = p < 0.05 are for the comparison of AAL (white bars) or GGAAF (patterned bars) activities in the presence of NO-avH-NOX to those activities in the presence of Fe(II)-avH-NOX.

FIGURE 4.

Interaction of H-NOX with HaCE domains. The top panel in A, B, and C illustrates the protein loading of the bait protein (GST or GST-tagged HaCE or HaCE domains) present in the pull-down assay, detected by either anti-GST immunoblotting or ponceau protein staining. The bottom panel illustrates the amount of H-NOX pulled down by the GST-tagged bait protein and detected by anti-His immunoblotting. (A) avH-NOX interacts with the full-length GST tagged avHaCE (lane 2). (B) avH-NOX interacts with avPDE (lane 1) and avDGC (lane 2); domain interactions shown schematically in the bottom panel. (C) swH-NOX interacts with swPAS (lane 1) and with swPDE (lane 3) but not with swDGC (lane 2); domain interactions shown schematically in the bottom panel.

FIGURE 5.

Biofilms of wild-type, ΔhnoX, and ΔhnoX/phnoX strains of A. vitis S4 grown in polystyrene plates for 48 h in AB media in the presence and absence of NO donor (20 μM DETA NONOate) and quantified with crystal violet (CV) staining. CV measurements at 570 nm were normalized to OD readings of the cultures before staining with CV. Error bars represent the standard deviation from the mean of at least three independent trials, and within each trial, each biofilm condition was run a minimum of 6 times. * = p < 0.05.

FIGURE 6.

Planktonic growth curves of wild-type, ΔhnoX, and ΔhnoX/phnoX strains of A. vitis S4 in AB minimal medium supplemented with or without 20 μM DETA NONOate at 25$°$C. The OD₆₀₀ was measured up to 72 h. Error bars represent the standard deviation from the mean of at least three independent trials. * = p < 0.05 in comparison to ΔhnoX/phnoX in the absence of NO.

FIGURE 7.

Schematic model for the H-NOX/HaCE system in A. vitis. H-NOX enhances the PDE activity of HaCE in response to NO leading to increased cyclic di-GMP degradation and reduced biofilm formation in A. vitis.