Functional analysis of cyclic diguanylate-modulating proteins in Vibrio fischeri

Abstract

As bacterial symbionts transition from a motile free-living state to a sessile biofilm state, they must coordinate behavior changes suitable to each lifestyle. Cyclic diguanylate (c-di-GMP) is an intracellular signaling molecule that can regulate this transition, and it is synthesized by diguanylate cyclase (DGC) enzymes and degraded by phosphodiesterase (PDE) enzymes. Generally, c-di-GMP inhibits motility and promotes biofilm formation. While c-di-GMP and the enzymes that contribute to its metabolism have been well studied in pathogens, considerably less focus has been placed on c-di-GMP regulation in beneficial symbionts. Vibrio fischeri is the sole beneficial symbiont of the Hawaiian bobtail squid (Euprymna scolopes) light organ, and the bacterium requires both motility and biofilm formation to efficiently colonize. c-di-GMP regulates swimming motility and cellulose exopolysaccharide production in V. fischeri. The genome encodes 50 DGCs and PDEs, and while a few of these proteins have been characterized, the majority have not undergone comprehensive characterization. In this study, we use protein overexpression to systematically characterize the functional potential of all 50 V. fischeri proteins. All 28 predicted DGCs and 10 of the 14 predicted PDEs displayed at least one phenotype consistent with their predicted function, and a majority of each displayed multiple phenotypes. Finally, active site mutant analysis of proteins with the potential for both DGC and PDE activities revealed potential activities for these proteins. This work presents a systems-level functional analysis of a family of signaling proteins in a tractable animal symbiont and will inform future efforts to characterize the roles of individual proteins during lifestyle transitions.IMPORTANCECyclic diguanylate (c-di-GMP) is a critical second messenger that mediates bacterial behaviors, and Vibrio fischeri colonization of its Hawaiian bobtail squid host presents a tractable model in which to interrogate the role of c-di-GMP during animal colonization. This work provides systems-level characterization of the 50 proteins predicted to modulate c-di-GMP levels. By combining multiple assays, we generated a rich understanding of which proteins have the capacity to influence c-di-GMP levels and behaviors. Our functional approach yielded insights into how proteins with domains to both synthesize and degrade c-di-GMP may impact bacterial behaviors. Finally, we integrated published data to provide a broader picture of each of the 50 proteins analyzed. This study will inform future work to define specific pathways by which c-di-GMP regulates symbiotic behaviors and transitions.

Keywords: Vibrio fischeri; biofilm; c-di-GMP; flagellar motility.

Fig 1

V. fischeri encodes 50 proteins across both chromosomes predicted to modulate c-di-GMP levels. The circular chromosomes are represented in a linear fashion for this representation. Numbers represent VF_ locus tags (e.g., VF_0087 and VF_A0056).

Fig 2

Many predicted V. fischeri DGCs and PDEs impact biofilm formation and swimming motility when overexpressed. (A) Quantification of Congo red binding for V. fischeri strains overexpressing the indicated proteins relative to the pRYI039 empty vector control. For each strain, n = 3–8 biological replicates (24 for controls). Congo red images are representative. (B) Quantification of migration through soft (0.3%) agar for V. fischeri strains overexpressing the indicated proteins relative to the pRYI039 empty vector control. For each strain, n = 4–11 biological replicates (33 for controls). For panels A and B, one-way analysis of variance was used for statistical analysis, each bar represents the means of biological replicates, error bars represent standard errors of the mean, asterisks represent significance relative to the pRYI039 empty vector control (*P < 0.05), and numbers represent VF_ locus tags (e.g., VF_0087 and VF_A0056); negative controls pRYI039 and pEVS143 as well as non-V. fischeri controls QrgB and VC1086 are also listed and indicated with a black dot.

Fig 3

Most V. fischeri proteins tested are expressed from the overexpression vector. Western blot of whole-cell lysates of V. fischeri expressing indicated FLAG-tagged proteins from the pEVS143 vector. Predicted band sizes (kDa) for each protein are indicated in parentheses. Anti-FLAG rabbit IgG was used as the primary antibody, and LI-COR IRDye 800CW goat anti-rabbit IgG was used as the secondary antibody. Anti-RpoA mouse IgG was used as a loading control, binding the RNAP α subunit, and LI-COR IRDye 680RD goat anti-mouse IgG was used as the secondary antibody of the loading control. Western blot is representative of n = 3 biological replicates.

Fig 4

Active site residues are required to modulate cellulose production and motility for selected DGCs and PDEs. (A) Quantification of Congo red binding for V. fischeri strains overexpressing the indicated proteins relative to the pRYI039 empty vector control. For each strain, n = 4–5 biological replicates (10 for controls). (B) Quantification of migration through soft (0.3%) agar for V. fischeri strains overexpressing the indicated proteins relative to the pRYI039 empty vector control. For each strain, n = 5 biological replicates (10 for controls). For panels A and B, unpaired t tests were used for statistical analysis, each bar represents the means of biological replicates, error bars represent standard errors of the mean, asterisks represent significance of a mutant relative to the corresponding wild-type protein (*P < 0.01), and numbers represent VF_ locus tags (e.g., VF_0087 and VF_A0056); negative controls pRYI039 and pEVS143 are indicated with a black dot.

Fig 5

VF_0985 is the only dual-function protein with strong active site-dependent phenotypes. (A) Quantification of Congo red binding for V. fischeri strains overexpressing the indicated proteins relative to the pRYI039 empty vector control. For each strain, n = 4–5 biological replicates (10 for controls). (B) Quantification of migration through soft (0.3%) agar for V. fischeri strains overexpressing the indicated proteins relative to the pRYI039 empty vector control. For each strain, n = 5 biological replicates (10 for controls). For panels A and B, unpaired t tests were used for statistical analysis, each bar represents the means of biological replicates, error bars represent standard errors of the mean, asterisks represent significance of a mutant relative to the corresponding wild-type protein (*P < 0.05), and numbers represent VF_ locus tags (e.g., VF_0087 and VF_A0056); negative controls pRYI039 and pEVS143 are indicated with a black dot.

Fig 6

Integration of phenotypic data for the V. fischeri DGCs and PDEs. Numbers represent VF_ locus tags (e.g., VF_0087 and VF_A0056); non-V. fischeri controls QrgB and VC1086 are also listed. Overexpression data are from this study; deletion data are integrated from reference . Blue coloring indicates phenotypes expected from elevated c-di-GMP, whereas pink indicates phenotypes expected from reduced c-di-GMP. White indicates no significant change. ^aThis study. ^bReference .