Quorum sensing mediates morphology and motility transitions in the model archaeon Haloferax volcanii

Abstract

Quorum sensing (QS) is a mechanism of intercellular communication that enables microbes to alter gene expression and adapt to the environment. This cell-cell signaling is necessary for intra- and interspecies behaviors such as virulence and biofilm formation. While QS has been extensively studied in bacteria, little is known about cell-cell communication in archaea. Here we established an archaeal model system to study QS. We showed that for Haloferax volcanii, the transition from motile rods to non-motile disks is dependent on a possibly novel, secreted small molecule present in cell-free conditioned medium (CM). Moreover, we determined that this putative QS molecule fails to induce the morphology transition in mutants lacking the regulatory factors, DdfA and CirA. Using quantitative proteomics of wild-type cells, we detected significant differential abundances of 236 proteins in the presence of CM. Conversely, in the ΔddfA mutant, addition of CM resulted in only 110 proteins of significant differential abundances. These results confirm that DdfA is involved in CM-dependent regulation. CirA, along with other proteins involved in morphology and swimming motility transitions, is among the proteins regulated by DdfA. These discoveries significantly advance our understanding of microbial communication within archaeal species.

Figure 1.. Hfx. volcanii cells are disk-shaped in early log in the presence of conditioned medium (CM).

Differential interference contrast (DIC) images of Hfx. volcanii cells from 5 mL cultures grown to OD₆₀₀ 0.05 with no additive (top) and with 50 μL of CM (1% v/v) added (bottom). Scale bar is 10 μm (a). Cell shape quantification of three biological replicates of each condition to determine lowest effective volume of CM from an Hfx. volcanii culture grown to OD₆₀₀ 1.5 (b) or of the effect of 1% CM from cultures grown to differing optical densities (c) by using the eccentricity parameter of CellProfiler^,, which outputs a value between 0 (circle) and 1 (line segment). Eccentricity values that range between 0.5 and 0.9 correspond to disks while values that range between 0.9 and 1.0 correspond to rods, and each data point represents a single counted cell. Median of each condition denoted by line. Statistical analysis performed via nested 1-way ANOVA with multiple comparisons to the “No additive” condition. Ns is p>0.05, * is p<0.05, ** is p<0.01, and **** is p<0.0001.

Figure 2.. CM incorporated into soft agar plates inhibits swimming motility.

Motility halos of Hfx. volcanii stab-inoculated onto soft agar plates of Hv-Cab (left) and Hv-Cab supplemented with 25% CM (right). Scale bar denotes 1 cm (a). Quantification of motility halo radius in millimeters of six biological replicates. Error bars depict mean with standard deviation. Statistical significance determined with unpaired parametric two-tailed t-test. **** is p<0.0001 (b).

Figure 3.. ΔddfA and ΔcirA form motile rod-shaped cells in the presence of CM.

Motility halos of wild type (WT), ΔddfA, and ΔcirA on soft agar plates supplemented with 25% CM (v/v). Scale bar denotes 1 cm (a, top). DIC microscopy images of WT, ΔddfA, and ΔcirA cells grown to early log in cultures containing 1% CM. Scale bar denotes 10 μm (a, bottom). Quantification of six replicate motility halo radii in millimeters. Error bars depict mean with standard deviation. Statistical significance determined with ordinary 1-way ANOVA with multiple comparisons to the WT condition (b). Cell shape quantification of three biological replicates of WT, ΔddfA, and ΔcirA grown to OD₆₀₀ 0.05 containing 1% CM as described in Figure 1. Median of each condition denoted by line. Statistical analysis performed via nested 1-way ANOVA with multiple comparisons to the WT. **** p-value <0.0001 (c).

Figure 4.. Quantitative proteomics analyses indicate proteins with altered abundances in the presence of CM.

Volcano plots depicting log₂ fold changes and their corresponding −log₁₀ p-values of protein abundances compared wild type between early and late log (a), wild type early log regular medium to 1% CM (b), and ΔddfA early log regular medium to 1% CM (d). Non-significant (p-value ≥ 0.05) protein changes correspond to gray circles; significant (p-value < 0.05) and positive protein fold changes correspond to maroon circles (higher abundance in second condition); significant (p-value < 0.05) and negative protein fold changes correspond to blue circles (higher abundance in first condition). Raw values can be found in the supplementary material. Interactive html files of the volcano plots are available at 10.5281/zenodo. 14510577 (a,b,d). Venn diagram of the overlap of proteins with significant differential abundance in the three proteomics comparisons (wild type early to late log, wild type no CM to 1% CM, and ΔddfA no CM to 1% CM). A list of proteins represented in each overlapping region can be found in Table S3 (c).

Figure 5.. DFS can be extracted from CM and is unaffected by DNase, RNase, or Proteinase K.

Cell shape quantification of three biological replicates of cultures containing final concentration of 1% of each of the following: autoclaved CM, flowthrough of CM passing through a 3 kDa centrifugal filter, elution from the filter using Tris buffer, and the Tris buffer as control (a), as well as eluates treated with RQ1 DNase, RNase A and Proteinase K. “Neg control” condition includes 5 μL RQ1 DNase 10X Reaction Buffer, 5 μL of 50% glycerol, and 5 μL RQ1 DNase Stop Solution to ensure that these additional components do not induce disk formation (b). Cell shape quantification as described in Figure 1. Line denotes median value. Statistical analysis performed via nested 1-way ANOVA with multiple comparisons to the respective “No additive” or “Neg control” condition. Ns is p>0.05 and **** is p<0.0001.

Figure 6.. Hfx. volcanii CM activates A. tumefaciens bioreporter strain similar to AHL, but DKPs do not induce Hfx. volcanii disk formation.

Miller Units of cleavage of ortho-Nitrophenyl-β-galactoside (ONPG) were measured in the bioreporter strain A. tumefaciens KYC55 (pJZ410)(pJZ384)(pJZ372). “No additive” condition corresponds to basal output of the bioreporter strain, “Media Ctrl” condition corresponds to the addition of 200 μL Hv-Cab extract, “CM” condition corresponds to 200 μL of CM extract, and “AHL” condition corresponds to final concentration 2.5 ng/μL of N-3-oxo-pentanoyl-L-Homoserine lactone. Statistical analysis performed via unpaired parametric two-tailed t-test (a). Cell shape quantification of three biological replicates comparing no additive and 1% CM to the addition of DKPs pooled together for final concentration 0.2 μg/mL. Because DKPs were solubilized in ethanol, similar amounts of ethanol (10 μL) were added to the “Neg Ctrl” and “CM” cultures. Cell shape quantification as described in Figure 1. Line denotes median value. Statistical analysis performed via nested 1-way ANOVA with multiple comparisons to the respective “No additive” or “Neg control” condition (b). Ns is p>0.05, * is p<0.05, and **** is p<0.0001.

Figure 7.. The initial roadmap for the QS-mediated regulatory network in Hfx. volcanii.

A subset of proteins discussed in this study is depicted as increasing or decreasing in abundance (vertical arrows) as a result of direct or indirect regulation (horizontal arrows), resulting in QS-mediated behaviors.