Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Phase variation of Clostridioides difficile colony morphology occurs via modulation of cell division

  • Anchal Mehra,

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Microbiology and Immunology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America

    ⨯
  • Elizabeth M. Garrett,

    Roles Conceptualization, Investigation, Methodology, Writing – review & editing

    Affiliation Department of Pathology and Laboratory Medicine, Penn State University Hershey Medical Center, Hershey, Pennsylvania, United States of America

    ⨯
  • Christopher J. Serody,

    Roles Formal analysis, Investigation, Methodology

    Affiliation Department of Biology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America

    ⨯
  • Rita Tamayo

    Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing

    rita_tamayo@med.unc.edu

    Affiliation Department of Microbiology and Immunology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States of America

    ⨯

Abstract

Phase variation of C. difficile colony morphology occurs via modulation of transcription of cmrRST, which encodes a three-protein signal transduction system. Response regulators CmrR and CmrT promote rough colony development, cell elongation and chaining, surface motility, and disease in the hamster model of infection, while impairing swimming motility and biofilm formation. Using RNA-Seq, we identified the CmrR and CmrT-dependent transcriptional differences in rough and smooth colonies. Further analysis showed that CmrT, but not CmrR, is required for differential expression of most of the genes. Two CmrT-regulated genes, herein named mrpA and mrpB, were together sufficient for restoring all CmrT-dependent in vitro phenotypes in a cmrT mutant and alleviating selection of cmr phase ON cells during growth on an agar surface. MrpA and MrpB are uncharacterized proteins with no known function but are highly conserved in C. difficile. Using immunoprecipitation and mass spectrometry to identify interacting partners, we found that MrpA interacts with the septum site-determining protein MinD and several other proteins involved in cell division and cell shape determination. Ectopic expression of mrpAB resulted in atypical cell division, consistent with MrpAB interference with MinD function. Our findings reveal a potential mechanism by which phase variation of CmrRST modulates colony morphology and motility: in cmr phase ON cells, CmrT-mediated expression of mrpAB interferes with normal cell division resulting elongated cells that enable expansion of the population across a surface while limiting swimming motility.

Author summary

C. difficile can reversibly switch between two colony morphology variants that differ in multiple additional phenotypes. Previous work determined that the phenotypic switch occurs through phase variation of the CmrRST signal transduction system, however the CmrRST-regulated genes that mediate these phenotypes were unknown. Here, we identified the CmrR- and CmrT-regulated genes that are differentially expressed between rough and smooth colonies and identified two genes, mrpA and mrpB, that together are sufficient to confer phenotypes associated with cmrRST expression. Protein interaction studies revealed that MrpA interacts with MinD, a protein that helps ensure symmetric cell division in bacteria. Based on these findings, we propose that MrpA disrupts MinD function leading to aberrant cell division, resulting in elongated cells that form rough colonies. This study reveals previously uncharacterized proteins that affect C. difficile cell division and broadly impact the physiology of this pathogen.

Citation: Mehra A, Garrett EM, Serody CJ, Tamayo R (2025) Phase variation of Clostridioides difficile colony morphology occurs via modulation of cell division. PLoS Pathog 21(12): e1013471. https://doi.org/10.1371/journal.ppat.1013471

Editor: Michael Otto, National Institutes of Health, UNITED STATES OF AMERICA

Received: August 18, 2025; Accepted: November 19, 2025; Published: December 1, 2025

Copyright: © 2025 Mehra et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The data used for RNA-Seq analysis is available at NCBI GEO (Accession GSE305463). Other data are included in supplemental information file S8 Table.

Funding: This work was supported by grants R01AI143638 and R21AI186221 to R.T. from the National Institute of Allergy and Infectious Disease. The UNC Proteomics Core Facility and the UNC Microscopy Services Laboratory (UNC Department of Pathology and Laboratory Medicine) are supported in part by Center Core Support Grant (P30CA016086) to the UNC Lineberger Comprehensive Cancer Center from the NIH National Cancer Institute. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Many bacterial species generate phenotypic heterogeneity to increase the fitness of a population as a whole. Some species benefit from phenotypic heterogeneity due to the division of labor among the distinct subpopulations that specialize in performing different tasks. Phenotypic diversity may also serve as a bet-hedging strategy, increasing the chances of a subpopulation surviving an environmental stress [1].

Clostridioides difficile is an intestinal pathogen with a significant global health burden, causing disease ranging from mild diarrhea to pseudomembranous colitis and sepsis [2,3]. Many C. difficile strains display colony dimorphism—on an agar surface, colonies appear rough with filamentous edges or smooth with rounded edges [4–6]. Prior work has established that C. difficile colony morphology is subject to phase variation, which generates a phenotypically heterogeneous population through stochastic, reversible ON/OFF modulation of gene expression [5,7]. Several other phenotypes are associated with each colony morphology. The rough colony variant forms elongated, chained cells and exhibits greater surface motility, while the smooth colony variant produces more biofilm and shows more swimming motility [5]. The rough colony variant also trends toward greater virulence than the smooth colony variant in the hamster model of acute C. difficile infection despite comparable toxin production.

Phase variation in C. difficile can occur through site-specific recombination that reversibly inverts regulatory DNA elements [7,8]. One of these elements, the cmr switch, controls colony dimorphism and the associated phenotypes [5]. The cmr switch contains a promoter that, when oriented toward cmrRST, drives transcription of the operon and leads to the cmr-ON state and rough colony development; when the cmr switch is in the inverse orientation, this transcription is lost, leading to the cmr-OFF state and smooth colonies (Fig 1A) [5,9].

thumbnail
Download:
Fig 1. RNA-Seq reveals an overlap in gene expression in wildtype rough and cmrΔ3-ON colonies.

(A) Diagram of the cmrRST operon with the ON or OFF cmr switch orientation (top and bottom, respectively) [5,7,9]. Arrows represent promoters; dashed lines represent transcription. The position of the c-di-GMP riboswitch that positively regulates expression is indicated. Created in BioRender. Mehra, A. (2025) https://BioRender.com/wqgy7oc. (B) Overview of study design. Wildtype rough, wildtype smooth, cmrΔ3-ON, cmrΔ3-OFF, and ΔcmrRΔcmrT strains were grown on BHIS-agar for 24 hours before collection of growth for RNA-Seq analysis. Arrows show the comparisons made. (C-D) Volcano plots of differences in transcript abundance between (C) wildtype rough versus smooth colonies and (D) cmrΔ3-ON vs. cmrΔ3-OFF mutants. Gray dotted lines demarcate cutoffs of > 2 log2 fold-change and/or FDR p-value < 0.01. Black dots indicate genes that met the cutoffs; gray dots indicate those that did not. Blue dots denote cmrRST. Orange dots highlight genes that were differentially expressed in the cmrΔ3-ON versus -OFF dataset and that met cutoffs.

https://doi.org/10.1371/journal.ppat.1013471.g001

The cmrRST operon encodes a signal transduction system: CmrS is a predicted histidine kinase for which the activating signal is unknown, and CmrR and CmrT are annotated as OmpR-family response regulators [10,11]. CmrR and CmrT each have an N-terminal phosphoreceiver domain and a C-terminal winged helix DNA binding domain, suggesting that they serve as transcriptional regulators [7,12]. These regulators appear to serve discrete roles. CmrR is required for intestinal colonization in the hamster model of infection, decreases biofilm production, and autoregulates the cmrRST operon, whereas CmrT is required for rough colony development, surface motility, cell chaining, and cell elongation [5,9]. Despite the broad effects of CmrRST on C. difficile physiology, the genes regulated by CmrR and/or CmrT that mediate cmr-associated phenotypes have not been identified. CmrRST may regulate a singular pathway that coordinately modulates all the cmr-associated phenotypes, or it may control multiple genes, each responsible for one or a subset of the phenotypes.

In this study, we used RNA-Seq to define the CmrR and CmrT regulons and their contributions to the rough and smooth colony transcriptomes. We identified two CmrRST-regulated genes, which we named modulators of rough phenotype A and B (mrpA and mrpB), that when co-expressed promote rough colony formation, surface motility, and cell elongation and chaining while impeding swimming motility. We found that MrpA interacts with septum site-determining protein MinD [13–15]. Consistent with this observation, expression of mrpAB induces abnormal cell division, with misplaced septa and elongated cells. Based on our results, we propose that phase variation of CmrRST, and thereby MrpAB, modulates cell division in C. difficile, with the cmr-ON state resulting in MrpAB-mediated disruption of proper cell division. The consequent elongated, chained cells may underlie the enhanced surface motility and reduced swimming motility of the rough colony variant.

Results

CmrRST accounts for some, but not all, of the transcriptional differences between rough and smooth colony variants

To determine the transcriptional differences that may drive colony dimorphism in C. difficile, we compared the transcriptomes of five strains grown on an agar surface. Specifically, rough and smooth colony isolates of wildtype R20291 were grown alongside phase-locked mutants, which were previously made by deleting three nucleotides in the right inverted repeat to prevent cmr switch inversion [9]. These cmrΔ3-ON and cmrΔ3-OFF mutants form exclusively rough and smooth colonies, respectively. Because cmrRST transcription is controlled by multiple promoters (Fig 1A) [9], a ΔcmrRΔcmrT mutant that forms only smooth colonies was also included. Fig 1B shows the five strains and the RNA-Seq comparisons conducted to identify differentially expressed genes (DEGs) that met cutoffs of log2 fold-change > 2 and FDR adjusted p-value < 0.01. Locus numbers formatted as CDR20291_XXXX in the R20291 genome (GenBank: FN545816.1) are abbreviated CDRXXXX below.

RNA-Seq analysis of the wildtype rough versus smooth colony variants revealed 49 DEGs (Fig 1C and S3 Table), 41 of which were more highly expressed in smooth colonies relative to rough colonies. Fifteen of these 41 genes are predicted to have phage-related functions, and ten are uncharacterized or hypothetical. Several metabolic genes had higher expression in smooth colonies including genes involved in the synthesis of riboflavin (ribH) and the utilization of ethanolamine (eutABCL and CDR1837–1838), mannitol (mtlFR), or pyruvate (ppdK). The remaining genes more highly expressed in smooth colonies have varying functions, including signaling, transport, and a protease. Of the eight genes more highly expressed in rough colonies, three have no predictive annotations (CDR1689, CDR1690, and CDR1914). CDR1911 encodes cell wall protein Cwp28 which contains three conserved cell wall binding domains shared amongst the Cwp family. Unlike many other proteins in the cell wall protein family, Cwp28 is not predicted to contain any additional functional domains [16–18]. The remaining four genes are involved in signaling (cmrRST and the c-di-GMP phosphodiesterase gene CDR2040) [19,20]. That the cmrRST operon has one of the highest fold-change increases in rough colonies is consistent with prior work showing greater cmrRST transcript in rough colonies than in smooth [5,9].

The cmrΔ3-ON and cmrΔ3-OFF transcriptomes had 18 DEGs (Fig 1D and S3 Table). Twelve genes that were differentially expressed between these strains were also differentially expressed among the rough versus smooth colony variants: cmrRST, CDR1689–1690, CDR1914, cwp28, CDR0921, eutBCL, and CDR1838 (S1 Fig).

Additional pairwise comparisons and principal component analyses illustrate the similarities between the smooth, cmrΔ3-OFF, and ΔcmrRΔcmrT colony transcriptomes and between the rough and cmrΔ3-ON colonies (S2 Fig). Notably, the only genes differentially expressed between ΔcmrRΔcmrT and either the smooth isolate or the cmrΔ3-OFF mutant were cmrR and cmrT, and phase variable gene cwpV, which can likely be attributed to stochastic inversion of the cwpV switch (S2A and S2B Fig and S5 Table) [7,21]. Comparing transcript abundance in cmrΔ3-ON versus ΔcmrRΔcmrT revealed 37 DEGs (S1D Fig and S4 Table), including all 18 DEGs identified in the cmrΔ3-ON versus -OFF comparison. Similarly, transcriptomic differences between the rough isolate and the ΔcmrRΔcmrT mutant encompassed all the DEGs between rough versus smooth colonies (S1C Fig and S4 Table). Overall, 33 of the 49 DEGs between rough and smooth colonies were unique to that RNA-Seq comparison. Collectively, these transcriptome analyses indicate that CmrRST accounts for many, but not all, transcriptional differences between rough and smooth colony variants despite the requirement for CmrRST in rough colony development.

The CmrR and CmrT regulons overlap with rough and cmr-ON regulons

We next sought to identify genes regulated by CmrRST. In the wildtype background, the individual cmrR and cmrT genes were ectopically expressed (“pCmrR” and “pCmrT”) under the control of an anhydrotetracycline (ATc)-inducible promoter, and RNA-Seq was used to compare the transcriptomes to similarly grown vector controls. In the pCmrR versus vector comparison, 15 genes were differentially expressed and met log2 fold-change > 2 and FDR p-value < 0.01 cutoffs (Fig 2A and Table 1). Six of these 15 DEGs were similarly differentially expressed between cmrΔ3-ON and cmrΔ3-OFF colonies (CDR1689, cwp28, CDR1914, and cmrRST). Increased transcript abundance of all three cmrRST genes in the pCmrR strain aligns with previous work establishing an autoregulatory role for CmrR (Fig 1A) [9]. All six of these genes, as well as pdcB, pdcC, CDR2040, and CDR2074, were similarly differentially expressed between rough and smooth colonies (S1 Fig). We note that c-di-GMP phosphodiesterases PdcB and PdcC undergo phase variation, so the observed differences in their expression are likely independent of CmrR/CmrT [20,22,23]. While c-di-GMP can impact cmrRST expression [5,9,24], we do not believe phase variation of PdcB or PdcC substantially affected CmrRST-mediated regulation because fold-changes in differentially expressed genes are not correlated with pdcB and pdcC expression levels (Table 1).

thumbnail
Download:
Table 1. Genes differentially expressed with overexpression of cmrR or cmrT relative to the vector control strain.

https://doi.org/10.1371/journal.ppat.1013471.t001

thumbnail
Download:
Fig 2. RNA-Seq identified genes regulated by CmrR and CmrT.

(A-B) Volcano plots of differences in transcript abundances in strains grown in BHIS-Tm10-ATc broth. (A) Wildtype R20291 overexpressing cmrR (pCmrR) vs. the vector control. (B) Wildtype R20291 overexpressing cmrT (pCmrT) vs. the vector control. Gray dotted lines demarcate cutoffs of > 2 log2 fold-change and/or FDR p-value < 0.01. Black dots indicate genes differentially expressed genes that met cutoffs, gray dots indicate genes that did not meet the cutoffs, and blue dots denote cmrRST. Purple dots highlight genes that were differentially expressed in all four RNA-Seq comparisons (rough vs. smooth, cmrΔ3-ON vs. cmrΔ3-OFF, pCmrR vs. vector, pCmrT vs. vector). (C) Venn diagram of genes regulated by CmrR, CmrT, or both.

https://doi.org/10.1371/journal.ppat.1013471.g002

Comparison of C. difficile pCmrT to the vector control yielded 12 genes meeting the fold-change and p-value cutoffs (Fig 2B and Table 1). Six genes were downregulated by cmrT overexpression: the phosphodiesterase-encoding gene pdcC, three metabolism-related genes (hcp, CDR2075, and CDR2932), and osmoprotectant transport system genes (CDR3074–3075). The six upregulated genes encode a putative membrane protein (CDR1913), Cwp28, three uncharacterized proteins (CDR1689, CDR1690, CDR1914), and CmrT itself (confirming the successful overexpression of cmrT). Eight of the 12 CmrT-regulated genes were also differentially expressed between pCmrR and vector strains, indicating an overlap between CmrR and CmrT-regulated genes. Additionally, eight of the 12 CmrT-regulated genes were also differentially expressed between rough and smooth colonies, and five were differentially expressed between cmrΔ3-ON and cmrΔ3-OFF colonies (S1 Fig). Of note are four genes that were more highly expressed in the rough, cmrΔ3-ON, pCmrR, and pCmrT strains: cmrT, cwp28, and uncharacterized genes CDR1689 and CDR1914.

CmrT is the predominant regulator of genes outside the cmrRST locus

Because prior work established an autoregulatory role for CmrR, but not CmrT [9], we assessed the contribution of each response regulator to differential expression of the genes identified by RNA-Seq. In contrast to our RNA-Seq analysis using the wildtype background, here we individually expressed either cmrR or cmrT in a ΔcmrRΔcmrT mutant so that only one regulator is produced. We then used quantitative reverse-transcriptase PCR (qRT-PCR) to measure transcript abundance of genes that appeared to be regulated by CmrR only, CmrT only, or by both regulators (Fig 2C and Table 1).

RNA-Seq analysis identified six genes as regulated by both CmrR and CmrT in the wildtype background (Fig 2C and Table 1): CDR1689, cwp28, CDR1914, CDR2932, and CDR3074–3075. CDR2932 showed opposite regulation by CmrR and CmrT, while the other genes had similar trends with each regulator. Examination of the raw read count RNA-Seq data indicated that CDR2932 transcript abundance was low in all strains, so this gene was not investigated further. For CDR1689, cwp28, and CDR1914, expression of cmrR alone in the ΔcmrRΔcmrT background did not alter transcript abundance, while cmrT expression alone resulted in significantly increased expression of all three genes (8.6-fold, 1.9-fold, and 9.2-fold, respectively) (Fig 3A-C). CDR3075 was used as a proxy for both CDR3074–3075 because of a putative shared transcriptional start site upstream of CDR3075 [25]. Although no significant differences in transcript abundance were observed when either cmrR or cmrT was expressed, transcript abundance was lowest when cmrT was expressed than in all other strains tested (Fig 3D). These genes are thus regulated by CmrT, with indirect regulation by CmrR autoregulation of cmrRST transcription.

thumbnail
Download:
Fig 3. Relative contributions of CmrR and CmrT to gene regulation.

(A-H) Relative transcript abundance determined by reverse-transcriptase qPCR for CDR1689 (A), cwp28 (B), CDR1914 (C), CDR3075 (D), CDR1913 (E), CDR2075 (F), cmrS (G), and CDR2040 (H) in ΔcmrRΔcmrT ectopically expressing cmrR or cmrT. Cultures were induced with ATc at 10 ng/mL for pCmrR and the respective vector control (blue bars), and 2 ng/mL for pCmrT and the respective vector control (orange bars). Shown are means and standard error for three to four biological replicates. *p < 0.05, ***p < 0.001, unpaired two-tailed t-test with Welch’s post-test.

https://doi.org/10.1371/journal.ppat.1013471.g003

CDR1913 and CDR2075 were identified by RNA-Seq as genes exclusively regulated by CmrT. In the ΔcmrRΔcmrT background, CDR1913 transcript abundance was significantly increased only when cmrT was expressed (2.8-fold), but not cmrR (Fig 3E). We note that CDR1913 may also be indirectly regulated by CmrR, as RNA-Seq showed a significant difference in CDR1913 between pCmrR and vector, though this did not meet our cut-off for fold-change. Neither cmrR nor cmrT expression significantly altered CDR2075 transcript abundance (Fig 3F), suggesting both regulators are necessary for altered expression. CDR2040 and cmrS were exclusively regulated by CmrR in the RNA-Seq analysis. Consistent with previous work defining an autoregulatory role for CmrR [9], qRT-PCR showed ~5.6-fold higher cmrS transcript upon the expression of cmrR. We also observed ~1.5x higher cmrS transcript when cmrT was expressed (Fig 3G). Neither CmrR nor CmrT alone was sufficient to alter transcript abundance of CDR2040 by qRT-PCR (Fig 3H), suggesting roles for both regulators. Overall, these data indicate that most CmrRST-regulated genes require only CmrT for altered expression, cmrRST regulation occurs via CmrR, and a subset of genes may require both CmrR and CmrT for altered expression.

CDR1689–1690, CDR1913–1914, and CDR1929 influence surface motility and rough colony development

To identify genes that mediate CmrRST-mediated rough colony development and surface motility on agar medium [5], we determined the impact of deleting or ectopically expressing CmrR- and CmrT-regulated genes on these two phenotypes. First, each of the CmrR- and CmrT-regulated genes highlighted in Fig 2C was cloned for ATc-inducible expression [26]; neighboring genes or genes that appear to be part of an operon were co-expressed. Because cmrT is required for rough colony development and surface motility, the genes were expressed in the ΔcmrT background to evaluate their sufficiency for these phenotypes [5]. An inducible cmrT-expressing strain was included as a control. As seen previously, low-level induction of cmrT (2 ng/mL ATc) yielded rough colonies and increased surface motility, whereas higher inducer concentrations resulted in toxicity (Fig 4A and 4B) [5].

thumbnail
Download:
Fig 4. CDR1689-1690, CDR1913-1914, and CDR1929 promote surface motility and/or rough colony development.

(A) Assay to determine the roles of CmrR- and/or CmrT-regulated genes in surface motility. Overnight cultures of ΔcmrT strains overexpressing CmrR- or CmrT-regulated genes or a vector control were spotted on BHIS-Tm10-agar and ATc 0-100 ng/mL. Surface motility was measured after seven days at 37°C. Shown are means and standard deviation for three biological replicates. ****p < 0.0001, one-way ANOVA with Dunnett’s post-test. (B) Colony morphology of ΔcmrT strains overexpressing CmrR- and/or CmrT-regulated genes or the vector control grown on BHIS-Tm10-agar and ATc 0-100 ng/mL. Images were taken at 2X magnification after 24 hours at 37°C. Shown are representative images from four biological replicates.

https://doi.org/10.1371/journal.ppat.1013471.g004

While most strains resembled the vector control, two expression strains showed increased surface motility and a rough colony phenotype, and one strain had altered colony morphology with no discernable effect on surface motility. Specifically, expression of CDR1689–1690 with 100 ng/mL ATc restored surface motility and induced formation of rough colonies (Fig 4A and 4B). CDR1929 expression with 100 ng/mL ATc significantly increased surface motility (Fig 4A), but this ATc concentration resulted in sparse growth on agar medium (Figs 4B and S3). However, induction with a lower ATc concentration (10 ng/mL) yielded rough colonies (Fig 4B). In contrast, ectopic expression of CDR1913–1914 did not affect surface motility with any ATc concentration but resulted in an intermediate colony phenotype with 100 ng/mL ATc (Fig 4A and 4B).

To determine the requirement of the CmrRST-regulated genes for surface motility or rough colony phenotype, we successfully deleted CDR1689–1690, cwp28, CDR1913–1914, and CDR3074–3075. All the deletion mutants retained wildtype levels of surface motility, indicating that none are required for the phenotype (S4A Fig). Additionally, all strains retained the ability to form both rough and smooth colonies (S4A and S4B Fig). Therefore, while CDR1689–1690 are sufficient to promote surface motility and rough colony development, they are not required, suggesting that C. difficile has other mechanisms mediating these phenotypes. Based on these phenotypes, we named CDR1689 and CDR1690 modulators of rough phenotype, mrpA and mrpB, respectively. The role of CDR1929 could not be assessed because we were unsuccessful in creating a deletion mutant.

Expression of genes that promote the rough colony morphology also decrease swimming motility

C. difficile cmr-ON variants have reduced swimming motility compared to cmr-OFF and ΔcmrRΔcmrT [5,9]. To identify the CmrRST-regulated genes that inhibit swimming motility, we assessed the expression and deletion strains described above for swimming in soft agar medium. The ΔcmrT pCmrT strain, which exhibits reduced swimming motility, was included as a positive control [5]. Ectopic expression of three sets of genes (mrpAB, CDR1913–1914, and CDR1929) reduced swimming motility relative to the vector control, though not to the extent of pCmrT (S5A Fig). These genes were determined to promote rough colony formation, suggesting a correlation between rough colonies and decreased swimming (Fig 4B).

Swimming motility of the mrpAB, cwp28, CDR1913–1914, and CDR3074–3075 deletion mutants were also assayed for swimming motility alongside cmrΔ3-ON and ΔcmrRΔcmrT controls. After 48 hours of growth, swimming motility of all mutants was more similar to cmrΔ3-ON than to ΔcmrRΔcmrT, indicating that these DEGs are not required for CmrRST-mediated inhibition of swimming motility (S5B Fig).

Expression of mrpAB restores surface motility in a cmr-OFF population

Prolonged incubation on an agar surface promotes C. difficile surface motility and accumulation of cmr-ON cells [5]. Specifically, a cmr-OFF wildtype isolate grown on an agar surface exhibits surface motility and concurrent shift to a population with the cmr switch predominantly in the ON orientation. How selective pressures encountered during growth on an agar surface favor a cmr-ON population is unknown but presumably involves one or more CmrRST-regulated factors that enable surface motility. To explore this possibility, we first asked whether cmr switch inversion occurs during surface growth in the absence of surface motility. Wildtype C. difficile and a ΔcmrR mutant, which are capable of surface motility, and ΔcmrT and ΔcmrRΔcmrT mutants, which lack surface motility, were grown on BHIS-agar for 3 days [5,9]. Quantitative PCR (qPCR) was then used to quantify the proportion of bacteria with the cmr-ON switch orientation before and after growth. The starting populations of all four strains were predominantly cmr-OFF (Fig 5A). After 3 days of surface growth, the wildtype and ΔcmrR populations showed surface motility and a significant shift to cmr-ON, from ~10% cmr-ON in the starting populations to 54% and 70% cmr-ON in the surface growth, respectively (Fig 5A). These results indicate that CmrR is not required for phase variation to the cmr-ON state. In contrast, the ΔcmrT and ΔcmrRΔcmrT populations remained in the cmr-OFF state (Fig 5A), which suggests the phase variation to the cmr-ON state requires CmrT or CmrT-dependent surface motility.

thumbnail
Download:
Fig 5. Surface motility conferred by mrpAB expression alleviates accumulation of cmr-ON cells.

(A) qPCR was used to measure cmr switch orientation in WT, ΔcmrR, ΔcmrT, or ΔcmrRΔcmrT populations before (T0, blue bars) and after (Tf, orange bars) three days of growth on BHIS-agar. (B) qPCR analysis of cmr switch orientation in the ΔcmrT mutant with vector or ectopically expressing mrpAB in the starting culture (T0, blue bars) and after seven days of growth on BHIS-Tm10-ATc100-agar (Tf, orange bars). (A, B) Below the x-axis are representative images of the strains after three (A) or seven (B) days of growth on BHIS-agar. Shown are the means and standard error for 4 – 5 biological replicates. Symbols are used to differentiate each biological replicate. Means are shown above bars as needed. ****p < 0.0001, two-way ANOVA with Tukey’s post-test. “ns”, not significant.

https://doi.org/10.1371/journal.ppat.1013471.g005

We next sought to distinguish whether the shift to a cmr-ON population on an agar surface requires CmrT or CmrT-dependent surface motility. We ectopically expressed mrpAB in the ΔcmrT background, then measured the proportion of cells with the cmr-ON switch orientation before and after growth on BHIS-agar. The ΔcmrT mutant bearing the vector served as a control. While ectopic expression of mrpAB restored surface motility as expected, the population remained cmr-OFF after seven days of growth (Fig 5B). These data indicate that the cmr-ON orientation is not required for surface motility, as mrpAB expression restored surface motility in a predominantly cmr-OFF population.

MrpA and MrpB are small proteins that are highly conserved in C. difficile

The role of MrpAB in several cmr-associated phenotypes led us to characterize these genes further, first by determining the contribution of the individual genes to surface motility and rough colony development. When ectopically expressed individually in the ΔcmrT background, neither mrpA nor mrpB were sufficient to confer surface motility (S6A Fig) or rough colony morphology (S6B Fig), suggesting that both of these genes work coordinately to produce phenotypes associated with the cmr-ON state. Consistent with this result, we detected a transcript that spanned the length of the two genes indicating that mrpA and mrpB can be co-transcribed (S6C Fig).

MrpA and MrpB are annotated as hypothetical proteins of 81 and 87 amino acids, respectively. To confirm that mrpA and mrpB are translated into proteins, we constructed C-terminal FLAG-tag translational fusions for mrpA and mrpB, then expressed them individually in E. coli DH5⍺ and assessed translation via western blotting. For both MrpA and MrpB, appropriately sized bands based on predicted molecular weight were detected, and the abundance of the proteins correlated with ATc concentration (Fig 6A). Three-dimensional modeling with ColabFold [27] predicts the structures of MrpA and MrpB with high confidence using a per-residue confidence metric (predicted local distance difference test (plDTT) of > 90). MrpA is predicted to consist of five antiparallel β-sheets and one ⍺-helix towards the C-terminus (Fig 6B). MrpB is predicted to contain four antiparallel β-sheets on one face of the protein, three antiparallel β-sheets on the other face, and one ⍺-helix (Fig 6C). Because mrpA and mrpB are both required to restore motility and rough colony morphology in the ΔcmrT mutant (S6A and S6B Fig), we also used ColabFold to predict the likelihood of heterodimer formation. Dimer modeling did not return a high confidence score where the two proteins are predicted to interact (plDTT ~ 70), suggesting low likelihood of heterodimer formation (S7A Fig). This prediction does not, however, account for multimers or additional protein partners that could influence protein conformation.

thumbnail
Download:
Fig 6. MrpA and MrpB encode small, uncharacterized proteins and are conserved in C. difficile.

(A) FLAG-tag translational fusions of MrpA and MrpB were ectopically expressed in E. coli DH5⍺ cells with ATc 0-500 ng/mL, as indicated. Translated proteins were detected via western blot with anti-FLAG antibodies. MrpA-FLAG is predicted to be 13.9 kilodaltons (kDa); MrpB-FLAG, 14.5 kDa. (B, C) Protein structures of MrpA (B) and MrpB (C) as predicted by ColabFold. Colors display the plDDT per-residue confidence metric indicated in the legend (bottom). (D, E) Amino acid conservation for MrpA (D) and MrpB (E) was determined across the 212 complete C. difficile genomes available on NCBI with RefSeq annotations. Along the x-axis are the amino acid sequences for each protein. Gray bars indicate the percentage of sequences that had the same amino acid as C. difficile R20291. Red and purple bars indicate the percentage of genomes that had a variant amino acid residue at that site; the substitutions are shown below the R20291 sequence in the respective red or purple text.

https://doi.org/10.1371/journal.ppat.1013471.g006

NCBI BLASTp revealed that both proteins are relatively restricted to C. difficile, with orthologs present in a limited number of closely related, unclassified Clostridioides, and the MrpB DUF1883 domain appearing in a range of bacterial genera [11,28]. To evaluate conservation of MrpA and MrpB across C. difficile strains, we searched for these amino acid sequences in the 212 complete genomes available on NCBI with RefSeq annotations. Both proteins were highly conserved in all 212 genomes that included strains from all five classical C. difficile clades, and most mutable residues had conservative substitutions (Fig 6D and 6E).

MrpA interacts with MinD, a septum site-determining protein

MrpA and MrpB influence cmr-associated phenotypes through an unknown mechanism. While MrpA has no predicted protein family motifs, MrpB has three: a domain of unknown function (DUF1883) that spans the majority of the length of the protein (residues 1–85), an overlapping bacterial pre-peptidase C-terminal domain (PPC, residues 5–70), and a short Src homology 3 domain (SH3_2, residues 8–23) [11]. DUF1883 and PPC domains have similar predicted functions, with the DUF1883 domain predicted to be a ligand-binding domain of a PPC-like protein [29]. In eukaryotes, SH3 domains are involved in protein-protein interactions that regulate the cytoskeleton [30,31]. Bacterial SH3 domains have been shown to mediate protein-protein and protein-peptidoglycan interactions [32–34]. Given the wide range of potential functions by these domains and small size of MrpA and MrpB, we postulated that MrpAB mediate C. difficile phenotypes by interacting with other proteins. To identify potential interacting partners, we used affinity purification and mass spectrometry with the C-terminally FLAG-tagged MrpA or MrpB as bait. The MrpA-FLAG and MrpB-FLAG translational fusions were expressed in a C. difficile rough colony variant background to increase the chances of capturing protein-protein interactions that occur in rough or cmr-ON colonies. Following immunoprecipitation and liquid chromatography-tandem mass spectrometry, enriched proteins were identified by comparing the peptides associated with the FLAG-tagged proteins to those in the vector control.

Pulldowns with MrpA identified 258 proteins meeting the criteria of log2 fold-change ≥ 1, q-value < 0.05, and sequence coverage > 50% (S6 Table). As expected, MrpA was the most abundantly detected protein, confirming successful overproduction of the bait protein. pMrpB-FLAG pulled down 197 proteins meeting the above criteria, with MrpB being the most abundant (S7 Table). Notably, MrpB pulled down MrpA with a log2 fold-change of 3.45 over the vector control. The reciprocal interaction was not observed; it is possible that the C-terminal FLAG-tag on MrpA interfered with its ability to pull down MrpB. Overall, 117 proteins uniquely interacted with MrpA, 56 proteins were unique to MrpB, and 141 proteins were found to interact with both MrpA and MrpB.

The most enriched proteins for MrpA are the septum site-determining protein MinD (CDR0987, DivIVB), a probable glycine dehydrogenase (CDR1556, GcvPB), and a phosphotransferase system (PTS) component (CDR2780, LicB) (S6 Table). MrpB also co-purified with MinD and GcvPB, though the proteins most enriched by MrpB were an uncharacterized protein (CDR2123) and two components of a PTS (CDR2902 and 2901), followed by MrpA (S7 Table). Given that CmrRST promotes cell elongation and chaining [5], we became interested in the interaction of MrpA with MinD. Of note, multiple other proteins involved in cell shape or structure were enriched by both MrpA and MrpB, including septum site-determining protein MinC, S-layer protein SlpA, cell-shape determining protein MreB2, peptidoglycan-crosslinking enzyme Cwp22, and late-stage cell division protein DivIVA (CDR2503). MrpA also interacted with cell division protein FtsZ. These findings suggest a role for MrpA and MrpB in cell morphology or division. None of the identified proteins were identified in the RNA-Seq analyses.

To validate the interaction of MrpA and MrpB with MinD, we used a bacterial two-hybrid system in E. coli [35,36]. We created N- and C-terminal fusions of MrpA, MrpB, and MinD to the T18 and T25 subunits of the adenylate cyclase CyaA, then examined pairwise interactions via CyaA-dependent β-galactosidase activity. Both N- and C-terminal fusions of MrpA interacted with the respective MinD fusions (Figs 7A and S7B and S7C). We were unable to confirm interactions between MrpA and MrpB or between MrpB and MinD, nor did we detect homodimerization of MrpA, MrpB, or MinD (Figs 7A and S7B and S7C). These negative results are inconclusive, since the fusion of the large CyaA subunits with these small C. difficile proteins may interfere with proper protein folding or binding, the proteins may not interact directly, or E. coli cell division protein homologs may compete for binding. Together, the pulldown experiments and bacterial two-hybrid assay reveal an interaction between MrpA and MinD.

thumbnail
Download:
Fig 7. MrpA interacts with MinD.

(A) Bacterial two-hybrid assay in which MrpA, MrpB, and MinD were translationally fused to the N-terminus of the T18 CyaA fragment and tested for interactions with MrpA, MrpB, or MinD fused to the N- or C-terminus of the CyaA T25 fragment. Overnight cultures of E. coli BTH101 carrying the indicated plasmid pair were spotted on LB-Kan50-Amp100-agar supplemented with 40 μg/mL X-gal and 0.5 mM IPTG. A T18 fragment with no translational fusion (“empty”) is included as a negative control. The black box indicates the zip-zip positive control. (B) FM 4-64 was used to stain membranes of ΔcmrT cells ectopically expressing mrpAB, cmrT, or a vector control. Overnight cultures were grown in TY-Tm10 broth supplemented with ATc 2 ng/mL (vector and pCmrT) or ATc 100 ng/mL (vector and pMrpAB). Cells were imaged at 100X. Shown are representative images from two biological replicates. White arrows indicate asymmetric septa. (C) Quantification of cell lengths from FM 4-64 staining. Cells were measured from at least 10 different images from two biological replicates. Vectors induced with ATc 2 and ATc 100 ng/mL were combined; vector n = 359 (ATc 2 ng/mL, red dots) and n = 262 (ATc 100 ng/mL, black dots), pMrpAB n = 478, pCmrT n = 444). ****p < 0.0001, one-way ANOVA with Dunnett’s post-test. (D) Scanning electron microscopy images of ΔcmrT ectopically expressing mrpAB, cmrT, or a vector control grown on BHIS-Tm10-ATc100-agar for 4 days. White arrows indicate asymmetric septa. Cells were imaged at 10,000X.

https://doi.org/10.1371/journal.ppat.1013471.g007

To determine if the MrpA-MinD interaction affects C. difficile cell division, we used membrane stain FM 4–64 to visualize cell morphology and septum formation in ΔcmrT cells ectopically expressing mrpAB, cmrT, or carrying vector controls. We found that mrpAB expression led to elongated cells, though fewer chained cells appeared compared to the pCmrT positive control (Fig 7B-D). Atypical cell division was apparent in both the pCmrT and pMrpAB strains; asymmetric or polar septa were present in both strains, and multiple septa also occurred, unlike in the vector control. These data corroborate the pulldowns and bacterial-two hybrid assays by providing phenotypic evidence that MrpA, and potentially MrpB, interact with MinD to disrupt C. difficile cell division.

Discussion

C. difficile colony morphology and multiple additional associated phenotypes undergo phase variation via the ON/OFF switching of CmrRST production. In this study, we used RNA-Seq analysis to characterize the transcriptional differences between the rough and smooth colony variants, determined the contribution of CmrRST to these differences, and identified two CmrRST-regulated genes, mrpAB, that promote the rough colony formation and surface motility.

Our multi-pronged RNA-Seq study revealed that CmrRST accounts for many, but not all, transcriptional differences between rough and smooth colony variants. Specifically, our results suggest that CmrRST activity drives development of rough colonies. Targeted analysis of these transcripts by qRT-PCR indicated that CmrT is the more immediate regulator of genes outside the cmrRST locus. For instance, expression of cmrT in the absence of cmrR was sufficient to alter mrpA, cwp28, CDR1913, and CDR1914 transcription. Whether CmrT directly activates transcription remains unknown, and there is no obvious consensus binding site upstream these genes. Expression of cmrR alone did not significantly affect transcription of any genes other than cmrRST. CmrR may strictly serve an autoregulatory role, indirectly regulating genes by controlling cmrT expression [9]. Because CmrT is a pseudoreceiver containing a glutamic acid residue at the usual response regulator phosphorylation site, modulation of CmrT levels may be sufficient to mediate cmr-associated phenotypes [5,9,37]. It remains possible that CmrR works in conjunction with CmrT for a subset of genes: expression of cmrR or cmrT individually was insufficient to alter expression of at least three genes (CDR2040, CDR2075, and CDR3075). We speculate that CmrR and CmrT may form heterodimers to control expression of a subset of genes, or there may be a more complex regulatory relationship between these two proteins and the CmrS sensor kinase [5,9,38].

Multiple genes were consistently associated with the rough colony/cmr-ON state, including mrpAB, cwp28, CDR1914, and CDR3074–3075. Through overexpression, we found that mrpAB and CDR1929 expression in a ΔcmrT mutant restored both rough colony formation and surface motility while inhibiting swimming motility. While CDR1929 was not further pursued due its toxicity when overexpressed, mrpAB was investigated for a role in CmrRST phase variation. We previously showed that growth of C. difficile on an agar surface selects for the cmr-ON variant [5]. We speculate that this selection is driven by the depletion of nutrients or accrual of a stress locally, such that C. difficile cells must expand outward. In the absence of cmrT, surface motility is lost along with the accumulation of cmr-ON cells. Expression of mrpAB in the ΔcmrT mutant restored surface motility in a population that remained primarily cmr-OFF, suggesting that MrpAB-mediated surface motility can occur independent of cmr-ON population expansion. These results suggest that expression of mrpAB and the restored surface motility ameliorates selection of cmr-ON cells during growth on BHIS-agar, though it remains possible that CmrT play a more complex role in modulation of cmr switch inversion.

Notably, none of the CmrRST-regulated genes were required for rough colony formation or surface motility. Redundancy is unlikely given the diverse gene functions. Our stringent criteria for differentially expressed genes may have excluded candidate genes. For example, cwlA, which encodes an endopeptidase necessary for cell separation [39,40], showed ~2-fold lower expression in the rough colony variant, the cmrΔ3-ON mutant, and the pCmrR strain compared to the respective smooth, cmrΔ3-OFF, or vector counterparts. Downregulation of cwlA could contribute to cell separation defects that result in chained cells in rough colonies and enhanced surface motility.

MrpA was found to interact with the septum site-determining protein MinD, revealing a potential mechanism for CmrRST-mediated phenotypes. The Min proteins are well characterized in E. coli and B. subtilis [13,14,41]. In E. coli, MinE causes oscillation of MinD between cell poles, and MinD interacts with MinC, an inhibitor of FtsZ protofilaments. The accumulation of MinCDE at the poles help localize FtsZ to the mid-cell for septum formation. In B. subtilis, DivIVA localizes MinCD to cell poles to create a stationary gradient that promotes FtsZ polymerization at the mid-cell [14,15,42]. C. difficile appears to have a hybrid of these two systems by encoding MinCDE and DivIVA, though the dynamics of these proteins are not well established [43,44]. In both B. subtilis and E. coli, minD mutants undergo abnormal cell division by forming both elongated and minicells [45,46]. Notably, a B. subtilis minCD mutant displays asymmetric, asynchronous cell division that results in elongated cells with irregularly spaced septa [47]. We observed a similar phenotype in C. difficile expressing mrpAB. We postulate that the MrpA-MinD interaction disrupts typical cell division to result in the observed elongated, asymmetrically dividing cells. MrpA may reduce MinD activity or interfere with interactions with other cell division proteins.

The role of MrpB is unclear. Its co-expression with MrpA is required to elicit rough colony development and surface motility. Pulldown experiments indicated that MrpA and MrpB may interact, though this latter result could not be corroborated with a bacterial two-hybrid assay. MrpB contains a bacterial SH3 domain, which frequently targets cell wall motifs and is often found fused to cell wall hydrolases [32–34], and a pre-peptidase domain. MrpB may thus be part of a larger complex influencing cell division. Supporting this hypothesis, we found that both MrpA and MrpB interacted with several proteins involved in cell shape and cell division that are already known to interact, including MinD, MinC, DivIVA, and FtsZ [41,43,44,48].

Based on these and previous results, we propose that the elongated, chained cells in the cmr-ON rough variant contribute to surface migration as non-separated cells are pushed along the axis of cell division and growth [5,49]. This process resembles sliding motility, in which expansion of a colony is driven by cell division and population growth [50–56]. In B. subtilis, sliding motility is the result of differentiated cell types; surfactin-producing cells facilitate colony expansion by matrix-producing cells that form long, filamentous loops called “van Gogh bundles” that exhibit sliding motility and have impaired swimming motility [52]. In C. difficile, inhibition of swimming motility may similarly be an indirect consequence as daughter cells remain tethered to mother cells, physically impeding flagellum-dependent motility. While no surfactant has been identified in C. difficile thus far, some bacteria exhibit sliding motility via alternate mechanisms. In Salmonella enterica serovar Typhimurium, sliding motility is thought to be facilitated by a surface protein [53,57]. Mycobacterium smegmatis, which also displays colony dimorphism correlated with motility and virulence, appears to rely on glycopeptidolipids for sliding motility [50,58,59].

Why certain environments favor the cmr-ON or cmr-OFF states is unknown. Presumably, the ability to migrate via different mechanisms allows access to nutrients (or avoidance of repellants), conferring a fitness advantage to the subpopulation that is more motile in the given environment. Consistent with this idea, many of the genes upregulated in smooth colonies function in metabolism, and computational modeling predicted the variants have distinct metabolic needs [60]. Specifically, the smooth colony variant, but not the rough, was predicted to rely on glucose via the pentose phosphate pathway—a prediction supported by experimental evidence that in the absence of glucose, C. difficile develops rough colonies. Ongoing studies will examine the contributions of CmrRST-regulated genes and the dynamics of cmr switch inversion during infection of a murine model of disease, which may elucidate the effects of C. difficile phase variation on its survival in a complex and changing intestinal environment.

Methods

Bacterial strains and growth conditions

C. difficile R20291 strains were grown statically at 37°C in an anaerobic chamber (Coy Laboratories) with an atmosphere of 85% N2, 5% CO2, and 10% H2 (Airgas). C. difficile was grown from freezer stocks on BHIS (37 g/L Brain Heart Infusion (BD 211059), 5 g/L yeast extract (Gibco 212750), and 1 g/L cysteine) with 1.5% agar. Overnight cultures (16–18 hours of growth) were grown in TY broth (30 g/L tryptone, 20 g/L yeast extract, 1 g/L sodium thioglycolate) and subcultured as indicated for each assay below. Media were supplemented with 10 μg/mL thiamphenicol (Tm10) for plasmid maintenance as needed. Ectopic expression of genes from the pRPF185 vector was induced with anhydrotetracycline (ATc, 0–100 ng/mL, as indicated). E. coli was grown aerobically in Luria-Bertani (LB) broth (Fisher BP1426 for broth, Fisher BP1425 for agar) at 37°C (for DH5⍺ and HB101(pRK24)) or 30°C (for BTH101). Antibiotics were included in the media as appropriate at the following concentrations: 10–20 μg/mL chloramphenicol (Cm10 or Cm20) and 100 μg/mL ampicillin (Amp100).

S1 Table lists all strains used in this study, and oligonucleotides are in S2 Table. Details on plasmid and strain construction can be found in S1 Text. Sequence-verified plasmids were transformed into HB101(pRK24) for conjugation into C. difficile R20291 as previously described [61]. C. difficile expression strains were confirmed by PCR, and deletion mutants were confirmed with sequencing.

RNA-Seq

Rough and smooth colony isolates of wildtype R20291, cmrΔ3-ON and cmrΔ3-OFF mutants, and the ΔcmrRΔcmrT mutant were spotted (10 μL) on BHIS-1.5% agar in triplicate. After 24-hours incubation at 37°C, the growth was collected in TRIzol for RNA extraction isopropanol precipitation as described previously [61,62]. Plasmid-bearing strains (pCmrR, pCmrT, or vector only) were grown from a single colony in TY-Tm10 broth overnight, then diluted 1:30 in BHIS-Tm10 broth. At an optical density at 600 nm (OD600) of ~ 0.3, expression was induced with ATc. C. difficile with pCmrR or the vector control were induced with 10 ng/mL ATc; because higher expression of cmrT is toxic to C. difficile [5], the pCmrT and vector control strains were induced with 2 ng/mL ATc. Strains with pCmrR or pCmrT were grown in triplicate, and vector controls were grown in duplicate. At OD600 ~ 1, 2 mL of each culture was collected for RNA purification. RNA was purified by TRIzol extraction and isopropanol precipitation [61,62].

Purified RNA was submitted to Genewiz (Azenta Life Sciences) for paired-end sequencing with kits used per manufacturer’s instructions. Briefly, rRNA was depleted using the Ribo Zero rRNA Removal Kit (Illumina). RNA sequencing libraries were prepared with the NEBNext Ultra RNA library prep kit for Illumina (NEB) and checked using a Qubit 2.0 fluorometer. The libraries were multiplexed for sequencing using a 2 × 150 paired-end configuration on an Illumina HiSeq 2500 instrument. Image analysis and base calling were done using the HiSeq control software. The resulting raw sequence data files (.bcl) were converted to the FASTQ format and demultiplexed with bcl2fastq 2.17 software (Illumina). One mismatch was permitted for index sequence identification. Sequencing files are available on NCBI GEO (Accession: GSE305463).

Differential expression analysis was done as previously described using established bioinformatics tools [63]. Briefly, Illumina reads were trimmed using Trimmomatic [64] then mapped to the C. difficile R20291 genome (GenBank: FN545816.1) using Bowtie2 [65]. Mapped reads were then assigned to genes via FADU [66]. DESeq2 analysis was used for pairwise comparison between the indicated strains [67].

Quantitative reverse transcriptase PCR

Overnight cultures of C. difficile were diluted 1:30 in BHIS-Tm10 and induced with ATc (2 ng/mL for pCmrT and vector control, 10 ng/mL for pCmrR and vector control) at OD600 ~ 0.3. At OD600 ~ 1, 2 mL of culture was pelleted and stored in 1:1 ethanol:acetone at -80°C. RNA was extracted as described previously [61,62]. Purified RNA was reverse transcribed and quantitative reverse transcriptase PCR (qRT-PCR) was carried out as previously described with normalization to reference gene rpoC [5,23,68]. No reverse transcriptase controls were run in parallel for all samples. Primer sequences can be found in S2 Table.

Orientation-specific qPCR

Overnight cultures were diluted 1:30 in 3 mL BHIS broth. At OD600 ~ 1, 5 μL of each culture was spotted on three BHIS-1.5% agar plates; 2 mL of the inoculating cultures were collected for genomic DNA extraction as previously described [61]. After three days, surface growth from three spots was collected and pooled, standardized by OD600 to collect roughly the same number of cells, and pelleted for genomic DNA extraction. For ectopic expression of mrpAB (CDR1689–1690) in the ΔcmrT mutant background, overnight cultures (5 μL) were spotted on BHIS-Tm10-ATc100-1.5% agar plates, and 2 mL of the inoculating culture were reserved for DNA isolation. After seven days, surface motility was measured, and each spot was collected independently for genomic DNA extraction.

All genomic DNA samples from the inoculating cultures and subsequent surface growth were assessed for cmr switch orientation with orientation-specific qPCR as described previously [5,23]. The percentage of the population with the cmr switch in the ON orientation was determined as described previously using rpoA as the reference gene [7]. Primer sequences can be found in S2 Table.

Multiple sequence analysis for MrpA and MrpB conservation

All Clostridioides difficile genomes available on NCBI were filtered to include only complete genomes that were annotated by NCBI RefSeq. The resultant 212 genomes and fasta files for proteins MrpA and MrpB were downloaded. FaToTwoBit Version v377 was used to convert each of the downloaded fna files in the genome directories to a 2bit format. BLAT version 36x2 [69] was then used to identify the DNA sequences corresponding to those proteins by querying the 2bit genomes for the protein sequence contained in the fasta files. The protein-coding sequences for each genome were then extracted into another fasta file, with the start and end points corrected to accommodate the start and stop codons for each sequence. Reverse-stranded sequences were base-complemented and reversed. Mafft version 7.525 [70] was used to align the resulting base sequences. Expasy’s Translate Tool [71] was then used to identify the correct reading frame for the base sequences, and then Python Version 3.13.3 and Biopython Version 1.85 [72] were used to convert the aligned DNA sequences to amino acid sequences. R version 4.4.0 and GraphPad Prism were used to produce the conservation plots from the amino acid sequences.

Affinity pulldowns

MrpA and MrpB with a C-terminus serine-glycine linker and 3X-FLAG tag were expressed in C. difficile using the pRPF185 tetracycline-inducible expression vector (S1 Table and S1 Text) [26]. To enrich for the rough colony morphotype, which expresses cmrRST at higher levels [5], overnights of C. difficile harboring pMrpA-FLAG, pMrpB-FLAG, or the vector control were spotted onto BHIS-Tm10-agar. After seven days of incubation at 37°C, rough tendrils were isolated and used for affinity co-purification.

Overnight cultures grown in 5 mL TY-Tm10 were diluted 1:100 into 500 mL BHIS-Tm10. Expression was induced with 100 ng/mL ATc when cultures reached an OD600 ~ 0.2-0.3. At OD600 ~ 1, cells were collected by centrifugation, supernatants were discarded, and pellets were stored overnight at -80°C. Once thawed, cell pellets were washed three times with 10 mL 1x PBS (137 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4, 1.5 mM KH2PO4) to remove residual culture medium, with centrifugation between washes occurring at 4,000 rpm, 10 minutes, 4°C. Washed cells were suspended in 3 mL of TNG buffer (50 mM Tris pH 6.8, 150 mM NaCl, 10% glycerol) supplemented with Pierce Protease Inhibitor Tablets (Thomas Scientific A32953) then lysed by bead beating (4x for 1 minute, 5 minutes on ice between rounds). Lysates were centrifuged at 10,000 x g for 15 minutes at 4°C, then supernatants were transferred to Protein LoBind tubes. Protein concentrations of the lysates were measured using the Pierce™️ BCA Protein Assay Kit (Thermo Scientific A55864). For each sample, 4 mg total protein was added to 50 μL of packed anti-FLAG M2 magnetic beads (Sigma M8823) equilibrated in TNG buffer. Samples were incubated on an end-over-end rotor at 4°C overnight. Beads were washed six times in 1x PBS, suspended in 70 μL PBS, then transferred to a new Protein LoBind tube. A 5% aliquot of each sample was saved for western blot analysis, and the remainder was stored at -80°C. Protein identification via LC-MC/MS is detailed in S1 Text.

Microscopy

For imaging whole colonies, strains were streaked from freezer stocks on BHIS-agar (and Tm10 as required) and incubated for one day. Isolated colonies were streaked again on BHIS-agar without 0.1% cysteine supplementation (with Tm10 as required). After 24 hours of incubation, colonies were imaged at 2X magnification on a BZ-X810 microscope (Keyence).

Microscopy to determine cell morphology was conducted on ΔcmrT vector, pCmrT, or pMrpAB strains grown overnight in TY-Tm10 broth supplemented with ATc 2 ng/mL (vector or pCmrT) or 100 ng/mL (vector and pMrpAB). For membrane staining, overnight cultures were suspended with 1 μg/mL FM 4–64 dye (N-(3-triethylammoniumpropyl)-4-(6-(4-(diethylamino) phenyl) hexatrienyl) pyridinium dibromide) (Invitrogen T3166) for 15–30 minutes then pipetted on microscope slides covered with agarose pads (1% agarose in water heated to dissolve, pipetted onto the microscope slide while warm, then cut to coverslip size when solidified). Cells were imaged at 100X magnification on a BZ-X810 microscope (Keyence). Cell lengths were measured using ImageJ software.

Scanning electron microscopy is detailed in S1 Text.

Bacterial two-hybrid

The mrpA, mrpB, and minD (CDR0987) sequences were codon-optimized for translation in E. coli and synthesized by Genewiz. The DNA fragments were cloned into pKT25, pKNT25, pUT18, and pUT18C to obtain N- and C-terminal fusions to CyaA domains (Euromedex) (S1 Text and S1 Table) [35,73]. Pairs of plasmids to be tested were co-transformed into chemically competent E. coli BTH101. Overnight cultures (5 μL) of these BTH101 strains were spotted on LB-agar supplemented with 40 μg/mL 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal), 0.5 mM isopropyl β-D-thiogalactopyranoside (IPTG), Amp100, and Kan50. Plates were imaged using standard photography after 24–36 hours of growth at 30°C.

Supporting information

S1 Fig. RNA-Seq reveals an overlap in transcriptional profiles of rough colony variants, cmrΔ3-ON, pCmrR, and pCmrT strains.

Venn diagram of differentially expressed genes (DEGs) as determined by RNA-Seq analysis. Genes in the circle labeled “Rough” were differentially expressed in the Rough vs. Smooth RNA-Seq comparison. Similarly, “cmrΔ3-ON” contains DEGs from the cmrΔ3-ON vs. -OFF comparison, “pCmrR” contains DEGs from the pCmrR vs. vector comparison, and “pCmrT” contains DEGs from the pCmrT vs. vector comparison. Genes in blue text had increased transcript abundance and genes in red text had decreased transcript abundance in the above comparisons. Green text indicates genes that appeared to be regulated in opposite directions (e.g., higher transcript abundance in one comparison and decreased transcript abundance in the other).

https://doi.org/10.1371/journal.ppat.1013471.s001

(TIF)

S2 Fig. RNA-Seq reveals that CmrR and CmrT play important roles in distinguishing rough and smooth colony transcriptomes.

(A-D) Volcano plots of differences in transcript abundance in strains grown on BHIS-agar for 24 hours and collected for RNA-Seq analysis. Gray dotted lines demarcate cutoffs of > 2 log2 fold-change and/or FDR p-value < 0.01. Black dots are genes that met cutoffs, gray dots indicate genes that did not meet the cutoffs, and blue dots denote cmrRST. (A) Wildtype smooth versus ΔcmrRΔcmrT colonies. (B) cmrΔ3-OFF versus ΔcmrRΔcmrT colonies. (C) Wildtype rough versus ΔcmrRΔcmrT colonies. Green dots are genes that met cutoffs and were also differentially expressed in the rough vs. smooth comparison (Fig 1C and Table 1). (D) cmrΔ3-ON versus ΔcmrRΔcmrT colonies. Green dots are genes that met cutoffs and were also differentially expressed in the cmrΔ3-ON vs. cmrΔ3-OFF comparison (Fig 1D and S3 Table). (E) A principal component analysis plot showing clustering of the wildtype rough, wildtype smooth, cmrΔ3-ON, cmrΔ3-OFF, and ΔcmrRΔcmrT colony transcriptomes (n = 2–3). Plot was generated via DESeq2 with regularized logarithm transformed data.

https://doi.org/10.1371/journal.ppat.1013471.s002

(TIF)

S3 Fig. Surface motility morphology of strains ectopically expressing CmrR- or CmrT-regulated genes in the ΔcmrT background.

Surface motility plates measured in Fig 4 were imaged with a Syngene G:Box after seven days at 37°C. Shown are representative images from three experiments. Images are not to scale.

https://doi.org/10.1371/journal.ppat.1013471.s003

(TIF)

S4 Fig. CDR1689–1690, cwp28, CDR1913–1914, and CDR3074–3075 are not required for surface motility and/or rough colony development.

(A) Surface motility assay of strains with CmrR- or CmrT-regulated gene deletions along with wildtype and ΔcmrRΔcmrT controls. Overnight cultures (5 μL) were spotted on BHIS-agar, and surface motility was measured after seven days at 37°C. Shown are means and standard deviation for three biological replicates. *p < 0.05, one-way ANOVA with Dunnett’s post-test. (B) Colony morphology of CmrR- and/or CmrT-regulated gene deletion strains grown on BHIS-agar along with wildtype rough, wildtype smooth, and ΔcmrRΔcmrT controls. Four colonies are shown per strain representing both rough and smooth colony morphologies. Images were taken at 2X magnification after 24 hours at 37°C. Shown are representative images from four biological replicates.

https://doi.org/10.1371/journal.ppat.1013471.s004

(TIF)

S5 Fig. Expression of mrpAB, CDR1913–1914, or CDR1929 decreases swimming motility.

(A-B) Swimming motility was assayed by inoculating a colony of each strain into soft agar (0.5X BHIS-0.3% agar), supplemented with Tm10 and ATc10 in (A). The diameter of motile growth was measured after 48 hours at 37°C. Shown are means and standard deviation for three biological replicates. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, one-way ANOVA with Dunnett’s post-test. (A) ΔcmrT strains ectopically expressing CmrR- or CmrT-regulated genes along with pCmrT and vector controls. (B) Strains with deletions of CmrR- or CmrT-regulated genes along with cmrΔ3-ON and ΔcmrRΔcmrT controls.

https://doi.org/10.1371/journal.ppat.1013471.s005

(TIF)

S6 Fig. mrpA and mrpB are co-transcribed and are both required to restore surface motility and rough colony morphology in ΔcmrT.

(A) Surface motility assay of ΔcmrT ectopically expressing mrpA, mrpB, or mrpAB along with a vector control. Overnight cultures were spotted on BHIS-Tm10-ATc100-agar, and surface motility was measured after seven days at 37°C. Shown are means and standard deviation for five biological replicates. ****p < 0.0001, one-way ANOVA with Dunnett’s post-test. (B) Colony morphology of ΔcmrT ectopically expressing mrpA, mrpB, or mrpAB along with a vector control. Strains were grown on BHIS-Tm10-ATc100-agar. Images were taken at 2X magnification after 24 hours at 37°C. Shown are representative images from four biological replicates. (C) mrpA and mrpB are co-transcribed. RNA was isolated from wildtype overexpressing cmrT and reverse transcribed to cDNA. Primers were designed to amplify regions within the mrpA and mrpB coding sequence, and a region bridging the intergenic sequence between these genes. Reactions with genomic DNA and no-reverse transcriptase templates were included as positive and negative controls, respectively.

https://doi.org/10.1371/journal.ppat.1013471.s006

(TIF)

S7 Fig. MrpA and MinD interact, but MrpA-MrpB or MrpB-MinD interactions could not be confirmed.

(A) Predicted dimer modeling of MrpA and MrpB by ColabFold, a derivative of AlphaFold2. Colors display the plDDT per-residue confidence metric indicated in the legend (bottom). (B-C) Bacterial two-hybrid results continued from Fig 7. The black square indicates the zip-zip positive control. (B) MrpA, MrpB, and MinD were translationally fused to the C-terminus of the T18 CyaA fragment and tested for interactions with MrpA, MrpB, or MinD fused to the N- or C-termini of the CyaA T25 fragment. A T18 fragment with no translational fusion (“empty”) was included as a negative control. (C) Negative controls for the bacterial two-hybrid assay. The T25 fragment with no translational fusion (“empty”) was tested for interactions with the T18 fragment.

https://doi.org/10.1371/journal.ppat.1013471.s007

(TIF)

S1 Table. Strains and plasmids used in this study.

https://doi.org/10.1371/journal.ppat.1013471.s008

(PDF)

S2 Table. Oligonucleotides used in this study.

https://doi.org/10.1371/journal.ppat.1013471.s009

(PDF)

S3 Table. Genes differentially expressed in rough vs. smooth colony variants and cmrΔ3-ON vs. cmrΔ3-OFF strains.

https://doi.org/10.1371/journal.ppat.1013471.s010

(XLSX)

S4 Table. Genes differentially expressed by wildtype rough and cmrΔ3-ON strains relative to a ΔcmrRΔcmrT mutant.

https://doi.org/10.1371/journal.ppat.1013471.s011

(XLSX)

S5 Table. Genes differentially expressed by wildtype smooth and cmrΔ3-OFF strains relative to a ΔcmrRΔcmrT mutant.

https://doi.org/10.1371/journal.ppat.1013471.s012

(XLSX)

S6 Table. Affinity-purification mass spectrometry results for FLAG-tagged MrpA pulldowns.

https://doi.org/10.1371/journal.ppat.1013471.s013

(XLSX)

S7 Table. Affinity-purification mass spectrometry results for FLAG-tagged MrpB pulldowns.

https://doi.org/10.1371/journal.ppat.1013471.s014

(XLSX)

S8 Table. Quantitative data file.

https://doi.org/10.1371/journal.ppat.1013471.s015

(XLSX)

S1 Text. Supplemental Materials and Methods.

https://doi.org/10.1371/journal.ppat.1013471.s016

(PDF)

Acknowledgments

This research is based in part upon work conducted using the UNC Proteomics Core Facility. Pulldown sample preparation and LC-MS/MS analysis was conducted by Thomas Webb and Dr. Aurora Cabrera, and data analysis was done by Angie Mordant and Dr. Laura Herring.

The UNC Bioinformatics and Analytics Research Collaborative determined the protein conservation of MrpAB. Dr. Corbin Jones, Dr. Austin Hepperla, and Tyler Interrante assisted Christopher J. Serody with the analysis.

Kristen White, Jillann Madren, and Victoria Madden at The Microscopy Services Laboratory, UNC Department of Pathology and Laboratory Medicine conducted the scanning electron microscopy.

References

  1. 1. Reyes Ruiz LM, Williams CL, Tamayo R. Enhancing bacterial survival through phenotypic heterogeneity. PLoS Pathog. 2020;16(5):e1008439. pmid:32437427
  2. 2. Smits WK, Lyras D, Lacy DB, Wilcox MH, Kuijper EJ. Clostridium difficile infection. Nat Rev Dis Primers. 2016;2:16020. pmid:27158839
  3. 3. Mylonakis E, Ryan ET, Calderwood SB. Clostridium difficile–Associated Diarrhea: A review. Arch Intern Med. 2001;161(4):525–33. pmid:11252111
  4. 4. Lipovsek S, Leitinger G, Rupnik M. Ultrastructure of Clostridium difficile colonies. Anaerobe. 2013;24:66–70. pmid:24120350
  5. 5. Garrett EM, Sekulovic O, Wetzel D, Jones JB, Edwards AN, Vargas-Cuebas G, et al. Phase variation of a signal transduction system controls Clostridioides difficile colony morphology, motility, and virulence. PLoS Biol. 2019;17(10):e3000379. pmid:31658249
  6. 6. Anjuwon-Foster BR, Tamayo R. Phase variation of Clostridium difficile virulence factors. Gut Microbes. 2018;9(1):76–83. pmid:28806147
  7. 7. Sekulovic O, Mathias Garrett E, Bourgeois J, Tamayo R, Shen A, Camilli A. Genome-wide detection of conservative site-specific recombination in bacteria. PLoS Genet. 2018;14(4):e1007332. pmid:29621238
  8. 8. Trzilova D, Tamayo R. Site-Specific Recombination - How Simple DNA Inversions Produce Complex Phenotypic Heterogeneity in Bacterial Populations. Trends Genet. 2021;37(1):59–72. pmid:33008627
  9. 9. Garrett EM, Mehra A, Sekulovic O, Tamayo R. Multiple Regulatory Mechanisms Control the Production of CmrRST, an Atypical Signal Transduction System in Clostridioides difficile. mBio. 2021;13(1):e0296921. pmid:35164558
  10. 10. Bateman A, Martin M-J, Orchard S, Magrane M, Adesina A, Ahmad S, et al. UniProt: the Universal Protein Knowledgebase in 2025. Nucleic Acids Res. 2025;53(D1):D609–17.
  11. 11. Kanehisa M, Furumichi M, Sato Y, Matsuura Y, Ishiguro-Watanabe M. KEGG: biological systems database as a model of the real world. Nucleic Acids Res. 2025;53(D1):D672–7. pmid:39417505
  12. 12. Galperin MY. Diversity of structure and function of response regulator output domains. Curr Opin Microbiol. 2010;13(2):150–9. pmid:20226724
  13. 13. de Boer PA, Crossley RE, Hand AR, Rothfield LI. The MinD protein is a membrane ATPase required for the correct placement of the Escherichia coli division site. EMBO J. 1991;10(13):4371–80. pmid:1836760
  14. 14. Bramkamp M, van Baarle S. Division site selection in rod-shaped bacteria. Curr Opin Microbiol. 2009;12(6):683–8. pmid:19884039
  15. 15. Wang N, Zhang T, Du S, Zhou Y, Chen Y. How Do MinC-D Copolymers Act on Z-Ring Localization Regulation? A New Model of Bacillus subtilis Min System. Front Microbiol. 2022;13:841171. pmid:35495694
  16. 16. Biazzo M, Cioncada R, Fiaschi L, Tedde V, Spigaglia P, Mastrantonio P, et al. Diversity of cwp loci in clinical isolates of Clostridium difficile. J Med Microbiol. 2013;62(Pt 9):1444–52. pmid:23722432
  17. 17. Bradshaw WJ, Roberts AK, Shone CC, Acharya KR. The structure of the S-layer of Clostridium difficile. J Cell Commun Signal. 2018;12(1):319–31. pmid:29170885
  18. 18. Fagan RP, Janoir C, Collignon A, Mastrantonio P, Poxton IR, Fairweather NF. A proposed nomenclature for cell wall proteins of Clostridium difficile. J Med Microbiol. 2011;60(Pt 8):1225–8. pmid:21252271
  19. 19. Gao X, Dong X, Subramanian S, Matthews PM, Cooper CA, Kearns DB, et al. Engineering of Bacillus subtilis strains to allow rapid characterization of heterologous diguanylate cyclases and phosphodiesterases. Appl Environ Microbiol. 2014;80(19):6167–74. pmid:25085482
  20. 20. Bordeleau E, Fortier L-C, Malouin F, Burrus V. c-di-GMP turn-over in Clostridium difficile is controlled by a plethora of diguanylate cyclases and phosphodiesterases. PLoS Genet. 2011;7(3):e1002039. pmid:21483756
  21. 21. Reynolds CB, Emerson JE, de la Riva L, Fagan RP, Fairweather NF. The Clostridium difficile cell wall protein CwpV is antigenically variable between strains, but exhibits conserved aggregation-promoting function. PLoS Pathog. 2011;7(4):e1002024. pmid:21533071
  22. 22. Dhungel BA, Govind R. Phase-variable expression of pdcB, a phosphodiesterase, influences sporulation in Clostridioides difficile. Mol Microbiol. 2021;116(5):1347–60. pmid:34606654
  23. 23. Reyes Ruiz LM, King KA, Agosto-Burgos C, Gamez IS, Gadda NC, Garrett EM, et al. Coordinated modulation of multiple processes through phase variation of a c-di-GMP phosphodiesterase in Clostridioides difficile. PLoS Pathog. 2022;18(7):e1010677. pmid:35789350
  24. 24. McKee RW, Harvest CK, Tamayo R. Cyclic Diguanylate Regulates Virulence Factor Genes via Multiple Riboswitches in Clostridium difficile. mSphere. 2018;3(5):e00423-18. pmid:30355665
  25. 25. Buels R, Yao E, Diesh CM, Hayes RD, Munoz-Torres M, Helt G, et al. JBrowse: a dynamic web platform for genome visualization and analysis. Genome Biol. 2016;17:66. pmid:27072794
  26. 26. Fagan RP, Fairweather NF. Clostridium difficile has two parallel and essential Sec secretion systems. J Biol Chem. 2011;286(31):27483–93. pmid:21659510
  27. 27. Mirdita M, Schütze K, Moriwaki Y, Heo L, Ovchinnikov S, Steinegger M. ColabFold: making protein folding accessible to all. Nat Methods. 2022;19(6):679–82. pmid:35637307
  28. 28. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215(3):403–10. pmid:2231712
  29. 29. Blum M, Andreeva A, Florentino LC, Chuguransky SR, Grego T, Hobbs E, et al. InterPro: the protein sequence classification resource in 2025. Nucleic Acids Res. 2025;53(D1):D444–56. pmid:39565202
  30. 30. Mehrabipour M, Jasemi NSK, Dvorsky R, Ahmadian MR. A Systematic Compilation of Human SH3 Domains: A Versatile Superfamily in Cellular Signaling. Cells. 2023;12(16):2054. pmid:37626864
  31. 31. Hancock LP, Palmer JS, Allwood EG, Smaczynska-de Rooij II, Hodder AJ, Rowe ML, et al. Competitive binding of actin and SH3 domains at proline-rich regions of Las17/WASP regulates actin polymerisation. Commun Biol. 2025;8(1):759. pmid:40374776
  32. 32. Kamitori S, Yoshida H. Structure-Function Relationship of Bacterial SH3 Domains. SH Domains. Springer International Publishing; 2015. p. 71–89.
  33. 33. Lu JZ, Fujiwara T, Komatsuzawa H, Sugai M, Sakon J. Cell wall-targeting domain of glycylglycine endopeptidase distinguishes among peptidoglycan cross-bridges. J Biol Chem. 2006;281(1):549–58. pmid:16257954
  34. 34. Xu Q, Mengin-Lecreulx D, Liu XW, Patin D, Farr CL, Grant JC, et al. Insights into Substrate Specificity of NlpC/P60 Cell Wall Hydrolases Containing Bacterial SH3 Domains. mBio. 2015;6(5):e02327-14. pmid:26374125
  35. 35. Olson MG, Goldammer M, Gauliard E, Ladant D, Ouellette SP. A Bacterial Adenylate Cyclase-Based Two-Hybrid System Compatible with Gateway® Cloning. Methods Mol Biol. 2018;1794:75–96. pmid:29855952
  36. 36. Karimova G, Ullmann A, Ladant D. A bacterial two-hybrid system that exploits a cAMP signaling cascade in Escherichia coli. Methods Enzymol. 2000;328:59–73. pmid:11075338
  37. 37. Bourret RB, Kennedy EN, Tamayo R, Foster CA. A clarifying perspective on bacterial pseudo-receiver domains. J Bacteriol. 2025;207(11):e0026125. pmid:41060066
  38. 38. Williams CL, Mehra A, Harrison EK, Thomas CG, Palmer JR, Cooper DK, et al. Characterization of two co-regulated response regulators in Clostridioides difficile. J Bacteriol. 2025;207(10):e0017725. pmid:40938651
  39. 39. Zhu D, Patabendige HMLW, Tomlinson BR, Wang S, Hussain S, Flores D, et al. Cwl0971, a novel peptidoglycan hydrolase, plays pleiotropic roles in Clostridioides difficile R20291. Environ Microbiol. 2021;23(9):5222–38. pmid:33893759
  40. 40. Garcia-Garcia T, Poncet S, Cuenot E, Douché T, Giai Gianetto Q, Peltier J, et al. Ser/Thr Kinase-Dependent Phosphorylation of the Peptidoglycan Hydrolase CwlA Controls Its Export and Modulates Cell Division in Clostridioides difficile. mBio. 2021;12(3):e00519-21. pmid:34006648
  41. 41. LaBreck CJ, Conti J, Viola MG, Camberg JL. MinC N- and C-Domain Interactions Modulate FtsZ Assembly, Division Site Selection, and MinD-Dependent Oscillation in Escherichia coli. J Bacteriol. 2019;201(4):e00374-18. pmid:30455283
  42. 42. Thomaides HB, Freeman M, El Karoui M, Errington J. Division site selection protein DivIVA of Bacillus subtilis has a second distinct function in chromosome segregation during sporulation. Genes Dev. 2001;15(13):1662–73. pmid:11445541
  43. 43. Valenčíková R, Krascsenitsová E, Labajová N, Makroczyová J, Barák I. Clostridial DivIVA and MinD interact in the absence of MinJ. Anaerobe. 2018;50:22–31. pmid:29408597
  44. 44. Makroczyová J, Jamroškovič J, Krascsenitsová E, Labajová N, Barák I. Oscillating behavior of Clostridium difficile Min proteins in Bacillus subtilis. Microbiologyopen. 2016;5(3):387–401. pmid:26817670
  45. 45. Parti RP, Biswas D, Wang M, Liao M, Dillon J-AR. A minD mutant of enterohemorrhagic E. coli O157:H7 has reduced adherence to human epithelial cells. Microb Pathog. 2011;51(5):378–83. pmid:21798335
  46. 46. Levin PA, Margolis PS, Setlow P, Losick R, Sun D. Identification of Bacillus subtilis genes for septum placement and shape determination. J Bacteriol. 1992;174(21):6717–28. pmid:1400224
  47. 47. Gregory JA, Becker EC, Pogliano K. Bacillus subtilis MinC destabilizes FtsZ-rings at new cell poles and contributes to the timing of cell division. Genes Dev. 2008;22(24):3475–88. pmid:19141479
  48. 48. Park K-T, Dajkovic A, Wissel M, Du S, Lutkenhaus J. MinC and FtsZ mutant analysis provides insight into MinC/MinD-mediated Z ring disassembly. J Biol Chem. 2018;293(16):5834–46. pmid:29414773
  49. 49. Lowrey LC, Gadda NC, Tamayo R. Maximizing bacterial survival: integrating sense-and-respond and bet-hedging mechanisms. Trends Microbiol. 2025. pmid:40537346
  50. 50. Martínez A, Torello S, Kolter R. Sliding motility in mycobacteria. J Bacteriol. 1999;181(23):7331–8. pmid:10572138
  51. 51. Agustí G, Astola O, Rodríguez-Güell E, Julián E, Luquin M. Surface spreading motility shown by a group of phylogenetically related, rapidly growing pigmented mycobacteria suggests that motility is a common property of mycobacterial species but is restricted to smooth colonies. J Bacteriol. 2008;190(20):6894–902. pmid:18689490
  52. 52. van Gestel J, Vlamakis H, Kolter R. From cell differentiation to cell collectives: Bacillus subtilis uses division of labor to migrate. PLoS Biol. 2015;13(4):e1002141. pmid:25894589
  53. 53. Hölscher T, Kovács ÁT. Sliding on the surface: bacterial spreading without an active motor. Environ Microbiol. 2017;19(7):2537–45. pmid:28370801
  54. 54. Tchoufag J, Ghosh P, Pogue CB, Nan B, Mandadapu KK. Mechanisms for bacterial gliding motility on soft substrates. Proc Natl Acad Sci U S A. 2019;116(50):25087–96. pmid:31767758
  55. 55. Burch AY, Shimada BK, Mullin SWA, Dunlap CA, Bowman MJ, Lindow SE. Pseudomonas syringae Coordinates Production of a Motility-Enabling Surfactant with Flagellar Assembly. J Bacteriol. 2012;194(6):1287–98.
  56. 56. McBride MJ, Zhu Y. Gliding motility and Por secretion system genes are widespread among members of the phylum bacteroidetes. J Bacteriol. 2013;195(2):270–8. pmid:23123910
  57. 57. Park S-Y, Pontes MH, Groisman EA. Flagella-independent surface motility in Salmonella enterica serovar Typhimurium. Proc Natl Acad Sci U S A. 2015;112(6):1850–5. pmid:25624475
  58. 58. Schorey JS, Sweet L. The mycobacterial glycopeptidolipids: structure, function, and their role in pathogenesis. Glycobiology. 2008;18(11):832–41. pmid:18723691
  59. 59. Recht J, Martínez A, Torello S, Kolter R. Genetic analysis of sliding motility in Mycobacterium smegmatis. J Bacteriol. 2000;182(15):4348–51. pmid:10894747
  60. 60. Jenior ML, Leslie JL, Powers DA, Garrett EM, Walker KA, Dickenson ME, et al. Novel Drivers of Virulence in Clostridioides difficile Identified via Context-Specific Metabolic Network Analysis. mSystems. 2021;6(5):e0091921. pmid:34609164
  61. 61. Bouillaut L, McBride SM, Sorg JA. Genetic manipulation of Clostridium difficile. Curr Protoc Microbiol. 2011;20:9A.2.1–17. pmid:21400677
  62. 62. Anjuwon-Foster BR, Tamayo R. A genetic switch controls the production of flagella and toxins in Clostridium difficile. PLoS Genet. 2017;13(3):e1006701. pmid:28346491
  63. 63. Furtado KL, Plott L, Markovetz M, Powers D, Wang H, Hill DB, et al. Clostridioides difficile-mucus interactions encompass shifts in gene expression, metabolism, and biofilm formation. mSphere. 2024;9(6):e0008124. pmid:38837404
  64. 64. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114–20. pmid:24695404
  65. 65. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9. pmid:22388286
  66. 66. Chung M, Adkins RS, Mattick JSA, Bradwell KR, Shetty AC, Sadzewicz L, et al. FADU: a Quantification Tool for Prokaryotic Transcriptomic Analyses. mSystems. 2021;6(1):e00917-20. pmid:33436511
  67. 67. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. pmid:25516281
  68. 68. Anjuwon-Foster BR, Maldonado-Vazquez N, Tamayo R. Characterization of Flagellum and Toxin Phase Variation in Clostridioides difficile Ribotype 012 Isolates. J Bacteriol. 2018;200(14):e00056-18. pmid:29735765
  69. 69. Kent WJ. BLAT—The BLAST-like alignment tool. Genome Res. 2002;12(4):656–64. pmid:11932250
  70. 70. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30(4):772–80. pmid:23329690
  71. 71. Gasteiger E, Gattiker A, Hoogland C, Ivanyi I, Appel RD, Bairoch A. ExPASy: The proteomics server for in-depth protein knowledge and analysis. Nucleic Acids Res. 2003;31(13):3784–8. pmid:12824418
  72. 72. Stajich JE, Block D, Boulez K, Brenner SE, Chervitz SA, Dagdigian C, et al. The Bioperl toolkit: Perl modules for the life sciences. Genome Res. 2002;12(10):1611–8. pmid:12368254
  73. 73. Karimova G, Pidoux J, Ullmann A, Ladant D. A bacterial two-hybrid system based on a reconstituted signal transduction pathway. Proc Natl Acad Sci U S A. 1998;95(10):5752–6. pmid:9576956