Impaired competence in flagellar mutants of Bacillus subtilis is connected to the regulatory network governed by DegU

The competent state is a developmentally distinct phase, in which bacteria are able to take up and integrate exogenous DNA into their genome. Bacillus subtilis is one of the naturally competent bacterial species and the domesticated laboratory strain 168 is easily transformable. In this study, we report a reduced transformation frequency of B. subtilis mutants lacking functional and structural flagellar components. This includes hag, the gene encoding the flagellin protein forming the filament of the flagellum. We confirm that the observed decrease of the transformation frequency is due to reduced expression of competence genes, particularly of the main competence regulator comK. The impaired competence is due to an increase in the phosphorylated form of the response regulator DegU, which is involved in regulation of both flagellar motility and competence. Altogether, our study identified a close link between motility and natural competence in B. subtilis suggesting that hindrance in motility has great impact on differentiation of this bacterium not restricted only to the transition towards sessile growth stage. Originality-Significance statement Understanding how versatile bacterial phenotypes influence each other is important for our basic understanding of microbial ecology. Our research highlights the novel intertwinement of bacterial differentiation and reveal how lack of single cell motility adjusts DNA exchange among bacterial strains.

The competent state is a developmentally distinct phase, in which bacteria are able to take up and 23 integrate exogenous DNA into their genome. Bacillus subtilis is one of the naturally competent 24 bacterial species and the domesticated laboratory strain 168 is easily transformable. In this study, 25 we report a reduced transformation frequency of B. subtilis mutants lacking functional and 26 structural flagellar components. This includes hag, the gene encoding the flagellin protein forming 27 the filament of the flagellum. We confirm that the observed decrease of the transformation 28 frequency is due to reduced expression of competence genes, particularly of the main competence 29 regulator comK. The impaired competence is due to an increase in the phosphorylated form of the 30 response regulator DegU, which is involved in regulation of both flagellar motility and competence. 31 Altogether, our study identified a close link between motility and natural competence in B. subtilis 32 suggesting that hindrance in motility has great impact on differentiation of this bacterium not 33 restricted only to the transition towards sessile growth stage. 34 35

Originality-Significance statement 36
Introduction 42 When facing stressful environmental conditions, bacteria can respond with a variety of post-43 exponential modifications including secretion of degradative enzymes, sporulation, or genetic 44 competence. Bacillus subtilis is one of the bacterial species that are able to take up free DNA from 45 the environment and incorporate it into its own genome, a phenomenon referred to as natural 46 competence (Dubnau, 1991). To import extracellular DNA into B. subtilis cells, a pseudopilus formed 47 6 Lack of competence in flagellar mutants is due to the reduced expression of competence genes. 136 To determine if the detected diminished transformation frequency of flagellar mutants was due to 137 altered competence gene expression, the fluorescent reporter P comG -gfp was introduced into these 138 strains. This reporter allows the detection of cells expressing the comG operon-encoding genes 139 required for pseudopilus formation and DNA uptake. In addition, this reporter provides a proxy on 140 the activity of the ComK protein, the master regulator of competence. Qualitative microscopy 141 analyses of cultures harboring the reporter, and which were grown in competence medium for 5 h, 142 showed indeed a decreased number of fluorescent (i.e. comG expressing) cells in the hag mutant 143 compared to the wild type, whereas a control strain lacking comK showed no fluorescence (Fig. 3A). Similarly, the motA and flgE mutants were analyzed microscopically as well as by using flow 149 cytometry. Both methods revealed fewer cells activated transcription of competence genes in these 150 mutants compared to the wild type ( Fig. 4; unpaired two-sample t-test with Welch Correction: P = 151 0.017 for WT -ΔmotA, P = 1.3·10 -9 for WT -ΔflgE, n = 3 for both; mean percentage of fluorescent 152 cells: 16.7% for wild type, 4.5% for ΔmotA, 4.2% for ΔflgE). Flow cytometry measurements at 153 different time points during growth in competence medium confirmed a similarly reduced fraction of 154 competent cells in the hag mutant compared to the wild type strain (Fig. S3). 155

156
Reduced competence in hag mutant can be rescued by overexpression of comK. 157 The reduced competence gene expression in the tested flagellar mutants suggested a regulatory link 158 between flagellar motility and competence. To investigate if regulatory elements upstream of comK 159 were responsible for our observations and if a bypass of those could therefore rescue 160 transformation frequency in the flagellar mutant, we examined a strain with an additional copy of 161 comK under the control of a xylose-inducible promoter (P xyl -comK). Indeed, in combination with P xyl -162 comK, the transformation level of the hag mutant increased back to a level that was statistically 163 indistinguishable from wild type levels (mean transformation frequency of 5.3·10 -6 for the wild type 164 and 8.5·10 -6 for Δhag P xyl -comK; Kruskal-Wallis test: P = 0.453, n = 9, Fig. 5A). Despite this observed 165 increase in the hag strain upon comK overexpression, the wild type strain, which contained an 166 inducible copy of comK showed a higher transformation frequency (Fig. 5A, Kruskal-Wallis test: P = 167 3.4·10 -4 for WT -WT P xyl -comK, P = 3.4·10 -4 for WT P xyl -comK -Δhag P xyl -comK, n = 9 for both), which 7 was probably due to higher levels of comK transcription at the native locus as previously observed 169 (Hahn et al., 1996). 170

171
Reduced competence in flagellar mutants is likely connected to unbalanced DegU phosphorylation. 172 As the above results suggested that regulatory elements in response to impaired flagellar motility 173 are responsible for the decreased comK expression, we investigated DegU as a likely candidate 174 causing the reduced competence in flagellar mutants. As non-phosphorylated DegU was implicated 175 to be required for comK transcription (Dahl et al., 1992;Hamoen et al., 2000), two variants of degU 176 were tested: degU32, which harbors a mutation resulting in an extended half-life and thus higher 177 stability of the phosphorylated form of the DegU protein (DegU~P), and degU146, which is cannot be 178 phosphorylated (Dahl et al., 1991;Dahl et al., 1992;Kunst et al., 1994). Both variants were tested in 179 wild type as well as the Δhag background to observe differences in transformability compared to the 180 wild type strain. The results of this experiment indicated that the transformation frequency of the 181 degU32 strain was slightly decreased ( Figure 5B), which is consistent with previous publications, 182 suggesting that non-phosphorylated DegU is required for priming comK transcription. The observed 183 difference, however, was only marginally significant in our experimental setup ( Figure 5B

Strains and cultivating conditions 275
The strains used in this study and their mutant derivatives are listed in Table S1. Mutants 276 constructed in this study were obtained by natural transformation of a B. subtilis receptor strain with 277 genomic DNA from a donor strain. Strain TB831 was created by transformation of strain 168 P xyl -278 comK with genomic DNA of strain GP902 (J. Stülke lab collection). To obtain strains TB926 and 279 TB925, genomic DNA of strain 168 P comG -gfp was used to transform strain TB710 and TB689, 280 respectively. Strain TB928 was obtained by transforming strain 168 P xyl -comK with genomic DNA of 281 GP901 (J. Stülke lab collection). To create strain TB935 and TB936, strain 168 was transformed with 282 genomic DNA obtained from strain QB4371 (Kunst et al., 1994) and QB4458 (Dahl et al., 1991), 283 respectively. Their derivatives harbouring also a mutation of hag (TB923 and TB924) were created by 284 transformation with genomic DNA, which was obtained from GP901. In-frame deletions of motA, 285 flgE, and cheA were created using plasmids pEC1, pDP306, and pDP338, respectively, as previously 286 described (Courtney et al., 2012;Chan et al., 2014;Calvo and Kearns, 2015). Strains were verified by 287 fluorescence microscopy (P comG -gfp reporter), PCR (hag mutants), or sequencing (degU variants), 288 using the oligonucleotides listed in Table S2 medium (LB-Lennox, Carl Roth; 10 g L -1 tryptone, 5 g L -1 yeast extract, and 5 g L -1 NaCl) for 16 h was 297 centrifuged for 2 min at 11,000 x g. The pellet was washed twice in de-ionized water and was re-298 suspended in 100 µl de-ionized water. The re-suspended culture was diluted (