What is the benefit of a nested rpoS promoter?

This study has provided a rare insight on the functionality of nlpD and how this gene interacts with rpoS. The high conservation of nlpD and rpoS as neighbouring genes has not previously been explained. Although the transcription start of rpoS has been frequently identified within nlpD across diverse species, the benefit of this nested promoter has not been identified.

The lack of an explanation for the nesting of PrpoS is surprising given the detailed research describing the complex regulation of rpoS, especially in E. coli, at both the transcriptional and translational levels (Landini et al. 2014). The 5’ untranslated region (UTR) of RpoS in E. coli is the longest described bacterial UTR. This length seems necessary to incorporate regions encoding antagonistic binding sites that promote or reduce translation. Deregulation of such a global regulator may have strongly deleterious consequences, and complex regulation of this gene is consistent with the expected pleiotropic consequences of deregulation of rpoS.

The benefit derived from nesting this complex UTR within the coding frame of nlpD has not been addressed from an evolutionary perspective. Disentanglement of the nested rpoS promoter from the ORF of nlpD is achieved by a simple duplication event (similar to that constructed in Section 5.3.3). Considering such mutational events could readily occur, why is the nesting of PrpoS within nlpD so widely conserved? The phenotypic consequences of mutations to nlpD suggests two, alternative, explanations for the conserved nesting of PrpoS within nlpD.

5.4.4.1 The nesting of PrpoS in nlpD may preserve the promoter

The nesting of PrpoS may prevent the loss of this promoter in environments where it remains inactive and rpoS is not expressed. This study has clearly demonstrated that a nonsense mutation to nlpD may have dramatic consequences at the cellular level. The deficiency in cell separation may confer a low fitness, especially in environments supporting exponential growth. Such a low fitness caused by loss of function of nlpD will prevent the increased frequency of genotypes with loss-of-function mutations to the rpoS promoter. This protection from loss-of-function PrpoS mutations would occur even in conditions in which the promoter is not expressed, allowing preservation of the

promoter without direct selection on the expression of rpoS. Nesting of PrpoS in nlpD may help conserve the complex regulation of rpoS from mutation in times of relaxed selection on rpoS expression, such as during exponential-phase growth.

While it is difficult to directly test this hypothesis, it may be possible to gain support for it by a bioinformatic search for similar transitively induced genes that have promoters nested in neighbouring genes. The nesting of these promoters may similarly prevent degeneration of such promoters during times of relaxed selection for the functionality of the downstream gene.

5.4.4.2 The nesting of PrpoS may cause a stochastic change in cell length

The evidence presented in this chapter provides an alternative explanation for the conserved nesting of PrpoS within nlpD. Assuming the transcription from the promoter results in a high mutation rate at the promoter of rpoS, the concomitant high frequency of loss-of-function mutations to nlpD – and the high frequency of reversions – may confer an adaptive benefit in causing the stochastic production of sessile chaining types. The benefit of stochastic switching in fluctuating environments has been well described (Acar et al. 2008; Veening et al. 2008). The stochastic generation of cell- chaining and motile types has been identified in populations of Bacillus subtilis (Kearns and Losick 2005). These chaining (sessile) types cause the biofilm to be initiated, which allows protection from environmental hazards such as antimicrobials or UV exposure (Hall-Stoodley et al. 2004) and colonisation of specific niches such as the plant surface (Vlamakis et al. 2013), whilst the stochastic generation of motile sub- populations helps disperse the biofilm. In the case of B. subtilis, the stochastic switching between biofilm chaining types and motile types is caused by an epigenetic switch (Norman et al. 2013). Perhaps the suspected high mutation rate of nlpD causes a similar stochastic switch between motile and sessile biofilm-forming types, via a reversible genetic switch.

A high mutation rate of nlpD, and a high frequency of cell-division mutants, may also allow protection from microfaunal predation. Predation by a variety of species, such as nematodes and protozoa, is believed to be a major selective force of diversity in bacterial species (Jousset 2012). Bacterial populations respond to the selective pressure of predation via a number of anti-grazing strategies. These strategies include

evasion of detection (Wildschutte et al. 2004) and the production of toxins (Vaitkevicius et al. 2006; Mazzola et al. 2009; Jousset et al. 2010). Predation can also be prevented by changes in cell shape and by aggregation, which prevent ingestion by predators. Such morphological changes can be produced by the production of biofilms, and exopolymers have been shown to limit ingestion (Hahn et al. 2004; Matz et al. 2004; Weitere et al. 2005). The production of cellular filaments has been shown to limit ingestion across diverse bacterial species (Hahn et al. 1999; Corno and Jurgens 2006; Queck et al. 2006). These filaments are visually similar to those caused by loss-of- function mutations of nlpD in SBW25.

The filamentous morphology of nlpD, combined with a high mutation rate at this locus, may be a stochastic switching mechanism to prevent grazing by microfaunal predators. The location of the nested promoter, which may also confer a high rate of mutation of nlpD, may provide a means of producing filamentous mutant types at a high rate, which are resistant to grazing. These filamentous types may be able to revert back to a non- filamentous WT morphotype, and not be locked in an evolutionary ‘dead end’ as a filamentous type. Predation may thus provide a means of selecting the nested promoter, by direct selection for genotypes that are able to stochastically switch between filamentous and non-filamentous types. Direct measurement of mutation rate at this locus in SBW25, and association of this mutation rate with fitness in predator- prey coevolutionary studies, may provide evidence that microfaunal predation selects for a high mutation rate in nlpD. This hypothesis may also apply to species outside of P. fluorescens, assuming the predicted high mutation rate in nlpD is found in other species.