Genes involved in acid response

It was postulated that fermentation during growth under the anaerobic conditions of this study have resulted in an acidic environment (257), which may have contributed to the observed mutation spectra. To further explore this possibility, gene expression data of genes previously shown to be induced in response to acidic environments were analysed. Studies suggest that while multiple metabolic processes are required for survival under acidic conditions, E. coli largely have four acid tolerance mechanisms

[reviewed in (281)]. Each mechanism confers varying degrees of tolerance depending on the growth medium, the growth phase and studies have also shown that these mechanisms may be affected by strain-specific differences (281-283). One such mechanism is acid resistance system 1 (AR1), where ATP synthase, also known as F0F1 synthase, catalyses the hydrolysis of ATP in cells at stationary phase in minimal glucose medium to generate an electrochemical proton gradient (281, 282, 284). In this study, the expression of many genes encoding proteins that contribute to AR1 were

143 significantly up-regulated under anaerobic conditions, consistent with our understanding of the mechanisms (Table A.6).

Studies have also shown that genes encoding proteins involved in DNA repair, cellular stress responses, membrane structure, membrane permeability, osmotic shock and ion transport are also generally induced in response to acidic cell conditions (285, 286). In this study, DNA repair genes were generally up-regulated under anaerobic conditions (section 4.2.3.1.1). Additionally, many of the genes encoding proteins previously shown to contribute to acid resistance were up-regulated under the anaerobic conditions of this study (Table A.2). Therefore, altogether, it seems feasible that the high level of

mutations observed in the anaerobically grown MA lineages could be a result of acidic conditions generated during anaerobic fermentation.

4.3Summary

The genome-wide spontaneous mutation rate for E. coli was greater in anaerobically

grown MA lineages as compared to aerobically grown MA lineages. To determine if the types of mutations that prevailed in the two environments differed, the mutation spectra of 24 aerobic and 24 anaerobic MA lineage genomes were determined and compared. In general, mutation rates per generation for BPSs, indels and GCRs were higher under anaerobic growth conditions than aerobic ones (Figure 4.1a), while mutation rates per

day for BPSs and indels were higher under aerobic growth conditions than anaerobic ones (Figure 4.1b). BPSs were the most prevalent mutation type in both environments

with a bias towards G Æ T transversions in aerobically grown cells, and biases towards C Æ A, T Æ G and A Æ C transversions in anaerobically grown ones (Figure 4.2). No

significant bias towards indels was detected in either environment (Figure 4.5) but

GCRs were generally more frequent under anaerobic growth conditions (Figure 4.6).

Generally, the frequencies of IS deletions and IS insertions underpinned the largest differences in GCR type frequency between the two environments; where there was a propensity for IS186 and IS150 transposition in aerobic and anaerobic cells,

respectively (Figure 4.7). Additionally, when considering mutation rates per generation

and per day together, it appears that GCRs and IS insertions, particularly IS150 and

144 Gene expression data from stationary phase cultures indicates that under anaerobic conditions, there was an overall increased expression of genes involved in repair and replication (section 4.2.3.1.1). In particular, there is evidence that genes involved in GCR repair were more highly expressed under anaerobic conditions (sections 4.2.3.1.1.2 to 4.2.3.1.1.4), consistent with observations that GCRs were more prevalent in anaerobic MA lineages. Expression data for genes involved in DNA repair and acid tolerance mechanisms (section 4.2.3.1.2) suggest that acidic conditions generated during anaerobic fermentation are potentially responsible for the reported increased activity of repair genes.

145

Chapter Five:The contribution of gross chromosomal rearrangements

to adaptive evolution

5.1Introduction

Repeat sequences, particularly IRs, in a DNA sequence are hotspots for genome instability because of their capacity to fold into secondary structures that can interfere with molecular processes such as transcription and DNA replication (53). In ssDNA, IRs can form hairpin structures while in dsDNA, IRs can form cruciform-like structures. Both forms of secondary structures can impede the action and progression of polymerases (53). To maintain genome integrity in E. coli, secondary structures can be

cleaved by nucleases, generating DSBs that are then repaired by homologous recombination (section 1.4.7.1) (53, 287). One such nuclease is SbcC (section 1.7.2), which forms a complex with SbcD and generates DSBs during DNA replication by hairpin cleavage. The DSBs can then be repaired by the RecBCD (section 1.4.7.1.1) or RecFOR pathways (section 1.4.7.1.2) (196, 288, 289). Thus, the SbcCD complex is thought to play an important role in maintaining genome stability (53, 104).

Improper repair of DBSs can lead to the generation of GCRs (described in section 1.3.3) (5). GCRs, such as deletions, translocations, inversions and IS element transposition, are of particular interest in evolution studies as they can modify gene expression, gene content and result in the formation of new DNA junctions (34). While GCRs have long been thought to play important roles in evolution, the impact of GCRs on population fitness and rate of adaptation in E. coli is not well understood (5). In our MA study of

E. coli, GCRs were found to occur at different rates in cells grown under aerobic and

anaerobic conditions, with rates being 2.6-fold greater under anaerobic conditions (section 4.2.1.3). Additionally, a microarray study of E. coli reported that sbcC was

differentially expressed between aerobic and anaerobic growth conditions, with increased expression under aerobic conditions (10). Therefore, the aims of this project were to use whole genome re-sequencing and experimental evolution techniques to:

x Determine the contribution of SbcC to the occurrence of GCRs in E. coli. x Determine the impact of GCRs on population fitness.

146

5.2Results and discussion