**A Recombination Puzzle Solved: Role for New DNA Repair Systems in**  *Helicobacter pylori* **Diversity/Persistence**

Ge Wang and Robert J. Maier *Department of Microbiology, University of Georgia, Athens Georgia* 

### **1. Introduction**

#### **1.1 Helicobacter pylori pathogenesis**

*Helicobacter pylori* is a gram-negative, slow-growing, microaerophilic, spiral bacterium. It is one of the most common human gastrointestinal pathogens, infecting almost 50% of the world's population [1]. Peptic ulcer disease is now approached as an infectious disease, and *H. pylori* is responsible for the majority of duodenal and gastric ulcers [2]. There is strong evidence that *H. pylori* infection increases the risk of gastric cancer [3], the second most frequent cause of cancer-related death. *H. pylori* infections are acquired by oral ingestion and is mainly transmitted within families in early childhood [2]. Once colonized, the host can be chronically infected for life, unless *H. pylori* is eradicated by treatment with antibiotics.

*H. pylori* is highly adapted to its ecologic niche, the human gastric mucosa. The pathogenesis of *H. pylori* relies on its persistence in surviving a harsh environment, including acidity, peristalsis, and attack by phagocyte cells and their released reactive oxygen species [4]. *H. pylori* has a unique array of features that permit entry into the mucus, attachment to epithelial cells, evasion of the immune response, and as a result, persistent colonization and transmission. Numerous virulence factors in *H. pylori* have been extensively studied, including urease, flagella, BabA adhesin, the vacuolating cytotoxin (VacA), and the cag pathogenicity island (cag-PAI) [5]. In addition to its clinical importance, *H. pylori* has become a model system for persistent host-associated microorganisms [6]. How *H. pylori* can adapt to, and persist in, the human stomach has become a problem of general interest in both microbial physiology and in pathogenesis areas.

#### **1.2 Genetic diversity of** *H. pylori*

*H. pylori* displays exceptional genetic variability and intra-species diversity [7]. Allelic diversity is obvious as almost every unrelated isolate of *H. pylori* has a unique sequence when a sequenced fragment of only several hundred base pairs is compared among strains for either housekeeping or virulence genes [8-10]. Approximately 5% nucleotide divergence is commonly observed at the majority of gene loci between pairs of unrelated *H. pylori* strains [11]. *H. pylori* strains also differ considerably in their gene contents, the genetic macro-diversity. The two sequenced strains 26695 and J99 share only 94% of their genes, whereas approximately 7% of the genes are unique for each strain [12, 13]. Supporting

A Recombination Puzzle Solved: Role for New

DNA Repair Systems in *Helicobacter pylori* Diversity/Persistence 5

only portions of them. *H. pylori* encodes the homologues of all four members of the nucleotide excision repair (NER) pathway; these are UvrA, UvrB, UvrC, and UvrD, all of which are well conserved in bacteria. NER deals with DNA-distorting lesions, in which an excinuclease removes a 12- to 13- nucleotide segment from a single strand centered around the lesion; the resulting gap is then filled in by repair synthesis [44]. Loss of *uvrB* in *H. pylori* was shown to confer sensitivity to UV light, alkylating agents and low pH, suggesting that the *H. pylori* NER pathway is functional in repairing a diverse array of DNA lesions [45]. *H. pylori* UvrD was shown to play a role in repairing DNA damage and limiting DNA

The methyl-directed mismatch repair system (MMR), consisting of MutS1, MutH, and MutL, is conserved in many bacteria and eukaryotes, and it plays a major role in maintaining genetic stability. MMR can liberate up to 1000 nucleotides from one strand during its function to correct a single mismatch arising during DNA replication [47]. Notably, MMR does not exist in *H. pylori*, contributing to the high mutation rates observed in *H. pylori* [17]. *H. pylori* has a MutS homologue that belongs to the MutS2 family. *H. pylori* MutS2 was shown to bind to DNA structures mimicking recombination intermediates and to inhibit DNA strand exchange, thus it may play a role in maintaining genome integrity by suppressing homologous and homeologous DNA recombination [48]. In addition, *H. pylori* MutS2 appears to play a role in

Damaged bases can be repaired by a variety of glycosylases that belong to the base excision repair (BER) pathway. All glycosylases can excise a damaged base resulting in an apurinic/apyrimidinic (AP) site, while some of them additionally nick the DNA deoxyribose-phosphate backbone (via an AP lyase activity). *H. pylori* harbors the glycosylase genes *ung, mutY*, *nth*, and *magIII*, whereas several other genes appear to be absent from the *H. pylori* genome, e.g. *tag*, *alkA*, and *mutM*. The *H. pylori* endonuclease III (*nth* gene product), which removes oxidized pyrimidine bases, was shown to be important in establishing longterm colonization in the host [50]. The *H. pylori* MutY glycosylase is functional in removing adenine from 8-oxoG:A mispair, and the loss of MutY leads to attenuation of the

To repair DNA double strand breaks and blocked replication forks, *H. pylori* is equipped with an efficient system of DNA recombinational repair, which is the main focus of this

Many bacteria encode a genetic program for a coordinated response to DNA damage called the SOS response. The best known *E. coli* SOS response is triggered when RecA binds ssDNA, activating its co-protease activity towards LexA, a transcriptional repressor [54]. Cleavage of LexA results in transcriptional induction of genes involved in DNA repair, lowfidelity polymerases, and cell cycle control. However, the *H. pylori* genome contains neither a gene for LexA homolog nor the genes for low-fidelity polymerases, and an SOS response

To define pathways for an *H. pylori* DNA damage response, Dorer et al. [55] used cDNA based microarrays to measure transcriptional changes in cells undergoing DNA damage. In both ciprofloxacin treated cells and the Δ*addA* (a major DNA recombination gene, see section 4.4 below) mutant cells, the same set of genes were induced which include genes required for energy metabolism, membrane proteins, fatty acid biosynthesis, cell division, and some translation factors, although the contribution of these genes to survival in the face

recombination, indicating it functions to ultimately maintain genome integrity [46].

repairing oxidative DNA damage, specifically 8-oxo-guanine [49].

colonization ability [51-53].

**2.2** *H. pylori* **response to DNA damage** 

pathway seems to be absent in *H. pylori* [12, 13].

review (See section 4).

studies using whole-genome microarray detected numerous genomic changes in the paired sequential isolates of *H. pylori* from the same patient [14, 15].

Mechanisms proposed to account for the observed genetic variability include mainly the high inherent mutation rate and high frequency of recombination [16]. The spontaneous mutation rate of the majority of *H. pylori* strains lies between 10-5 and 10-7 [17]. This is several orders of magnitude higher than the average mutation rate of *Escherichia coli*, and similar to that of *E. coli* strains defective in mismatch repair functions (mutator strains) [18]. While mutation is essential for introducing sequence diversity into the species, a key role in generating diversity is played by recombination.

*H. pylori* is naturally competent for DNA transformation, and has a highly efficient system for recombination of short-fragment involving multiple recombination events within a single locus [19, 20]. A special apparatus homologous to type IV secretion system (T4SS, encoded by *comB* locus) is dedicated to a DNA uptake role [21, 22] and a composite system involving proteins at the *comB* locus and ComEC mediates two-step DNA uptake in *H. pylori* [23]. T4SS systems are known to transport DNA and proteins in other bacteria, but *H. pylori* is the only species known to use a T4SS for natural competence [24]. Unlike several other bacterial species, *H. pylori* does not require specific DNA sequences for uptake of related DNA [25]. Instead, numerous and efficient restriction modification systems take over the function as a barrier to horizontal gene transfer from foreign sources [26, 27].

Population genetic analyses of unrelated isolates of *H. pylori* indicated that recombination was extremely frequent in *H. pylori* [9, 28]. There is evidence that humans are occasionally infected with multiple genetically distinct isolates and that recombination between *H. pylori* strains can occur in humans [29, 30]. Using mathematical modeling approaches on sequence data from 24 pairs of sequential *H. pylori* isolates, Falush et al. [31] estimated that the mean size of imported fragments was only 417 bp, much shorter than that observed for other bacteria. The recombination rate per nucleotide was estimated as 6.9 x 10-5, indicating that every pair of strains differed on average by 114 recombination events. Compared to other bacteria studied in this way [32-34], the recombination frequency within *H. pylori* is extraordinarily high. The *H. pylori* genome also has extensive repetitive DNA sequences that are targets for intragenomic recombination [35].

#### **2. Overview of DNA repair in** *H. pylori*

Oxidative DNA damage represents a major form of DNA damage. Among the many oxidized bases in DNA, 8-oxo-guanine is a ubiquitous biomarker of DNA oxidation [36]. In addition, acid (low pH) conditions may result in DNA damage via depurination [37]. *H. pylori* survives on the surface of the stomach lining for the lifetime of its host and causes a chronic inflammatory response. Several lines of evidence suggest that *H. pylori* is exposed to oxidative damage soon after infection [38, 39]. Under physiological conditions, *H. pylori* is thought to frequently suffer oxidative and acid stress [40, 41]. In addition to diverse oxidant detoxification enzymes (e.g. superoxide dismutase, catalse, and peroxiredoxins) [42] and potent acid avoidance mechanisms (mainly urease) [43], efficient DNA repair systems are required for *H. pylori* to survive in the host.

#### **2.1 DNA repair systems in** *H. pylori*

The whole genome sequences of *H. pylori* revealed it contains several DNA repair pathways that are common to many bacterial species, while it lacks other repair pathways or contains

studies using whole-genome microarray detected numerous genomic changes in the paired

Mechanisms proposed to account for the observed genetic variability include mainly the high inherent mutation rate and high frequency of recombination [16]. The spontaneous mutation rate of the majority of *H. pylori* strains lies between 10-5 and 10-7 [17]. This is several orders of magnitude higher than the average mutation rate of *Escherichia coli*, and similar to that of *E. coli* strains defective in mismatch repair functions (mutator strains) [18]. While mutation is essential for introducing sequence diversity into the species, a key role in

*H. pylori* is naturally competent for DNA transformation, and has a highly efficient system for recombination of short-fragment involving multiple recombination events within a single locus [19, 20]. A special apparatus homologous to type IV secretion system (T4SS, encoded by *comB* locus) is dedicated to a DNA uptake role [21, 22] and a composite system involving proteins at the *comB* locus and ComEC mediates two-step DNA uptake in *H. pylori* [23]. T4SS systems are known to transport DNA and proteins in other bacteria, but *H. pylori* is the only species known to use a T4SS for natural competence [24]. Unlike several other bacterial species, *H. pylori* does not require specific DNA sequences for uptake of related DNA [25]. Instead, numerous and efficient restriction modification systems take over the

Population genetic analyses of unrelated isolates of *H. pylori* indicated that recombination was extremely frequent in *H. pylori* [9, 28]. There is evidence that humans are occasionally infected with multiple genetically distinct isolates and that recombination between *H. pylori* strains can occur in humans [29, 30]. Using mathematical modeling approaches on sequence data from 24 pairs of sequential *H. pylori* isolates, Falush et al. [31] estimated that the mean size of imported fragments was only 417 bp, much shorter than that observed for other bacteria. The recombination rate per nucleotide was estimated as 6.9 x 10-5, indicating that every pair of strains differed on average by 114 recombination events. Compared to other bacteria studied in this way [32-34], the recombination frequency within *H. pylori* is extraordinarily high. The *H. pylori* genome also has extensive repetitive DNA sequences that

Oxidative DNA damage represents a major form of DNA damage. Among the many oxidized bases in DNA, 8-oxo-guanine is a ubiquitous biomarker of DNA oxidation [36]. In addition, acid (low pH) conditions may result in DNA damage via depurination [37]. *H. pylori* survives on the surface of the stomach lining for the lifetime of its host and causes a chronic inflammatory response. Several lines of evidence suggest that *H. pylori* is exposed to oxidative damage soon after infection [38, 39]. Under physiological conditions, *H. pylori* is thought to frequently suffer oxidative and acid stress [40, 41]. In addition to diverse oxidant detoxification enzymes (e.g. superoxide dismutase, catalse, and peroxiredoxins) [42] and potent acid avoidance mechanisms (mainly urease) [43], efficient DNA repair systems are

The whole genome sequences of *H. pylori* revealed it contains several DNA repair pathways that are common to many bacterial species, while it lacks other repair pathways or contains

function as a barrier to horizontal gene transfer from foreign sources [26, 27].

sequential isolates of *H. pylori* from the same patient [14, 15].

generating diversity is played by recombination.

are targets for intragenomic recombination [35].

**2. Overview of DNA repair in** *H. pylori*

required for *H. pylori* to survive in the host.

**2.1 DNA repair systems in** *H. pylori*

only portions of them. *H. pylori* encodes the homologues of all four members of the nucleotide excision repair (NER) pathway; these are UvrA, UvrB, UvrC, and UvrD, all of which are well conserved in bacteria. NER deals with DNA-distorting lesions, in which an excinuclease removes a 12- to 13- nucleotide segment from a single strand centered around the lesion; the resulting gap is then filled in by repair synthesis [44]. Loss of *uvrB* in *H. pylori* was shown to confer sensitivity to UV light, alkylating agents and low pH, suggesting that the *H. pylori* NER pathway is functional in repairing a diverse array of DNA lesions [45]. *H. pylori* UvrD was shown to play a role in repairing DNA damage and limiting DNA recombination, indicating it functions to ultimately maintain genome integrity [46].

The methyl-directed mismatch repair system (MMR), consisting of MutS1, MutH, and MutL, is conserved in many bacteria and eukaryotes, and it plays a major role in maintaining genetic stability. MMR can liberate up to 1000 nucleotides from one strand during its function to correct a single mismatch arising during DNA replication [47]. Notably, MMR does not exist in *H. pylori*, contributing to the high mutation rates observed in *H. pylori* [17]. *H. pylori* has a MutS homologue that belongs to the MutS2 family. *H. pylori* MutS2 was shown to bind to DNA structures mimicking recombination intermediates and to inhibit DNA strand exchange, thus it may play a role in maintaining genome integrity by suppressing homologous and homeologous DNA recombination [48]. In addition, *H. pylori* MutS2 appears to play a role in repairing oxidative DNA damage, specifically 8-oxo-guanine [49].

Damaged bases can be repaired by a variety of glycosylases that belong to the base excision repair (BER) pathway. All glycosylases can excise a damaged base resulting in an apurinic/apyrimidinic (AP) site, while some of them additionally nick the DNA deoxyribose-phosphate backbone (via an AP lyase activity). *H. pylori* harbors the glycosylase genes *ung, mutY*, *nth*, and *magIII*, whereas several other genes appear to be absent from the *H. pylori* genome, e.g. *tag*, *alkA*, and *mutM*. The *H. pylori* endonuclease III (*nth* gene product), which removes oxidized pyrimidine bases, was shown to be important in establishing longterm colonization in the host [50]. The *H. pylori* MutY glycosylase is functional in removing adenine from 8-oxoG:A mispair, and the loss of MutY leads to attenuation of the colonization ability [51-53].

To repair DNA double strand breaks and blocked replication forks, *H. pylori* is equipped with an efficient system of DNA recombinational repair, which is the main focus of this review (See section 4).

### **2.2** *H. pylori* **response to DNA damage**

Many bacteria encode a genetic program for a coordinated response to DNA damage called the SOS response. The best known *E. coli* SOS response is triggered when RecA binds ssDNA, activating its co-protease activity towards LexA, a transcriptional repressor [54]. Cleavage of LexA results in transcriptional induction of genes involved in DNA repair, lowfidelity polymerases, and cell cycle control. However, the *H. pylori* genome contains neither a gene for LexA homolog nor the genes for low-fidelity polymerases, and an SOS response pathway seems to be absent in *H. pylori* [12, 13].

To define pathways for an *H. pylori* DNA damage response, Dorer et al. [55] used cDNA based microarrays to measure transcriptional changes in cells undergoing DNA damage. In both ciprofloxacin treated cells and the Δ*addA* (a major DNA recombination gene, see section 4.4 below) mutant cells, the same set of genes were induced which include genes required for energy metabolism, membrane proteins, fatty acid biosynthesis, cell division, and some translation factors, although the contribution of these genes to survival in the face

A Recombination Puzzle Solved: Role for New

**4. DNA recombinational repair factors in** *H. pylori*

**4.1 The central recombination protein RecA** 

required for full activity of DNA repair [67].

recombination [64].

years.

DNA Repair Systems in *Helicobacter pylori* Diversity/Persistence 7

homologous recombination. RecN was shown to play an important role in repairing DSBs, probably coordinating alignment of the broken segments with intact duplexes to facilitate

While some genes that are predicted to be involved in DNA recombinational repair, including *recA, recG, recJ, recR, recN, and ruvABC,* were annotated from the published *H. pylori* genome sequences, many genes coding for the components that are involved in the pre-synapsis stage, such as RecBCD, RecF, RecO, and RecQ, were missing. Considering that *H. pylori* is highly genetic diverse with a high recombination frequency, this has been a big puzzle over the past decade. Recent studies revealed the existence of both pathways, AddAB (RecBCD-like) and RecRO, for initiation of DNA recombinational repair in *H. pylori*. In the following sections we will summarize the current understanding of DNA recombinational repair in *H. pylori* by reviewing the literature accumulated in recent

The RecA protein is a central component of the homologous recombination machinery and of the SOS system in most bacteria. The relatively small RecA protein contains many functional domains including different DNA-binding sites and an ATP-binding site. *E. coli* RecA has also coprotease activities for the LexA repressor and other factors involved in SOS response. However, *H. pylori* genome does not contain a LexA homolog and an SOS response pathway is likewise absent in *H. pylori*. Thus, a coprotease activity may be dispensable for the *H. pylori* RecA protein. Nevertheless, RecA is required for DNA damage

Before the genome era, the roles of *H. pylori* RecA in DNA recombination and repair have been studied genetically [65, 66]. *H. pylori* RecA (37.6 kDa protein) is highly similar to known bacterial RecA proteins. The *H. pylori* recA mutants were severely impaired in their ability to survive treatment with DNA damaging agents such as UV light, methyl methanesulfonate, ciprofloxacin, and metronidazole. *H. pylori* RecA also played a role in survival at low pH in a mechanism distinct from that mediated by urease [66]. Disruption of *recA* in *H. pylori* abolished general homologous recombination [65]. Interestingly, *H. pylori* RecA protein is subject to posttranslational modifications that result in a slight shift in its electrophoretic mobility [67]. One putative mechanism for RecA modification is protein glycosylation. *H. pylori* RecA protein was shown to be membrane associated, but this association is not dependent on the posttranslational modification. The RecA modification is

In recent years, the phenotypes of *H. pylori recA* mutants have been further characterized in comparison with other mutants. Among the mutants of DNA recombination and repair genes, *recA* mutants displayed the most severe phenotypes. For example, *recA* mutants were much more sensitive to UV or Gamma radiation than the *recB* or *recO* single mutants, and were similar to the *recBO* double mutant [68-70]. The *recA* mutants completely lost the ability to undergo natural transformation [68-70]. The intra-genomic recombination frequency of the *recA* mutant was also much lower than that of the *recR* or *recB* single mutants [68, 71]. Finally, the *recA* mutants completely lost the ability to colonize mouse stomachs [69]. In competition experiments (mixed infection with wild type and mutant

response observed in *H. pylori*, although the underlying mechanism is unclear [55].

of DNA damage is not understood. No DNA repair genes, a hallmark of the SOS response, were induced in either the antibiotic-treated cells or the recombination gene deleted strain. Surprisingly, several genes involved in natural competence for DNA transformation (*com* T4SS components *comB3*, *comB4* and *comB9*) were induced significantly. Indeed, natural transformation frequency was shown to be increased under DNA damage conditions. Another DNA damage-induced gene was a lysozyme-encoding gene. Experimental evidence was provided that a DNA damage-induced lysozyme may target susceptible cells in culture and provide a source of DNA for uptake [55]. Taken together, DNA damage (mainly DSBs in their experiments) induces the capacity for taking up DNA segments from the neighboring cells of the same strain (homologous) or co-colonizing strain (homeologous) that may be used for recombinational DNA repair.
