Helicobacter pylori activation of PARP-1

Abstract

Chronic infection of the human stomach by Helicobacter pylori is an important risk factor for gastric cancer. H. pylori produces a cache of virulence factors that promote colonization and persistence, which, in turn, contributes to a robust inflammatory response at the host-pathogen interface. Recently, we reported that H. pylori activates the abundant nuclear regulator poly(ADP-ribose) polymerase (PARP)-1, resulting in the production of the catabolite poly(ADP-ribose) (PAR). PARP-1 is emerging as a key player in establishing homeostasis at the host-pathogen interface. In this article, we summarize the discovery of H. pylori-dependent PARP-1 activation, and discuss potential roles for PARP-1 in H. pylori-mediated gastric disease. In light of the remarkable successes that have reported for treating inflammatory disorders and cancers with PARP-1 inhibitors, we discuss the prospects of targeting PARP-1 for treatment of H. pylori-associated gastric disease.

This article refers to:

Prologue

Successful pathogens often remodel host cells and tissues to create a more suitable niche for colonization.1 Active remodeling at the site of colonization brings about changes that can promote bacterial adherence, invasion and nutrient acquisition, while at the same time subverting host mechanisms for pathogen clearance. A common and effective strategy for pathogen-initiated remodeling is to specifically target key regulators of host cell function for modulation.

Helicobacter pylori, a human gastric pathogen, is an example of a bacterium that is very successful at remodeling the host-pathogen interface to promote long-term persistence.2 However, persistent infection with Helicobacter pylori is a well-established risk factor for the development of serious gastric disease in humans, including stomach cancer, which is the second leading cause of cancer-related death worldwide.3 Thus, there is considerable interest in identifying the molecular mechanisms that underlie gastric remodeling during chronic H. pylori infection, as well as how these changes contribute to persistence and disease.

Recently, we reported that the abundant nuclear protein, poly(ADP-ribose) polymerase (PARP)-1, becomes activated in H. pylori infected gastric epithelial cells and, moreover, that purified recombinant PARP-1 is directly activated by a protein factor present within the H. pylori culture broth after removal of intact bacterial cells (e.g., culture filtrate).4 PARP-1, which is the most abundant of the PARP super-family,5 is a 113-kDa protein with three well-defined functional domains: a DNA-binding domain that recognizes DNA strand breaks, a central automodification domain that is an acceptor for PAR and a NAD-binding catalytic domain.6,7 The catalytic domain of PARP-1 is typically activated by sensing and binding damaged DNA, which can result in a 500-fold increase over low basal levels of the synthesis of long, branched, negatively charged polymers of the metabolite poly-ADP ribose (PAR).8,9

In the presence of damaged DNA, PAR synthesis facilitates PARP-1 mediated base excision repair (BER), which is normally important for the long-term maintenance of genomic stability. However, under certain conditions, inappropriate PARP-1 activation, by alternative mechanisms, such as interaction with other proteins,10–12 may promote necrotic or apoptotic cell death, as well enhanced inflammatory signaling and secondary damage to host tissues.13 In this addendum, we discuss H. pylori-induced PARP-1 activation as a potential mechanism by which this pathogen might contribute to gastric remodeling by inducing inflammatory and apoptotic signaling pathways.

Discovery of the H. pylori PARP-1-activating Factor

The PARP-1 activating factor produced by H. pylori was discovered during a screen of H. pylori culture filtrates for enzymatic activities normally associated with bacterial toxins or effectors. The screen yielded a high molecular protein within the lysates of several mammalian cell lines, identified as PARP-1,4 that had been covalently modified in a manner consistent with ADP-ribosylation.14 Because ADP-ribosylation is an important modifying activity associated with several potent exotoxins (e.g., diphtheria and cholera toxins) and type III effectors (e.g., Pseudomonas aeruginosa ExoS and ExoT), these findings suggested that H. pylori might generate an ADP-ribosylating factor that specifically targets PARP-1 for modification.

However, further studies revealed that our initial hypothesis was incorrect. Efforts to map the ADP-ribose acceptor domain of PARP-1 indicated that all three domains of PARP-16,7 were required for H. pylori-dependent modification.4 Moreover, an active PARP-1 catalytic domain was required for H. pylori-dependent PARP-1 modification.4 As described above, the catalytic domain of PARP-1 synthesizes large quantities of PAR in response to activation, which can result in auto-modification or trans-modification of other nuclear proteins.6,9 Indeed, we found that purified recombinant PARP-1 was modified with PAR in an H. pylori dependent fashion. Thus, rather than generating a toxin that ADP-ribosylates PARP-1, H. pylori produces a novel protein factor that directly activates PARP-1, resulting in the synthesis of PAR. These studies are the first to describe a bacterial factor that directly activates PARP-1.4

Importantly, PAR was detected not only under highly defined conditions using recombinant proteins, but also within intact gastric epithelial cells that had been infected with H. pylori.4 Intracellular PAR production was nearly eliminated in PARP-1 knockdown cells and, moreover, there was little evidence of H. pylori-dependent ADP-ribosylation within lysates prepared from parp-1^-/- mouse embryonic fibroblasts.4 These data strongly suggested that PARP-1 is the target for H. pylori-mediated PAR synthesis and, moreover, suggested that other members of the PARP super-family (e.g., PARP-2) are not likely to be significant contributors to H. pylori-dependent PAR synthesis.

The Relationship between Persistent H. pylori Infection, Chronic Inflammation and Apoptosis at the Epithelial Barrier

A variety of diseases, including a number of cancers, are linked to chronic inflammation.15 Although both strain-specific virulence traits and host responses contribute to the risk that H. pylori infections will progress to disease, chronic H. pylori-induced gastric inflammation is the single most important risk factor for the development of stomach cancer. Subsequent to initial H. pylori infection, a well-characterized inflammatory response develops,16,17 which includes the secretion of pro-inflammatory cytokines by epithelial17–20 or mucosal antigen presenting cells,21 increased recruitment of neutrophils,22 and an IgA response.23 However, this robust inflammatory response is not protective, as H. pylori infection both promotes and limits the extent of inflammation, in part because regulatory Th cells (Tregs) are induced.24–28 The Tregs appear to protect the host against excessive tissue damage, while at the same time promoting persistent infection,28,29 thereby allowing H. pylori to exist in a long-term, dynamic equilibrium with its human host. Persistent inflammation contributes directly to tissue damage and increased apoptotic cell death within the gastric mucosa, both of which are hallmarks of persistent H. pylori infection.30–34 For a subset of individuals, the biological cost of the H. pylori-human long-term relationship is the development of one of several gastric diseases, including cancer.

The Very Talented PARP-1

PARP-1 was originally discovered as an important contributor to the maintenance of genome integrity, with functional roles in the modulation of chromatin compaction and decondensation, repair of DNA single strand breaks (base excision repair), repair of DNA double strand breaks, DNA replication and telomere homeostasis or integrity.6 More recently, PARP-1 has been functionally linked to several processes central to cell division, proliferation, differentiation and death.6 The normal physiological functions of PARP-1 continue to expand and the full range of PARP-1 activities has not likely been identified.35

In addition to the potential roles for PARP in normal physiology and function, PARP-1 has also been associated with a surprising number of disease states. These pathologies include, but are not limited to, carcinogenesis, tumor progression and hypoxia.6 As discussed below, a unifying feature is that many of these PARP-1 associated pathologies are inflammatory in nature.

PARP-1 as a Regulator of Inflammation

The importance of PARP-1 in inflammatory-mediated pathologies is accentuated by the success of using PARP-1 inhibitors in treating a number of inflammatory disorders, including ischemia/reperfusion,36,37 heart failure,38 hypertension,39 arthritis,40,41 atherosclerosis,42 and LPS-induced inflammation.43 PARP-1 inhibitors decrease the in vivo production of inflammatory cytokines and chemokines, decrease antigen-specific Th1-cell expansion and induce the production of the anti-inflammatory cytokine IL-10.40 PARP-1 inhibition also reduced neutrophil recruitment and multiple organ failure in a rodent model of inflammation.44 Finally, monocytes/macrophage-mediated release of TNFalpha, IL-1 and IL-6 was reduced by PARP-1 inhibition.45

Causal associations between H. pylori-mediated PARP-1 activation and host inflammatory responses have not yet been explored, although potential mechanistic links do exist. For example, overactivation of PARP-1 is associated with increased activity of pro-inflammatory factors like activator protein-1 (AP-1), mitogen-activated protein (MAP) kinases (MAPKs) and nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB).46 At the same time, activation of AP-1, MAPK and NFκB are hallmarks of H. pylori infection, and are thought to contribute to increased inflammatory cytokine expression, an increased rate of apoptosis, an increased proliferation rate and altered cell cycle in gastric epithelial cells.47

PARP-1-mediated inflammation may occur via indirect mechanisms, such as necrotic cell death that occurs as a result of NAD depletion.48 However, direct mechanistic roles for PARP-1 in regulating pro-inflammatory processes at the transcriptional level have also emerged. PARP-1 can directly interact with NFκB and promote NFκB interactions with other proteins as well as DNA to promote the expression of proinflammatory cytokines as well as the activity of nitric oxide synthases.49–51 Moreover, PARP-1/NFκB interactions also regulate cellular differentiation, proliferation and death.52 PARP-1/NFκB interactions are regulated at several levels, and requires PARP-1 acetylation by a p300/cAMP response element-binding-binding protein.53,54 Acetylation-dependent PARP-1 trans-activation functions are balanced by the sumoylation of PARP-1. It remains to be seen which, if any, of these pathways are involved in mediating the consequences of H. pylori-dependent PARP-1 activation.

PARP-1 as a Regulator of Cell Death

While PARP-1 activation under mild genotoxic stress normally promotes cell survival, inappropriate PARP-1 activation and excessive PAR production can instead promote cell death. The importance of PARP-1-mediated cell death is accentuated by genetic evidence in some animal models that over-activation of PARP-1 contributes significantly to the cell death that accompanies stroke, diabetes, inflammation and neurodegeneration.55

PARP-1 activation can induce cell death by several overall mechanisms. Necrotic cell death can result from the depletion of cellular NAD that accompanies excessive PAR production.48 PARP-1 activation has also been linked to a more direct pathway of cell death called parthanatos, which is derived from “PAR” and “thanatos,” which is the Greek reference to death and mortality.55–57 “Parthanatic” cell death shares characteristics with both apoptosis and necrosis.56 However, one of the distinguishing characteristics of parthanatos is that this pathway is triggered by excessive PARP-1 activation and PAR production. A key event in parthanatic cell death is the release of the mitochondrial localized apoptosis-inducing factor (AIF) by a PAR-dependent mechanism.58,59 Several lines of evidence suggest that PAR and AIF directly interact,60 and that both of these factors have important roles in PARP-1-dependent cell death,61,62 AIF is a mitochondrial-localized flavoprotein63 that normally has an important role in cellular bioenergetics.64,65 During parthanatos, PARP-1 synthesized PAR translocates from the nucleus to the mitochondria and, by a poorly understood mechanism, promotes release of AIF to the cytosol.61,62 AIF in turn, localizes back to the nucleus and participates in killing of the cell.

Although it is currently unclear whether the characteristic increase in cell death within the gastric epithelium accompanying H. pylori infection involves PAR-dependent mechanisms of cell death, it is noteworthy that AIF release from mitochondria during H. pylori infection has been reported in both cultured gastric epithelial cells66 and T and B cells.67 In preliminary studies,4 we discovered that within H. pylori-infected gastric epithelial cells, AIF release was significantly inhibited in the presence of a PARP-1 inhibitor, suggesting a potential link between H. pylori-mediated PARP-1 activation and parthanatos.

PARP-1 as a Modulator of Host Responses to Pathogens

Contributions of PARP-1 towards the outcome of human infections with pathogenic microorganisms are only beginning to be explored. However, the importance of PARP-1 in mediating important physiological responses to cellular stress suggests that this nuclear factor could potentially have a functional role in the outcome of some host infections with pathogenic microbes. Several studies suggest that PARP-1 may have multiple and sometimes paradoxical roles in both regulating and promoting infections.

PARP-1 was recently reported to function as a transcriptional repressor of the anti-inflammatory cytokine interleukin (IL)-1068 in humans with a single nucleotide polymorphism in the promoter region associated with increased susceptibility to septic shock during pneumococcal infection.69 In another recent study, PARP-1 was implicated in a mouse model for enterocolitis induced by S. typhimurium.70 Specifically, in response to Salmonella infection, PARP-1 deficient animals demonstrated reduced infiltration of immune cells into the gut, along with severely delayed inflammation.

PARP-1 has been demonstrated to be important for the life cycle of several viruses. PARP-1 is important for retroviral integration into the host genome,71 including efficient human immunodeficiency virus type 1 (HIV-1) integration.72 HIV-1 infection was nearly abolished using PARP-1 knockout fibroblasts.72 PARP-1 is rapidly activated following Sindbis virus infection and contributes to apoptotic cell death.73 For Epstein-Barr virus (EBV), PARP-1 has apparently two roles, both to protect EBV from excess DNA damage, while exerting antiviral activity to protect the cell from excessive EBV proliferation.74

PARP-1 as a Pathogen Target?

H. pylori is the first pathogenic bacteria reported to directly modulate PARP-1 activity. However, given the increasing evidence for the significance of PARP-1 during host-pathogen interactions, it is perhaps not surprising that there is an expanding body of evidence indicating that PARP-1 may be an important target for modulation by some pathogens. In fact, several pathogenic viruses have been recently reported to produce proteins that directly interact with and activate PARP-1. For example, two capsid proteins of the Simian virus 40 (SV40) directly interact with PARP-1.75 PARP-1 interaction with the minor capsid protein VP3 directly activates PARP-1 enzymatic activity. VP3-mediated PARP activation was linked to necrotic death of infected cells and the release of SV40 to the culture supernatant, suggesting a possible role for PARP activation in SV40 dissemination.75 The polyomavirus capsid protein viral protein 1 (VP1) has been demonstrated to activate PAR synthesis by interacting directly with PARP-1 and knocking down or knocking PARP-1 in target cells reduced early viral transcription.76 The Sindbis virus non-structural protein 3 interacts with PARP-1 within infected neuronal cells.77

Other pathogens may act to prevent PARP-1 function. For example, the host immune response to viral infections was blocked by hepatitis B virus-mediated inhibition of PARP-1-dependent type I interferon receptor 1 transcription.78 A second example is the prevention of PARP-1 nuclear localization by the Vpr protein of HIV-1, in complex with the glucocorticoid receptor, resulting in suppression of NFκB.79 These few examples seem to suggest that some pathogens specifically target PARP-1 for activation, while others prevent PARP-1 activation as virulence strategies.

Exploiting H. pylori-activation of PARP-1 Therapeutically?

PARP-1 inhibitors have emerged as promising and effective therapeutics for breast, ovarian, melanoma and lymphoid cancers,80 but the use of PARP-1 inhibitors for the treatment of H. pylori-associated stomach cancers remains to be evaluated clinically. However, using pre-clinical murine models of H. felis induced atrophic gastritis, epithelial hyperplasia and metaplasia, the small PARP-1 inhibitor PJ34 was recently reported to prevent the formation of Helicobacter-induced precancerous lesions.81 Perhaps more exciting, however, was that PJ34, when combined with antibiotics, effectively reversed existing Helicobacter-induced pre-cancerous lesions.81 These results are potentially highly significant because while antibiotics often effectively eliminate H. pylori infections in patients presenting with clinical symptoms, eradication therapy alone does not reverse H. pylori-associated metaplasia or dysplasia, which are strong predictors for the development of gastric cancer.82

As mentioned above, PARP-1 is an important component of the BER pathway to repair damaged DNA. Thus, impairment of BER through PARP-1 inhibition may enhance the cytotoxicity of radiation or chemotherapy drugs that generate single-strand breaks in DNA.83–85 In addition, PARP-1 inhibitors are emerging as “stand-alone” cancer therapeutics in their own right by inducing “synthetic lethality” in cancer cells in which DNA repair mechanisms that normally rescue BER-deficient cells are themselves impaired.86 Some PARP-associated cancers have been correlated to mutations in the genes encoding the tumor suppressors BRCA1 and BRCA2 (breast cancer 1 and 2, early onset), which are important components of the homologous recombination repair pathway. Damage to DNA in BRCA1- or BRCA2-deficient cells is often not sufficiently repaired and can thus evolve into cytotoxic double stranded breaks. Thus, inhibition of a second, alternative DNA repair pathway, such as the BER pathway with PARP-1 inhibitors, can be highly effective at inducing selective lethal cytotoxicity in the tumor cells of BRCA1- or BRCA2-deficient individuals. This principle is thought to underlie the highly encouraging anti-tumor activity resulting from PARP-1 inhibition in breast cancer, as well as other BRCA deficient cancers such as ovarian cancer.86,87 Interestingly, BRCA mutations have now also been linked to increased risk in gastric cancers.88,89 It remains to be tested whether PARP-1 inhibition might also be an effective strategy for killing tumor cells and reversing pre-neoplastic lesions in patients with BRCA-deficient stomach cancers.

Future Prospects

There are a number of key issues surrounding the discovery of H. pylori-dependent PARP-1 activation that have yet to be explored. Detailed mechanistic studies will require that the PARP-1 activating factor be identified, likely via a combination of biochemical and genetic approaches. Consistent with the idea that the factor is released, direct contact of H. pylori with epithelial cells is not required for PARP-1 activation, as the application of sterilized culture filtrates alone, in the absence of bacteria cells, was sufficient for inducing activation.4 PARP-1 was not activated when mammalian cells were incubated with bacteria that had been seperated from culture filtrate and then killed by mild heat treatment,4 further suggesting that the activating factor is not normally associated with the bacterial outer membrane surface. Currently, it is not known whether the PARP-1 activating factor is specifically secreted or released non-specifically, although very little in vitro PARP-1 activating activity was detected within whole cell bacterial lysates,14 suggesting that the activity does not accumulate intracellularly. The PARP-1 activating factor is not likely to be a small peptide, as activity is retained in dialysis membranes with molecular mass cutoffs of either 8 or 20 kDa (unpublished observation).

The mechanism of H. pylori-mediated PARP-1 activation is unknown, although preliminary studies suggest that a protein factor released by the bacterium may interact with the DNA-binding domain of PARP-1.4 Consistent with the notion that the activating factor must access intracellular PARP-1 for activation to occur, the presence of a protein transfection reagent that promotes protein entry into cells resulted in significantly more PAR production than cells incubated with culture filtrate alone. Nothing is known about the pathways involved in the uptake and intracellular transport of the H. pylori PARP-1 activating factor, but it is likely that localization to the nucleus will be required for the bacterial factor to access PARP-1.

Ultimately, the importance of the PARP-1 activating factor for H. pylori colonization, persistence or disease pathogenesis will be determined using human epidemiologic studies coupled with animal studies employing an isogenic bacterial mutant lacking only the gene encoding the PARP-1 activating factor. The in vivo expression patterns of the PARP-1 activating factor in animal models of H. pylori colonization and disease, as well as during human infections may also provide clues as to the importance of PARP-1 modulation in establishing and maintaining homeostasis at the host-pathogen interface. Finally, identifying inflammatory and apoptotic signaling affected by PARP-1 activation will provide insights into the functional roles for PARP-1 during chronic infection.

Conclusions

PARP-1 is emerging as a key player in establishing homeostasis at the host-pathogen interface. Because of increasing evidence of the importance of PARP-1 in regulating host responses to infection, it is perhaps not surprising that a number of pathogens, primarily viruses to date, have been discovered to specifically target PARP-1 for modulation. The discovery of H. pylori-mediated PARP-1 activation suggests a novel pathway by which a bacterium may modulate host responses to establish and maintain the long-term dynamic relationship within its gastric environment. The use of PARP-1 inhibitors in successfully treating a broad spectrum of inflammatory diseases and cancers raises the interesting possibility of PARP-1 targeted therapeutic intervention for treating H. pylori-mediated gastric diseases.

Acknowledgements

This work was supported from a grant from the National Institutes of Health R01 AI045928 (S.R.B.).

Addendum to: