While most well known for its role in base excision repair, PARP-1 also serves to regulate other signaling pathways, such as cell death and inflammation. PARP-1 inhibition has been shown to be beneficial in a milieu of animal models of inflammatory diseases, such as diabetes, asthma and atherosclerosis [38]. The role of PARP-1 in inflammatory signaling is becoming more evident as the focus of study shifts from solely examining its involvement in DNA repair to other processes including mechanisms of inflammation.

PARP has been shown to contribute to inflammatory processes through a variety of means, including the regulation of transcription factors, cytokines, adhesion factors and inflammatory mediators [39,40]. The well-known and thoroughly described transcription factor, NFκB was one of the first mediators of inflammation to be identified as a target for PARP activity [41]. NFκB has been shown to be a binding partner and also to be PARylated by PARP-1, as previously discussed. PARP-1 is known to be a co-activator of NFκB [41,42]. These studies established in PARP^−/− mice and cell lines, that NFκB activity was greatly abrogated independent of upstream activation of NFκB [41,42]. PARP^−/− mice were protected from LPS-induced endotoxic shock, due to a strong downregulation of downstream signaling components that propagate the inflammatory signal, including inducible nitric oxide synthase (iNOS) and TNFα [42]. Recently, this has been further confirmed as it has been shown that systemic inflammation serves to upregulate several members of the PARP family in mouse hippocampus, including PARP-1,-3, -9, -12 and -14 [43]. Additionally, the levels of PARylation were increased, and importantly, levels of iNOS were upregulated. PARP inhibitor served to protect cells against increased PARylation and increased iNOS gene expression which is an integral player in ROS/NOS-induced stress in systemic inflammatory response [43]. In both of these cases, knockout and inhibition of PARP served to downregulate the inflammatory response by abrogating NFκB signaling.

The role of PARP-induced inflammation is also multifaceted. It serves as a co-factor for NFκB, and also serves to switch cell death pathways from the well-controlled and highly regulated apoptosis to the more inflammatory necrosis in the event of high levels of ROS species [44,45,46]. In the inflammatory response, the cellular environment is subjected to increased DNA insults carried out by inflammation-induced ROS. This then can lead to the upregulation of PARP-1; as expression of PARP-1 is dramatically increased with DNA strand breaks [47]. The upregulation of PARP-1, and subsequent elevated PARP-1 activity, can deplete cellular levels of ATP and NAD, forcing the cell into necrosis [44,45]. While BER/SSB processes are upregulated, the increase in PARP-1 will also allow PARP-1 to serve as a co-activator of NFκB. If the delicate balance of insult-repair ratio is disturbed, rapid energy depletion and necrotic cell death can occur, resulting in increased inflammatory signaling [45]. By serving as both an NFκB coactivator and pushing ROS-exposed cells to necrosis, PARP-1 functions to further propagate the inflammatory signal. Combined with the regulation of cytokines and inflammatory mediators such as iNOS, it is clear that PARP-1 is a potentially substantial contributor to inflammation.

In support of the pro-inflammatory functions of PARP-1, PARP inhibitors have been demonstrated to be an effective treatment of many inflammatory-related diseases in preclinical models. The wide array of inflammatory conditions that have shown improvement by PARP inhibition ranges from chronic diseases such as diabetic neuropathy and atherosclerosis to more acute events such as sunburn-related inflammation and cisplatin-induced kidney inflammation [48,49,50,51,52,53,54]. These inflammatory states are wide in breadth and demonstrate that PARP may contribute to several levels of regulation in a wide range of inflammatory processes and stimuli. These mechanisms are yet to be fully determined and provide subject for further investigation. An overview of the multiple cellular processes regulated by PARP-1 is depicted in Figure 1.

In the updated version of the landmark description of the hallmarks of cancer, Hanahan and Weinberg describe two “enabling characteristics” that allow acquisition of the traditional cancer hallmarks such as resisting cell death and inducing angiogenesis. These enabling characteristics are: (1) genome instability and mutation and (2) tumor-promoting inflammation [55]. Interestingly, PARP-1 sits at the nexus of both of these enabling characteristics. As previously discussed, PARP-1 plays a large role in genome maintenance (DNA repair, chromatin remodeling, transcription factor regulation), and also contributes to the propagation of the inflammatory phenotype. Clinically, the focus of PARP-1 is as a target for the treatment of familial cancers, such as BRCA1/2 deficient breast and ovarian tumors [56]. However, based on its many functions, PARP inhibition may be applicable as a mode of therapy beyond familial cancers.

The majority of the literature to this point has not addressed the possibility of PARP as a pro-tumorigenic factor. This interesting paradigm may be related to its interaction with NFκB, its role in inflammation and subsequent ROS generation, or its pro-survival functions. Mice deficient in PARP have shown decreased susceptibility to skin cancer due to reduced NFκB signaling [57]. In an in vitro model of melanoma, PARP-1 driven NFκB signaling led to an increase in the secretion of pro-metastatic cytokines during senescence, termed the PARP-1 NFκB-Associated Secretome, or PNAS. Inhibition of PARP or NFκB in this system abrogates the proinvasive phenotype conferred by PARP-1/NFκB signaling [58]. The extracellular signal-activated kinase, or Erk, has been shown to modulate PARP-1 activity, including the phosphorylation of PARP-1 and the PARylation of NFκB (p65) in an inflammatory response model. This suggests that Erk may also regulate the PARP-1-dependent activation of NFκB [40], further points to a role for this interaction in cancer, as dysregulated Erk pathways are thought to be a present in an estimated one-third of cancers [59]. Additionally, in the mouse model where NFκB signaling was disrupted, a reduction in mammary tumor formation was observed, indicating NFκB signaling as a requirement for tumorigenesis for HER2+ tumors [60]. Interestingly, our laboratory has also shown HER2+ breast cancer cells are sensitive to PARP inhibition independent of HR repair mechanisms but rather due to suppression of NFκB signaling. This further suggests PARP and NFκB together play a crucial role in carcinogenesis [61] and provides more evidence for the broader utility of PARP inhibitors in cancer therapy.

While these studies are not conclusive evidence that PARP-1 plays a role in tumorigenesis, it is worth considering that PARP-1 regulated signaling may contribute to the initiation or progression of tumors. Interestingly, PARP-1 is highly expressed in a variety of cancers, including breast and hepatocellular carcinoma (HCC), and is correlated with poor prognosis in early breast cancers [62,63,64]. In studies of breast tumors, PARP-1 expression was correlated with higher grade and estrogen receptor (ER) negative status, suggesting that PARP-1 was over-expressed in more aggressive tumor subtypes [63,64]. In the HCC study by Shimizu et al., it was found that PARP-1 expression was significantly increased in HCC samples as compared to uninvolved liver tissue, and that PARP-1 expression was also increased in cirrhosis tissues, further implicating PARP-1 in the inflammatory response and in the susceptibility to tumorigenesis [62]. Recently, it was also shown that PARP-1 expression as well as PARylation levels were increased in non-small cell lung cancer cell lines resistant to cisplatin, as compared to their cisplatin-susceptible counterparts, indicating PARP-1 expression and activity may be correlated with tumor resistance to therapy [65]. In addition, PARP-1 and its interaction with the cell cycle control protein p21 (CDKN1A) may also contribute to tumorigenesis and the tumor phenotype. Mechanisms published by Cazzalini et al. [66] demonstrate that p21 interacts directly with PARP-1 in the DNA repair process, and knockdown of p21 lead to increased PARP-1 PARylation activity. p21 is often downregulated in cancers due to its regulation by p53 (reviewed in [67]), and based on this study could lead to increased PARP-1 activity. Along these lines, downregulated p21 and subsequent increased PARP-1 activity may provide an additional mechanism through which PARP-1 may play a role in tumorigenesis.

Together, these studies present an interesting perspective on PARP-1 as a tumorigenic factor that is yet to be fully explored. Is chronic activation of PARP-1 in highly oxidative environments responsible for the increased inflammation leading to genome instability and cancer, or are the DNA damages causing an upregulation of PARP-1? One potential mechanism through which PARP-1 and the inflammation-DNA damage feedback loop could be propagated is through the development and attempted repair of oxidative clustered DNA lesions.

Endogenous or spontaneous OCDLs occur normally within cells of humans and animals in the range of 200–500 clusters/Gbp [78,79,80]; however, several tumor types have been shown to harbor an increased level of OCDLs [81]. This suggests that there may be correlation between a higher risk of tumorigenesis in tissues and accumulated OCDLs. Additionally, oxidative stress is thought to contribute to tumorigenesis as many chronic inflammatory states lead to increase risk of cancer, such as H. pylori infections, hepatitis, and ulcerative colitis [1]. This is due in part to the large production of free radicals in the inflammatory environment leading to increased oxidative DNA damage and genome instability [82]. Decreased clusters were found in cells treated with selenium, a known anti-oxidant, and has been shown to protect against oxidative-induced DNA damage and cancer [79,83]. Functionally, OCDLs may contribute to tumorigenesis as they have been shown to display increased risk for mutagenicity and genetic instability due to the inherent challenges associated with their repair [76,84,85]. Furthermore, IR-induced lesions were shown to persist in mouse tissue 20 weeks post treatment, suggesting these lesions are long lasting and difficult to repair, which could lead to an increased risk of mutagenicity, and tumorigenesis [86].

The potential roles of PARP-1 and OCLDs in tumorigenesis provide a thought-provoking paradigm where in “normal” conditions, ROS induces DNA damage, which is then repaired via the BER process that involves PARP-1 recruitment and enzymatic activity. However, in highly oxidative environments or conditions of chronic inflammation, the sustained upregulation of PARP-1 from the repair process may lead to more inflammation and ROS generation, contributing to the formation of OCDLs. This introduces an interesting question. Is PARP-1 responsible for the increased inflammation leading to genome instability (i.e., through potentiating OCDLs), or are the DNA damages causing an upregulation of PARP-1, and thereby permitting inflammation to influence the cellular environment? This resulting inflammatory feedback loop propagates the phenotype that may be contributing to tumorigenesis. There has been no work to our knowledge that addresses these questions as of yet, but it does present an intriguing area for new research. Additionally, it may provide a feasible avenue for the novel use of PARP-1 inhibitors to prevent cancer outside that of BRCA-associated tumors.

While clinical trials have been conducted for the use of PARP-1 inhibitors in the treatment of cancer, the focus of PARP-1 inhibition was concentrated on its use to obtain synthetic lethality in familial cancers, such as those with a BRCA1/2 mutated phenotypes [87,88,89]. However, this potential connection to OCDLs and inflammation may provide a wider therapeutic application for PARP-1 inhibitors as a potential preventative agent in cancers highly associated with inflammatory phenotypes. By treating with PARP-1 inhibitors, one could reduce the inflammatory process, and thereby also decrease the induction of oxidative induced lesions (like OCDLs) that may lead to cancers.

The principle behind PARP inhibitor treatment of familial cancers deficient in homologous recombination (HR) repair is that by inhibiting the BER pathway in conjunction with endogenous failure of the HR pathway, the cell will be incapable of repairing DNA damage leading to cell death. However, our lab has demonstrated that PARP-1 inhibition also kills HER2+ breast cancer cells despite proficient HR repair, indicating PARP-1 may be an effective treatment for non-familial, sporadic cancers as well as familial cancers [61]. Interestingly, we found that susceptibility of non-familial cancers to PARP-1 inhibition may be through altered NFκB signaling [61]. While the mechanisms behind this are not yet determined, it does provide evidence that the applications of PARP-1 inhibitors in the treatment of cancers may be further extended than originally thought. Therefore, it may be possible to use of PARP-1 inhibitors to not only treat familial cancers, but also to treat and even prevent cancers of non-familial origin, especially cancers associated with chronic inflammation. Chronic PARP-1 activation in tissues undergoing high metabolic activity or tissues subjected to continual ROS-induced stress may be at increased risk for developing cancer driven by PARP-1 signaling, which may be the case for the study in hepatocellular carcinoma mentioned previously which show elevated PARP-1 levels in cirrhotic patients [62]. Together this points to PARP inhibition as an attractive target not only for familial cancers, but also as a treatment, and potentially a preventative agent, for cancers in tissues or organs experiencing chronic ROS-induced damage or high levels of metabolism. A schematic of this potential divergent role of PARP-1 is shown in Figure 2.