PARP activation was first described in relation to DNA damage response, during which PARP1 cleaves NAD⁺ to yield ADP-ribose residues that it transfers to target proteins, building up poly (ADP-ribose) (PAR) chains (Chambon et al., 1963; Durkacz et al., 1980; Miller, 1975). A couple of decades later, a second PARP enzyme, PARP2 was shown to exert similar catalytic activity (Ame et al., 1999; Schreiber et al., 2002). In the early 2000s, it was recognized that tumour cells with defective homologous recombination DNA repair pathways, such as BRCA1/2-deficient cancer cells, are highly sensitive to the inhibition of PARP activity (Bryant et al., 2005; Farmer et al., 2005). That realization gave a significant boost to research and development efforts with PARP inhibitors, and currently olaparib, rucaparib, niraparib and talazoparib have FDA approval as anti-cancer agents. All of these PARP inhibitors display simultaneous inhibition of PARP1 and PARP2 (Knezevic et al., 2016; Thorsell et al., 2017), the two isoforms with highest activity, as PARP1 covers 85%–95%, while PARP2 covers 5%–15% of total cellular PARP activity depending on the model used (Ame et al., 1999; Schreiber et al., 2002).

As a result of intense studies conducted on the PARP field during the past decade, now ample evidence demonstrates that PARP activation and PAR synthesis are involved in a plethora of biological processes beyond DNA repair. These processes involve energy metabolism and mitochondrial biology (Szanto and Bai, 2020), lipid metabolism (Szanto et al., 2021), transcriptional and translational regulation (Jubin et al., 2017), cell death (Fatokun et al., 2014; Hong et al., 2004), differentiation and inflammation (Fehr et al., 2020; Kunze and Hottiger, 2019). In accordance, the clinically used PARP inhibitors displayed beneficial effects in several models of non-oncological, inflammatory pathologies, such as asthma (Ghonim et al., 2015), sepsis (Ahmad et al., 2019), alcoholic hepatitis (Gariani et al., 2017; Mukhopadhyay et al., 2017), acute pancreatitis (Ahmad et al., 2020), burn injury (Ahmad et al., 2018), and acute lung injury (Martins et al., 2022). Hence, it has been postulated that PARP inhibitors could be repurposed for the treatment of non-oncological diseases (Berger et al., 2018).

Psoriasis is one the most common inflammatory skin diseases, to which currently there is no cure. Psoriasis manifests under specific genetic and environmental factors, and it is characterized by the hyperproliferation and aberrant terminal differentiation of epidermal keratinocytes. This results in the formation of thickened, red plaques on various skin areas. (Griffiths et al., 2021). Psoriatic skin lesions contain a robust immune cell infiltrate, and psoriasis is considered primarily an immune-mediated disorder. It is widely accepted that T helper (Th)17 cells and the IL17 cytokine family has a prominent role in activating those signalling circuits in keratinocytes that sustain chronic inflammation in psoriasis (Antal et al., 2022; Lowes et al., 2007).

Psoriasis has several animal models, among them, the most widely utilized in psoriasis-related studies is the imiquimod (IMQ)-induced psoriasiform dermatitis (van der Fits et al., 2009). We previously studied this model in PARP1 ^−/− and PARP2 ^−/− mice. Surprisingly, we found that the dermatitis upon IMQ treatment was more severe in PARP1 ^−/− mice than in wild-type mice (Kiss et al., 2020), while PARP2 ^−/− mice displayed a protected phenotype against the IMQ-induced symptoms (Antal et al., 2023). In the current study, we aimed to test the effect of PARP inhibitors, which target PARP1 and PARP2 in equal measure, in the IMQ-induced psoriasis model.

The distribution of data was analysed by Shapiro-Wilk test. If two groups were compared, we used independent t-test (two-tailed), as the Shapiro–Wilk test showed normal distribution. If the distribution was not normal, Mann-Whitney test was used. When we compared more than two groups and if the distribution was normal, we used ANOVA followed by Tukey’s post hoc test. In case the data did not show a normal distribution, Kruskal–Wallis test were applied complemented by Dunn’s post hoc test. The significance level was set at 0.05. Statistical tests were performed using GraphPad Prism 9.1.2 software (GraphPad Software, Boston, MA, United States).

To test if PARP inhibition can affect IMQ-induced psoriasis-like dermatitis in mice, first we performed a pilot study with the four clinically available inhibitors. We included only male mice in this study for several reasons. Multiple studies show that psoriasis changes in women during pregnancy, postpartum and with menopause, indicating that psoriasis is influenced by estrogen hormones (Adachi and Honda, 2022; Lin and Huang, 2016). In addition, a recent study provided evidence that estradiol suppresses IMQ-induced psoriasis in mice (Adachi et al., 2022). Since female mice have estrous cycle, it is possible that the hormonal status of female mice could interfere with the development and severity of IMQ-induced symptoms. To note, studies suggest that the effects of PARP inhibitors also depend on the sex of the animals, with male mice being more responsive than females (Mabley et al., 2005; McCullough et al., 2005).

We found in the pilot experiment that, in the applied doses, only rucaparib and talazoparib reduced the extent of induration, erythema and scaling of the skin caused by IMQ treatment, while olaparib and niraparib did not affect the visual appearance of dermatitis compared to the cohort that received only vehicle (VEH) in i. p. injection before treatment of their skin with IMQ (VEH IMQ) (Figure 1A).

It is interesting that not all PARP inhibitors influenced the IMQ-induced psoriasiform lesions. This discordance may have several explanations. For one, we tested the inhibitors in a dosage selected based on oncological studies. However, it has been suggested that the optimal dose of PARP inhibitors in non-oncological conditions may be lower than the doses used in oncological indications (Ahmad et al., 2018), and that the inhibitors may have bell-shaped dose-responses (Teng et al., 2016). Another feasible explanation may be the known difference in the pharmacological and pharmacokinetic profiles of the compounds (Zeng et al., 2024). For example, niraparib and talazoparib have longer half-lives than olaparib and rucaparib, hence only niraparib and talazoparib are used in a once-daily dosing in therapy (Zeng et al., 2024). Talazoparib goes through minimal hepatic metabolism but its dosage should be reduced in patients with impaired renal function, while niraparib requires dose adjustments in patients with hepatic impairment (Zeng et al., 2024). In our experiments we formulated the PARP inhibitors for i. p. injection. Administration via oral gavage may have been closer to the tablet formulation how PARP inhibitors are applied in clinical therapy currently, and may have yielded different results.

Nevertheless, based on our findings and considering that talazoparib is the most potent PARP inhibitor (Curtin and Szabo, 2020), we chose the talazoparib-treated cohort for mechanistic studies. Figure 1B shows the back skin of mice that received either VEH or talazoparib daily i. p. injection prior to topical IMQ treatment (VEH IMQ and TAL IMQ, respectively) as compared to mice that received VEH i. p. and were treated with a control cream (VEH CTL) on day 5 of the experiment. The IMQ-evoked epidermal thickening was less pronounced in the skin of mice that received talazoparib (Figure 1C), as is visible on skin sections after hematoxylin and eosin (H&E) staining (Figure 1D).

Immunohistochemical analyses of skin samples revealed premature and smeared expression of the terminal differentiation marker involucrin (IVL) in those mice that received only VEH in i. p. injection before IMQ treatment (VEH IMQ) (Figure 2A). In talazoparib-administered mice (TAL IMQ), although the staining zone broadened, involucrin expression displayed a sharp appearance in the granular layer, more closely resembling normal skin (VEH CTL) (Figure 2A). In addition, analysis of nuclear BrdU incorporation showed fewer proliferating basal keratinocytes in mice that received talazoparib prior to IMQ application (Figure 2A). Next, we performed cytokine array from whole skin lysates. Since visually, the severity of lesions seemed reduced in the talazoparib-treated cohort, it was surprising to find that expressions of inflammatory markers did not change in the skin of these mice compared to the IMQ reference cohort (VEH IMQ). Moreover, inflammation-related chemokines CXCL1 and CCL12 could only have been detected in the samples of talazoparib-treated mice (TAL IMQ), and not in the skin of VEH IMQ mice, suggesting a slight pro-inflammatory effect of talazoparib in the skin (Figure 2B). In accordance, talazoparib did not lessen IMQ-evoked NF-κB activation as determined by detection of phosphorylated p65 positive nuclei in skin sections (Figure 2C).

Given that PARP1 has much higher average activity than PARP2 (85%–90% compared to 5%–15% of total cellular PARP activity), it may be interesting how the phenotype resulted from talazoparib application in the IMQ model channels more aspects of the PARP2 ^−/− phenotype than the PARP1 ^−/− phenotype in the same model. Namely, the skin of PARP2 ^−/− mice similarly displayed maintained terminal differentiation, reduced proliferation and overall alleviated symptoms, but also reduced inflammation after IMQ application compared to wild-type mice (Antal et al., 2023). Conversely, PARP1 ^−/− mice displayed more severe psoriasis-like dermatitis and enhanced inflammatory marker expression than the wild-type mice in the IMQ model (Kiss et al., 2020). In this regard, it has to be noted that not all functions of PARPs depend on their catalytic activity, some of their functions rely only on their physical presence and interactions with other proteins (Thomas and Tulin, 2013). The latter functions can only be observed in knock-out model systems and are out of the scope of PARP inhibitors. Characterization of enzymatic activity dependent and independent roles of PARP1 and PARP2 in psoriasis should be elucidated in future studies in order to better understand what signalling mechanisms could be targeted by PARP inhibitors that are relevant in psoriasis.

The phenotype observed upon talazoparib treatment can be explained by its pro-apoptotic effect (Das et al., 2023; Lin et al., 2023), as it has been reported that apoptosis has a crucial role in balancing proliferation and terminal differentiation of epidermal keratinocytes (Raj et al., 2006), and that psoriasis features dysfunctional keratinocyte apoptosis (Kastelan et al., 2009). As the apoptotic effector caspase-3 was shown to be activated during keratinocyte apoptosis and terminal differentiation (Allombert-Blaise et al., 2003; McGill et al., 2005; Raj et al., 2006), we studied caspase-3 expression in the skin of mice. In VEH CTL mice we detected caspase-3 positivity in the basal and lower spinous layers, which was similar in IMQ-treated mice receiving VEH i. p. injection (Figure 3A). In the TAL IMQ cohort, caspase-3 showed strong and even expression in epidermal layers (Figure 3A), indicating enhanced apoptosis in the skin upon talazoparib administration.

Next, we aimed to study the role of talazoparib in human keratinocyte differentiation. Differentiation of HPV-Ker human keratinocyte cultures in talazoparib-containing media resulted in higher expression of terminal differentiation marker keratin 10, which was accompanied by increased caspase-3 expression (Figure 3B). In addition, Annexin V/propidium iodide dual staining of differentiated keratinocytes revealed an increment in late apoptotic population when the cells were under talazoparib stimulation either alone, or in combination with psoriasis-mimicking cytokines IL17A and TNFα (Figure 3C). These data indicate that talazoparib improves keratinocyte differentiation by activating caspase-mediated programmed cell death. Then we further studied the possible underlying mechanisms of these findings. We have previously shown that PARP1 and PARP2 regulate mitochondrial biogenesis (Bai et al., 2011a; Bai et al., 2011b). Since higher mitochondrial content correlates with greater tendency of cells to undergo caspase-mediated apoptosis (Marquez-Jurado et al., 2018), we studied mitochondrial parameters in keratinocytes upon PARP inhibition with talazoparib. We detected higher MitoTracker (Figures 4A, B) and MitoSox (Figure 4C) fluorescence in talazoparib-treated HPV-Ker keratinocytes, indicative of increased mitochondrial mass and higher mitochondrial production of reactive oxygen species (ROS). Also, we found that talazoparib treatment caused a significant upregulation in pro-inflammatory cytokine mRNA expressions in human keratinocytes, which was even more pronounced when talazoparib was applied together with the combination of IL17A and TNFα (Figure 4D). Although this pro-inflammatory effect of PARP inhibition is somewhat surprising, it is not without precedent. We have previously reported that a combined treatment of HPV-Ker cells with olaparib and IMQ caused higher increase in IL6 mRNA expression than IMQ alone (Kiss et al., 2020). In a different study, olaparib enhanced UVB-induced TNFα and IL8 mRNA expression in human keratinocytes (Chiu et al., 2021). Moreover, olaparib treatment increased ROS production in human tumour macrophages, which essentially promoted pro-inflammatory and anti-tumour function of macrophages (Wang et al., 2022). The exact molecular mechanisms through which talazoparib increased both terminal differentiation and inflammation in keratinocytes is not yet clarified. In this regard, it may be important that we found a marked elevation in the expression of the stress-responsive transcription factor activating transcription factor 3 (ATF3) in talazoparib-treated keratinocytes (Figure 4E). ATF3, depending on context, can either positively or negatively regulate various signalling pathways involved in apoptosis and inflammation (Liu et al., 2024). Mitochondrial ROS-induced ATF3 has been associated with the induction of caspase-3-mediated apoptosis and pro-inflammatory response (Aung et al., 2016; Mashima et al., 2001; Nyunt et al., 2019). We assume that higher ROS production in talazoparib-treated keratinocytes caused the upregulation of ATF3, and consequently, the activation of caspase-mediated apoptosis and induction of inflammatory gene expression. Ultimately, higher apoptotic rate led to better terminal differentiation of keratinocytes, which, at least partly, could explain the amelioration of IMQ-induced symptoms in mice. The pro-inflammatory response observed upon talazoparib treatment calls into question the overall beneficial effect of talazoparib in the IMQ-induced psoriasis model. However, the afore-presented mode of action of talazoparib is very similar to what was described in the case of the commonly used anti-psoriatic topical drug dithranol (anthralin). That is, dithranol was shown to act by inducing mitochondria-driven apoptotic pathways (George et al., 2013; Holstein et al., 2017; McGill et al., 2005), while also producing skin inflammation, due to the generation of ROS (Lange et al., 1998; Peus et al., 2004).

Taken together, this study raises the notion that PARP inhibitors may have an anti-psoriatic effect, but several questions remain unanswered. First, from a translational point of view, it may be a limitation to our study that we used a co-treatment approach with PARP inhibitors and IMQ, through which talazoparib inhibited the full development of psoriatic symptoms, but that does not answer the question if PARP inhibitors could be effective in alleviating an already developed psoriasis, which is when treatment of the disease is typically initiated. We used this approach mainly because one of our aims with this study was to compare the application of PARP inhibitors in mice with the application of PARP1 ^−/− and PARP2 ^−/− mice that we used in our previous studies in the IMQ model. This approach may more closely mimic the potential preventive effects of PARP inhibitors on psoriasis flares and on the development of more serious symptoms.

Second, it has to be stressed that in this work only keratinocyte-related effects were studied, despite the fact that talazoparib was administered in i. p. injection in mice. Psoriasis is considered primarily an immune-mediated disease. Since PARP1 and PARP2 regulate various aspects of innate and adaptive immune responses (Rosado et al., 2013; Szanto et al., 2024), the skin phenotype we have seen in the psoriasis model in mice upon talazoparib treatment probably involves both immune-modulatory and the hereby presented keratinocyte-specific effects of talazoparib. How PARP inhibition affects immune responses in psoriatic conditions has to be determined in further studies in order to fully evaluate the potential of PARP inhibitor repurposing in psoriasis.