Gallic acid diminishes pro-inflammatory interferon-γ- and interleukin-17-producing sub-populations in vitro in patients with psoriasis

Psoriasis is characterized by chronic inflammation of the skin. The interleukin (IL-)-23/T helper (Th) l17/IL-17 pathway has been recognized as crucial for the development of the disease. Skewing of CD4⁺ T cells towards a Th17 phenotype is in part mediated by tumor necrosis factor (TNF-) α and is characterized by signal transducer and activator of transcription 3 (STAT3) and retineic-acid-receptor-related orphan nuclear receptor gamma (ROR-γt) transcription factor upregulation [1]. Increased IL-17 production stimulates pro-inflammatory mediators in IL-17-R-bearing cell types, like keratinocytes. Another crucial component of the IL-17-mediated induction of psoriasis has been proposed to be nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) [2]. Moreover, in psoriasis, additional cell subsets, like Th1 or Th1-like Th17, have been found in increased fractions compared to HCs [3]. Interferon-γ (IFN-γ), which is produced by such cell populations [4], has been suggested to induce a transcriptomic signature similar to that of psoriatic skin in unaffected skin [5].

We have previously demonstrated that delphinidin, which is an anthocyanin, effectively diminished in vitro IL-17 and IFN-γ production by phorbol 12-myristate 13-acetate (PMA)-activated peripheral blood mononuclear cells (PBMCs) isolated by psoriasis patients [6]. The degradation product of delphinidin has been found to be a phenolic acid, namely gallic acid (GA) [7, 8]. GA is a phenolic acid compound [9], detected both free and as part of hydrolysable tannins, widely in fruits (i.e., pomegranate) and vegetables [10]. It is also found in red wine or green tea [11]. Biologic properties of non-flavonoid polyphenols such as phenolic acids, ranging from antioxidant, anticarcinogenic, to antimicrobial or anti-inflammatory, have been presented in numerous studies (reviewed by [12]). GA has also been examined for similar activities [13‐16]. A plethora of patented applications of GA and its esters in medicine has been reported. Indeed, various properties in industry as an anti-oxidant, an anti-angiogenic, or a cosmetic agent have been attributed to GA [11]. Intriguingly, the anti-oxidant activity of GA, when compared to other known anti-oxidant compounds, has been reported to be one of the highest [17, 18]. On the other hand, both GA and its esters have been reported to induce cell death in tumor cell lines [19, 20] suggesting that GA may have both anti-oxidant and pro-oxidant capacities [13]. GA’s anticancer activity has thus been investigated in many studies [21‐23]. Further beneficial properties of GA include anti-microbial and anti-viral activities [13].

Based on the aforementioned observations regarding the significant in vivo or in vitro effect of GA, an investigation of the modulation of PBMCs isolated by psoriasis patients by GA was considered of particular interest. Of relevance to the present study, mounting evidence also supports the immunomodulatory and anti-inflammatory properties of GA. In murine models, researchers showed that GA treatment significantly inhibited NF-κB activation. Moreover, GA treatment compared to mock treatment resulted in lower levels of interleukin IL-1β and TNF-α expression in the small intestine and also lower levels of both cytokines in mouse sera [24]. Furthermore, in a murine chronic constriction injury model, GA reduced the levels of TNF-α, NF-kB, and phosphorylated STAT 3 (p-STAT3) in mouse sera [25]. Intriguingly, in an allergic rhinitis mouse model, GA treatment alleviated symptoms. Moreover, GA treatment suppressed levels of IL-17 and Th17 transcription factor, ROR-γt in the nasal lavage fluid [26]. The inhibitory effect of GA on pro-inflammatory molecules has also been investigated in vitro. Suppression of NF-κB phosphorylation has also been reported in LPS-challenged IPEC-J2 cell cultures when pre-treated with GA. Importantly, GA also abrogated TNF-α and IL-8 gene expression in IPEC-J2 cells [27]. In another study, poly (GA)—a compound enzymatically produced by GA—pre-treatment reduced IL-1β and IL-6 levels in supernatants of PMA-stimulated THP-1 cell cultures [28]. In A549 lung cancer cell cultures, GA treatment resulted in attenuated phosphorylation of both JAK and STAT3 [29]. In a HT-29 cell line culture that was challenged with human recombinant IL-17 and TNF-α, IL-17A expression was significantly suppressed when cells were co-cultured with GA [30].

In view of the above, we were interested to assess the in vitro effect of GA in an immune-mediated disease, using psoriasis as a reference disease. Due to the high instability of delphinidin, we intended to investigate whether the in vitro or in vivo anti-inflammatory effects exerted by delphinidin could potentially be expanded by the subsequent activity of GA after delphinidin’s rapid degradation. Researchers have already supported that both delphinidin’s metabolites and GA promote Treg differentiation when co-cultured with CD4⁺ naïve T cells isolated from mice [7]. However, the immunomodulatory effect of GA on PBMCs from individuals diagnosed with psoriasis has to our knowledge not been explored. The lack of sufficient information regarding this matter has led us to investigate by multicolor flow cytometry the effect of GA on IL-17 and/or IFN-γ expression by PMA-stimulated PBMCs isolated by individuals with psoriasis and healthy controls (HCs).

The effect of various concentrations of GA on IL-17 and IFN-γ expression in PBMCs isolated by both participants with psoriasis (n = 28) and HCs (n = 12) was examined. Based on previous in vitro experiments involving GA of the literature, the final concentrations of GA tested were 5, 10, 30, 50, and 100 μg/mL for either 0.5 or 2 h. The strongest effect of GA on cytokine abrogation that was accompanied by cell viability preservation was observed in the presence of 30 μg/mL GA 0.5 h prior stimulation. Viability of cells at this concentration was maintained, while higher concentrations of GA diminished cell viability. The optimal concentration of GA for our experiments was thus considered to be 30 μg/mL (Online Resource 3). This concentration corresponds to that used for in vitro experiments using cell lines in published series [32‐34].

PBMCs were effectively sub-gated utilizing both surface and intracellular staining. Figure 1 presents gating strategy used to analyze our data. Cell viability for each sample used was assessed after thawing of cells. Percentage of cells that were stained with Annexin V, which identifies apoptotic cells, was consistently ≤ 10%. GA treatment did not affect cell viability. CD3⁺CD4⁺ and CD3⁺CD4⁻ cell sub-population percentages were not affected by GA treatment. The CD3⁺CD4⁺ and CD3⁺CD4⁻ cell population percentages did not significantly differ between the GA-treated and -untreated PBMCs, as depicted in Online Resource 4.

To our knowledge, this is the first study to explore the effect of GA supplementation on Th17-, Th1-, and IFN-γ-producing NK cells in patients diagnosed with psoriasis vulgaris. We hereby present a remarkable decrease in the frequencies of the aforementioned PMA-stimulated cell subsets upon in vitro GA exposure. Interestingly, the sharp reduction in IFN-γ-producing cell percentages mediated by GA included Tc cells too. Our in vitro results clearly demonstrate direct effects of GA on pro-inflammatory cytokines, which remarkably are observed in psoriasis patients, but not in HCs. Whether GA affects cytokine production in a disease-specific manner remains to be explored in future studies enrolling both larger cohorts and other disease-control patients. Nevertheless, we are confident that our data are significant, as the inhibitory effect of GA was consistent in all psoriasis patients tested, while lack of such an inhibitory effect was noted in every single healthy control assessed. Our data could be intriguing given that such a ramification noted in psoriasis patients but not in HCs was not evident in our previous in vitro study using the naturally GA-enriched compound of delphinidin.

The GA-mediated decrease on IL-17- and IFN-γ-producing cells provides indications that GA may be a compound of interest in any effort to explore new anti-inflammatory agents against the disease. IL-17 has long been characterized as the crucial cytokine in the pathogenesis of psoriasis [35]. High efficacy anti-IL-17 biologic therapies are used in the management of the disease [36]. The mechanisms responsible for the inhibitory effect of GA on IL-17 production by CD4⁺ cells are not clear. Transcription factor ROR-γt has been identified as a lineage defining factor of Th17 cells [37]. In nasal lavage fluid collected by ovalbumin-induced murine allergic rhinitis murine model, GA treatment suppressed levels of ROR-γt [26]. Furthermore, STAT3 has also been characterized as a Th17 transcription factor [38]. Inhibition of p-STAT3 has been reported to be mediated by GA [25, 29]. Additional to the inhibitory effect of GA on IL-17 production, the compound has been reported to diminish the activity of IL-17-stimulated downstream signal transduction pathways. The cytokine-driven feed-forward loop facilitated in psoriasis is transduced by certain molecules, including NF-κB. Mutations in genes encoding activators of NF-κB have been associated with psoriasis development [39]. Interestingly, GA has been found to suppress NF-κB phosphorylation in LPS-challenged IPEC-J2 cell cultures [27].

We also observed an effect of GA pre-treatment on the percentage of IFN-γ-producing NK cells following PMA stimulation. GA significantly decreased the fraction of CD3⁻CD56⁺IFN-γ⁺ cell subset. NK cells have been suggested to participate in the establishment of the inflammation loop of psoriasis [40]. Specifically, NK cells have been reported to accumulate in psoriatic skin and promote inflammation [41, 42].

Of relevance, recent in vitro evidence suggests that GA inhibits additional immune pathways that are known to be actively involved in the disease. For example, IL-6 has been identified as a key cytokine that participates in pro-inflammatory mechanisms in skin and eventually leads to development of psoriatic skin lesions. In psoriatic skin, dendritic cells produce high levels of IL-6, which promotes escape from Treg suppression [43] and skewing of CD4⁺ T cells towards a Th17 phenotype [44]. GA has been shown to mediate inhibition of IL-6 production in IL-33-activated basophilic KU812 cells in vitro [45]. In another study, poly(GA) reduced IL-6 levels in supernatants of PMA-stimulated THP-1 cells [28]. Furthermore, dendritic cells produce TNF-α [46], a central molecule in the inflammatory milieu of psoriasis, which directly affects keratinocyte gene expression [47]. In murine models, GA treatment has been associated with decreased levels of TNF-α in mouse sera [24].

The effect of GA on immune cell populations has been partly investigated [9]. In a recent in vitro study, GA was shown to enhance Foxp3⁺ cells (iTreg differentiation when C57BL/6 mice spleen-isolated CD4⁺ T cells were co-cultured in the presence of GA with TGF-β and IL-2). Moreover, GA co-culture with iTregs suppressed Teff proliferation compared to cultures where GA was absent. In an in vivo study, intraperitoneal administration of GA in an allograft rejection mouse model expanded Tregs and decreased activated T cells [7]. In a murine model of arthritis, GA decreased local inflammation, and inhibited the in vitro release of neutrophil TNF-a [48]. Furthermore, in monosodium urate crystal stimulated swelling of joints in mice, GA mitigated symptoms and inhibited IL-1β expression [49]. In the human setting, GA treatment of fibroblast-like synoviocytes isolated from patients with rheumatoid arthritis increases the levels of pro-apoptotic caspase-3 and decreases the expression of pro-inflammatory molecules in vitro [50]. Whether GA’s immunoregulating activity could also be exploited for the management of inflammatory joint diseases remains to be seen, but early in vitro or in vivo data suggest that such a probability should not be underestimated.

In our study, GA-mediated inhibition of IFNγ production was significantly higher in CD3⁺CD4⁻ T cells isolated from patients receiving bDMARDs compared to PBMCs isolated from treatment-naïve patients. A synergistic effect of GA with biotechnological drugs has been documented. In a melanoma-bearing mouse model, combination of GA and an IDO inhibitor exerted an in vitro effect in CD8⁺ and Tregs [51]. An in vitro combined effect of GA with doxorubicin was stronger in suppressing growth of DU145 prostate cancer cells [52] and the combined effect was also noted in human squamous carcinoma [53] and glioma cell lines [33]. To our knowledge, our in vitro study is the first to indicate an indirect synergistic effect of GA with a bDMARD. A hypothesis regarding the molecular mechanisms that could explain such a synergism could be formed. While bDMARDs inhibit the end product of the pro-inflammatory feed-forward loop, such as IL-23 or IL-17, GA has been reported to inhibit NF-κB and STAT3, and thus effectively decrease the proportion of pro-inflammatory cells such as the circulating Th17 cells. As shown in this study, GA inhibits the proportion of PBMCs producing IFNγ and IL-17. GA has also been suggested to have high binding affinities with the receptors of INF-a2, IL-4, and IL-6 [54]. Our in vitro approach prevented us from speculative arguments that could exploit this further but future research will shed a light in revealing such a possible synergism.

In our study, GA’s inhibitory effect on PBMC pro-inflammatory subpopulations was significant in patients with psoriasis and not in HCs. Our data cannot provide a mechanistic elucidation for such a finding. We could only hypothesize that the selective effect of GA in activated PBMCs is mediated by STAT3 and/or NF-κB hyperactivation; STAT3 hyperactivation significantly contributes to cytokine production and lesion development in patients with psoriasis [55], and GA is able to inhibit phosphorylation of STAT3. A “STAT3 hyperactivated state” in patients with psoriasis but not in HCs may account for the exerted inhibitory effect of GA and that may also be the case for NF-κB. Both molecules have been targeted in the past for the management of the disease [56]. Such speculative scenarios must be addressed experimentally in vitro and in vivo.

Whether GA could be utilized as a therapeutical agent against psoriasis or as a supplement to conventional treatment needs to be determined in proper in vivo studies and well-designed randomized control clinical trials. Important tackles to overcome relate to the absorption and bioavailability characteristics of GA. A few studies have investigated the absorption and bioavailability characteristics of GA. Orally administered concentrated tea brew (about three times concentrated compared to normal tea brew) in individuals led to a maximum plasma GA concentration of 2 μmol/L after approximately 1 h of tea consumption[57]. The reported relatively low plasma concentration is of limited relevance (up to 80 times lower) to the concentration used in our in vitro study, suggesting that the observed effects on PBMC subsets could result from bioenhanced pharmacologically relevant GA concentrations in order to be of clinical significance. The relatively low plasma concentrations achieved from oral compounds may limit its therapeutic potential. Such limitations have been observed in other diet supplements such as curcumin and led to the development of more bioactive formulas. Repeated dosing [58], usage of phospholipid complexation [59], or colloidal delivery system development [9] as a way to enhance the therapeutic effect of GA have also been investigated with favorable results. Finally, the biological effects of GA may be hampered by the rapid metabolism in humans. Poor stability and solubility may also explain the observed inconsistent oral bioavailability of GA [9].

GA appears to be very safe and significant issues regarding its toxicity have not been raised. Acute toxicity of GA is reported; only enormous doses are administered in mice (≥ 2000 mg/kg) [60]. No adverse events were observed in F344 rats fed with 120 mg/kg/day for 13 weeks [61]. Furthermore, in a murine model of pelvic inflammatory disease, administration of 210 mg/kg had no toxic effect [62]. In another study investigating the toxicity of GA, researchers reported that even high doses of up to 900 mg/kg/day GA for 28 days were not toxic [63]. Based on these results, GA’s administration could be evaluated as clinically relevant, since toxicity is reported only in extremely high doses. Nevertheless, toxicity and safety issues of more bioactive formulas must be assessed.

We consider that our results could constitute the impetus for future in vitro or in vivo studies that assess the pharmaceutical effect of GA in any form—ranging from the dietary intake to orally supplemented GA as a concentrated compound or even to topically applied creams or lotions containing GA alone or in conjunction to other anti-inflammatory agents—on psoriatic lesions.