**3. Discussion**

Our laboratory has been studying mechanisms leading to endometriosis and pain experienced by endometriosis patients [55–58]. This study stemmed from our previous investigations into the miRNA profile of endometriosis tissues and PF treated cells [33]. Nineteen percent of differentially expressed miRNAs in endo tissues targeted JARID2. Despite the global downregulation seen in the micronome of endometriotic tissues [33], miRNAs that targeted JARID2 were highly expressed in the eutopic tissues of endometriosis patients who also experienced pain as a symptom. The overexpression of miR-148a, miR-29a [33], and miR-155 in endo tissues (Figure 1B) seemed to further support this theory. As shown in Figure 1A, there was an increased expression of PRC2 complex proteins such as *EED* (0.0067), as well as a noticeable trend in overexpression of corresponding genes in ectopic tissues from endometriosis patients, particularly in *EZH2*. This correlates with the findings of Colon-Caraballo and colleagues [8,23] and supports the characterization of EZH2 as a contributor to transcriptional repression and progression of the disease.

Although miR-155 was not originally identified based on the micronome array (*p* > 0.05), its relationship with JARID2 has recently drawn the attention of researchers in the field of inflammatory diseases [34,35]. miR-155 seems to play a key intermediate that regulates the crosstalk between JARID2 and PRC2 complex. miR-155 also plays a role in inflammation by working with FOXP3 to promote an inflammatory environment, since it has been shown that FOXP3 induces miR-155 expression [38,42]. Hence, miR-155 is a potential therapeutic target. This study explored the role of miR-155 in endometriosis by studying its interactions with the PRC2 complex, JARID2 and FOXP3. Endometrial cells were transfected with a miR-155 mimic or antagonist and then exposed to endo or control PF treatments. All PRC2 complex proteins examined showed an increase in expression in endo PF treated cells when compared to media alone treated cells. However, the cells transfected with a miR-155 mimic showed a downregulation of PRC2 complex proteins when exposed to either control or endo PF. The effect of gain- or- loss- of function of miR-155 on *JARID2* expression was interesting. miR-155 mimic transfected cells treated with control PF showed an increase in *JARID2* expression, while endo PF showed no change in expression. When transfected with the miR-155 inhibitor, no statistical difference in expression was seen in cells treated with control or endo PF. These results were unexpected and suggest that the miR-155 regulation of JARID2 is not sufficient to alter its expression. Hence, other transcription factors and/or epigenetic mediators could play a role in its aberrant expression in endometriosis.

FOXP3 showed a significant increase in expression in both control and endo PF treated cells when transfected with a miR-155 mimic (Figure 7C), which paralleled the results seen for mRNA expression (Figure 6C). Such interactions between miR-155 and FOXP3 has been observed earlier. In diffuse B-cell lymphoma (DLBCL), high FOXP3 expression was correlated with a poor prognosis in patients and when miR-155 was silenced in these cells, there was a parallel decrease in FOXP3 levels [59]. In breast cancer, it was found that FOXP3 and miR-155 work together to down regulate ZEB2, resulting in reduced invasion [42].

Methylation of the FOXP3 promoter could be partly responsible for pain that women with endometriosis may experience based on the trend of increased methylation in cells treated with PF from endo patients, particularly those reporting pain (Figure S2). This has been seen in both biliary atresia and prostatitis [60,61]. Bamidele and colleagues looked at the interaction of EZH2 and FOXP3 in inflammatory bowel disease and found that a mutation in FOXP3 disrupted EZH2 recruitment and its co-repressive function. They also showed that IL-6 voided the FOXP3-EZH2 interaction and that this destabilized interaction may drive the gastrointestinal inflammation [62]. This disruption in interaction may also be true in endometriosis, since we and others have shown that IL-6 is increased in patients with endometriosis [55,63]. While *FOXP3* mRNA expression in endo PF-treated cells trended to be higher than that of cells treated with control PF, there was no statistical

significance observed between the two treatments. These results suggest that FOXP3 is working alongside miR-155 to modulate the expression of EZH2.

*EZH2* mRNA expression in cells treated with both control and endo PF were higher when compared to media alone cells, but there was no statistical difference seen. However, an upregulation was seen in H3K27me3 in cells treated with endo PF. This is also significant as H3K27me3 is the downstream target of EZH2 and performs the transcriptional repression in cells [64]. The benefit of studying the PRC2 complex proteins in tissues and treated cells gave us the ability to compare short-term (in vitro) and long-term (in vivo) effects of peritoneal fluid on endometrial cells. This difference is likely to contribute to explaining the disparities in the observed results.

ChIP-qPCR was used to better understand the regulatory roles of JARID2 and EZH2 and their cross-interactions in endometriosis. By observing how it binds to regulatory elements of various genes, a sense of how the mechanisms described above differ between PF from patients with and without endometriosis was gained. The data presented in Table 1 showed that the pull-down expression of JARID2 by EZH2 IP was by over 5-fold higher in cells treated with endo PF compared to control PF. It is interesting to note that, while not significant, JARID2 IP has a fold-change greater than 1 for EZH1, while for EZH2 it is less than 1 for endo PF treated cells. This suggests that the JARID2 interaction with EZH2 may not be as strong as it is with EZH1, which can also methylate H3K27 to contribute to transcriptional repression. Although it is typically associated with active domains, EZH1 can actually achieve repressive results similar to EZH2 via additional histone modifications [65–67]. It is interesting to note that when comparing genes after immunoprecipitations by the two antibodies (JARID2 and EZH2) in the two PF (endo or control PF) treated cells (Table 1B), the endo PF treated cells showed a trend of having a fold-change less than 1 when comparing JARID2 vs. EZH2 IP. This suggests that EZH2 in endo PF treated cells is having more of an effect on the expression of all genes in the array (PRC2 complex core, alternate and binding partners) when compared to JARID2 further supporting a role for EZH2 in endometriosis.

One gene that should be noted and that was shown to have higher fold-change post EZH2 IP in endo PF treated cells vs. control PF treated cells was PHF19. PHF19 is a gene silencer and co-factor that can bind H3K36me3, which allows it to act as a recruiter for the PCR2 complex [68,69]. PHF19 also promotes tumorigenesis by the enhancement of the deposition of H3K27me3 and when PHF19 is depleted, this led to a loss of H3K27me3 domains [70]. This suggests that another mechanism which may be at play in transcriptional repression involves PHF19. PHF19 has also been deemed to play a role in the switch from proliferative to invasive states in melanoma cells [71]. Thus there are studies suggesting targeting PHF19 as an alternate strategy to inhibit EZH2 [72]. This study found that *PHF19* mRNA expression was significantly upregulated in cells treated with endo PF (Figure 8). This may suggest that miR-155 and PHF19 may be working together to bring the PRC2 complex to its targets in endometriosis. Putting these results together with miR-155 transfection studies, this study suggests that while miR-155 and PHF19 may be the main helper in regulating the PRC2 complex in endometriosis, JARID2 may be taking up the slack when miR-155 is inhibited.

The findings presented here, as summarized in Figure 9, provide potential mechanisms that may be at play in endometriosis patients. This study shows that in the presence of endometrial PF, all the components of the PRC2 complex, along with JARID2, FOXP3 and miR-155 are increased in expression when compared to control PF. Gain-or loss-of function of miR-155 showed an effect on PRC2 complex proteins but not on JARID2 levels. This suggested that other epigenetic regulators may be involved. ChiP-qPCR pull-down studies using JARID2 or EZH2 antibodies in PF treated cells showed alterations in epigenetic proteins associated with either of these complexes. In addition to the known binding partners such as EZH1, DNMT3B etc, the expression of PHF19 (a PRC2 complex co-factor) was highly upregulated in EZH2 compared to JARID2 pull-down assay. This finding, in addition to what is known in the literature [54,69,70,72], it is presumable that in women

with endometriosis, FOXP3/miR-155, in conjunction with PHF19, co-localizes with the PRC2 complex to promote its interaction and function with its targets. This leads to the increased H3K27me3 deposition thus modulating gene transcription. In contrast, this complex has a reverse effect on JARID2, thus preventing its association with the PRC2 complex, unless miR-155 is altered. This novel crosstalk among key epigenetic regulators leads to an increase in inflammation and growth of endometriotic lesions. This opens the door for testing newer targets in addition to the EZH2 inhibitors and miRNA mimics/antagonists currently being tested in endometriosis. For example, although histone demethylase inhibitors are thought to be ineffective against JARID2 due to its lack of true demethylase activity, additional investigations into the role of JARID2 in endometriosis could uncover alternate options to therapeutically regulate it, such as dihydroartemisinin which has been used in prostate cancer [73]. Additionally, the role for PHF19 as the master-regulator of the miR-155-PRC2 complex-JARID2 crosstalk is also a viable candidate for therapy and should be further explored in endometriosis. *Int. J. Mol. Sci.* **2021**, *22*, 3492 13 of 19 door for testing newer targets in addition to the EZH2 inhibitors and miRNA mimics/antagonists currently being tested in endometriosis. For example, although histone demethylase inhibitors are thought to be ineffective against JARID2 due to its lack of true demethylase activity, additional investigations into the role of JARID2 in endometriosis could uncover alternate options to therapeutically regulate it, such as dihydroartemisinin which has been used in prostate cancer [73]. Additionally, the role for PHF19 as the master-regulator of the miR-155-PRC2 complex-JARID2 crosstalk is also a viable candidate for therapy and should be further explored in endometriosis.

**Figure 9.** Proposed schematic of the epigenetic crosstalk playing a role in endometriosis. Mechanism proposed in normal women vs. women with endometriosis. Arrows indicate activation or general **Figure 9.** Proposed schematic of the epigenetic crosstalk playing a role in endometriosis. Mechanism proposed in normal women vs. women with endometriosis. Arrows indicate activation or general targeting while "T" bars indicate inhibition.

#### targeting while "T" bars indicate inhibition. **4. Materials and Methods**
