Causal association of cathepsins and endometriosis: A Mendelian randomization study

On the basis of our study, Cathepsin H increases the risk of overall endometriosis, endometriosis of ovary, endometriosis of pelvic peritoneum, and deep endometriosis. Overall endometriosis may lead to increased Cathepsin H levels.

Endometriosis, a complex and debilitating gynecological condition, is characterized by a range of distressing symptoms, including dysmenorrhea, chronic pelvic pain, and dyspareunia. [ 1 , 2 ] Globally, approximately 70 million women of reproductive age are affected by various forms of endometriosis. [ 3 ] Interestingly, endometriosis shares certain characteristics with malignant neoplasms, such as the ability to invade tissues locally and metastasize to distant sites. [ 4 ] The process of local invasion relies on enzymes that break down the extracellular matrix of surrounding connective tissue. [ 5 ] The activity of the proteolytic system plays a crucial role in maintaining local homeostasis in both health and disease. However, the specific role of this group of enzymes in the development of endometriosis remains largely unknown. It is possible that the activity and altered expression of these enzymes contribute to the clinical profile of endometriosis. [ 6 ] Cathepsins, which encompass a group of lysosomal proteolytic enzymes, [ 7 ] are prominently found in humans and are part of the papain superfamily of cysteine proteases. [ 8 ] These enzymes play crucial roles in numerous physiological and pathophysiological cellular processes. They are involved in protein and lipid metabolism, autophagy, antigen presentation, recycling of growth factor receptors, cellular stress signaling, degradation of the extracellular matrix, and lysosome-mediated cell death. [ 9 ] Given their involvement in these essential processes, various cathepsins have been implicated in the development of different diseases, including endometriosis. [ 10 ] Recent studies have shed light on the roles of several cathepsins in endometriosis. These include Cathepsin B, [ 11 – 13 ] Cathepsin D, [ 12 , 14 ] Cathepsin G, [ 12 , 15 ] Cathepsin K, [ 16 ] and Cathepsin L. [ 16 ] Emerging evidence reveals that cathepsins (particularly types B, D, L, S, and E) contribute to endometriosis through a multifaceted pathogenic network. These proteolytic enzymes facilitate ectopic lesion establishment by degrading extracellular matrix components [ 17 ] while simultaneously activating vascular endothelial growth factor to promote angiogenesis. [ 18 ] Beyond tissue invasion, cathepsins drive disease progression through pain modulation – with Cathepsin S activating PAR-2 in sensory neurons to induce neurogenic inflammation [ 19 ] and Cathepsin B amplifying TNF-α/IL-1β signaling to exacerbate pelvic pain. [ 20 ] The enzymes also promote fibrosis via TGF-β1-mediated myofibroblast differentiation [ 21 ] and disrupt immune surveillance by impairing NK cell cytotoxicity [ 22 ] and macrophage phagocytosis. [ 23 ] Notably, cathepsins participate in hormonal dysregulation, as progesterone resistance upregulates Cathepsin L which further degrades progesterone receptors, [ 24 ] creating a vicious cycle. These findings have therapeutic implications, with Cathepsin B inhibitors like CA-074 showing efficacy in primate models. [ 25 ] The cumulative evidence positions cathepsins as central players in endometriosis pathogenesis and promising therapeutic targets. However, there is still a lack of sufficient research exploring the causal relationship between specific types of cathepsins and the risk of different histological subtypes of endometriosis. Therefore, further investigation is necessary to better understand the causal associations between different cathepsins and the risk of specific endometriosis subtypes. With the advancements in genomics, there is growing evidence highlighting the role of heritability in the etiology of diseases. [ 26 ] Mendelian randomization (MR) is an analytical method used in epidemiological studies to infer disease etiology based on the Mendel laws. For MR to be considered valid, it is essential for the causal sequence to be plausible and supported by evidence. [ 27 ] Previous observational studies have identified associations between cathepsin traits and endometriosis, suggesting a potential correlation between them. In this specific study, a comprehensive 2-sample MR analysis was conducted to establish a causal link between specific cathepsin signatures and endometriosis. By utilizing this approach, the researchers aimed to provide stronger evidence for the causal relationship between cathepsins and the development of endometriosis.

Data curation: Na Aru, Yuntian Chen. Funding acquisition: Jiaming Liu. Investigation: Na Aru, Congyu Yang. Methodology: Congyu Yang. Software: Congyu Yang, Yuntian Chen. Supervision: Jiaming Liu. Writing – original draft: Na Aru, Yuntian Chen. Writing – review & editing: Jiaming Liu.

In our study, we conducted a 2-sample MR analysis to evaluate the causal relationship between cathepsins and endometriosis. To ensure reliable results, 3 key assumptions must be met when performing MR analysis: a strong association between genetic variants and exposure factors, no correlation between genetic variants and confounding variables, and the influence of genetic variants on the outcome is solely mediated through the exposure factors, excluding other pathways. All datasets were derived from the European population. The Genome-wide association study (GWAS) statistics for endometriosis were sourced from FinnGen Research data release (gs://finngen-public-data-r9/summary_stats/finngen_R9_N14_ENDOMETRIOSIS.gz). The GWAS statistics encompassed 15,088 cases and 107,564 controls. Summary statistics for endometriosis of ovary were obtained from a comprehensive GWAS study conducted by FinnGen Research (gs://finngen-public-data-r9/summary_stats/finngen_R9_N14_ENDOMETRIOSIS_OVARY.gz), involving a cohort of 5867 cases and 107,564 controls. Summary statistics of endometriosis of pelvic peritoneum was derived from a GWAS encompassing a cohort of 5628 cases and 107,564 controls of European ancestry (gs://finngen-public-data-r9/summary_stats/finngen_R9_N14_ENDOMETRIOSIS_PELVICPERITONEUM.gz). Lastly, GWAS data for deep endometriosis was extracted from FinnGen Research (gs://finngen-public-data-r9/summary_stats/finngen_R9_N14_ENDOMETRIOSIS_DEEP.gz), encompassing 2856 cases and 203,296 controls of European ancestry. Furthermore, to ensure result stability, the instrumental variables (IVs) for endometriosis were obtained through the aforementioned methodology and subjected to repeated analysis using summary data from another GWAS study involving 21,059 individuals of Asian descent. [ 28 ] Genetic instruments for evaluating the levels of different cathepsins (μg/L) were obtained from the INTERVAL study, which involved 3301 individuals of European descent. [ 29 ] The summary data can be accessed at https://gwas.mrcieu.ac.uk . In 2-sample MR, rigorous quality control is essential to ensure valid causal inference. Key steps include: single nucleotide polymorphism (SNP) selection ( P   10 to avoid weak instruments) [ 30 ] ; harmonization of effect alleles between exposure and outcome datasets, excluding palindromic SNPs with intermediate allele frequencies (0.42 < EAF < 0.58) [ 31 ] ; linkage disequilibrium (LD) pruning ( r ² < 0.001) to ensure independence of instruments. [ 32 ] For missing data, SNPs lacking summary statistics (β, SE, or P -value) in either dataset are excluded. Allele mismatches are resolved by strand flipping or removal, while ambiguous SNPs are filtered. Horizontal pleiotropy is assessed via MR-Egger intercept and MR-PRESSO (outlier removal). Sensitivity analyses (weighted median, leave-one-out) validate robustness. Confounders were detected using PhenoScanner, SNPs associated with both exposure and outcomes were removed from the study. [ 33 ] The GWAS data on exposures and outcomes used in this paper are from different studies, so there is no population overlap. For our 2-sample MR analysis, SNPs were selected as IVs based on stringent criteria to satisfy the 3 key MR assumptions. Inclusion criteria required: genome-wide association with cathepsin levels ( P   10); low LD ( r ² < 0.001); and successful harmonization between exposure (cathepsins) and outcome (endometriosis) datasets to ensure allele alignment. Exclusion criteria removed: palindromic SNPs with intermediate allele frequencies (EAF > 0.42); SNPs associated with confounders (e.g., BMI, inflammation) via PhenoScanner; and outliers detected by MR-PRESSO (global test P  < .05). Sensitivity analyses (MR-Egger intercept, leave-one-out) confirmed robustness against pleiotropy. These steps ensured IVs were robust, unconfounded, and exclusively linked to endometriosis risk through cathepsin pathways. The MR analysis in this study utilized the Two-sample MR package of R software. We employed several MR approaches, namely the inverse-variance weighted (IVW) method, MR-Egger, MR-PRESSO, maximum likelihood, and weighted median. The main finding of our study was to estimate the association between cathepsins and endometriosis using the traditional IVW method. [ 34 ] However, to ensure the validity and robustness of the IVW results, we tested their effectiveness using other methods, such as MR-Egger, MR-PRESSO and weighted median. In order to minimize the impact of horizontal pleiotropy on the results (which occurs when a genetic variant influences multiple traits), we employed the MR-PRESSO method to detect and remove any outliers associated with extensive horizontal pleiotropy in all analyses. The outliers identified by MR-PRESSO were subsequently discarded, and the MR analyses were re-conducted. Furthermore, for the statistically significant results, we performed heterogeneity and pleiotropy tests, including the MR-Egger intercept test and modified Cochran Q statistics, to evaluate the robustness of the findings. A P -value >0.05 in heterogeneity and pleiotropy analyses would indicate that our study is unbiased. [ 35 ] To further assess the significance of the MR results, we also conducted funnel plots and leave-one-out analyses. Symmetrical funnel plots would indicate the absence of outliers, while significant and similar estimates remaining after each SNP is removed would establish the robustness of the findings. In addition, we employed multivariable MR, an extension of standard univariable MR, to consider multiple cathepsins when analyzing their causal effects on different subtypes of endometriosis. This approach also allowed us to estimate the direct causal effects of each exposure in a single analysis using the “MendelianRandomization” package. [ 36 ] Reverse MR analyses, treating endometriosis as the exposure and cathepsins as the outcomes, were performed to evaluate reverse causality and explore bidirectional causality. To assess the potential impact of directional pleiotropy, we examined each SNP for potential associations with secondary phenotypes using the GWAS Catalog ( https://www.phenoscanner.medschl.cam.ac.uk/ ). Subsequently, we repeated the MR analyses, excluding SNPs associated with other phenotypes. Furthermore, we conducted leave-one-out sensitivity analyses on significant findings to determine if a single SNP was responsible for the observed causal relationship. The overall research design is illustrated in Figure 1 . The MR analyses were conducted using the “TwoSampleMR” package (version 0.5.7) in the R software environment. [ 35 , 37 ] Research overview and design of Mendelian randomization analysis.

The findings of the univariable MR analysis revealed that genetically predicted Cathepsin H were positively correlated with the risk of overall endometriosis (OR [95%]: 1.037[1.007 to 1.067], P  = .013) in the IVW method. Similar results were observed by using Weighted median (OR [95%]: 1.036 [1.005 to 1.068], P  = .022), and Weighted mode (OR [95%]: 1.037 [1.005 to 1.069], P  = .049). We noticed that Cathepsin H were suggestively associated with a higher risk of the rest of 3 histological subtypes of endometriosis in IVW method. OR [95%]: 1.022 [1.001 to 1.042], P  = .046 for endometriosis of ovary; OR [95%]: 1.046 [1.002 to 1.089], P  = .047 for endometriosis of pelvic peritoneum; OR [95%]: 1.050 [1.002 to 1.099], P  = .048 for deep endometriosis (Fig. 2 ). No other types of cathepsins exhibited significant association with the risk of overall endometriosis or its different histological subtypes (Table S1, Supplemental Digital Content, https://links.lww.com/MD/P28 and Fig. 3 ). Similar outcomes were obtained when repeating the analysis using an alternate GWAS dataset, as presented in Table S2, Supplemental Digital Content, https://links.lww.com/MD/P28 . The effect of Cathepsin H on overall endometriosis and its different histological subtypes by univariable Mendelian randomization analysis. nsnp = number of single nucleotide polymorphisms, OR = odds ratio, CI = confidence interval. The effect of cathepsins on endometriosis with IVW method by univariable Mendelian randomization analysis. nsnp = number of single nucleotide polymorphisms, OR = odds ratio, CI = confidence interval. Moreover, we conducted multivariable MR to assess the genetic predisposition involving cathepsins in relation to the risk of different histological subtypes of endometriosis. The results revealed that even after adjusting for other types of cathepsins, elevated Cathepsin H levels retained a robust association with an increased risk of overall endometriosis (IVW: OR [95%]: 1.034 [1.002 to 1.066], P  = .049), endometriosis of ovary (IVW: OR [95%]: 1.022 [1.005 to 1.039], P  = .047), endometriosis of pelvic peritoneum (IVW: OR [95%]: 1.051 [1.009 to 1.096], P  = .042), and deep endometriosis (IVW: OR [95%]: 1.099 [1.045 to 1.143], P  = .043). However, no statistically significant causal association was observed between other types of cathepsins and overall endometriosis or its different histological subtypes (Table S3, Supplemental Digital Content, https://links.lww.com/MD/P28 and Fig. 4 ). The effect of cathepsins on overall endometriosis and its different histological subtypes with IVW method by multivariable Mendelian randomization analysis. OR = odds ratio, CI = confidence interval. As detailed in Table S4, Supplemental Digital Content, https://links.lww.com/MD/P28 , we evaluated the potential associations of cathepsins and endometriosis using the reverse MR analyses. Overall endometriosis was suggestively associated with increased Cathepsin H (IVW: OR [95%]: 1.017 [1.003, 1.073], P  = .041). We did not find statistically significant associations between endometriosis and other types of cathepsins, the results were stable across sensitivity analyses, which are listed in Table S5, Supplemental Digital Content, https://links.lww.com/MD/P28 . Moreover, we conducted a repeated analysis using an alternative GWAS dataset, and the outcomes yielded similar conclusions (Table S6, Supplemental Digital Content, https://links.lww.com/MD/P28 ). And the MR-Egger intercept test and MR-PRESSO global test results indicated no evidence of heterogeneity or horizontal pleiotropy, the results are presented in Table S7, Supplemental Digital Content, https://links.lww.com/MD/P28 . The MR-Egger intercept test and MR-PRESSO global test results indicated no evidence of heterogeneity or horizontal pleiotropy in the associations between cathepsins and endometriosis, and the same is true of repeated analysis. These results are presented in Tables S8 and S9, Supplemental Digital Content, https://links.lww.com/MD/P28 . Furthermore, the leave-one-out analysis demonstrated the robustness of the MR results (Figs. S1–S5, Supplemental Digital Content, https://links.lww.com/MD/P29 ). Excluding any single SNP associated with cathepsins and endometriosis did not significantly alter the overall findings. To account for potential directional pleiotropy, we conducted an analysis using the GWAS Catalog to identify SNPs linked to cathepsins and endometriosis. Two SNPs were discovered to exhibit associations with other traits, as detailed in Tables S10, Supplemental Digital Content, https://links.lww.com/MD/P28 . After excluding these pleiotropic SNPs, the associations between cathepsins and endometriosis remained stable, as shown in Table S11, Supplemental Digital Content, https://links.lww.com/MD/P28 .

Our analysis presents suggestive evidence that cathepsins may affect the risk of endometriosis based on a comprehensive genetic approach utilizing large-scale GWAS summary data. To the best of our knowledge, this is the first MR analysis to examine the causal relationship between multiple cathepsins and endometriosis. By using SNPs as IVs and incorporating various 2-sample MR methods, we confirmed a significant association between Cathepsin H and the risk of endometriosis. Additionally, in the reverse MR analysis, we observed a suggestive association between endometriosis and increased levels of Cathepsin H. Endometriosis is a severe reproductive system disorder often referred to as “benign cancer.” Symptoms of endometriosis, such as dysmenorrhea, menorrhagia, and infertility, are likely a result of the abnormal implantation of endometrial tissues, leading to the formation of lesions on various sites including the peritoneal wall, uterus, and ovaries. [ 38 , 39 ] While current treatments can alleviate moderate pain symptoms, their effectiveness in preventing lesion formation is inconsistent and varies among patients. [ 40 ] The occurrence of retrograde menstruation, where endometrial fragments travel upward through the oviducts into the peritoneal cavity, is strongly associated with the development of the disease. However, this alone cannot explain the disorder since retrograde menstruation occurs in approximately 90% of women, [ 41 ] whereas only 6% to 10% of women are diagnosed with endometriosis. [ 42 ] Hence, we have valid grounds to propose the involvement of other vital physiological and pathological mechanisms in the genesis of endometriosis. Previous research has revealed that the infiltration of estrogen-dependent cells assumes a central role in the formation of endometriosis lesions. [ 43 ] Enzymes responsible for the degradation of extracellular matrix proteins, including collagen, fibronectin, and laminin, which constitute the peritoneal and uterine walls, may facilitate this invasive process. [ 44 , 45 ] Cathepsin H, influenced by estrogen, significantly promotes cellular invasion. Moreover, Bergqvist et al discovered considerably higher levels of cathepsin in endometriosis tissues compared to eutopic endometrium, suggesting its potential importance in implantation and invasive growth. Protopapas` study demonstrated a significant correlation between cathepsin expression and higher stages of the disease, indicating its role in the advanced phases of ovarian endometriosis development. [ 14 ] Cathepsin levels were found to increase further in advanced stages of endometriosis, thereby accelerating protease-induced degradation of the basement membrane and extracellular matrix. [ 46 ] Consequently, it is our belief that the continuous elevation of cathepsin in ectopic endometrium of endometriosis patients contributes to the implantation and development of ectopic endometrial tissue within the abdominal cavity or any other region of the human body. Cathepsin H (CatH, EC 3.4.22.16) is a lysosomal cysteine protease that was initially discovered in rat liver lysosomes in 1976 and later isolated from human liver. [ 47 , 48 ] Cysteine cathepsins are involved in regulating growth, migration, angiogenesis, and metastasis. [ 49 – 51 ] In the context of endometrium, Cathepsin H expression is higher in the proliferative phase compared to the secretory phase, suggesting a potential link to estrogenic control, either directly or indirectly. [ 52 ] Elevated levels of Cathepsin H have also been detected in the serum and tissue extracts of breast cancer patients, which is also tightly associated with estrogen. [ 53 , 54 ] Cathepsin H plays multiple roles, including intracellular and extracellular proteolytic functions, as well as promoting cell proliferation and migration. It facilitates cell invasion through degradation of the extracellular matrix, both extracellularly and intracellularly. [ 55 , 56 ] However, its relatively weak endopeptidase activity suggests that Cathepsin H may not be a primary contributor to direct extracellular matrix degradation. While the exact role of Cathepsin H in cell progression remains unclear, it has been identified in early endosomes and implicated in talin processing within focal adhesions of migrating PC-3 cells, suggesting an intracellular proteolytic role in regulating cell migration. [ 57 ] Additionally, upregulation of extracellular Cathepsin H mediated by T3 has been shown to activate MMPs or extracellular signal-regulated kinases, promoting cell migration and invasion. [ 58 ] Estrogen-dependent cell invasion is central to endometriosis lesion establishment, and we believe that this might be the possible mechanism of Cathepsin H promoting endometriosis. The genetic associations identified between Cathepsin H and various endometriosis subtypes (ORs ranging from 1.022 for ovarian endometriosis to 1.050 for deep infiltrating disease) provide compelling evidence for its involvement in disease pathogenesis through multiple interconnected biological mechanisms. As a member of the cysteine cathepsin family, Cathepsin H exhibits unique dual exopeptidase and endopeptidase activity that enables its participation in critical pathogenic processes. [ 59 ] Its ability to degrade key extracellular matrix components, including laminin and type IV collagen, [ 60 , 61 ] facilitates the initial attachment and subsequent invasion of ectopic endometrial tissue. This proteolytic function is particularly relevant in deep infiltrating endometriosis, where the observed stronger association may reflect the enzyme’s role in tissue penetration beyond the peritoneal surface. The estrogen-responsive nature of Cathepsin H expression provides a crucial link to endometriosis pathophysiology. [ 62 , 63 ] Estrogen not only upregulates Cathepsin H production in endometrial tissue but also creates a positive feedback loop, as established lesions produce local estrogen that further stimulates protease expression. [ 17 ] This hormonal regulation helps explain the enzyme’s involvement across all disease subtypes. Additionally, emerging research highlights Cathepsin H immunomodulatory functions, including its ability to promote M2 macrophage polarization and contribute to the immune-tolerant microenvironment characteristic of endometriosis lesions. [ 64 , 65 ] While the modest effect sizes suggest Cathepsin H operates as part of a broader proteolytic network rather than as a sole pathogenic driver, [ 17 ] its strategic position in protease activation cascades and hormonal pathways makes it an attractive therapeutic target. The consistent associations across subtypes, supported by tissue-based studies showing elevated Cathepsin H in ectopic lesions, [ 66 ] underscore its potential as both a disease biomarker and a target for combination therapies that simultaneously inhibit multiple proteolytic pathways. [ 67 , 68 ] Future research should explore tissue-specific expression patterns and evaluate Cathepsin H inhibitors in preclinical models, particularly in combination with other targeted agents. [ 66 , 69 ] The genetic association between Cathepsin H and endometriosis offers clinically actionable insights for advancing treatment strategies. While current hormonal therapies remain first-line, the role of Cathepsin H in extracellular matrix degradation (particularly in deep infiltrating subtypes) and estrogen-mediated invasion suggests protease inhibition could benefit refractory cases. [ 67 ] Cathepsin H could be targeted as part of combination therapy, such as with Matrix Metalloproteinases (MMP inhibitors) to block invasion or anti-angiogenics to counteract vascular endothelial growth factor activation. [ 61 ] Patient stratification may be key – elevated Cathepsin H levels could identify candidates for protease-directed therapies. [ 70 ] Challenges include optimizing selective inhibitors to avoid disrupting related cathepsins (e.g., CatB/L) [ 7 ] and managing hormonal crosstalk, as Cathepsin H estrogen responsiveness may necessitate concurrent aromatase inhibition. [ 62 ] Preclinical studies in endometriosis models [ 71 ] and biomarker-driven pilot trials [ 72 ] are critical next steps to evaluate Cathepsin H therapeutic potential while addressing safety concerns. These findings exemplify how genetic insights can transition endometriosis management toward precision approaches, complementing existing hormonal strategies with pathway-specific interventions. Our findings provide evidence supporting a potential causal relationship between Cathepsin H and endometriosis. This discovery has implications for guiding prognosis and treatment decisions, as well as for the development of new drugs. However, it is important to acknowledge that endometriosis is a complex condition with multifaceted causes, and there are variations in clinical presentation among different histological subtypes. Therefore, a personalized treatment approach may be more effective than a one-size-fits-all approach. Further research is needed to delve into the intricate interplay between Cathepsin H and endometriosis, which will contribute to a deeper understanding of the condition and potentially lead to more targeted and efficient therapeutic interventions. It is essential to acknowledge the inherent limitations of our study, which should not be overlooked. Firstly, it is important to highlight that MR analysis, used in our study, cannot replace the need for clinical trials in establishing causality between exposure and outcome. MR analysis serves as a valuable tool for assessing potential causal relationships, but further investigations, including experimental studies and randomized controlled trials, are required to validate the association between Cathepsin H and the risk of endometriosis. Furthermore, it is crucial to recognize that our MR analysis was conducted exclusively within the European population due to limited availability of genetic data resources. Genetic heterogeneity exists among different ethnic groups, and therefore, it is possible that the results may vary in other populations. Future studies should consider conducting subgroup analyses that include diverse populations to obtain a more comprehensive and generalizable conclusion. Our results need to be verified by further clinical and basic research, exploring the mechanisms by which Cathepsin H influences endometriosis risk and evaluating potential therapeutic targets. To build upon these genetic findings, future studies should prioritize experimental validation of Cathepsin H mechanistic role in endometriosis progression. In vitro, models using patient-derived endometrial stromal cells could assess how Cathepsin H knockdown or inhibition affects invasion, angiogenesis, and cytokine secretion. In vivo, studies in conditional Cathepsin H knockout mouse models of endometriosis would clarify its tissue-specific contributions to lesion establishment and pain sensitization. Additionally, spatial transcriptomics of human endometriosis lesions could map Cathepsin H expression patterns across disease subtypes and microenvironments. Given the modest effect sizes, systems biology approaches may identify synergistic pathways (e.g., MMPs, TGF-β) for combination targeting. Lastly, prospective cohort studies measuring serum Cathepsin H levels before/after treatment could evaluate its utility as a predictive biomarker for therapeutic response. These efforts would bridge genetic associations to actionable biological insights and clinical applications. In conclusion, while our findings offer valuable insights into the potential association between Cathepsin H and the risk of endometriosis using MR analysis, it is important to interpret the results in light of these limitations. Continued research efforts, involving different study designs and populations, are necessary to confirm and expand upon our findings.