The effect of circulating proteins and their role in mediating adiposity’s effect on endometrioid and non-endometrioid endometrial cancer risk: Mendelian randomisation and colocalization analyses 

The effect of circulating proteins and their role in mediating adiposity’s effect on endometrioid and non-endometrioid endometrial cancer risk: Mendelian randomisation and colocalization analyses | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The effect of circulating proteins and their role in mediating adiposity’s effect on endometrioid and non-endometrioid endometrial cancer risk: Mendelian randomisation and colocalization analyses Sabrina E Wang, Vanessa Y Tan, James Yarmolinsky, Yadi Zheng, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5815826/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Proteomics could enhance our understanding of endometrial carcinogenesis. However, addressing confounding in traditional observational studies remains challenging, especially given the strong impact of adiposity on the plasma proteome and endometrial cancer risk. The role of circulating proteins in mediating adiposity’s effect on endometrial cancer risk is also not fully elucidated. Methods: Using Mendelian randomization (MR) and colocalization analyses, we examined the causal association between 2,751 unique proteins from UK Biobank (N Olink proteins=2,031; N=52,363) and deCODE (N SomaScan proteins=1,667; N=35,559) and endometrial cancer risk [overall (N cases=12,270; N controls=46,126), endometrioid (N cases=8,758), and non-endometrioid (N cases=1,230) in the meta-analysed Endometrial Cancer Association Consortium and Epidemiology of Endometrial Cancer Consortium data]. We performed enrichment analyses to explore pathways overrepresented among plasma proteins in endometrioid and non-endometrioid cancer subtypes. Additionally, we assessed the role of circulating proteins in mediating the effect of body mass index (BMI) on endometrial cancer risk using univariable and multivariable MR. Results: We identified 20 associations between circulating proteins and endometrial cancer risk in MR and colocalization analyses. GSTO1-1 and SKAP1 were positively associated and MMP10 was negatively associated with both overall and endometrioid endometrial cancer; DTYMK and ABO were positively associated and TSSC4 was negatively associated with overall endometrial cancer; IGF2R was positively associated with endometrioid cancer; MAPK9 was positively associated and DNAJB14, IFI16, LCN2, and SCT were negatively associated with non-endometrioid endometrial cancer. Distinct pathways were overrepresented in endometrioid (e.g., PDGF signalling and PTEN gene regulation) and non-endometrioid (e.g., non-canonical NF-kB signalling) cancer subtypes. GSTO1-1 and IGF2R were identified as potential mediators for the effect of BMI on endometrioid cancer risk in univariable MR, but evidence for mediation was not observed in multivariable MR analyses. Conclusions: We observed distinct plasma proteins and pathways associated with endometrioid and non-endometrioid endometrial cancers. These findings highlight candidate proteins for further mechanistic investigations, which could support the development of non-invasive methods to differentiate endometrial cancer subtypes and guide clinical intervention strategies. There was limited evidence that the effect of adiposity on endometrial cancer risk was mediated by circulating proteins examined in our study. Endometrial cancer plasma proteomics adiposity Mendelian randomisation colocalization Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1 Background Endometrial cancer incidence has been rising in successive generations across continents over the past few decades. 1 Adiposity is strongly associated with endometrial cancer risk, 2 with 34% of global endometrial cancer cases attributable to increased BMI. 3 The endometrioid subtype is more common and more consistently associated with adiposity, whereas non-endometrioid endometrial cancer, including serous carcinomas, clear cell carcinomas, and carcinosarcomas, generally have poorer prognosis and less established risk factors. 4 Plasma proteomics is a non-invasive method to better understand potential pathogenic pathways associated with different endometrial cancer subtypes, which may be useful in aiding diagnosis, identifying targets for clinical interventions, and exploring the mechanisms behind known risk factors. Traditional observational studies have reported associations between circulating biomarkers and endometrial cancer risk, including inverse associations with sex hormone binding globulin (SHBG) 5 and insulin-like growth factor 1 (IGF1) 6 and positive associations with insulin and inflammatory cytokines such as interleukin (IL-6) and tumour necrosis factor-α (TNF-α). 7 – 10 Addressing residual and unmeasured confounding remains a challenge for traditional observational studies, particularly given the strong effect of adiposity on the circulating proteome and endometrial cancer risk. 2 , 11 Mendelian randomisation (MR) and genetic colocalization are complimentary statistical methods that can be applied to genome-wide association study (GWAS) summary data to assess causal associations between traits of interest. 12 MR analyses leverage genetic variants as instrumental variables to infer causality. 13 Given a sufficiently large population, MR is akin to the randomisation process in randomised controlled trials and theoretically minimises risk of confounding. Multivariable MR (MVMR) can estimate the effect of multiple exposures on an outcome, allowing mediation analysis in the MR framework. 14 Genetic colocalization estimates the probability that two or more traits are associated with the same genetic variant(s), thereby sharing a common genetic cause. 12 Genetic colocalization analysis cannot distinguish between exposure and outcome traits, but it does not rely on instrumental variable assumptions required in MR analyses. 12 Consistent evidence from both MR and colocalization analysis can strengthen evidence for a causal effect. MR analyses of plasma proteins have reported causal associations with endometrial cancer risk, including reduced SHBG 15 and elevated insulin, 16 whereas no association was observed for IGF1 15 or a panel of 66 circulating inflammatory markers. 17 A study that examined molecular mediators for BMI’s effect on endometrial cancer risk reported potential mediating roles for SHBG, insulin, and bioavailable testosterone. 18 We aimed to investigate the effect of circulating proteins on endometrial cancer risk. To capture a broad spectrum of plasma proteins, we used GWAS summary data from two independent studies. 6 , 19 We examined 2,751 unique proteins (2,031 Olink proteins and 1,667 SomaScan aptamers) and their association with overall, endometrioid, and non-endometrioid endometrial cancer risk by performing univariable MR (UVMR) and colocalization analyses. We further examined the potential role of plasma proteins in mediating the effect of BMI on endometrial cancer risk using UVMR and MVMR. 2 Methods The main analysis involved four parts (Figure 1). We first examined the causal relationship between plasma proteins and endometrial cancer risk, by performing cis variant UVMR and colocalization (Part a). We then examined the causal relationship between adiposity (i.e., BMI) and endometrial cancer risk (Part b) and between adiposity and plasma proteins using UVMR (Part c). To exclude bi-directional effects, we performed both forward and reverse MR analyses for Part (a) – (c). For each potential protein mediator, we performed MVMR to explore effect mediation (Part d). 2.1 GWAS sources and study populations We used summary results from GWAS published by 1) Sun et al. (2023) 19 on up to 2,922 plasma proteins measured with Olink Explore 3072 platform in a sub-cohort of 52,363 UK-Biobank participants of European ancestries; 2) Ferkingstad et al. (2021) 6 on up to 4,907 aptamers for 4,719 plasma proteins measured with SomaScan multiplex aptamer assay (version 4) in 35,559 Icelanders from deCODE; 3) O’Mara et al. (2018) 20 on endometrial cancer (12,270 cases and 46,126 controls), endometrioid cancer subtype (8,758 cases and 46,126 controls), and non-endometrioid cancer subtype (1,230 cases and 35,447 controls of European ancestries) from the meta-analysed Endometrial Cancer Association Consortium and Epidemiology of Endometrial Cancer Consortium data excluding UK-Biobank participants; 4) Pulit et al. (2019) 21 on female-specific BMI, which includes 434,794 women of European ancestries from the GIANT consortium and UK-Biobank. 2.2 Genetic instruments To select a genetic instrument for each plasma protein, a 1Mb region was defined around each cis variant [i.e., a variant ≤1Mb from the transcription start site of the protein coding gene (discovery) or ≤1Mb from the gene encoding the measured protein (replication)] reaching the genome-wide significance threshold (P <1.8×10 − 9 in deCODE; P< 1.7×10 − 11 in UK-Biobank). Starting with the variant with the lowest P-value, any overlapping regions were merged until no overlapping regions remained. Linkage disequilibrium (LD) based clumping was then used to merge regions for variants in high LD (r 2 ≥0.8). The variant with the lowest P-value after merging was considered the sentinel variant and was used as the genetic instrument for the protein. We obtained the list of instruments for plasma proteins directly from the respective GWAS. 6,19 A total of 2,031 Olink proteins and 1,667 SomaScan aptamers (1,447 proteins) were instrumented with a cis variant, representing 2,751 unique proteins, with 727 proteins measured by both platforms. Genetic instruments for BMI were selected by extracting variants that reach a genome-wide significance threshold of P <5×10 -9 (to account for a wider coverage of data sequenced 22 ) and clumping at a LD independence threshold of r 2 <0.001 using the 1000 Genomes Project phase 3 European population. Genetic instruments for liability to endometrial cancers were selected by searching variants that reach a genome-wide significance threshold of P <5×10 -8 (P <5×10 -7 for non-endometrioid cancer as no variant reached the threshold) and clumping at a LD independence threshold of r 2 10 considered as less likely to suffer from weak instrument bias. Data on BMI, proteins, and aptamers were inverse rank normally transformed in their respective GWAS. Assuming the distribution of each trait was normal prior to transformation, the measures of association obtained from MR analyses could be approximately interpreted as the change in per normalized standard deviation unit change in the exposure. 2.3 Statistical analysis 2.3.1 Univariable Mendelian randomization UVMR analysis was performed to examine protein-cancer, adiposity-cancer, adiposity-protein (forward MR analyses), cancer-protein, cancer-adiposity, and protein-adiposity (reverse MR analyses) associations. For the primary analysis, we estimated causal associations using Wald ratio (where single cis variants were used as instruments for specific proteins) or inverse variance weighted multiplicative random effects (IVW-MRE) model (where ≥2 variants were used as instruments for adiposity or cancer). The IVW-MRE model assumes that pleiotropic effects of individual instruments sum to zero and that pleiotropic effects are independent of instrument strength across all variants. 23,24 As sensitivity analyses for the IVW-MRE model, we used MR-Egger, weighted-median, and weighted-mode models. MR-Egger can help in detection and adjustment for directional horizontal pleiotropy by providing an intercept term of the linear regression for the measured association. 25 The weighted-median can provide consistent estimates when at least 50% of the instruments are valid. 25 The weighted-mode can provide consistent estimates as long as the largest cluster of similar estimates are derived from valid instruments. 26 In addition, we performed leave-one-out and single-variant MR analyses to assess the influence of individual variants on observed associations. To correct for multiple testing for analyses involving plasma proteins, we used PhenoSpD to estimate the phenotypic correlations and the number of independent tests. 27 A total of 505 SomaScan and 1,026 Olink independent proteins were identified, corresponding to a PhenoSpD corrected P-value of 1.02×10 -4 and 5.00×10 -5 respectively. We considered a causal association from MR analysis as one with 1) a P-value below the multiple comparison corrected threshold from the primary analysis (Wald ratio or IVW-MRE); 2) consistent direction of association across all sensitivity analysis models (MR-Egger, weighted-median, weighted-mode) if applicable; 3) no consistent association observed in the reverse MR analysis. Furthermore, a plasma protein was considered a potential mediator for the effect of adiposity on endometrial cancer risk if there was consistent direction of effect from adiposity to protein and from protein to cancer risk. 2.3.2 Colocalization We performed colocalization for plasma protein and endometrial cancer risk assuming a single causal variant. 28 The cis variant instrument used in MR for each plasma protein was extracted along with a 1 Mb window around the variant from the protein and the endometrial cancer GWAS. We assigned a prior of 1×10 -6 for p1 and p2, and 1×10 -7 for p12 using https://chr1swallace.shinyapps.io/coloc-priors/ (accessed: 07/03/2024), based on an average of approximately 3,000 variants available for colocalization within the 1mb window. 29 We used P(h 4 ) ≥0.8 as evidence for colocalization. 2.3.3 Multivariable Mendelian randomization We performed MVMR analysis to estimate the natural direct effect of adiposity on endometrial cancer risk conditional on each potential plasma protein mediator. 14 To construct instruments for adiposity conditioned on a plasma protein mediator, we searched the gene region (±1Mb) for variants from the plasma protein GWAS that reach P <5×10 -8 . We clumped at a relaxed LD independence threshold of r 2 <0.1 as an attempt to improve instrument strength (a separate sensitivity analysis using LD independence threshold of r 2 <0.001 was also performed). We then extracted these variants from the BMI GWAS, combined them with instruments for adiposity, and clumped to remove variants in LD. The effect of adiposity on endometrial cancer risk conditioned on each plasma protein was estimated using an inverse variance weighted model. We compared this direct effect with the total effect of adiposity on endometrial cancer risk estimated from analysis Part (b). We did not calculate the proportion of total effect mediated by each protein as, when using the difference method (i.e., comparing effects estimates from regression model of adiposity on endometrial cancer without and with the potential mediator) with binary outcomes, the indirect effect could be underestimated due to non-collapsibility property of odds ratios (OR). 14,30 2.3.4 Pathway enrichment We performed pathway enrichment analysis for proteins associated with endometrioid or non-endometrioid endometrial cancer at P <0.05 in the UVMR analysis. We used an active-subnetwork-oriented approach, which maps each gene onto a protein-protein interaction network to better account for gene interaction information. 31 Active subnetworks were identified using a greedy algorithm, which were then included in enrichment analysis. 31 The process was repeated for ten iterations to cross validate the greedy algorithm selections. We used the Biogrid protein-protein interaction repository and the pathway-based Reactome gene sets database. Hierarchical clustering was performed on identified pathways to select representatives (i.e., pathways with the lowest multiple testing adjusted P-value) among groups of related pathways. 31 2.3.5 Expression quantitative trait loci To explore whether proteins associated with endometrial cancer risk might reflect gene expression and biological processes in tissues, we examined whether the cis protein quantitative trait loci (pQTL) were overlapping or in high LD (r 2 > 0.8) with cis expression quantitative trait loci (eQTL). We used cis -eQTL data from 49 non-diseased tissue sites in the Genotype-Tissue Expression Project (GTEx) Release V8. A cis eQTL was defined as a variant within the gene region with P <5×10 -8 . We calculated the LD using the 1000 Genomes Project phase 3 European population. We further used this analysis to investigate possible epitope or aptamer binding effect, given both Olink and SomaScan rely on the binding of antibody or aptamer to proteins for quantification. If a variant also influences the expression of the gene encoding the protein, then it is considered less likely that the observed association is due to epitope or aptamer binding effect. 6,17 3 Results Of the 2,031 Olink proteins and 1,667 SomaScan aptamers (1,447 proteins) that had an available cis variant instrument, we extracted 1,648 and 901 variants or proxies respectively from the endometrial cancers GWAS and examined for protein-cancer association in UVMR and colocalization (Figure 2). All 2,031 Olink proteins and 1,667 SomaScan aptamers were examined for BMI-protein association. Details of genetic variants used to instrument adiposity measures, plasma proteins, and endometrial cancers are presented in Supplementary Table 1-4. F statistics indicated strong instruments for BMI (mean=67.7), endometrial cancer (44.2), endometrioid cancer (43.0), and non-endometrioid cancer subtype (26.9). F statistics for plasma proteins instrumented using a single cis variant ranged from 36.2 to 117,180, with a mean of 3,202. 3.1 Protein-endometrial cancer There were 20 protein-endometrial cancer associations between 12 plasma proteins and the three endometrial cancer outcomes with evidence of colocalization [P(h 4 ) > 0.8], of which 5 pairs reached the multiple testing corrected P-value threshold for UVMR (Figure 3a, Table 1). GSTO1-1 and SKAP1 were positively associated and MMP10 was negatively associated with both overall and endometrioid endometrial cancer; DTYMK and ABO were positively associated and TSSC4 was negatively associated with overall endometrial cancer; IGF2R was positively associated with endometrioid endometrial cancer; MAPK9 was positively associated and DNAJB14, IFI16, LCN2, and SCT were negatively associated with non-endometrioid endometrial cancer. Of those associated with endometrial cancers risk, four proteins (MMP10, IGF2R, MAPK9, and DNAJB14) were analysed in both Olink and SomaScan platforms with highly consistent results. LCN2 was also analysed in both platforms, with UVMR results showing consistent inverse association with non-endometrioid endometrial cancer risk but only the Olink assay showed evidence for colocalization [Olink P(h4)=0.84; SomaScan P(h4)=0.43]. The remaining proteins associated with endometrial cancers were examined only in one of the two platforms. Full results from the protein-cancer MR and the cancer-protein reverse MR analyses are available in Supplementary Table 5-6. Table 1. Summary of proteins causally associated with endometrial cancer risk, their functions and possible link to endometrial cancer development. Outcome Protein BMI-protein 2 Protein-cancer 2 Functions and possible link to cancer Endometrial, Endometrioid GSTO1 + + Mediates pro-inflammatory signalling through TLR4 and NLRP3 inflammasome, leading to consequent release of TNF-α, IL-6 and IL-1β. 32,33 SKAP1 + 3 + Regulates T-cell signalling and enhances adhesion of T-cells with antigen-presenting cells. SKAP1 expression in whole blood identified as a candidate susceptibility gene for endometrioid cancer. 34 MMP10 1 ∆/− 4 − Initiates extracellular matrix remodelling during physiological (e.g., menstrual cycle) and pathological (e.g., metastasis) processes. It is specifically expressed in the endometrium and is thought to be down regulated by progesterone. 35,36 Endometrial DTYMK + 4 + Catalyses the conversion of dTMP to dTDP, essential for DNA synthesis and repair. Found to be upregulated in several tumours, including liver, lung, breast, and colon tumour cells. 37 TSSC4 + 4 − Tumour suppressing sub-transferable candidate 4 located in the imprinted gene domain 11p15.5. Alteration in this region have been associated with several cancers. 38 ABO ∆ 4 + ABO blood group. Expression of A or B antigen has been linked with increased risk of several cancers, including endometrial cancer. 39 Endometrioid IGF2R 1 + + Soluble IGF2R transports and inhibits IGF2 mediated growth. It also has binding site for M6P and various ligands. Elevated plasma IGF2R has been observed in individuals with obesity, type 2 diabetes, colorectal cancer, and breast cancer, but the underlying mechanism remains unclear. 40 Non-endometrioid MAPK9 1 ∆/+ 4 + Phosphorylates p53 protein to increase its stability under physiological states but can promote cancer cell survival through crosstalk with other pathways including NF-κB. 41 DNAJB14 1 + 4 − A molecular chaperone that supports proper protein folding and responsible for degradation of misfolded proteins. 42 IFI16 + − Modulation of p53 function and inhibition of cell growth via the Ras/Raf signalling pathway. 43 LCN2 + − Important for innate immunity against bacterial growth but can promote epithelial-to-mesenchymal transition to facilitate tumour invasion and metastasis. 44 It has been associated with aggressiveness of endometrial cancer. 45 SCT ∆ 4 − Stimulates the secretion of bile and bicarbonate. Overexpression has been observed in pancreatic tumours, 46 but no clear link with other tumours. 1 Causally associated with endometrial cancer risk in both SomaScan and Olink platforms 2 Direction of causal association: + =positively associated; − =inversely associated; ∆ =failed to demonstrate consistent direction across MR sensitivity models 3 Failed to rule out an association in the protein-BMI reverse MR analysis 4 P-value did not reach the multiple testing corrected threshold in the BMI-protein MR analysis 3.2 Adiposity-endometrial cancer BMI was associated with overall endometrial [OR per normalized standard deviation change in BMI =1.79; 95% confidence interval (CI) 1.55-2.01], endometrioid (1.81; 1.58-2.07), and non-endometrioid cancer (1.63; 1.23-2.17; Figure 3b). Full results from the adiposity-cancer MR and the cancer-adiposity reverse MR analyses are available in Supplementary Table 7-8. 3.3 Adiposity-protein Figure 3c shows the association between BMI and plasma proteins associated with risk of endometrial cancer or subtypes, and their reasons for exclusion from the MVMR analysis. We identified two plasma proteins that had consistent causal association direction between adiposity-protein and protein-cancer: GSTO1-1 measured by SomaScan was a potential mediator for the effect of BMI on overall and endometrioid endometrial cancer risk; and IGF2R measured by both SomaScan and Olink was a potential mediator for the effect of BMI on endometrioid cancer risk. Full results from the adiposity-protein MR and the protein-adiposity reverse MR analyses are available in Supplementary Table 9-10. 3.4 Adiposity-protein-cancer effect mediation Figure 4 compares the total effect of BMI on overall and endometrioid endometrial cancer from UVMR analysis, and the direct effects after conditioning on potential protein mediators from MVMR analysis. There was little observed difference between the total and the direct effect for all potential protein mediators. The conditional F-statistics for GSTO1-1 (F=1.9), SomaScan IGF2R (1.6), and Olink IGF2R (1.5) were <10, indicating potential weak instrument bias in MVMR analyses. Similar results were observed in sensitivity analyses using a more stringent LD independence threshold of r 2 <0.001 (Supplementary Table 11). 3.5 Pathway enrichment In UVMR analyses, a total of 153 and 147 plasma proteins were associated with endometrioid and non-endometrioid endometrial cancer respectively at P <0.05, and they were included in the pathway enrichment analysis. We identified 65 and 69 pathways respectively for the endometrioid and non-endometrioid subtypes, which were grouped into 30 clusters each using hierarchical clustering (Supplementary Table 12-13). Pathways in the same cluster had similar upregulation and down regulation of genes, with some clusters containing only one pathway. Figure 5 shows the representative pathways of top ten clusters from the enrichment analysis based on multiple testing corrected P-values. For clusters with ≥2 pathways, we showed the top two pathways. We observed pathways that were distinct to each endometrial cancer subtype, including the top five pathway clusters for the endometrioid subtype (PDGF signalling, PTEN regulation, tRNA aminoacylation, and extracellular matrix degradation) and the top four pathway clusters for the non-endometrioid subtype (non-canonical NF-kB pathways, chemokines signalling, and EGFR signalling). The two subtypes also shared some similar pathways, including post-translational protein phosphorylation, PI3K/AKT signalling, and IGF/IGFBP regulation. 3.6 Expression quantitative trait loci Of the 12 plasma proteins associated with one of the endometrial cancer outcomes, four proteins (MMP10, IGF2R, MAPK9, and DNAJB14) were analysed in both platforms, of which two (MMP10 and DNAJB14) used the same cis pQTL and two (IGF2R and MAPK9) used different cis pQTL. This led to 14 cis pQTL examined for overlapping or in LD with eQTL in tissues. We found eight cis pQTL [GSTO1, IFI16, MAPK9 (SomaScan), IGF2R (SomaScan), IGF2R (Olink), TSSC4, DNAJB14, and ABO] overlapping or in high LD (r 2 > 0.8) with a cis eQTL in one or more tissues, including eQTL were found in whole blood [MAPK9 (SomaScan) and ABO], adipose tissues (GSTO1, DNAJB14, and ABO), and reproductive tissues (ABO) (Supplementary Table 14). 4 Discussion Of the 2,751 unique plasma proteins examined for their association with overall, endometrioid, and non-endometrioid endometrial cancer, we identified 20 pairs of potential causal associations with evidence from UVMR and colocalization. Endometrioid (GSTO1-1, MMP10, IGF2R, and SKAP1) and non-endometrioid (MAPK9, DNAJB14, IFI16, LCN2, and SCT) cancer subtypes were causally associated with different plasma proteins. Distinct pathways were overrepresented in endometrioid (e.g., PDGF signalling and PTEN gene regulation) and non-endometrioid (e.g., non-canonical NF-kB signalling) cancer subtypes. In UVMR, GSTO1-1 and IGF2R were identified as potential mediators for the effect of BMI on endometrioid endometrial cancer risk. In MVMR analyses, we observed little evidence of a mediating role of plasma proteins in the effect of BMI on endometrial cancer risk. The MR approach reduces risk of traditional observational confounding and improves causal inference. By using cis variants for proteins, we were more confident with the exclusion restriction assumption given cis variants are less likely to have pleiotropic effects. 47 However, by using only a lead cis variant for each protein, our analysis might have less power to detect an association. Several proteins showed evidence for colocalization but did not reach the multiple testing corrected P-value threshold in UVMR. Colocalization is usually considered the more conservative out of the two methods. 12 Furthermore, multiple testing correction can minimise false positives, but the large number of proteins tested may have led to incorrectly accepting the null hypothesis for some plasma proteins that were weakly associated with endometrial cancer risk. We make these results available for further analysis for interested readers. We used sex-combined GWAS data for plasma proteins, as female-specific GWAS data were not available. Emerging evidence suggests genetic effects on plasma protein level do not differ substantially by sex. 48 There was some sample overlap of UK Biobank participants in the adiposity GWAS including GIANT consortium and UK-Biobank participants and the Olink proteins GWAS in a subset of UK-Biobank participants. Given the strong instruments used in UVMR analyses, bias as a result of sample overlap is likely negligible. 49 However, conditional F-statistics for plasma proteins in MVMR analyses were < 10, and sample overlap of UK-Biobank participants may further exacerbate weak instruments bias. The role plasma proteins in mediating the effect of adiposity on endometrial cancer risk thus warrants further investigation using alternative methods, such as mediation analysis in traditional observational studies. Table 1 summarises functions of plasma proteins associated with endometrial cancer risk and their possible link with carcinogenesis. The function of some proteins is consistent with carcinogenesis and pathways identified from the enrichment analysis. For instance, SKAP1 expression in whole blood has been identified as a candidate susceptibility gene for endometrioid endometrial cancer in a transcriptome-wide association study and colocalization. 34 MMP10 initiates extracellular matrix remodelling in the endometrium, 35 , 36 which was also identified as one of the enriched pathways for endometrioid endometrial cancer. MAPK9 can promote cancer cell survival through crosstalk with other pathways including NF-κB, the top enriched pathways identified for non-endometrioid endometrial cancer. 41 We did not find any overlap of proteins between the endometrioid and non-endometrioid subtypes, suggesting that circulating proteins causally associated with the two endometrial cancer subtypes are different. This is consistent with the pathway enrichment results, where we identified different pathways for each endometrial cancer subtypes. Top pathways for endometrioid endometrial cancer included PDGF signalling in tumorigenesis, which could be stimulated by oestrogen and inhibited by progesterone, 4 , 50 and the established PTEN mutation pathway. 4 Top pathways for non-endometrioid endometrial cancer includes non-cardinal NF-κB pathways and chemokine bindings important for immune functions. 51 Mutant p53, present more commonly in non-endometrioid cancer, has been reported to fail in suppressing NF-κB mediated apoptotic signal, leading to overexpression of NF-κB in various tumours. 52 We also observed some common pathways for the endometrioid and non-endometrioid subtypes, including the PI3K/AKT pathway that is often deregulated in endometrial cancer. 4 The pathway enrichment analysis warrants replication, as we could only use plasma proteins associated with endometrial cancer risk at P < 0.05. Examining overlaps with eQTL in various tissues allowed us to explore how tissue-specific regulatory processes might relate to plasma protein levels. Only ABO, which is widely expressed in many tissues, had an eQTL in uterine tissue in high LD with pQTL. This suggests the plasma proteins identified in our analyses might not have a direct carcinogenic effect on the uterus. It is also possible that protein profiles in healthy tissues used in GTEx differ from cancerous tissues. GSTO1-1 and IGF2R were identified as potential mediators for the effect of BMI on endometrioid endometrial cancer in the UVMR analyses, but no evidence of mediation was observed in MVMR analyses. A recent study on circulating anti-GSTO1-1 antibodies reported elevated levels in various inflammatory conditions, suggesting it as a marker for acute and chronic inflammation. 53 GSTO1-1 activates NLRP3 inflammasome and TLR4 mediated release of pro-inflammatory cytokines including IL-6, TNF-α, and IL-1β. 32 , 54 These cytokines have been associated with low-grade adiposity-related chronic inflammation, as well as an increased risk of type 2 diabetes due to impaired insulin signalling in adipose tissue. 55 – 57 Membrane-bound IGF2R is considered a tumour suppressant by negatively regulates IGF2 activities. 58 In contrast, observational studies have reported elevated circulating IGF2R in individuals with obesity, type 2 diabetes, colorectal cancer, and breast cancer compared with healthy individuals. 40 , 59 , 60 However, the mechanisms linking elevated soluble IGF2R and diseases remain poorly understood. Previous MR study on potential molecular mediators for the effect of BMI on endometrial cancer identified fasting insulin, bioavailable testosterone, and SHBG using trans variants as instruments. 18 We did not examine insulin or testosterone. Our results for SHBG using single cis variant as instrument suggest an inverse association between BMI-SHBG and SHBG-endometrial cancer at P < 0.05 (Supplementary Table 5). We additionally observed an inverse association for IGFBP7 with overall and endometrioid endometrial cancer risk at P < 0.05, but not for other IGF binding proteins and receptors. We identified plasma proteins that might be involved in carcinogenic pathways for endometrioid and non-endometrioid endometrial cancer using complimentary approaches of MR and colocalization. These findings offer potential targets for further mechanistic studies, which could support the development of non-invasive methods to differentiate endometrial cancer subtypes and guide therapeutic strategies for clinical intervention. We identified two potential plasma proteins mediators, GSTO1-1 and IGF2R, for the effect of adiposity on endometrioid endometrial cancer risk in UVMR analysis. Alternative methods to examine effect mediation and future research integrating known pathways through a multiple-mediators mediation analysis could offer a more thorough understanding of how adiposity affects endometrial cancer risk. Abbreviations BMI body mass index CI confidence interval EC endometrial cancer EEC endometrioid endometrial cancer eQTL expression quantitative trait loci GWAS genome-wide association study IGF1 insulin-like growth factor 1 IL-6 interleukin 6 IVW-MRE inverse variance weighted multiplicative random effects LD linkage disequilibrium MR Mendelian randomisation MVMR multivariable Mendelian randomisation NEEC non-endometrioid endometrial cancer OR odds ratio pQTL protein quantitative trait loci SHBG sex hormone binding globulin TNF-α tumour necrosis factor-α UVMR univariable Mendelian randomisation WHR waist-hip ratio Declarations Ethics approval and consent to participate All studies contributing data to the analyses received the relevant institutional review board approval from each country, in accordance with the Declaration of Helsinki. All participants provided informed consent. Consent for publication Not applicable Availability of data and materials The UK Biobank Pharma Proteomics Project GWAS summary data are publicly available from UK Biobank (https://doi.org/10.7303/syn51364943). The GWAS summary data for SomaScan from the deCODE program are publicly available at https://download.decode.is/form/folder/proteomics. The meta-analysed Endometrial Cancer Association Consortium (ECAC) and Epidemiology of Endometrial Cancer Consortium (E2C2) GWAS summary data were obtained directly from the GWAS corresponding author (doi:10.1038/s41467-018-05427-7). They are available upon request from the corresponding author with the permission of the corresponding author of the ECAC GWAS. The combined GIANT consortium and UK-Biobank genome-wide association study (GWAS) summary data are publicly available at https://zenodo.org/records/1251813. The Genotype-Tissue Expression (GTEx) Project was supported by the Common Fund of the Office of the Director of the National Institutes of Health, and by NCI, NHGRI, NHLBI, NIDA, NIMH, and NINDS. The data used for the analyses described in this manuscript were obtained from the GTEx Portal on 16/09/2024. All analyses were performed using R (Vienna, Austria) version 4.1.2. Key R packages for the statistical analysis include: TwoSampleMR (version 0.6.8), coloc (version 5.2.2) 29 , MVMR (version 0.4) 61 , PathfindR, (version 2.4.1) 31 , and LDlinkR (version 1.4.0.9000) 62 . Statistical analysis codes used for this study are available at sabrinawang113/adiposity-proteins-endometrial (github.com). Competing interests The authors declare that they have no competing interests. Funding Funding for IIG_FULL_2021_008 was obtained from Wereld Kanker Onderzoek Fonds (WKOF), as part of the World Cancer Research Fund International grant programme; Funding for INCA_15849 was obtained from Institut national du cancer (INCa). NJT works within the CRUK Integrative Cancer Epidemiology Programme (C18281/A29019) and with support from CRUK (PRCPJT-May22\100028). TOM is supported by a National Health and Medical Research Council of Australia Emerging Leader Fellowship (APP1173170). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Contributors SEW: conceptualization, data curation, formal analysis, methodology, visualization, writing (original draft), writing (review & editing) VYT: methodology, writing (review & editing) JY: methodology, writing (review & editing) YZ: data curation, writing (review & editing) TOM: conceptualization, funding acquisition, data curation, resources, writing (review & editing) NJT: methodology, writing (review & editing) MJG: conceptualization, funding acquisition, resources, writing (review & editing) LD: conceptualization, funding acquisition, supervision, writing (review & editing) ML: conceptualization, data curation, methodology, visualization, writing (review & editing) Disclaimer Where authors are identified as personnel of the International Agency for Research on Cancer / World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy, or views of the International Agency for Research on Cancer / World Health Organization. References Lortet-Tieulent J, Ferlay J, Bray F, Jemal A. International Patterns and Trends in Endometrial Cancer Incidence, 1978–2013. JNCI: Journal of the National Cancer Institute 2017;110:354-61. Lauby-Secretan B, Scoccianti C, Loomis D, Grosse Y, Bianchini F, Straif K. Body Fatness and Cancer — Viewpoint of the IARC Working Group. New England Journal of Medicine 2016;375:794-8. Arnold M, Pandeya N, Byrnes G, Renehan AG, Stevens GA, Ezzati M, et al. Global burden of cancer attributable to high body-mass index in 2012: a population-based study. The Lancet Oncology 2015;16:36-46. Makker V, MacKay H, Ray-Coquard I, Levine DA, Westin SN, Aoki D, et al. Endometrial cancer. Nature Reviews Disease Primers 2021;7:88. Lukanova A, Lundin E, Micheli A, Arslan A, Ferrari P, Rinaldi S, et al. Circulating levels of sex steroid hormones and risk of endometrial cancer in postmenopausal women. International Journal of Cancer 2004;108:425-32. Ferkingstad E, Sulem P, Atlason BA, Sveinbjornsson G, Magnusson MI, Styrmisdottir EL, et al. Large-scale integration of the plasma proteome with genetics and disease. Nature genetics 2021;53:1712-21. Dossus L, Becker S, Rinaldi S, Lukanova A, Tjønneland A, Olsen A, et al. Tumor necrosis factor (TNF)‐α, soluble TNF receptors and endometrial cancer risk: The EPIC study. International journal of cancer 2011;129:2032-7. Dossus L, Rinaldi S, Becker S, Lukanova A, Tjonneland A, Olsen A, et al. Obesity, inflammatory markers, and endometrial cancer risk: a prospective case–control study. Endocrine-related cancer 2010;17:1007. Wang SE, Viallon V, Lee M, Dimou N, Hamilton F, Biessy C, et al. Circulating inflammatory and immune response proteins and endometrial cancer risk: a nested case-control study and Mendelian randomization analyses. eBioMedicine 2024;108:105341. Gunter MJ, Hoover DR, Yu H, Wassertheil-Smoller S, Manson JE, Li J, et al. A Prospective Evaluation of Insulin and Insulin-like Growth Factor-I as Risk Factors for Endometrial Cancer. Cancer Epidemiology, Biomarkers & Prevention 2008;17:921-9. Painter JN, O'Mara TA, Marquart L, Webb PM, Attia J, Medland SE, et al. Genetic Risk Score Mendelian Randomization Shows that Obesity Measured as Body Mass Index, but not Waist:Hip Ratio, Is Causal for Endometrial Cancer. Cancer Epidemiology, Biomarkers & Prevention 2016;25:1503-10. Zuber V, Grinberg NF, Gill D, Manipur I, Slob EAW, Patel A, et al. Combining evidence from Mendelian randomization and colocalization: Review and comparison of approaches. Am J Hum Genet 2022;109:767-82. Yarmolinsky J, Wade KH, Richmond RC, Langdon RJ, Bull CJ, Tilling KM, et al. Causal Inference in Cancer Epidemiology: What Is the Role of Mendelian Randomization? Cancer Epidemiology, Biomarkers & Prevention 2018;27:995-1010. Sanderson E. Multivariable Mendelian Randomization and Mediation. Cold Spring Harb Perspect Med 2021;11. Mullee A, Dimou N, Allen N, O’Mara T, Gunter MJ, Murphy N. Testosterone, sex hormone-binding globulin, insulin-like growth factor-1 and endometrial cancer risk: observational and Mendelian randomization analyses. British Journal of Cancer 2021;125:1308-17. Nead KT, Sharp SJ, Thompson DJ, Painter JN, Savage DB, Semple RK, et al. Evidence of a Causal Association Between Insulinemia and Endometrial Cancer: A Mendelian Randomization Analysis. JNCI: Journal of the National Cancer Institute 2015;107. Yarmolinsky J, Robinson JW, Mariosa D, Karhunen V, Huang J, Dimou N, et al. Association between circulating inflammatory markers and adult cancer risk: a Mendelian randomization analysis. eBioMedicine 2024;100. Hazelwood E, Sanderson E, Tan VY, Ruth KS, Frayling TM, Dimou N, et al. Identifying molecular mediators of the relationship between body mass index and endometrial cancer risk: a Mendelian randomization analysis. BMC Medicine 2022;20:125. Sun BB, Chiou J, Traylor M, Benner C, Hsu Y-H, Richardson TG, et al. Plasma proteomic associations with genetics and health in the UK Biobank. Nature 2023;622:329-38. O’Mara TA, Glubb DM, Amant F, Annibali D, Ashton K, Attia J, et al. Identification of nine new susceptibility loci for endometrial cancer. Nature Communications 2018;9:3166. Pulit SL, Stoneman C, Morris AP, Wood AR, Glastonbury CA, Tyrrell J, et al. Meta-analysis of genome-wide association studies for body fat distribution in 694 649 individuals of European ancestry. Human molecular genetics 2019;28:166-74. Pulit SL, de With SAJ, de Bakker PIW. Resetting the bar: Statistical significance in whole-genome sequencing-based association studies of global populations. Genetic Epidemiology 2017;41:145-51. Bowden J, Del Greco M F, Minelli C, Davey Smith G, Sheehan N, Thompson J. A framework for the investigation of pleiotropy in two‐sample summary data Mendelian randomization. Statistics in medicine 2017;36:1783-802. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genetic epidemiology 2013;37:658-65. Burgess S, Bowden J, Fall T, Ingelsson E, Thompson SG. Sensitivity Analyses for Robust Causal Inference from Mendelian Randomization Analyses with Multiple Genetic Variants. Epidemiology 2017;28:30-42. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. International Journal of Epidemiology 2017;46:1985-98. Zheng J, Richardson TG, Millard LAC, Hemani G, Elsworth BL, Raistrick CA, et al. PhenoSpD: an integrated toolkit for phenotypic correlation estimation and multiple testing correction using GWAS summary statistics. Gigascience 2018;7. Giambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, et al. Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using Summary Statistics. PLOS Genetics 2014;10:e1004383. Wallace C. Eliciting priors and relaxing the single causal variant assumption in colocalisation analyses. PLoS Genet 2020;16:e1008720. VanderWeele TJ. Mediation Analysis: A Practitioner's Guide. Annual Review of Public Health 2016;37:17-32. Ulgen E, Ozisik O, Sezerman OU. pathfindR: An R Package for Comprehensive Identification of Enriched Pathways in Omics Data Through Active Subnetworks. Frontiers in Genetics 2019;10. Hughes MM, Hooftman A, Angiari S, Tummala P, Zaslona Z, Runtsch MC, et al. Glutathione Transferase Omega-1 Regulates NLRP3 Inflammasome Activation through NEK7 Deglutathionylation. Cell Reports 2019;29:151-61.e5. Menon D, Innes A, Oakley AJ, Dahlstrom JE, Jensen LM, Brüstle A, et al. GSTO1-1 plays a pro-inflammatory role in models of inflammation, colitis and obesity. Sci Rep 2017;7:17832. Kho PF, Wang X, Cuéllar-Partida G, Dörk T, Goode EL, Lambrechts D, et al. Multi-tissue transcriptome-wide association study identifies eight candidate genes and tissue-specific gene expression underlying endometrial cancer susceptibility. Communications Biology 2021;4:1211. Osteen KG, Bruner-Tran KL, Keller NR, Eisenberg E. Progesterone-Mediated Endometrial Maturation Limits Matrix Metalloproteinase (MMP) Expression in an Inflammatory-like Environment. Annals of the New York Academy of Sciences 2002;955:37-47. Lee Y, Sooranna SR, Terzidou V, Christian M, Brosens J, Huhtinen K, et al. Interactions between inflammatory signals and the progesterone receptor in regulating gene expression in pregnant human uterine myocytes. Journal of Cellular and Molecular Medicine 2012;16:2487-503. Heydari N, Mahdizadeh M, Jafari SM. The evolving landscape of involvement of DTYMK enzymes in cancer. Med Oncol 2023;40:213. National Center for Biotechnology Information (NCBI). Gene: TSSC4 tumor suppressing subtransferable candidate 4 [Homo sapiens (human)]. NCBI. Accessed July 11, 2024, from https://www.ncbi.nlm.nih.gov/gene/10078. Cui H, Qu Y, Zhang L, Zhang W, Yan P, Yang C, et al. Epidemiological and genetic evidence for the relationship between ABO blood group and human cancer. International Journal of Cancer 2023;153:320-30. Scott CD, Kiess W. Soluble M6P/IGFIIR in the circulation. Best Practice & Research Clinical Endocrinology & Metabolism 2015;29:723-33. Wu Q, Wu W, Fu B, Shi L, Wang X, Kuca K. JNK signaling in cancer cell survival. Medicinal Research Reviews 2019;39:2082-104. Purificação ADd, Debbas V, Tanaka LY, Gabriel GVdM, Wosniak Júnior J, De Bessa TC, et al. DNAJB12 and DNJB14 are non-redundant Hsp40 redox chaperones involved in endoplasmic reticulum protein reflux. Biochimica et Biophysica Acta (BBA) - General Subjects 2024;1868:130502. National Center for Biotechnology Information (NCBI). Gene: IFI16 interferon gamma inducible protein 16 [Homo sapiens (human)]. NCBI. Accessed July 11, 2024, from https://www.ncbi.nlm.nih.gov/gene/3428. Rodvold JJ, Mahadevan NR, Zanetti M. Lipocalin 2 in cancer: when good immunity goes bad. Cancer letters 2012;316:132-8. Mannelqvist M, Stefansson IM, Wik E, Kusonmano K, Raeder MB, Øyan AM, et al. Lipocalin 2 expression is associated with aggressive features of endometrial cancer. BMC Cancer 2012;12:169. Ding W-Q, Cheng Z-J, McElhiney J, Kuntz SM, Miller LJ. Silencing of Secretin Receptor Function by Dimerization with a Misspliced Variant Secretin Receptor in Ductal Pancreatic Adenocarcinoma1. Cancer Research 2002;62:5223-9. Yarmolinsky J, Tzoulaki I, Gunter MJ, Travis RC, Davey Smith G, Smith-Byrne K. RE: Exploring the cross-cancer effect of circulating proteins and discovering potential intervention targets for 13 site-specific cancers. JNCI: Journal of the National Cancer Institute 2024;116:764-5. Mine K, Eleanor W, Nicola DK, Spiros D, Julia C-Z, Chloe MO, et al. Similar and different: systematic investigation of proteogenomic variation between sexes and its relevance for human diseases. medRxiv 2024:2024.02.16.24302936. Sadreev II, Elsworth BL, Mitchell RE, Paternoster L, Sanderson E, Davies NM, et al. Navigating sample overlap, winner’s curse and weak instrument bias in Mendelian randomization studies using the UK Biobank. medRxiv 2021:2021.06.28.21259622. Gentilini D, Busacca M, Di Francesco S, Vignali M, Viganò P, Di Blasio AM. PI3K/Akt And ERK1/2 signalling pathways are involved in endometrial cell migration induced by 17β-estradiol and growth factors. Molecular Human Reproduction 2007;13:317-22. Sun S-C. The non-canonical NF-κB pathway in immunity and inflammation. Nature Reviews Immunology 2017;17:545-58. Gulati AP, Yang Y-M, Harter D, Mukhopadhyay A, Aggarwal BA, Benzil DL, et al. Mutant human tumor suppressor p53 modulates the activation of mitogen-activated protein kinase and nuclear factor-κB, but not c-Jun N-terminal kinase and activated protein-1. Molecular Carcinogenesis 2006;45:26-37. Piaggi S, Lorenzini E, Pratesi F, Migliorini P, Pompella A, Bruschi F, et al. Anti-glutathione S-transferase omega 1-1 (GSTO1-1) antibodies are increased during acute and chronic inflammation in humans. Clin Exp Immunol 2022;209:305-10. Menon D, Innes A, Oakley AJ, Dahlstrom JE, Jensen LM, Brüstle A, et al. GSTO1-1 plays a pro-inflammatory role in models of inflammation, colitis and obesity. Scientific Reports 2017;7:17832. Maachi M, Piéroni L, Bruckert E, Jardel C, Fellahi S, Hainque B, et al. Systemic low-grade inflammation is related to both circulating and adipose tissue TNFα, leptin and IL-6 levels in obese women. International Journal of Obesity 2004;28:993-7. Wei S, Schmidt MD, Dwyer T, Norman RJ, Venn AJ. Obesity and menstrual irregularity: associations with SHBG, testosterone, and insulin. Obesity 2009;17:1070-6. Gao D, Madi M, Ding C, Fok M, Steele T, Ford C, et al. Interleukin-1β mediates macrophage-induced impairment of insulin signaling in human primary adipocytes. American Journal of Physiology-Endocrinology and Metabolism 2014;307:E289-E304. Livingstone C. IGF2 and cancer. Endocrine-Related Cancer 2013;20:R321-R39. Chanprasertyothin S, Jongjaroenprasert W, Ongphiphadhanakul B. The Association of Soluble IGF2R and IGF2R Gene Polymorphism with Type 2 Diabetes. Journal of Diabetes Research 2015;2015:216383. Jeyaratnaganthan N, Højlund K, Kroustrup JP, Larsen JF, Bjerre M, Levin K, et al. Circulating levels of insulin-like growth factor-II/mannose-6-phosphate receptor in obesity and type 2 diabetes. Growth Hormone & IGF Research 2010;20:185-91. Sanderson E, Spiller W, Bowden J. Testing and correcting for weak and pleiotropic instruments in two-sample multivariable Mendelian randomization. Statistics in medicine 2021;40:5434-52. Machiela MJ, Chanock SJ. LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics 2015;31:3555-7. Additional Declarations No competing interests reported. Supplementary Files APEMRsupptablesbmcmedicine.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5815826","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":407237087,"identity":"d169d71d-2763-4392-99c4-445c76213943","order_by":0,"name":"Sabrina E Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzUlEQVRIiWNgGAWjYNCCCgkwxQzEPERqOYOshY0YHYxtDHAtDAS1GBw/e/jlz3kW8ubtB9ikCyoYZMzlGwhoOZOXZs27TcJwzpkENukZZxh4LNsI2GJ2IMfMmHGbBOMMCQa227xtDDwGxwhpOf/GzPDnHAl7ErTcyDF+wNsgkUi8Fvsbb8yYeY5JJM/gSWz/zXNGAqglAb8Wyf4c448/aupsZ7AfPmzMU2Fjb3D4AAFrgPEAiUcGxgYgIUFQOQgwfyBK2SgYBaNgFIxcAADHxzmQXhsTmQAAAABJRU5ErkJggg==","orcid":"","institution":"International Agency for Research on Cancer, World Health Organization","correspondingAuthor":true,"prefix":"","firstName":"Sabrina","middleName":"E","lastName":"Wang","suffix":""},{"id":407237088,"identity":"37b6b187-c14d-47ea-834f-ca4244b75dca","order_by":1,"name":"Vanessa Y Tan","email":"","orcid":"","institution":"University of Bristol","correspondingAuthor":false,"prefix":"","firstName":"Vanessa","middleName":"Y","lastName":"Tan","suffix":""},{"id":407237089,"identity":"1e2fff07-1d21-4489-9169-034552e04fc9","order_by":2,"name":"James Yarmolinsky","email":"","orcid":"","institution":"Imperial College London","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"Yarmolinsky","suffix":""},{"id":407237090,"identity":"69fda4aa-8664-4e44-bac6-eaabf8de3a5b","order_by":3,"name":"Yadi Zheng","email":"","orcid":"","institution":"International Agency for Research on Cancer, World Health Organization","correspondingAuthor":false,"prefix":"","firstName":"Yadi","middleName":"","lastName":"Zheng","suffix":""},{"id":407237091,"identity":"92eff10c-4f10-4cdd-bf4a-475aa8e5ad80","order_by":4,"name":"Tracy A O’Mara","email":"","orcid":"","institution":"QIMR Berghofer Medical Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Tracy","middleName":"A","lastName":"O’Mara","suffix":""},{"id":407237092,"identity":"e0ff57bf-51d2-4363-9982-fe47d6e1e068","order_by":5,"name":"Nicholas J Timpson","email":"","orcid":"","institution":"University of Bristol","correspondingAuthor":false,"prefix":"","firstName":"Nicholas","middleName":"J","lastName":"Timpson","suffix":""},{"id":407237093,"identity":"5a10f720-6ddc-41c9-9b78-06d1b8e9839c","order_by":6,"name":"Marc J Gunter","email":"","orcid":"","institution":"International Agency for Research on Cancer, World Health Organization","correspondingAuthor":false,"prefix":"","firstName":"Marc","middleName":"J","lastName":"Gunter","suffix":""},{"id":407237094,"identity":"ce17ccd6-1923-4973-9b8e-6502c893fb47","order_by":7,"name":"Laure Dossus","email":"","orcid":"","institution":"International Agency for Research on Cancer, World Health Organization","correspondingAuthor":false,"prefix":"","firstName":"Laure","middleName":"","lastName":"Dossus","suffix":""},{"id":407237095,"identity":"377c28bf-cee0-4141-bf4a-a5de469d2b14","order_by":8,"name":"Matthew A Lee","email":"","orcid":"","institution":"International Agency for Research on Cancer, World Health Organization","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"A","lastName":"Lee","suffix":""}],"badges":[],"createdAt":"2025-01-13 01:23:13","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5815826/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5815826/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":75405474,"identity":"ef9c6734-5ca3-424a-85ab-adf37ecf179c","added_by":"auto","created_at":"2025-02-04 08:47:23","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":61537,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic diagram of the statistical analysis\u003c/p\u003e\n\u003cp\u003eSolid line = forward MR analysis; dotted line = reverse MR analysis\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5815826/v1/df3874a8f7d9a38e9ce044cf.png"},{"id":75405477,"identity":"08ebda86-55ca-40db-84da-9a2cecd75ce7","added_by":"auto","created_at":"2025-02-04 08:47:23","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":178689,"visible":true,"origin":"","legend":"\u003cp\u003eFlow diagram for plasma proteins included in analyses\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e1 \u003c/sup\u003eWald ratio P-value below multiple testing corrected threshold or evidence for colocalization\u003cbr\u003e\n\u0026nbsp; No association in the reverse MR\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e2 \u003c/sup\u003eIVW-MRE P-value below multiple testing corrected threshold\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; Consistent direction of effect across all sensitivity analysis models\u003cbr\u003e\n\u0026nbsp; No association in the reverse MR\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e3 \u003c/sup\u003eConsistent directions of causal association between BMI-Protein and Protein-EC\u003c/p\u003e\n\u003cp\u003eBMI=body mass index; EC=endometrial cancer; MVMR=multivariable Mendelian randomisation; EEC=endometrioid endometrial cancer; NEEC=non-endometrioid endometrial cancer\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5815826/v1/85ee480d0af2e17ea1f4c6e2.png"},{"id":75407782,"identity":"10cdacf5-5e63-4f51-ac7a-8ef07e7fcdc5","added_by":"auto","created_at":"2025-02-04 08:55:23","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":430622,"visible":true,"origin":"","legend":"\u003cp\u003eSummary results from analysis Part (a)-(c). (a) Plasma proteins associated with endometrial cancer risk in MR and colocalization analysis; (b) Causal association between BMI and endometrial cancer risk; (c) Causal association between BMI and selected proteins associated with endometrial cancer risk, and their reason for exclusion from the MVMR.\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e1 \u003c/sup\u003eSolid circle indicates p-value below multiple testing corrected threshold.\u003c/p\u003e\n\u003cp\u003eBMI=body mass index; CI=confidence interval; MR=Mendelian randomisation; MVMR=multivariable Mendelian randomisation; OR=odds ratio; WHR=waist-hip ratio\u003c/p\u003e\n\u003cp\u003e_o = Olink platform; _s = SomaScan platform\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5815826/v1/20992480221393807bbe0bf2.png"},{"id":75405479,"identity":"7bcc7a2e-a91e-4f07-984c-a00db3ae6b3d","added_by":"auto","created_at":"2025-02-04 08:47:23","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":102524,"visible":true,"origin":"","legend":"\u003cp\u003eResults from MVMR mediation analysis comparing the total effect of BMI on (a) endometrial and (b) endometrioid cancer risk with the direct affect after conditioning on potential protein mediators.\u003c/p\u003e\n\u003cp\u003eBMI=body mass index; CI=confidence interval; OR=odds ratio\u003c/p\u003e\n\u003cp\u003e_o = Olink platform; _s = SomaScan platform\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5815826/v1/aa228ebb783dc35695734714.png"},{"id":75405482,"identity":"4a0e40d0-1248-4fc7-8f8f-7ec0ed6fbd89","added_by":"auto","created_at":"2025-02-04 08:47:23","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":375227,"visible":true,"origin":"","legend":"\u003cp\u003ePathway enrichment analysis for proteins associated with endometrioid and non-endometrioid cancer subtypes at P \u0026lt;0.05 in UVMR\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-5815826/v1/b5ab6b95c0a625fc89e39090.png"},{"id":75411581,"identity":"6fd4dbd3-030b-41b1-8618-a99fc1255f41","added_by":"auto","created_at":"2025-02-04 09:11:23","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1850097,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5815826/v1/55af69d9-0457-45b3-876c-ef80d64ece14.pdf"},{"id":75405481,"identity":"a52b6a28-b90a-4a78-92e2-e5b06b4e4b7d","added_by":"auto","created_at":"2025-02-04 08:47:23","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":6225641,"visible":true,"origin":"","legend":"","description":"","filename":"APEMRsupptablesbmcmedicine.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5815826/v1/ea15ae37cccf12e8200cc589.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"The effect of circulating proteins and their role in mediating adiposity’s effect on endometrioid and non-endometrioid endometrial cancer risk: Mendelian randomisation and colocalization analyses ","fulltext":[{"header":"1 Background","content":"\u003cp\u003eEndometrial cancer incidence has been rising in successive generations across continents over the past few decades.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e Adiposity is strongly associated with endometrial cancer risk,\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e with 34% of global endometrial cancer cases attributable to increased BMI.\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e The endometrioid subtype is more common and more consistently associated with adiposity, whereas non-endometrioid endometrial cancer, including serous carcinomas, clear cell carcinomas, and carcinosarcomas, generally have poorer prognosis and less established risk factors.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003ePlasma proteomics is a non-invasive method to better understand potential pathogenic pathways associated with different endometrial cancer subtypes, which may be useful in aiding diagnosis, identifying targets for clinical interventions, and exploring the mechanisms behind known risk factors. Traditional observational studies have reported associations between circulating biomarkers and endometrial cancer risk, including inverse associations with sex hormone binding globulin (SHBG)\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e and insulin-like growth factor 1 (IGF1)\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e and positive associations with insulin and inflammatory cytokines such as interleukin (IL-6) and tumour necrosis factor-α (TNF-α).\u003csup\u003e\u003cspan additionalcitationids=\"CR8 CR9\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e Addressing residual and unmeasured confounding remains a challenge for traditional observational studies, particularly given the strong effect of adiposity on the circulating proteome and endometrial cancer risk.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eMendelian randomisation (MR) and genetic colocalization are complimentary statistical methods that can be applied to genome-wide association study (GWAS) summary data to assess causal associations between traits of interest.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e MR analyses leverage genetic variants as instrumental variables to infer causality.\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e Given a sufficiently large population, MR is akin to the randomisation process in randomised controlled trials and theoretically minimises risk of confounding. Multivariable MR (MVMR) can estimate the effect of multiple exposures on an outcome, allowing mediation analysis in the MR framework.\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e Genetic colocalization estimates the probability that two or more traits are associated with the same genetic variant(s), thereby sharing a common genetic cause.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e Genetic colocalization analysis cannot distinguish between exposure and outcome traits, but it does not rely on instrumental variable assumptions required in MR analyses.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e Consistent evidence from both MR and colocalization analysis can strengthen evidence for a causal effect.\u003c/p\u003e \u003cp\u003eMR analyses of plasma proteins have reported causal associations with endometrial cancer risk, including reduced SHBG\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e and elevated insulin,\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e whereas no association was observed for IGF1\u003csup\u003e15\u003c/sup\u003e or a panel of 66 circulating inflammatory markers.\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e A study that examined molecular mediators for BMI\u0026rsquo;s effect on endometrial cancer risk reported potential mediating roles for SHBG, insulin, and bioavailable testosterone.\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e We aimed to investigate the effect of circulating proteins on endometrial cancer risk. To capture a broad spectrum of plasma proteins, we used GWAS summary data from two independent studies.\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e We examined 2,751 unique proteins (2,031 Olink proteins and 1,667 SomaScan aptamers) and their association with overall, endometrioid, and non-endometrioid endometrial cancer risk by performing univariable MR (UVMR) and colocalization analyses. We further examined the potential role of plasma proteins in mediating the effect of BMI on endometrial cancer risk using UVMR and MVMR.\u003c/p\u003e"},{"header":"2 Methods","content":"\u003cp\u003eThe main analysis involved four parts (Figure 1). We first examined the causal relationship between plasma proteins and endometrial cancer risk, by performing \u003cem\u003ecis\u003c/em\u003e variant UVMR and colocalization (Part a). We then examined the causal relationship between adiposity (i.e., BMI) and endometrial cancer risk (Part b) and between adiposity and plasma proteins using UVMR (Part c). To exclude bi-directional effects, we performed both forward and reverse MR analyses for Part (a) – (c). For each potential protein mediator, we performed MVMR to explore effect mediation (Part d). \u003c/p\u003e\n\u003ch2\u003e2.1 GWAS sources and study populations\u003c/h2\u003e\n\u003cp\u003eWe used summary results from GWAS published by 1) Sun et al. (2023)\u003csup\u003e19\u003c/sup\u003e on up to 2,922 plasma proteins measured with Olink Explore 3072 platform in a sub-cohort of 52,363 UK-Biobank participants of European ancestries; 2) Ferkingstad et al. (2021)\u003csup\u003e6\u003c/sup\u003e on up to 4,907 aptamers for 4,719 plasma proteins measured with SomaScan multiplex aptamer assay (version 4) in 35,559 Icelanders from deCODE; 3) O’Mara et al. (2018)\u003csup\u003e20\u003c/sup\u003e on endometrial cancer (12,270 cases and 46,126 controls), endometrioid cancer subtype (8,758 cases and 46,126 controls), and non-endometrioid cancer subtype (1,230 cases and 35,447 controls of European ancestries) from the meta-analysed Endometrial Cancer Association Consortium and Epidemiology of Endometrial Cancer Consortium data excluding UK-Biobank participants; 4) Pulit et al. (2019)\u003csup\u003e21\u003c/sup\u003e on female-specific BMI, which includes 434,794 women of European ancestries from the GIANT consortium and UK-Biobank.\u003c/p\u003e\n\u003ch2\u003e2.2 Genetic instruments\u003c/h2\u003e\n\u003cp\u003eTo select a genetic instrument for each plasma protein, a 1Mb region was defined around each \u003cem\u003ecis\u003c/em\u003e variant [i.e., a variant ≤1Mb from the transcription start site of the protein coding gene (discovery) or ≤1Mb from the gene encoding the measured protein (replication)] reaching the genome-wide significance threshold (P \u0026lt;1.8×10\u003csup\u003e− 9\u003c/sup\u003e in deCODE; P\u0026lt; 1.7×10\u003csup\u003e− 11\u003c/sup\u003e in UK-Biobank). Starting with the variant with the lowest P-value, any overlapping regions were merged until no overlapping regions remained. Linkage disequilibrium (LD) based clumping was then used to merge regions for variants in high LD (r\u003csup\u003e2 \u003c/sup\u003e≥0.8). The variant with the lowest P-value after merging was considered the sentinel variant and was used as the genetic instrument for the protein. We obtained the list of instruments for plasma proteins directly from the respective GWAS.\u003csup\u003e6,19\u003c/sup\u003e A total of 2,031 Olink proteins and 1,667 SomaScan aptamers (1,447 proteins) were instrumented with a \u003cem\u003ecis\u003c/em\u003e variant, representing 2,751 unique proteins, with 727 proteins measured by both platforms. \u003c/p\u003e\n\u003cp\u003eGenetic instruments for BMI were selected by extracting variants that reach a genome-wide significance threshold of P \u0026lt;5×10\u003csup\u003e-9\u003c/sup\u003e (to account for a wider coverage of data sequenced\u003csup\u003e22\u003c/sup\u003e) and clumping at a LD independence threshold of r\u003csup\u003e2\u003c/sup\u003e \u0026lt;0.001 using the 1000 Genomes Project phase 3 European population. Genetic instruments for liability to endometrial cancers were selected by searching variants that reach a genome-wide significance threshold of P \u0026lt;5×10\u003csup\u003e-8\u003c/sup\u003e (P \u0026lt;5×10\u003csup\u003e-7\u003c/sup\u003e for non-endometrioid cancer as no variant reached the threshold) and clumping at a LD independence threshold of r\u003csup\u003e2\u003c/sup\u003e \u0026lt;0.001. Where genetic variants were not available in the outcome GWAS, proxy variants with LD r\u003csup\u003e2\u003c/sup\u003e ≥0.8 were used. The relevance assumption was evaluated by F-statistics, with an F-statistic \u0026gt;10 considered as less likely to suffer from weak instrument bias.\u003c/p\u003e\n\u003cp\u003eData on BMI, proteins, and aptamers were inverse rank normally transformed in their respective GWAS. Assuming the distribution of each trait was normal prior to transformation, the measures of association obtained from MR analyses could be approximately interpreted as the change in per normalized standard deviation unit change in the exposure.\u003c/p\u003e\n\u003ch2\u003e2.3 Statistical analysis\u003c/h2\u003e\n\u003ch3\u003e2.3.1 Univariable Mendelian randomization\u003c/h3\u003e\n\u003cp\u003eUVMR analysis was performed to examine protein-cancer, adiposity-cancer, adiposity-protein (forward MR analyses), cancer-protein, cancer-adiposity, and protein-adiposity (reverse MR analyses) associations. For the primary analysis, we estimated causal associations using Wald ratio (where single \u003cem\u003ecis\u003c/em\u003e variants were used as instruments for specific proteins) or inverse variance weighted multiplicative random effects (IVW-MRE) model (where ≥2 variants were used as instruments for adiposity or cancer). The IVW-MRE model assumes that pleiotropic effects of individual instruments sum to zero and that pleiotropic effects are independent of instrument strength across all variants.\u003csup\u003e23,24\u003c/sup\u003e As sensitivity analyses for the IVW-MRE model, we used MR-Egger, weighted-median, and weighted-mode models. MR-Egger can help in detection and adjustment for directional horizontal pleiotropy by providing an intercept term of the linear regression for the measured association.\u003csup\u003e25\u003c/sup\u003e The weighted-median can provide consistent estimates when at least 50% of the instruments are valid.\u003csup\u003e25\u003c/sup\u003e The weighted-mode can provide consistent estimates as long as the largest cluster of similar estimates are derived from valid instruments.\u003csup\u003e26\u003c/sup\u003e In addition, we performed leave-one-out and single-variant MR analyses to assess the influence of individual variants on observed associations.\u003c/p\u003e\n\u003cp\u003eTo correct for multiple testing for analyses involving plasma proteins, we used PhenoSpD to estimate the phenotypic correlations and the number of independent tests.\u003csup\u003e27\u003c/sup\u003e A total of 505 SomaScan and 1,026 Olink independent proteins were identified, corresponding to a PhenoSpD corrected P-value of 1.02×10\u003csup\u003e-4\u003c/sup\u003e and 5.00×10\u003csup\u003e-5\u003c/sup\u003e respectively. We considered a causal association from MR analysis as one with 1) a P-value below the multiple comparison corrected threshold from the primary analysis (Wald ratio or IVW-MRE); 2) consistent direction of association across all sensitivity analysis models (MR-Egger, weighted-median, weighted-mode) if applicable; 3) no consistent association observed in the reverse MR analysis. Furthermore, a plasma protein was considered a potential mediator for the effect of adiposity on endometrial cancer risk if there was consistent direction of effect from adiposity to protein and from protein to cancer risk.\u003c/p\u003e\n\u003ch3\u003e2.3.2 Colocalization\u003c/h3\u003e\n\u003cp\u003eWe performed colocalization for plasma protein and endometrial cancer risk assuming a single causal variant.\u003csup\u003e28\u003c/sup\u003e The \u003cem\u003ecis\u003c/em\u003e variant instrument used in MR for each plasma protein was extracted along with a 1 Mb window around the variant from the protein and the endometrial cancer GWAS. We assigned a prior of 1×10\u003csup\u003e-6 \u003c/sup\u003efor p1 and p2, and 1×10\u003csup\u003e-7\u003c/sup\u003e for p12 using https://chr1swallace.shinyapps.io/coloc-priors/ (accessed: 07/03/2024), based on an average of approximately 3,000 variants available for colocalization within the 1mb window.\u003csup\u003e29\u003c/sup\u003e We used P(h\u003csub\u003e4\u003c/sub\u003e) ≥0.8 as evidence for colocalization.\u003c/p\u003e\n\u003ch3\u003e2.3.3 Multivariable Mendelian randomization\u003c/h3\u003e\n\u003cp\u003eWe performed MVMR analysis to estimate the natural direct effect of adiposity on endometrial cancer risk conditional on each potential plasma protein mediator.\u003csup\u003e14\u003c/sup\u003e To construct instruments for adiposity conditioned on a plasma protein mediator, we searched the gene region (±1Mb) for variants from the plasma protein GWAS that reach P \u0026lt;5×10\u003csup\u003e-8\u003c/sup\u003e. We clumped at a relaxed LD independence threshold of r\u003csup\u003e2\u003c/sup\u003e \u0026lt;0.1 as an attempt to improve instrument strength (a separate sensitivity analysis using LD independence threshold of r\u003csup\u003e2\u003c/sup\u003e \u0026lt;0.001 was also performed). We then extracted these variants from the BMI GWAS, combined them with instruments for adiposity, and clumped to remove variants in LD. The effect of adiposity on endometrial cancer risk conditioned on each plasma protein was estimated using an inverse variance weighted model. We compared this direct effect with the total effect of adiposity on endometrial cancer risk estimated from analysis Part (b). We did not calculate the proportion of total effect mediated by each protein as, when using the difference method (i.e., comparing effects estimates from regression model of adiposity on endometrial cancer without and with the potential mediator) with binary outcomes, the indirect effect could be underestimated due to non-collapsibility property of odds ratios (OR).\u003csup\u003e14,30\u003c/sup\u003e \u003c/p\u003e\n\u003ch3\u003e2.3.4 Pathway enrichment\u003c/h3\u003e\n\u003cp\u003eWe performed pathway enrichment analysis for proteins associated with endometrioid or non-endometrioid endometrial cancer at P \u0026lt;0.05 in the UVMR analysis. We used an active-subnetwork-oriented approach, which maps each gene onto a protein-protein interaction network to better account for gene interaction information.\u003csup\u003e31\u003c/sup\u003e Active subnetworks were identified using a greedy algorithm, which were then included in enrichment analysis.\u003csup\u003e31\u003c/sup\u003e The process was repeated for ten iterations to cross validate the greedy algorithm selections. We used the Biogrid protein-protein interaction repository and the pathway-based Reactome gene sets database. Hierarchical clustering was performed on identified pathways to select representatives (i.e., pathways with the lowest multiple testing adjusted P-value) among groups of related pathways.\u003csup\u003e31\u003c/sup\u003e\u003c/p\u003e\n\u003ch3\u003e2.3.5 Expression quantitative trait loci \u003c/h3\u003e\n\u003cp\u003eTo explore whether proteins associated with endometrial cancer risk might reflect gene expression and biological processes in tissues, we examined whether the \u003cem\u003ecis \u003c/em\u003eprotein quantitative trait loci (pQTL) were overlapping or in high LD (r\u003csup\u003e2\u003c/sup\u003e \u0026gt; 0.8) with \u003cem\u003ecis\u003c/em\u003e expression quantitative trait loci (eQTL). We used \u003cem\u003ecis\u003c/em\u003e-eQTL data from 49 non-diseased tissue sites in the Genotype-Tissue Expression Project (GTEx) Release V8. A \u003cem\u003ecis\u003c/em\u003e eQTL was defined as a variant within the gene region with P \u0026lt;5×10\u003csup\u003e-8\u003c/sup\u003e. We calculated the LD using the 1000 Genomes Project phase 3 European population. We further used this analysis to investigate possible epitope or aptamer binding effect, given both Olink and SomaScan rely on the binding of antibody or aptamer to proteins for quantification. If a variant also influences the expression of the gene encoding the protein, then it is considered less likely that the observed association is due to epitope or aptamer binding effect.\u003csup\u003e6,17\u003c/sup\u003e\u003c/p\u003e"},{"header":"3 Results","content":"\u003cp\u003eOf the 2,031 Olink proteins and 1,667 SomaScan aptamers (1,447 proteins) that had an available \u003cem\u003ecis\u003c/em\u003e variant instrument, we extracted 1,648 and 901 variants or proxies respectively from the endometrial cancers GWAS and examined for protein-cancer association in UVMR and colocalization (Figure 2). All 2,031 Olink proteins and 1,667 SomaScan aptamers were examined for BMI-protein association. Details of genetic variants used to instrument adiposity measures, plasma proteins, and endometrial cancers are presented in Supplementary Table 1-4. F statistics indicated strong instruments for BMI (mean=67.7), endometrial cancer (44.2), endometrioid cancer (43.0), and non-endometrioid cancer subtype (26.9). F statistics for plasma proteins instrumented using a single \u003cem\u003ecis\u0026nbsp;\u003c/em\u003evariant ranged from 36.2 to 117,180, with a mean of 3,202.\u003c/p\u003e\n\u003ch2\u003e3.1 Protein-endometrial cancer\u003c/h2\u003e\n\u003cp\u003eThere were 20 protein-endometrial cancer associations between 12 plasma proteins and the three endometrial cancer outcomes with evidence of colocalization [P(h\u003csub\u003e4\u003c/sub\u003e) \u0026gt; 0.8], of which 5 pairs reached the multiple testing corrected P-value threshold for UVMR (Figure 3a, Table 1). GSTO1-1 and SKAP1 were positively associated and MMP10 was negatively associated with both overall and endometrioid endometrial cancer; DTYMK and ABO were positively associated and TSSC4 was negatively associated with overall endometrial cancer; IGF2R was positively associated with endometrioid endometrial cancer; MAPK9 was positively associated and DNAJB14, IFI16, LCN2, and SCT were negatively associated with non-endometrioid endometrial cancer.\u003c/p\u003e\n\u003cp\u003eOf those associated with endometrial cancers risk, four proteins (MMP10, IGF2R, MAPK9, and DNAJB14) were analysed in both Olink and SomaScan platforms with highly consistent results. LCN2 was also analysed in both platforms, with UVMR results showing consistent inverse association with non-endometrioid endometrial cancer risk but only the Olink assay showed evidence for colocalization [Olink P(h4)=0.84; SomaScan P(h4)=0.43]. The remaining proteins associated with endometrial cancers were examined only in one of the two platforms. Full results from the protein-cancer MR and the cancer-protein reverse MR analyses are available in Supplementary Table 5-6.\u003c/p\u003e\n\u003cp\u003eTable 1. Summary of proteins causally associated with endometrial cancer risk, their functions and possible link to endometrial cancer development.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOutcome\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProtein\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBMI-protein\u003csup\u003e2\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProtein-cancer\u003csup\u003e2\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFunctions and possible link to cancer\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eEndometrial, Endometrioid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eGSTO1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eMediates pro-inflammatory signalling through TLR4 and NLRP3 inflammasome, leading to consequent release of TNF-\u0026alpha;, IL-6 and IL-1\u0026beta;.\u003csup\u003e32,33\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eSKAP1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+ \u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eRegulates T-cell signalling and enhances adhesion of T-cells with antigen-presenting cells. SKAP1 expression in whole blood identified as a candidate susceptibility gene for endometrioid cancer.\u003csup\u003e34\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eMMP10\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e∆/\u0026minus; \u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026minus;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eInitiates extracellular matrix remodelling during physiological (e.g., menstrual cycle) and pathological (e.g., metastasis) processes. It is specifically expressed in the endometrium and is thought to be down regulated by progesterone.\u003csup\u003e35,36\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eEndometrial\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eDTYMK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+ \u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eCatalyses the conversion of dTMP to dTDP, essential for DNA synthesis and repair. Found to be upregulated in several tumours, including liver, lung, breast, and colon tumour cells.\u003csup\u003e37\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eTSSC4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+ \u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026minus;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eTumour suppressing sub-transferable candidate 4 located in the imprinted gene domain 11p15.5. Alteration in this region have been associated with several cancers.\u003csup\u003e38\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eABO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e∆ \u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eABO blood group. Expression of A or B antigen has been linked with increased risk of several cancers, including endometrial cancer.\u003csup\u003e39\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eEndometrioid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eIGF2R\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eSoluble IGF2R transports and inhibits IGF2 mediated growth. It also has binding site for M6P and various ligands. Elevated plasma IGF2R has been observed in individuals with obesity, type 2 diabetes, colorectal cancer, and breast cancer, but the underlying mechanism remains unclear.\u003csup\u003e40\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003eNon-endometrioid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eMAPK9\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e∆/+ \u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003ePhosphorylates p53 protein to increase its stability under physiological states but can promote cancer cell survival through crosstalk with other pathways including NF-\u0026kappa;B.\u003csup\u003e41\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eDNAJB14\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+ \u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026minus;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eA molecular chaperone that supports proper protein folding and responsible for degradation of misfolded proteins.\u003csup\u003e42\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eIFI16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026minus;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eModulation of p53 function and inhibition of cell growth via the Ras/Raf signalling pathway.\u003csup\u003e43\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eLCN2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e+\u003csup\u003e\u0026nbsp;\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026minus;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eImportant for innate immunity against bacterial growth but can promote epithelial-to-mesenchymal transition to facilitate tumour invasion and metastasis.\u003csup\u003e44\u003c/sup\u003e It has been associated with aggressiveness of endometrial cancer.\u003csup\u003e45\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 123px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 95px;\"\u003e\n \u003cp\u003eSCT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e∆ \u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 76px;\"\u003e\n \u003cp\u003e\u0026minus;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 562px;\"\u003e\n \u003cp\u003eStimulates the secretion of bile and bicarbonate. Overexpression has been observed in pancreatic tumours,\u003csup\u003e46\u003c/sup\u003e but no clear link with other tumours.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003csup\u003e1\u0026nbsp;\u003c/sup\u003eCausally associated with endometrial cancer risk in both SomaScan and Olink platforms\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e2\u0026nbsp;\u003c/sup\u003eDirection of causal association:\u0026nbsp;+\u0026nbsp;=positively associated; \u0026minus; =inversely associated; ∆ =failed to demonstrate consistent direction across MR sensitivity models\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e3\u0026nbsp;\u003c/sup\u003eFailed to rule out an association in the protein-BMI reverse MR analysis\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e4\u0026nbsp;\u003c/sup\u003eP-value did not reach the multiple testing corrected threshold in the BMI-protein MR analysis\u003c/p\u003e\n\u003ch2\u003e3.2 Adiposity-endometrial cancer\u003c/h2\u003e\n\u003cp\u003eBMI was associated with overall endometrial [OR per normalized standard deviation change in BMI =1.79; 95% confidence interval (CI) 1.55-2.01], endometrioid (1.81; 1.58-2.07), and non-endometrioid cancer (1.63; 1.23-2.17; Figure 3b). Full results from the adiposity-cancer MR and the cancer-adiposity reverse MR analyses are available in Supplementary Table 7-8.\u003c/p\u003e\n\u003ch2\u003e3.3 Adiposity-protein\u003c/h2\u003e\n\u003cp\u003eFigure 3c shows the association between BMI and plasma proteins associated with risk of endometrial cancer or subtypes, and their reasons for exclusion from the MVMR analysis. We identified two plasma proteins that had consistent causal association direction between adiposity-protein and protein-cancer: GSTO1-1 measured by SomaScan was a potential mediator for the effect of BMI on overall and endometrioid endometrial cancer risk; and IGF2R measured by both SomaScan and Olink was a potential mediator for the effect of BMI on endometrioid cancer risk. Full results from the adiposity-protein MR and the protein-adiposity reverse MR analyses are available in Supplementary Table 9-10.\u003c/p\u003e\n\u003ch2\u003e3.4 Adiposity-protein-cancer effect mediation\u003c/h2\u003e\n\u003cp\u003eFigure 4 compares the total effect of BMI on overall and endometrioid endometrial cancer from UVMR analysis, and the direct effects after conditioning on potential protein mediators from MVMR analysis. There was little observed difference between the total and the direct effect for all potential protein mediators. The conditional F-statistics for GSTO1-1 (F=1.9), SomaScan IGF2R (1.6), and Olink IGF2R (1.5) were \u0026lt;10, indicating potential weak instrument bias in MVMR analyses. Similar results were observed in sensitivity analyses using a more stringent LD independence threshold of r\u003csup\u003e2\u003c/sup\u003e \u0026lt;0.001 (Supplementary Table 11).\u003c/p\u003e\n\u003ch2\u003e3.5 Pathway enrichment\u003c/h2\u003e\n\u003cp\u003eIn UVMR analyses, a total of 153 and 147 plasma proteins were associated with endometrioid and non-endometrioid endometrial cancer respectively at P \u0026lt;0.05, and they were included in the pathway enrichment analysis. We identified 65 and 69 pathways respectively for the endometrioid and non-endometrioid subtypes, which were grouped into 30 clusters each using hierarchical clustering (Supplementary Table 12-13). Pathways in the same cluster had similar upregulation and down regulation of genes, with some clusters containing only one pathway.\u003c/p\u003e\n\u003cp\u003eFigure 5 shows the representative pathways of top ten clusters from the enrichment analysis based on multiple testing corrected P-values. For clusters with \u0026ge;2 pathways, we showed the top two pathways. We observed pathways that were distinct to each endometrial cancer subtype, including the top five pathway clusters for the endometrioid subtype (PDGF signalling, PTEN regulation, tRNA aminoacylation, and extracellular matrix degradation) and the top four pathway clusters for the non-endometrioid subtype (non-canonical NF-kB pathways, chemokines signalling, and EGFR signalling). The two subtypes also shared some similar pathways, including post-translational protein phosphorylation, PI3K/AKT signalling, and IGF/IGFBP regulation.\u003c/p\u003e\n\u003ch2\u003e3.6 Expression quantitative trait loci\u003c/h2\u003e\n\u003cp\u003eOf the 12 plasma proteins associated with one of the endometrial cancer outcomes, four proteins (MMP10, IGF2R, MAPK9, and DNAJB14) were analysed in both platforms, of which two (MMP10 and DNAJB14) used the same \u003cem\u003ecis\u003c/em\u003e pQTL and two (IGF2R and MAPK9) used different \u003cem\u003ecis\u003c/em\u003e pQTL. This led to 14 \u003cem\u003ecis\u003c/em\u003e pQTL examined for overlapping or in LD with eQTL in tissues. We found eight \u003cem\u003ecis\u0026nbsp;\u003c/em\u003epQTL [GSTO1, IFI16, MAPK9 (SomaScan), IGF2R (SomaScan), IGF2R (Olink), TSSC4, DNAJB14, and ABO] overlapping or in high LD (r\u003csup\u003e2\u003c/sup\u003e \u0026gt; 0.8) with a \u003cem\u003ecis\u0026nbsp;\u003c/em\u003eeQTL in one or more tissues, including eQTL were found in whole blood [MAPK9 (SomaScan) and ABO], adipose tissues (GSTO1, DNAJB14, and ABO), and reproductive tissues (ABO) (Supplementary Table 14).\u003c/p\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eOf the 2,751 unique plasma proteins examined for their association with overall, endometrioid, and non-endometrioid endometrial cancer, we identified 20 pairs of potential causal associations with evidence from UVMR and colocalization. Endometrioid (GSTO1-1, MMP10, IGF2R, and SKAP1) and non-endometrioid (MAPK9, DNAJB14, IFI16, LCN2, and SCT) cancer subtypes were causally associated with different plasma proteins. Distinct pathways were overrepresented in endometrioid (e.g., PDGF signalling and PTEN gene regulation) and non-endometrioid (e.g., non-canonical NF-kB signalling) cancer subtypes. In UVMR, GSTO1-1 and IGF2R were identified as potential mediators for the effect of BMI on endometrioid endometrial cancer risk. In MVMR analyses, we observed little evidence of a mediating role of plasma proteins in the effect of BMI on endometrial cancer risk.\u003c/p\u003e \u003cp\u003eThe MR approach reduces risk of traditional observational confounding and improves causal inference. By using \u003cem\u003ecis\u003c/em\u003e variants for proteins, we were more confident with the exclusion restriction assumption given \u003cem\u003ecis\u003c/em\u003e variants are less likely to have pleiotropic effects.\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e However, by using only a lead \u003cem\u003ecis\u003c/em\u003e variant for each protein, our analysis might have less power to detect an association. Several proteins showed evidence for colocalization but did not reach the multiple testing corrected P-value threshold in UVMR. Colocalization is usually considered the more conservative out of the two methods.\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e Furthermore, multiple testing correction can minimise false positives, but the large number of proteins tested may have led to incorrectly accepting the null hypothesis for some plasma proteins that were weakly associated with endometrial cancer risk. We make these results available for further analysis for interested readers.\u003c/p\u003e \u003cp\u003eWe used sex-combined GWAS data for plasma proteins, as female-specific GWAS data were not available. Emerging evidence suggests genetic effects on plasma protein level do not differ substantially by sex.\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e There was some sample overlap of UK Biobank participants in the adiposity GWAS including GIANT consortium and UK-Biobank participants and the Olink proteins GWAS in a subset of UK-Biobank participants. Given the strong instruments used in UVMR analyses, bias as a result of sample overlap is likely negligible.\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e However, conditional F-statistics for plasma proteins in MVMR analyses were \u0026lt;\u0026thinsp;10, and sample overlap of UK-Biobank participants may further exacerbate weak instruments bias. The role plasma proteins in mediating the effect of adiposity on endometrial cancer risk thus warrants further investigation using alternative methods, such as mediation analysis in traditional observational studies.\u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarises functions of plasma proteins associated with endometrial cancer risk and their possible link with carcinogenesis. The function of some proteins is consistent with carcinogenesis and pathways identified from the enrichment analysis. For instance, SKAP1 expression in whole blood has been identified as a candidate susceptibility gene for endometrioid endometrial cancer in a transcriptome-wide association study and colocalization.\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e MMP10 initiates extracellular matrix remodelling in the endometrium,\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e which was also identified as one of the enriched pathways for endometrioid endometrial cancer. MAPK9 can promote cancer cell survival through crosstalk with other pathways including NF-κB, the top enriched pathways identified for non-endometrioid endometrial cancer.\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eWe did not find any overlap of proteins between the endometrioid and non-endometrioid subtypes, suggesting that circulating proteins causally associated with the two endometrial cancer subtypes are different. This is consistent with the pathway enrichment results, where we identified different pathways for each endometrial cancer subtypes. Top pathways for endometrioid endometrial cancer included PDGF signalling in tumorigenesis, which could be stimulated by oestrogen and inhibited by progesterone,\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e and the established PTEN mutation pathway.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e Top pathways for non-endometrioid endometrial cancer includes non-cardinal NF-κB pathways and chemokine bindings important for immune functions.\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e Mutant p53, present more commonly in non-endometrioid cancer, has been reported to fail in suppressing NF-κB mediated apoptotic signal, leading to overexpression of NF-κB in various tumours.\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e We also observed some common pathways for the endometrioid and non-endometrioid subtypes, including the PI3K/AKT pathway that is often deregulated in endometrial cancer.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e The pathway enrichment analysis warrants replication, as we could only use plasma proteins associated with endometrial cancer risk at P\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e \u003cp\u003eExamining overlaps with eQTL in various tissues allowed us to explore how tissue-specific regulatory processes might relate to plasma protein levels. Only ABO, which is widely expressed in many tissues, had an eQTL in uterine tissue in high LD with pQTL. This suggests the plasma proteins identified in our analyses might not have a direct carcinogenic effect on the uterus. It is also possible that protein profiles in healthy tissues used in GTEx differ from cancerous tissues.\u003c/p\u003e \u003cp\u003eGSTO1-1 and IGF2R were identified as potential mediators for the effect of BMI on endometrioid endometrial cancer in the UVMR analyses, but no evidence of mediation was observed in MVMR analyses. A recent study on circulating anti-GSTO1-1 antibodies reported elevated levels in various inflammatory conditions, suggesting it as a marker for acute and chronic inflammation.\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e GSTO1-1 activates NLRP3 inflammasome and TLR4 mediated release of pro-inflammatory cytokines including IL-6, TNF-α, and IL-1β.\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e These cytokines have been associated with low-grade adiposity-related chronic inflammation, as well as an increased risk of type 2 diabetes due to impaired insulin signalling in adipose tissue.\u003csup\u003e\u003cspan additionalcitationids=\"CR56\" citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e Membrane-bound IGF2R is considered a tumour suppressant by negatively regulates IGF2 activities.\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e In contrast, observational studies have reported elevated circulating IGF2R in individuals with obesity, type 2 diabetes, colorectal cancer, and breast cancer compared with healthy individuals.\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e,\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e,\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e However, the mechanisms linking elevated soluble IGF2R and diseases remain poorly understood.\u003c/p\u003e \u003cp\u003ePrevious MR study on potential molecular mediators for the effect of BMI on endometrial cancer identified fasting insulin, bioavailable testosterone, and SHBG using \u003cem\u003etrans\u003c/em\u003e variants as instruments.\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e We did not examine insulin or testosterone. Our results for SHBG using single \u003cem\u003ecis\u003c/em\u003e variant as instrument suggest an inverse association between BMI-SHBG and SHBG-endometrial cancer at P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 (Supplementary Table\u0026nbsp;5). We additionally observed an inverse association for IGFBP7 with overall and endometrioid endometrial cancer risk at P\u0026thinsp;\u0026lt;\u0026thinsp;0.05, but not for other IGF binding proteins and receptors.\u003c/p\u003e \u003cp\u003eWe identified plasma proteins that might be involved in carcinogenic pathways for endometrioid and non-endometrioid endometrial cancer using complimentary approaches of MR and colocalization. These findings offer potential targets for further mechanistic studies, which could support the development of non-invasive methods to differentiate endometrial cancer subtypes and guide therapeutic strategies for clinical intervention. We identified two potential plasma proteins mediators, GSTO1-1 and IGF2R, for the effect of adiposity on endometrioid endometrial cancer risk in UVMR analysis. Alternative methods to examine effect mediation and future research integrating known pathways through a multiple-mediators mediation analysis could offer a more thorough understanding of how adiposity affects endometrial cancer risk.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eBMI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ebody mass index\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003econfidence interval\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eendometrial cancer\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEEC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eendometrioid endometrial cancer\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eeQTL\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eexpression quantitative trait loci\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGWAS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003egenome-wide association study\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eIGF1\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003einsulin-like growth factor 1\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eIL-6\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003einterleukin 6\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eIVW-MRE\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003einverse variance weighted multiplicative random effects\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003elinkage disequilibrium\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMendelian randomisation\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMVMR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003emultivariable Mendelian randomisation\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNEEC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003enon-endometrioid endometrial cancer\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eOR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eodds ratio\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003epQTL\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eprotein quantitative trait loci\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eSHBG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003esex hormone binding globulin\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eTNF-α\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003etumour necrosis factor-α\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eUVMR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eunivariable Mendelian randomisation\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eWHR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ewaist-hip ratio\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll studies contributing data to the analyses received the relevant institutional review board approval from each country, in accordance with the Declaration of Helsinki. All participants provided informed consent.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe UK Biobank Pharma Proteomics Project GWAS summary data are publicly available from UK Biobank (https://doi.org/10.7303/syn51364943). The GWAS summary data for SomaScan from the deCODE program are publicly available at https://download.decode.is/form/folder/proteomics. The meta-analysed Endometrial Cancer Association Consortium (ECAC) and Epidemiology of Endometrial Cancer Consortium (E2C2) GWAS summary data were obtained directly from the GWAS corresponding author (doi:10.1038/s41467-018-05427-7). They are available upon request from the corresponding author with the permission of the corresponding author of the ECAC GWAS. The combined GIANT consortium and UK-Biobank genome-wide association study (GWAS) summary data are publicly available at https://zenodo.org/records/1251813. The Genotype-Tissue Expression (GTEx) Project was supported by the Common Fund of the Office of the Director of the National Institutes of Health, and by NCI, NHGRI, NHLBI, NIDA, NIMH, and NINDS. The data used for the analyses described in this manuscript were obtained from the GTEx Portal on 16/09/2024.\u003c/p\u003e\n\u003cp\u003eAll analyses were performed using R (Vienna, Austria) version 4.1.2. Key R packages for the statistical analysis include: TwoSampleMR (version 0.6.8), coloc (version 5.2.2)\u003csup\u003e29\u003c/sup\u003e, MVMR (version 0.4)\u003csup\u003e61\u003c/sup\u003e, PathfindR, (version 2.4.1)\u003csup\u003e31\u003c/sup\u003e, and LDlinkR (version 1.4.0.9000)\u003csup\u003e62\u003c/sup\u003e. Statistical analysis codes used for this study are available at sabrinawang113/adiposity-proteins-endometrial (github.com).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFunding for IIG_FULL_2021_008 was obtained from Wereld Kanker Onderzoek Fonds (WKOF), as part of the World Cancer Research Fund International grant programme; Funding for INCA_15849 was obtained from Institut national du cancer (INCa).\u003c/p\u003e\n\u003cp\u003eNJT works within the CRUK Integrative Cancer Epidemiology Programme (C18281/A29019) and with support from CRUK (PRCPJT-May22\\100028).\u003c/p\u003e\n\u003cp\u003eTOM is supported by a National Health and Medical Research Council of Australia Emerging Leader Fellowship (APP1173170).\u003c/p\u003e\n\u003cp\u003eThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributors\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSEW: conceptualization, data curation, formal analysis, methodology, visualization, writing (original draft), writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eVYT: methodology, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eJY: methodology, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eYZ: data curation, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eTOM: conceptualization, funding acquisition, data curation, resources, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eNJT: methodology, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eMJG: conceptualization, funding acquisition, resources, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eLD: conceptualization, funding acquisition, supervision, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003eML: conceptualization, data curation, methodology, visualization, writing (review \u0026amp; editing)\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDisclaimer\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhere authors are identified as personnel of the International Agency for Research on Cancer / World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy, or views of the International Agency for Research on Cancer / World Health Organization.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eLortet-Tieulent J, Ferlay J, Bray F, Jemal A. International Patterns and Trends in Endometrial Cancer Incidence, 1978\u0026ndash;2013. JNCI: Journal of the National Cancer Institute 2017;110:354-61.\u003c/li\u003e\n \u003cli\u003eLauby-Secretan B, Scoccianti C, Loomis D, Grosse Y, Bianchini F, Straif K. Body Fatness and Cancer \u0026mdash; Viewpoint of the IARC Working Group. New England Journal of Medicine 2016;375:794-8.\u003c/li\u003e\n \u003cli\u003eArnold M, Pandeya N, Byrnes G, Renehan AG, Stevens GA, Ezzati M, et al. Global burden of cancer attributable to high body-mass index in 2012: a population-based study. The Lancet Oncology 2015;16:36-46.\u003c/li\u003e\n \u003cli\u003eMakker V, MacKay H, Ray-Coquard I, Levine DA, Westin SN, Aoki D, et al. Endometrial cancer. Nature Reviews Disease Primers 2021;7:88.\u003c/li\u003e\n \u003cli\u003eLukanova A, Lundin E, Micheli A, Arslan A, Ferrari P, Rinaldi S, et al. Circulating levels of sex steroid hormones and risk of endometrial cancer in postmenopausal women. International Journal of Cancer 2004;108:425-32.\u003c/li\u003e\n \u003cli\u003eFerkingstad E, Sulem P, Atlason BA, Sveinbjornsson G, Magnusson MI, Styrmisdottir EL, et al. Large-scale integration of the plasma proteome with genetics and disease. Nature genetics 2021;53:1712-21.\u003c/li\u003e\n \u003cli\u003eDossus L, Becker S, Rinaldi S, Lukanova A, Tj\u0026oslash;nneland A, Olsen A, et al. Tumor necrosis factor (TNF)‐\u0026alpha;, soluble TNF receptors and endometrial cancer risk: The EPIC study. International journal of cancer 2011;129:2032-7.\u003c/li\u003e\n \u003cli\u003eDossus L, Rinaldi S, Becker S, Lukanova A, Tjonneland A, Olsen A, et al. Obesity, inflammatory markers, and endometrial cancer risk: a prospective case\u0026ndash;control study. Endocrine-related cancer 2010;17:1007.\u003c/li\u003e\n \u003cli\u003eWang SE, Viallon V, Lee M, Dimou N, Hamilton F, Biessy C, et al. Circulating inflammatory and immune response proteins and endometrial cancer risk: a nested case-control study and Mendelian randomization analyses. eBioMedicine 2024;108:105341.\u003c/li\u003e\n \u003cli\u003eGunter MJ, Hoover DR, Yu H, Wassertheil-Smoller S, Manson JE, Li J, et al. A Prospective Evaluation of Insulin and Insulin-like Growth Factor-I as Risk Factors for Endometrial Cancer. Cancer Epidemiology, Biomarkers \u0026amp; Prevention 2008;17:921-9.\u003c/li\u003e\n \u003cli\u003ePainter JN, O\u0026apos;Mara TA, Marquart L, Webb PM, Attia J, Medland SE, et al. Genetic Risk Score Mendelian Randomization Shows that Obesity Measured as Body Mass Index, but not Waist:Hip Ratio, Is Causal for Endometrial Cancer. Cancer Epidemiology, Biomarkers \u0026amp; Prevention 2016;25:1503-10.\u003c/li\u003e\n \u003cli\u003eZuber V, Grinberg NF, Gill D, Manipur I, Slob EAW, Patel A, et al. Combining evidence from Mendelian randomization and colocalization: Review and comparison of approaches. Am J Hum Genet 2022;109:767-82.\u003c/li\u003e\n \u003cli\u003eYarmolinsky J, Wade KH, Richmond RC, Langdon RJ, Bull CJ, Tilling KM, et al. Causal Inference in Cancer Epidemiology: What Is the Role of Mendelian Randomization? Cancer Epidemiology, Biomarkers \u0026amp; Prevention 2018;27:995-1010.\u003c/li\u003e\n \u003cli\u003eSanderson E. Multivariable Mendelian Randomization and Mediation. Cold Spring Harb Perspect Med 2021;11.\u003c/li\u003e\n \u003cli\u003eMullee A, Dimou N, Allen N, O\u0026rsquo;Mara T, Gunter MJ, Murphy N. Testosterone, sex hormone-binding globulin, insulin-like growth factor-1 and endometrial cancer risk: observational and Mendelian randomization analyses. British Journal of Cancer 2021;125:1308-17.\u003c/li\u003e\n \u003cli\u003eNead KT, Sharp SJ, Thompson DJ, Painter JN, Savage DB, Semple RK, et al. Evidence of a Causal Association Between Insulinemia and Endometrial Cancer: A Mendelian Randomization Analysis. JNCI: Journal of the National Cancer Institute 2015;107.\u003c/li\u003e\n \u003cli\u003eYarmolinsky J, Robinson JW, Mariosa D, Karhunen V, Huang J, Dimou N, et al. Association between circulating inflammatory markers and adult cancer risk: a Mendelian randomization analysis. eBioMedicine 2024;100.\u003c/li\u003e\n \u003cli\u003eHazelwood E, Sanderson E, Tan VY, Ruth KS, Frayling TM, Dimou N, et al. Identifying molecular mediators of the relationship between body mass index and endometrial cancer risk: a Mendelian randomization analysis. BMC Medicine 2022;20:125.\u003c/li\u003e\n \u003cli\u003eSun BB, Chiou J, Traylor M, Benner C, Hsu Y-H, Richardson TG, et al. Plasma proteomic associations with genetics and health in the UK Biobank. Nature 2023;622:329-38.\u003c/li\u003e\n \u003cli\u003eO\u0026rsquo;Mara TA, Glubb DM, Amant F, Annibali D, Ashton K, Attia J, et al. Identification of nine new susceptibility loci for endometrial cancer. Nature Communications 2018;9:3166.\u003c/li\u003e\n \u003cli\u003ePulit SL, Stoneman C, Morris AP, Wood AR, Glastonbury CA, Tyrrell J, et al. Meta-analysis of genome-wide association studies for body fat distribution in 694 649 individuals of European ancestry. Human molecular genetics 2019;28:166-74.\u003c/li\u003e\n \u003cli\u003ePulit SL, de With SAJ, de Bakker PIW. Resetting the bar: Statistical significance in whole-genome sequencing-based association studies of global populations. Genetic Epidemiology 2017;41:145-51.\u003c/li\u003e\n \u003cli\u003eBowden J, Del Greco M F, Minelli C, Davey Smith G, Sheehan N, Thompson J. A framework for the investigation of pleiotropy in two‐sample summary data Mendelian randomization. Statistics in medicine 2017;36:1783-802.\u003c/li\u003e\n \u003cli\u003eBurgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genetic epidemiology 2013;37:658-65.\u003c/li\u003e\n \u003cli\u003eBurgess S, Bowden J, Fall T, Ingelsson E, Thompson SG. Sensitivity Analyses for Robust Causal Inference from Mendelian Randomization Analyses with Multiple Genetic Variants. Epidemiology 2017;28:30-42.\u003c/li\u003e\n \u003cli\u003eHartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. International Journal of Epidemiology 2017;46:1985-98.\u003c/li\u003e\n \u003cli\u003eZheng J, Richardson TG, Millard LAC, Hemani G, Elsworth BL, Raistrick CA, et al. PhenoSpD: an integrated toolkit for phenotypic correlation estimation and multiple testing correction using GWAS summary statistics. Gigascience 2018;7.\u003c/li\u003e\n \u003cli\u003eGiambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, et al. Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using Summary Statistics. PLOS Genetics 2014;10:e1004383.\u003c/li\u003e\n \u003cli\u003eWallace C. Eliciting priors and relaxing the single causal variant assumption in colocalisation analyses. PLoS Genet 2020;16:e1008720.\u003c/li\u003e\n \u003cli\u003eVanderWeele TJ. Mediation Analysis: A Practitioner\u0026apos;s Guide. Annual Review of Public Health 2016;37:17-32.\u003c/li\u003e\n \u003cli\u003eUlgen E, Ozisik O, Sezerman OU. pathfindR: An R Package for Comprehensive Identification of Enriched Pathways in Omics Data Through Active Subnetworks. Frontiers in Genetics 2019;10.\u003c/li\u003e\n \u003cli\u003eHughes MM, Hooftman A, Angiari S, Tummala P, Zaslona Z, Runtsch MC, et al. Glutathione Transferase Omega-1 Regulates NLRP3 Inflammasome Activation through NEK7 Deglutathionylation. Cell Reports 2019;29:151-61.e5.\u003c/li\u003e\n \u003cli\u003eMenon D, Innes A, Oakley AJ, Dahlstrom JE, Jensen LM, Br\u0026uuml;stle A, et al. GSTO1-1 plays a pro-inflammatory role in models of inflammation, colitis and obesity. Sci Rep 2017;7:17832.\u003c/li\u003e\n \u003cli\u003eKho PF, Wang X, Cu\u0026eacute;llar-Partida G, D\u0026ouml;rk T, Goode EL, Lambrechts D, et al. Multi-tissue transcriptome-wide association study identifies eight candidate genes and tissue-specific gene expression underlying endometrial cancer susceptibility. Communications Biology 2021;4:1211.\u003c/li\u003e\n \u003cli\u003eOsteen KG, Bruner-Tran KL, Keller NR, Eisenberg E. Progesterone-Mediated Endometrial Maturation Limits Matrix Metalloproteinase (MMP) Expression in an Inflammatory-like Environment. Annals of the New York Academy of Sciences 2002;955:37-47.\u003c/li\u003e\n \u003cli\u003eLee Y, Sooranna SR, Terzidou V, Christian M, Brosens J, Huhtinen K, et al. Interactions between inflammatory signals and the progesterone receptor in regulating gene expression in pregnant human uterine myocytes. Journal of Cellular and Molecular Medicine 2012;16:2487-503.\u003c/li\u003e\n \u003cli\u003eHeydari N, Mahdizadeh M, Jafari SM. The evolving landscape of involvement of DTYMK enzymes in cancer. Med Oncol 2023;40:213.\u003c/li\u003e\n \u003cli\u003eNational Center for Biotechnology Information (NCBI). Gene: TSSC4 tumor suppressing subtransferable candidate 4 [Homo sapiens (human)]. NCBI. Accessed July 11, 2024, from https://www.ncbi.nlm.nih.gov/gene/10078.\u003c/li\u003e\n \u003cli\u003eCui H, Qu Y, Zhang L, Zhang W, Yan P, Yang C, et al. Epidemiological and genetic evidence for the relationship between ABO blood group and human cancer. International Journal of Cancer 2023;153:320-30.\u003c/li\u003e\n \u003cli\u003eScott CD, Kiess W. Soluble M6P/IGFIIR in the circulation. Best Practice \u0026amp; Research Clinical Endocrinology \u0026amp; Metabolism 2015;29:723-33.\u003c/li\u003e\n \u003cli\u003eWu Q, Wu W, Fu B, Shi L, Wang X, Kuca K. JNK signaling in cancer cell survival. Medicinal Research Reviews 2019;39:2082-104.\u003c/li\u003e\n \u003cli\u003ePurifica\u0026ccedil;\u0026atilde;o ADd, Debbas V, Tanaka LY, Gabriel GVdM, Wosniak J\u0026uacute;nior J, De Bessa TC, et al. DNAJB12 and DNJB14 are non-redundant Hsp40 redox chaperones involved in endoplasmic reticulum protein reflux. Biochimica et Biophysica Acta (BBA) - General Subjects 2024;1868:130502.\u003c/li\u003e\n \u003cli\u003eNational Center for Biotechnology Information (NCBI). Gene: IFI16 interferon gamma inducible protein 16 [Homo sapiens (human)]. NCBI. Accessed July 11, 2024, from https://www.ncbi.nlm.nih.gov/gene/3428.\u003c/li\u003e\n \u003cli\u003eRodvold JJ, Mahadevan NR, Zanetti M. Lipocalin 2 in cancer: when good immunity goes bad. Cancer letters 2012;316:132-8.\u003c/li\u003e\n \u003cli\u003eMannelqvist M, Stefansson IM, Wik E, Kusonmano K, Raeder MB, \u0026Oslash;yan AM, et al. Lipocalin 2 expression is associated with aggressive features of endometrial cancer. BMC Cancer 2012;12:169.\u003c/li\u003e\n \u003cli\u003eDing W-Q, Cheng Z-J, McElhiney J, Kuntz SM, Miller LJ. Silencing of Secretin Receptor Function by Dimerization with a Misspliced Variant Secretin Receptor in Ductal Pancreatic Adenocarcinoma1. Cancer Research 2002;62:5223-9.\u003c/li\u003e\n \u003cli\u003eYarmolinsky J, Tzoulaki I, Gunter MJ, Travis RC, Davey Smith G, Smith-Byrne K. RE: Exploring the cross-cancer effect of circulating proteins and discovering potential intervention targets for 13 site-specific cancers. JNCI: Journal of the National Cancer Institute 2024;116:764-5.\u003c/li\u003e\n \u003cli\u003eMine K, Eleanor W, Nicola DK, Spiros D, Julia C-Z, Chloe MO, et al. Similar and different: systematic investigation of proteogenomic variation between sexes and its relevance for human diseases. medRxiv 2024:2024.02.16.24302936.\u003c/li\u003e\n \u003cli\u003eSadreev II, Elsworth BL, Mitchell RE, Paternoster L, Sanderson E, Davies NM, et al. Navigating sample overlap, winner\u0026rsquo;s curse and weak instrument bias in Mendelian randomization studies using the UK Biobank. medRxiv 2021:2021.06.28.21259622.\u003c/li\u003e\n \u003cli\u003eGentilini D, Busacca M, Di Francesco S, Vignali M, Vigan\u0026ograve; P, Di Blasio AM. PI3K/Akt And ERK1/2 signalling pathways are involved in endometrial cell migration induced by 17\u0026beta;-estradiol and growth factors. Molecular Human Reproduction 2007;13:317-22.\u003c/li\u003e\n \u003cli\u003eSun S-C. The non-canonical NF-\u0026kappa;B pathway in immunity and inflammation. Nature Reviews Immunology 2017;17:545-58.\u003c/li\u003e\n \u003cli\u003eGulati AP, Yang Y-M, Harter D, Mukhopadhyay A, Aggarwal BA, Benzil DL, et al. Mutant human tumor suppressor p53 modulates the activation of mitogen-activated protein kinase and nuclear factor-\u0026kappa;B, but not c-Jun N-terminal kinase and activated protein-1. Molecular Carcinogenesis 2006;45:26-37.\u003c/li\u003e\n \u003cli\u003ePiaggi S, Lorenzini E, Pratesi F, Migliorini P, Pompella A, Bruschi F, et al. Anti-glutathione S-transferase omega 1-1 (GSTO1-1) antibodies are increased during acute and chronic inflammation in humans. Clin Exp Immunol 2022;209:305-10.\u003c/li\u003e\n \u003cli\u003eMenon D, Innes A, Oakley AJ, Dahlstrom JE, Jensen LM, Br\u0026uuml;stle A, et al. GSTO1-1 plays a pro-inflammatory role in models of inflammation, colitis and obesity. Scientific Reports 2017;7:17832.\u003c/li\u003e\n \u003cli\u003eMaachi M, Pi\u0026eacute;roni L, Bruckert E, Jardel C, Fellahi S, Hainque B, et al. Systemic low-grade inflammation is related to both circulating and adipose tissue TNF\u0026alpha;, leptin and IL-6 levels in obese women. International Journal of Obesity 2004;28:993-7.\u003c/li\u003e\n \u003cli\u003eWei S, Schmidt MD, Dwyer T, Norman RJ, Venn AJ. Obesity and menstrual irregularity: associations with SHBG, testosterone, and insulin. Obesity 2009;17:1070-6.\u003c/li\u003e\n \u003cli\u003eGao D, Madi M, Ding C, Fok M, Steele T, Ford C, et al. Interleukin-1\u0026beta; mediates macrophage-induced impairment of insulin signaling in human primary adipocytes. American Journal of Physiology-Endocrinology and Metabolism 2014;307:E289-E304.\u003c/li\u003e\n \u003cli\u003eLivingstone C. IGF2 and cancer. Endocrine-Related Cancer 2013;20:R321-R39.\u003c/li\u003e\n \u003cli\u003eChanprasertyothin S, Jongjaroenprasert W, Ongphiphadhanakul B. The Association of Soluble IGF2R and IGF2R Gene Polymorphism with Type 2 Diabetes. Journal of Diabetes Research 2015;2015:216383.\u003c/li\u003e\n \u003cli\u003eJeyaratnaganthan N, H\u0026oslash;jlund K, Kroustrup JP, Larsen JF, Bjerre M, Levin K, et al. Circulating levels of insulin-like growth factor-II/mannose-6-phosphate receptor in obesity and type 2 diabetes. Growth Hormone \u0026amp; IGF Research 2010;20:185-91.\u003c/li\u003e\n \u003cli\u003eSanderson E, Spiller W, Bowden J. Testing and correcting for weak and pleiotropic instruments in two-sample multivariable Mendelian randomization. Statistics in medicine 2021;40:5434-52.\u003c/li\u003e\n \u003cli\u003eMachiela MJ, Chanock SJ. LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics 2015;31:3555-7.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Endometrial cancer, plasma proteomics, adiposity, Mendelian randomisation, colocalization","lastPublishedDoi":"10.21203/rs.3.rs-5815826/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5815826/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Proteomics could enhance our understanding of endometrial carcinogenesis. However, addressing confounding in traditional observational studies remains challenging, especially given the strong impact of adiposity on the plasma proteome and endometrial cancer risk. The role of circulating proteins in mediating adiposity’s effect on endometrial cancer risk is also not fully elucidated.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e Using Mendelian randomization (MR) and colocalization analyses, we examined the causal association between 2,751 unique proteins from UK Biobank (N Olink proteins=2,031; N=52,363) and deCODE (N SomaScan proteins=1,667; N=35,559) and endometrial cancer risk [overall (N cases=12,270; N controls=46,126), endometrioid (N cases=8,758), and non-endometrioid (N cases=1,230) in the meta-analysed Endometrial Cancer Association Consortium and Epidemiology of Endometrial Cancer Consortium data]. We performed enrichment analyses to explore pathways overrepresented among plasma proteins in endometrioid and non-endometrioid cancer subtypes. Additionally, we assessed the role of circulating proteins in mediating the effect of body mass index (BMI) on endometrial cancer risk using univariable and multivariable MR.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eWe identified 20 associations between circulating proteins and endometrial cancer risk in MR and colocalization analyses. GSTO1-1 and SKAP1 were positively associated and MMP10 was negatively associated with both overall and endometrioid endometrial cancer; DTYMK and ABO were positively associated and TSSC4 was negatively associated with overall endometrial cancer; IGF2R was positively associated with endometrioid cancer; MAPK9 was positively associated and DNAJB14, IFI16, LCN2, and SCT were negatively associated with non-endometrioid endometrial cancer. Distinct pathways were overrepresented in endometrioid (e.g., PDGF signalling and PTEN gene regulation) and non-endometrioid (e.g., non-canonical NF-kB signalling) cancer subtypes. GSTO1-1 and IGF2R were identified as potential mediators for the effect of BMI on endometrioid cancer risk in univariable MR, but evidence for mediation was not observed in multivariable MR analyses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e We observed distinct plasma proteins and pathways associated with endometrioid and non-endometrioid endometrial cancers. These findings highlight candidate proteins for further mechanistic investigations, which could support the development of non-invasive methods to differentiate endometrial cancer subtypes and guide clinical intervention strategies. There was limited evidence that the effect of adiposity on endometrial cancer risk was mediated by circulating proteins examined in our study.\u003c/p\u003e","manuscriptTitle":"The effect of circulating proteins and their role in mediating adiposity’s effect on endometrioid and non-endometrioid endometrial cancer risk: Mendelian randomisation and colocalization analyses ","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-04 08:47:18","doi":"10.21203/rs.3.rs-5815826/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"406a7722-8018-4f48-9019-feab8850f9f7","owner":[],"postedDate":"February 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-02-04T08:47:18+00:00","versionOfRecord":[],"versionCreatedAt":"2025-02-04 08:47:18","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5815826","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5815826","identity":"rs-5815826","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}