PO.PS01.06 · 人群科学
整合组学数据揭示饮食相关炎症与结直肠癌风险之间的潜在机制
Integrating omics data unravels potential mechanisms linking diet-related inflammation and colorectal cancer risk
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
尽管饮食相关炎症已被与结直肠癌（CRC）风险联系起来，但这一关联背后的生物学机制仍不明确。我们在一个前瞻性队列中对饮食、临床和多组学数据进行了整合分析，以研究连接饮食相关炎症与CRC风险的潜在分子通路。我们使用能量调整膳食炎症指数（EDII）作为203,770名至少有一次24小时膳食回顾的UK Biobank参与者饮食相关炎症的替代指标。该指数已在先前研究中得到广泛验证，我们在本研究中进一步考察了其与炎症蛋白白介素-6（IL6）、白介素-18（IL18）、TNF受体超家族成员1B（TNFRSF1B）和脂联素的关联，并校正了年龄、性别、BMI、吸烟和非甾体抗炎药使用。我们使用稳定性选择和交叉验证弹性网络建模推导出EDII的代谢（mEDII）和蛋白质组（pEDII）特征，并使用回归模型考察它们与CRC风险的关联，校正上述混杂因素。通过多性状GWAS分析识别这些特征的遗传决定因素，并通过Ensembl的变异效应预测器（Variant Effect Predictor）对信号进行优先排序。我们使用共定位分析探索潜在的共享变异，整合mEDII和pEDII的GWAS数据、结肠和内脏脂肪组织中的基因表达（GTEx v8）以及CRC风险（185,616名参与者，包括78,473例CRC病例）。EDII与血浆IL-6（每增加1个SD的beta = 0.039；95%CI：0.025, 0.052）、IL-18（0.060；0.046, 0.075）和TNFRSF1B（0.057；0.043, 0.072）呈正相关，与脂联素呈负相关（-0.025；-0.037, -0.013）。mEDII包含26种代谢物，pEDII包含21种蛋白质，各自解释了约11.5%的EDII方差。在中位随访9.8年期间，发生1,427例CRC病例（42.9%为女性），其中934例（65.4%）位于结肠。在男性中，EDII与结肠癌风险之间存在提示性正相关（每增加1个SD的OR = 1.11；95%CI：1.01, 1.22）。mEDII在两性中均与结肠癌风险呈正相关，独立于EDII（每增加1个SD的OR = 1.14；95%CI：1.01, 1.29）。GWAS为mEDII和pEDII识别出19个位点，对102个基因具有潜在功能影响。我们发现有证据表明两种特征、CRC风险和结肠组织中的基因表达存在共享的遗传基础［位于脂肪酸去饱和酶1（FADS1）、Ras相互作用蛋白1（RASIP1）和MEF2激活基序及SAP结构域含转录调节因子（MAMSTR）区域］，以及在内脏脂肪中［位于FADS1和跨膜蛋白258（TMEM258）区域］。饮食相关炎症可能通过涉及脂质代谢、内皮功能和肠道炎症的机制促进CRC风险。
查看英文原文 English abstract
Although diet-related inflammation has been linked to colorectal cancer (CRC) risk, the biological mechanisms underlying this association remain unclear. We conducted integrative analyses of diet, clinical, and multi-omics data in a prospective cohort to investigate potential molecular pathways linking diet-related inflammation to CRC risk. We used the Energy-adjusted Dietary Inflammatory Index (EDII) to proxy diet-related inflammation in 203,770 UK Biobank participants with at least one 24-hour dietary recall. The index has been extensively validated in previous studies, and we further examined its associations with inflammatory proteins interleukin-6 (IL6), interleukin-18 (IL18), TNF receptor superfamily member 1B (TNFRSF1B), and adiponectin in our study, adjusting for age, sex, BMI, smoking, and non-steroidal anti-inflammatory drug use. We derived metabolic (mEDII) and proteomic (pEDII) signatures of the EDII using stability selection and cross-validated elastic net modelling, and examined their associations with CRC risk using regression models, adjusting for the aforementioned confounders. Genetic determinants of the signatures were identified through multi-trait GWAS analyses, and signals were prioritized via Ensembl's Variant Effect Predictor. We explored potentially shared variants using colocalization analyses, integrating GWAS data for mEDII and pEDII, gene expression in colon and visceral fat tissue (GTEx v8), and CRC risk (185,616 participants, including 78,473 CRC cases). The EDII showed positive associations with plasma IL-6 (beta per SD increase = 0.039; 95%CI: 0.025, 0.052), IL-18 (0.060; 0.046, 0.075), and TNFRSF1B (0.057; 0.043, 0.072), and an inverse association with adiponectin (-0.025; -0.037, -0.013). The mEDII comprised 26 metabolites and the pEDII comprised 21 proteins, each explaining approximately 11.5% of EDII variance. Over a median follow-up of 9.8 years, 1,427 CRC cases occurred (42.9% in women), of which 934 (65.4%) in the colon. There was a suggestive positive association between EDII and colon cancer risk in men (OR per SD increase = 1.11; 95%CI: 1.01, 1.22). The mEDII was positively associated with colon cancer risk across both sexes, independently of EDII (OR per SD increase = 1.14; 95%CI: 1.01, 1.29). GWAS identified 19 loci for mEDII and pEDII, with potential functional consequences for 102 genes. We found evidence for a shared genetic basis across both signatures, CRC risk, and gene expression in colon tissue [in the Fatty Acid Desaturase 1 ( FADS1 ), Ras Interacting Protein 1 ( RASIP1 ), and MEF2 Activating Motif and SAP Domain Containing Transcriptional Regulator ( MAMSTR ) regions], and in visceral fat [in the FADS1 and Transmembrane Protein 258 ( TMEM258 ) regions]. Diet-related inflammation may contribute to CRC risk through mechanisms involving lipid metabolism, endothelial function, and intestinal inflammation.
利益披露 Disclosure
E. Bouras, None..
J. Li, None..
F. J. B. Van Duijnhoven, None..
U. Peters, None..
E. L. Giovannucci, None..
K. K. Tsilidis, None..
M. J. Gunter, None.