PO.CH02.01 · 化学
在两个大型前瞻性队列中与早发性结直肠癌相关的诊断前血浆蛋白质组学特征
Pre-diagnostic plasma proteomic signatures associated with early-onset colorectal cancer in two large prospective cohorts
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
目的：早发性结直肠癌（EOCRC）在全球范围内一直在上升，但其分子通路和机制尚未得到充分探索。大规模蛋白质组学研究表明，循环蛋白可以阐明关键生物学通路和癌变的早期标志物，但关于EOCRC特异性诊断前蛋白生物标志物的可靠证据仍然稀缺。
设计：我们使用Olink Explore平台在两项前瞻性研究——护士健康研究II（NHSII，1989-2015；43对；98.8%为白人）和南方社区队列研究（SCCS，2002-2015；109对；78.9%为黑人）——的EOCRC（诊断时年龄<55岁）病例和匹配对照中测量了3072种诊断前血浆蛋白，同时在NHSII中测量了59对晚期腺瘤（年龄<50岁）。对照按采血时的年龄和年份、性别和种族进行匹配。使用队列特异性多变量logistic回归模型估计每种蛋白每标准差（SD）增加的比值比（OR）和95%置信区间（CI）。逆方差加权固定效应meta分析估计合并OR。我们还生成了基于血浆蛋白的器官特异性生物学衰老估计，以评估病例与对照之间加速衰老的差异。
结果：NHSII中采血时的平均（SD）年龄为43.7（4.7）岁，SCCS中为45.5（3.6）岁。Meta分析识别出13种与EOCRC相关且符合复制标准（效应方向一致，I² ≤ 40%，FDR < 0.25）的蛋白。复制度最高的蛋白涉及免疫调节（IL7 [每SD增加OR：1.67；95% CI 1.25-2.25]；ANK2 [1.35；1.02-1.80]；PTX3 [1.33；1.03-1.72]）、细胞结构通路（PRSS53 [1.31；1.00-1.71]；FSTL1 [0.67；0.50-0.92]；MMP13 [0.75；0.56-0.99]）和代谢通路（GHRL [1.33；1.02-1.74]；GIPR [0.73；0.55-0.96]）。队列特异性分析揭示了额外的独特蛋白关联。在NHSII中，TLR4、VEGFA、PGLYRP2、PXDNL、KRT17和HRC与EOCRC风险相关，提示宿主-微生物组相互作用紊乱、血管生成信号和上皮生物学改变可能是潜在通路。部分关联（KRT17和HRC）在晚期腺瘤与对照的比较中也被观察到，突显了它们在癌前病变起始中的潜在作用。在SCCS中，与EOCRC风险正相关的蛋白提示基质和细胞骨架重塑改变（AFAP1、CORO6、ARAF）和免疫激活（WAS）。器官特异性衰老分析提示EOCRC病例中多个器官（包括大脑、免疫系统和肠道）的加速衰老水平更高。
结论：在这项前瞻性、多队列蛋白质组学分析研究中，我们识别了与EOCRC风险相关的诊断前循环蛋白，突显免疫和炎症信号以及代谢失调为关键生物学通路。
查看英文原文 English abstract
Objective: Early-onset colorectal cancer (EOCRC) has been rising globally, yet its molecular pathways and mechanisms have not been fully explored. Large-scale proteomic studies suggest circulating proteins can illuminate key biological pathways and early markers of carcinogenesis, but robust evidence on pre-diagnostic protein biomarkers specific to EOCRC is still scarce.
Design: We measured 3,072 pre-diagnostic plasma proteins using the Olink Explore platform in EOCRC (<age 55 at diagnosis) cases and matched controls, in two prospective studies, the Nurses' Health Study II (NHSII, 1989-2015; 43 pairs; 98.8% White) and Southern Community Cohort Study (SCCS, 2002-2015; 109 pairs; 78.9% Black), as well as 59 pairs of advanced adenoma (<age 50) in NHSII. Controls were matched on age and year at blood draw, sex, and race. Cohort-specific multivariable logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) per standard deviation (SD) increase by protein. Inverse variance-weighted fixed-effect meta-analyses estimated pooled ORs. We additionally generated plasma protein-based estimates of organ-specific biological aging to assess differences in accelerated aging between cases and controls.
Results: The mean (SD) age at blood draw was 43.7 (4.7) years in NHSII and 45.5 (3.6) years in SCCS. Meta-analysis identified 13 proteins associated with EOCRC that met replication criteria (consistent effect direction, I² ≤ 40%, FDR < 0.25). Top replicated proteins implicate immune regulation (IL7 [OR per SD increase: 1.67; 95% CI 1.25-2.25; ANK2 [1.35; 1.02-1.80]; PTX3 [1.33; 1.03-1.72]), cellular structural pathways (PRSS53 [1.31; 1.00-1.71]; FSTL1 [0.67; 0.50-0.92]; MMP13 [0.75; 0.56-0.99]), and metabolic pathways (GHRL [1.33; 1.02-1.74]; GIPR [0.73; 0.55-0.96]). Cohort-specific analyses revealed additional distinct protein associations. In NHSII, TLR4, VEGFA, PGLYRP2, PXDNL, KRT17, and HRC were associated with EOCRC risk, implicating disrupted host-microbiome interactions, angiogenic signaling and altered epithelial biology as potential pathways. Some associations (KRT17 and HRC) were also observed in advanced adenoma compared with controls, highlighting their potential roles in initiation of precancerous changes. In SCCS, proteins positively associated with EOCRC risk suggested alterations in matrix and cytoskeletal remodeling (AFAP1, CORO6, ARAF) and immune activation (WAS). Organ-specific aging analyses suggested higher accelerated aging levels in EOCRC cases for multiple organs, including brain, the immune system, and intestine.
Conclusion: In this prospective, multi-cohort proteomic profiling study, we identified pre-diagnostic circulating proteins associated with EOCRC risk, highlighting immune and inflammatory signaling, and metabolic dysregulation as key biological pathways.
利益披露 Disclosure
M. Shi, None..
X. Zong, None..
X. Zhang, None..
Y. Sun, None..
G. Yang, None..
W. Zheng, None.