肿瘤中的非整倍体：83,862例肿瘤的单细胞数据分析

Aneuploidy in neoplasia: Single-cell data on 83,862 tumors

DOI 原文链接

用sci-hub下载

INTERNATIONAL JOURNAL OF CANCER

影响因子:4.7

分区:医学2区 / 肿瘤学2区

发表日期:2025 Jan 01

作者: Fredrik Mertens, Jakob Hofvander, Nils Mandahl, Felix Mitelman

DOI: 10.1002/ijc.35163

摘要

染色体非整倍体，即数目染色体异常，是癌症的分子标志之一。然而，当采用测序和芯片阵列方法研究肿瘤时，染色体数目和倍性状态通常是从整体DNA数据推断得出的。此外，已发表的关于肿瘤相关非整倍体的分子估计常常也包括由各种结构重排引起的基因组不平衡，这些可能由除数目染色体异常之外的其他机制引起。因此，我们利用来自83,862个肿瘤的单细胞细胞遗传学数据分析染色体数目，显示良恶性肿瘤在偏离正常二倍体状态方面高度异质。我们关注112种特定肿瘤类型的染色体数目，这些类型由精确的形态学诊断和器官位置定义，且数据样本≥50例，发现两大簇：一类以近二倍体肿瘤为主，另一类为具有广泛非整倍体和一种或多种全基因组倍增的肿瘤。前者包括大多数良性实体瘤、髓系肿瘤和与基因融合相关的恶性实体瘤，而后者主要是恶性实体瘤和淋巴瘤。在16种恶性肿瘤类型中，染色体数目的分布与TCGA的倍性水平数据进行了比较。细胞遗传学和分子数据相关性良好，但前者显示出更高的克隆异质性。研究结果提示某些肿瘤类型具有共同的发病机制，并为分子分析提供参考。

Abstract

Chromosomal aneuploidy, that is, numerical chromosome aberrations, is one of the molecular hallmarks of cancer. However, when neoplasms are studied with sequencing- and array-based approaches, chromosome numbers and ploidy states are typically inferred from bulk DNA data. Furthermore, published molecular estimates of neoplasia-associated aneuploidy often also include genomic imbalances resulting from various types of structural rearrangement, which likely result from other mechanisms than numerical chromosome aberrations. We thus analyzed chromosome numbers using single-cell cytogenetic data from 83,862 tumors, and show that both benign and malignant tumors are highly heterogeneous with regard to deviations from the normal, diploid state. Focusing on the chromosome numbers in 112 specific tumor types, defined by both exact morphologic diagnosis and organ location and from which data from ≥50 cases were available, we found two major clusters: one predominated by near-diploid neoplasms and one by neoplasms with extensive aneuploidy and one or more whole genome doublings. The former cluster included most benign solid tumors, myeloid neoplasms, and malignant gene fusion-associated solid tumors, whereas the latter was predominated by malignant solid tumors and lymphomas. For 16 malignant tumor types, the distribution of chromosome numbers could be compared to TCGA ploidy level data. Cytogenetic and molecular data correlated well, but the former indicates a higher level of clonal heterogeneity. The results presented here suggest shared pathogenetic mechanisms in certain tumor types and provide a reference for molecular analyses.