Department of Dermatology, Venereology and Allergology, University of Leipzig, Philipp-Rosenthal-Str. 23-25, 04103, Leipzig, Germany.
Interdisciplinary Centre for Bioinformatics, University of Leipzig, Härtelstrasse 16-18, 04107, Leipzig, Germany.
Oncogene. 2018 Nov;37(47):6136-6151. doi: 10.1038/s41388-018-0385-y. Epub 2018 Jul 11.
Recent studies revealed trajectories of mutational events in early melanomagenesis, but the accompanying changes in gene expression are far less understood. Therefore, we performed a comprehensive RNA-seq analysis of laser-microdissected melanocytic nevi (n = 23) and primary melanoma samples (n = 57) and characterized the molecular mechanisms of early melanoma development. Using self-organizing maps, unsupervised clustering, and analysis of pseudotime (PT) dynamics to identify evolutionary trajectories, we describe here two transcriptomic types of melanocytic nevi (N1 and N2) and primary melanomas (M1 and M2). N1/M1 lesions are characterized by pigmentation-type and MITF gene signatures, and a high prevalence of NRAS mutations in M1 melanomas. N2/M2 lesions are characterized by inflammatory-type and AXL gene signatures with an equal distribution of wild-type and mutated BRAF and low prevalence of NRAS mutations in M2 melanomas. Interestingly, N1 nevi and M1 melanomas and N2 nevi and M2 melanomas, respectively, cluster together, but there is no clustering in a stage-dependent manner. Transcriptional signatures of M1 melanomas harbor signatures of BRAF/MEK inhibitor resistance and M2 melanomas harbor signatures of anti-PD-1 antibody treatment resistance. Pseudotime dynamics of nevus and melanoma samples are suggestive for a switch-like immune-escape mechanism in melanoma development with downregulation of immune genes paralleled by an increasing expression of a cell cycle signature in late-stage melanomas. Taken together, the transcriptome analysis identifies gene signatures and mechanisms underlying development of melanoma in early and late stages with relevance for diagnostics and therapy.
最近的研究揭示了早期黑色素瘤发生过程中的突变事件轨迹，但伴随的基因表达变化却知之甚少。因此，我们对激光显微切割的黑素细胞痣（n=23）和原发性黑素瘤样本（n=57）进行了全面的 RNA-seq 分析，并对早期黑色素瘤发展的分子机制进行了描述。使用自组织图、无监督聚类和伪时间（PT）动力学分析来识别进化轨迹，我们在这里描述了两种黑素细胞痣（N1 和 N2）和原发性黑素瘤（M1 和 M2）的转录组类型。N1/M1 病变的特征是色素型和 MITF 基因特征，并且 M1 黑素瘤中 NRAS 突变的发生率很高。N2/M2 病变的特征是炎症型和 AXL 基因特征，BRAF 野生型和突变型分布均匀，M2 黑素瘤中 NRAS 突变的发生率较低。有趣的是，N1 痣和 M1 黑素瘤以及 N2 痣和 M2 黑素瘤分别聚类在一起，但没有按阶段聚类。M1 黑素瘤的转录特征具有 BRAF/MEK 抑制剂耐药的特征，M2 黑素瘤具有抗 PD-1 抗体治疗耐药的特征。痣和黑素瘤样本的伪时间动力学提示黑色素瘤发展中存在类似于免疫逃避的机制，免疫基因下调，晚期黑素瘤中细胞周期特征的表达增加。总之，转录组分析确定了早期和晚期黑色素瘤发展的基因特征和机制，对诊断和治疗具有重要意义。
