Cyclin C-CDK3/8/19 kinases play a tumor-suppressive role in vivo

Cyclin C was cloned as a growth-promoting G1 cyclin, and several studies postulated a role for cyclin C in driving cell proliferation. Moreover, cyclin C, together with its kinase partner, the cyclin-dependent kinase CDK8, is believed to represent an essential component of basal transcriptional machinery where it globally represses gene expression. However, the function of cyclin C in vivo has never been addressed. Here we show that in the living organism cyclin C acts as a haploinsufficient tumor suppressor, through its function of controlling Notch1 oncogene levels. Cyclin C activates an “orphan” CDK19 kinase, as well as CDK8 and CDK3. These cyclin C-CDK complexes phosphorylate Notch1 intracellular domain (ICN1), which allows binding of ICN1 to Fbw7 and triggers ICN1 polyubiquitination. Genetic ablation of cyclin C blocks ICN1 phosphorylation, disrupts Fbw7 binding, and decreases ICN1 ubiquitination in vivo, thereby strongly elevating ICN1 levels in several compartments of cyclin C knockout mice. Ablation of cyclin C, or cyclin C heterozygosity collaborate with other oncogenic lesions and accelerate development of T-cell acute lymphoblastic leukemia (T-ALL) in cyclin Cdeficient mice. Furthermore, the locus encoding cyclin C is heterozygously deleted in a significant fraction of human T-ALL, and these tumors express reduced cyclin C levels. In addition, we describe point mutations in human T-ALL tumors that render cyclin C-CDK unable to phosphorylate ICN1. These studies reveal that in sharp contrast to all other cyclin proteins, cyclin C functions as a growth-suppressor in vivo, and suggest that human tumor cells develop different strategies to evade cyclin C inhibitory function. Comparison of wild-type mouse embryonic fibroblasts (n=3 biological replicates) versus cyclin C knockout MEFs (n=3), wild-type mouse embryonic stem cells (n=3) versus cyclin C knockout ESC (n=3), wild-type mouse embryonic brain (n=3) versus cyclin C knockout embryonic brain (n=3)细胞周期蛋白C（Cyclin C）最初作为促生长型G1周期蛋白被克隆，多项研究曾推测其在驱动细胞增殖中发挥作用。此外，细胞周期蛋白C与其激酶伴侣细胞周期蛋白依赖性激酶CDK8（cyclin-dependent kinase CDK8）共同被认为是基础转录机器的核心组分，可全局性抑制基因表达。然而，此前尚无研究阐明细胞周期蛋白C在活体中的功能。本研究证实，在活体生物中，细胞周期蛋白C通过调控Notch1癌基因的表达水平，发挥单倍剂量不足型肿瘤抑制因子的功能。细胞周期蛋白C可激活“孤儿”激酶CDK19，同时也能激活CDK8与CDK3。这些细胞周期蛋白C-CDK复合物可磷酸化Notch1胞内结构域（ICN1），促使ICN1与Fbw7结合并启动ICN1的多泛素化过程。细胞周期蛋白C的基因敲除会在活体中阻断ICN1的磷酸化、破坏Fbw7与ICN1的结合并降低ICN1的泛素化水平，从而显著提升Cyclin C敲除小鼠多个组织分区内的ICN1水平。细胞周期蛋白C的基因敲除或单等位基因缺失，可与其他致癌损伤协同作用，加速Cyclin C缺陷小鼠体内T细胞急性淋巴细胞白血病（T-cell acute lymphoblastic leukemia, T-ALL）的发生发展。此外，在相当比例的人类T-ALL病例中，编码细胞周期蛋白C的基因位点发生了单等位基因缺失，且此类肿瘤中的细胞周期蛋白C表达水平显著降低。本研究还发现，人类T-ALL肿瘤中存在的点突变可使细胞周期蛋白C-CDK复合物丧失磷酸化ICN1的能力。本研究证实，与所有其他周期蛋白截然不同，细胞周期蛋白C在活体中发挥生长抑制因子的功能，同时提示人类肿瘤细胞可通过多种途径逃逸细胞周期蛋白C的抑癌作用。本研究对野生型小鼠胚胎成纤维细胞（mouse embryonic fibroblasts, MEFs，n=3生物学重复）与细胞周期蛋白C敲除小鼠胚胎成纤维细胞（MEFs，n=3）、野生型小鼠胚胎干细胞（embryonic stem cells, ESCs，n=3）与细胞周期蛋白C敲除小鼠胚胎干细胞（ESCs，n=3）、野生型小鼠胚胎脑组织（n=3）与细胞周期蛋白C敲除小鼠胚胎脑组织（n=3）进行了对比分析。