食管鳞状细胞癌（Esophageal squamous cell carcinoma，ESCC）是最具侵袭性的鳞状细胞癌之一，排全世界癌症相关死亡原因的第六位。我国是ESCC的高发区，每年大约有30万新发病例，死亡率排全国癌症的第四位。由于缺乏特异性靶向治疗和精准医疗方法，ESCC患者的预后较差。

甲基苄基亚硝胺(NMBzA)可特异性地诱导食管上皮发生癌变，利用NMBzA构建的大鼠食管癌模型是ESCC病因学和预防研究常用的动物模型，在组织病理学上与人类ESCC经历相似的几个阶段，包括从正常上皮-增生-乳头状瘤-中到重度不典型增生，最终进展到鳞癌。但这一动物模型存在周期长，诱癌率低等缺点。本研究在应用NMBzA作为起始致癌剂的基础上，增加促癌剂索拉非尼的使用，最终建立周期短，成功率高的大鼠食管癌动物模型。

目前尚不清楚NMBzA诱导的大鼠食管癌变是否与人类ESCC的基因组改变相同。在本研究中，我们对这一模型中的乳头瘤和鳞癌两个阶段的样本进行了全基因组测序分析。我们发现，大鼠食管肿瘤存在RAS的高频突变，而RAS在人类ESCC中并不常见。不过除此之外，大鼠食管肿瘤的突变特征与人类ESCC高度相似，都是以C>T的突变为主，并且发现人类ESCC中的一些其它显著突变基因比如TP53、NOTCH家族、FAT家族、PIK3CA、CREBBP、PLEC等，在大鼠食管肿瘤中同样存在。大鼠食管乳头瘤与鳞癌在基因突变和突变负荷上没有显著差异，不过鳞癌表现出更多的拷贝数变异（Copy number variation，CNV）。这些结果为将来在大鼠食管癌模型中进行功能性研究提供了基因组背景基础。

利用我们构建的大鼠食管癌动物模型，本研究在体外培养中建立了大鼠食管肿瘤，包括乳头状瘤和鳞癌组织的永生化细胞，并且成功利用这些细胞建立了3D类器官培养体系。通过体外的2D和3D培养体系，我们发现良性的乳头瘤细胞与恶性的鳞癌细胞相比保留了较多正常上皮细胞的特征，比如无法在免疫缺陷的小鼠体内形成肿瘤，细胞核型正常，形成的类器官在结构形态上与正常食管上皮细胞形成的类器官相似等。

FAT家族作为大鼠和人类ESCC中最为显著突变的基因之一，其失去功能（Loss of function，LOF）的突变可以激活Hippo-YAP通路。我们通过转录组测序分析发现乳头瘤和鳞癌相较于正常大鼠食管上皮均存在YAP1信号的激活，而且从乳头瘤进展到癌伴随着YAP1的进一步活化。我们通过siRNA或YAP1抑制剂抑制大鼠食管肿瘤细胞的YAP1活性，可以抑制大鼠ESCC在体内外的生长和类器官形成。综上，我们的研究发现YAP1的激活可能是ESCC的强力驱动因素，而针对YAP1可以成为治疗ESCC的一个新的策略。

Esophageal squamous cell carcinoma（ESCC）is one of the most malignant squamous cell carcinomas and the sixth leading cause of cancer-related mortality worldwide. ESCC is highly prevalent in China, with an estimated 300,000 new cases per year, and the mortality rate ranks the fourth among all cancers. Due to the lack of targeted therapy and precision medicines, the prognosis of ESCC patients is poor.

N-nitrosomethylbenzylamine (NMBzA) can specifically induce carcinogenesis of esophageal epithelium. The NMBzA-induced rat ESCC model is a commonly used animal model in the study of etiology and prevention of ESCC. The histopathology of tumor induced by NMBzA in the rat esophagus are similar to human ESCC origin, with stages from normal epithelium, hyperplasia, papilloma, dysplasia to squamous cell carcinoma. However, this method has long carcinogenesis cycle and low rate. So, in this study, we use NMBzA as the initial carcinogen and sorafenib as a promoting agent to establish a multiple-stage chemical carcinogenesis model which has short cycle and high success rate. It remains unclear whether NMBzA-induced esophageal carcinogenesis in rats is driven by the same spectrum of genomic alterations found in humans. In this study, whole-genome sequencing (WGS) was performed on samples from papilloma and squamous cell carcinoma in this model. We find that rat esophageal tumors present recurrent mutations in RAS, which is not common in human ESCC. The rat ESCC mutational landscape is similar to that of human ESCC, with C>T mutations dominate the spectra. Some significantly mutated genes in human ESCC, such as TP53、Notch、FAT、PIK3CA、 CREBBP、PLEC, etc., are also present in rat esophageal tumors. There are no significant differences in mutated genes and mutation burden between papilloma and squamous cell carcinoma, but squamous cell carcinoma present substantial copy number variation (CNV). These results provide a genomic background for future functional studies in rat ESCC model.

 Based on this model, immortalized cells from rat esophageal tumors, including papilloma and squamous cell carcinoma, were established in vitro. We also successfully establish a 3D organoids culture system with these cells. Through 2D and 3D culture system in vitro, we find that benign papilloma cells retain more characteristics of normal epithelial cells than malignant squamous cell carcinoma cells. For example, they could not form tumor in immunodeficient mice, the karyotype of cells is normal, and the organoids formed are similar to those formed by normal esophageal epithelial cells in structure and morphology.

As one of the most significant mutant genes in rat and human ESCC, the loss of function (LOF) mutations of FAT family can activate the Hippo-YAP pathway. Using RNA sequencing, we uncover that both papilloma and squamous cell carcinoma showing activation of YAP1 signal compared with normal rat esophageal epithelium, and the progression from papilloma to carcinoma is accompanied by further activation of YAP1. Strikingly, siRNA or pharmacological inhibition of YAP1 can inhibit the growth and organoid formation of rat ESCC cells in vitro and vivo. In conclusion, our study find that YAP1 activation may be a potent driver of ESCC, and targeting YAP1 can be a new strategy for the treatment of ESCC.