Development of a patient-derived tumoroid culture system to explore novel medical treatments for refractory prolactinomas

摘要 催乳素瘤（PRL瘤）是最常见的分泌型垂体瘤， 性腺功能减退和生育能力低下。虽然多巴胺受体激动剂治疗最初效果良好，1/3的患者 治疗2年后出现长期缓解，1/3的患者需要长期D2激动剂治疗， 维持正常的PRL水平，另外1/3的患者对DA治疗无效或不耐受。 副作用.虽然手术切除可以提供在一些，缓解率在大的局部浸润性 肿瘤<50%。由于缺乏任何人类PRL瘤细胞模型，转化研究受到阻碍。 使用目前的方法，人PRL瘤培养物在体外仅存活7-10天。目前迫切 对建立保持其高度分化的人PRL分泌肿瘤模型的未满足的需求 表型，并允许足够的长期细胞扩增性用于临床前转化研究， 开发新的治疗方案。我们连续比较了全球转录组的格局， 使用批量RNA-seq从两个人泌乳素分泌型垂体瘤原代培养物中传代， 了解导致体外人催乳素瘤期间激素分泌丧失的分子事件 文化由于垂体瘤细胞失去激素分泌，我们观察到血管生成减少，存活率降低， 信号和免疫反应与增加的胶原蛋白catalysts，细胞粘附和细胞外 矩阵组织在这些发现的指导下，我们开发了一种独特的三维（3D）PRL瘤培养方法， 系统和第一次，我们已经产生了长期（> 6个月）可扩展的PRL分泌的3D人体 垂体瘤培养（>6个月）。 在两个目标中的第一个，我们将使用我们独特的患者来源的3D PRL瘤培养模型， 通量筛选以鉴定PRL分泌和增殖的小分子抑制剂。我们已经 证明了这种方法的可行性，并在试点屏幕中确定了一种感兴趣的化合物， 我希望进行一项扩展的HTS，以确定难治性PRL瘤患者的新疗法。我们 初步筛选将包括药理学验证和再利用-，靶向-，铅样-和多样化- 图书馆.初始命中将使用强大的z得分统计进行选择，并过渡到aim 2，以进一步 发展目标2将采用一系列后续试验来验证潜在的命中化合物。这些将 包括大鼠和人催乳素瘤细胞的剂量反应曲线、评估和排除非特异性 有毒化合物。以及对各种鼠和人的激素分泌的特异性试验 神经内分泌肿瘤此外，硅靶预测将与RNA-seq转录组相结合 分析以分离药物靶途径并解卷积所得命中的MOA。最后，潜在的行动 通过使用短发夹RNA破坏途径， 在大鼠和人PRL瘤中使用shRNA和CRISPR文库来定义一组一流的命中。

ABSTRACT Prolactinomas (PRL-omas), the most common secreting pituitary tumors, frequently cause hypogonadism and subfertility. Although dopamine agonist therapy often works well initially and 1/3 of patients experience long-term remission after 2 years of therapy, 1/3 of patients require long-term D2 agonist therapy to maintain normal PRL levels and a further 1/3 of patients are either refractory to DA therapy or intolerant of their side effects. Although surgical resection can be offered in some, remission rates in large locally invasive tumors is <50%. Translational research has been hampered by the lack of any human PRL-oma cell models. Using current methodologies, human PRL-oma cultures only survive for 7-10 days in vitro. There is an urgent unmet need for establishment of a human PRL-secreting tumor model that retains its highly differentiated phenotype and allows sufficient long-term expandability of cells for use in preclinical translational research and development of novel treatment options. We compared the global transcriptome landscape in consecutive passages from two human prolactin-secreting pituitary tumor primary cultures using bulk RNA-seq to understand the molecular events leading to loss of hormone secretion during in vitro human prolactinoma culture. As pituitary tumor cells lost hormone secretion, we observed a reduction in angiogenesis, survival signals and immune responses in parallel with increased collagen catabolism, cell adhesion and extracellular matrix organization. Guided by these findings, we developed a unique 3-dimensional (3D) PRL-oma culture system and for the first time, we have generated long-term (> 6 months) expandible PRL-secreting 3D human pituitary tumor cultures (>6 months). In the first of two aims, we will use our unique patient-derived 3D PRL-oma culture model in a high throughput screen to identify small molecule inhibitors of PRL secretion and proliferation. We have already demonstrated the feasability of this approach and identified a compound of interest in a pilot screen and now wish to conduct an expanded HTS to identify novel therapies for patients with refractory PRL-omas. Our primary screen will include pharmacological validation and repurposing-, targeted-, lead-like- and diverse- libraries. Initial hits will be selected using robust z-score statistics and transitioned to aim 2 for further development. Aim 2 will employ a cascade of follow-up assays to validate potential hit compounds. These will include dose-response curves in both rat and human PRL-oma cells, assessment and exclusion of non-specific overly toxic compounds. And tests of specificity on hormone secretion in various murine and human neuroendocrine tumors. Additionally, in-silica target prediction will be combined with RNA-seq transcriptome profiling to segregate drug target pathways and deconvolute the MOA of resultant hits. Finally, potential actions of hits on identified pathway targets will be corroborated by pathway disruption using short hairpin RNA (shRNA) and CRISPR libraries in both rat and human PRL-omas to define a group of first-in-class hits.