Neural Circuits, Kinetics and Energetics HTS of Human iPSC-Neurons, -Microglia, and -Astrocytes: AI-Enabled Platform for Target ID, and Drug Discovery and Toxicity (e.g., Cancer Chemo & HIV ARTs)

神经毒性是临床试验期间药物失败的最常见原因之一，说明目前 以整个动物或动物细胞为特征的临床前测试模型在预测人类方面非常差 神经毒性。此外，许多目前FDA批准的治疗药物，如抗癌化疗药物， 是神经毒性的，这限制了它们的治疗剂量，并经常导致在治疗过程中和之后的认知障碍 治疗。FDA的“黑匣子”警告，描述神经副作用，是常见的，目前也是处方 艾滋病的治疗可能会导致HIV相关的神经认知障碍(HAND)。相应地，瓦拉 Science Inc.提出了一个快速跟踪SBIR项目，以开发一种利用神经元、星形胶质细胞 和小胶质细胞，来自人类诱导的多能干细胞(HiPSC-神经元/星形胶质细胞/小胶质细胞)来测试 神经毒性的候选疗法。细胞将在384孔的培养皿中培养，以实现高产量 复合作用于神经元活动的多重分析(通过动态图像细胞术[KIC]- 获取/分析用钙和电压指示剂标记的细胞的视频)，以及突触、线粒体、 内质网和细胞活力。由于化验将以人类细胞为特征，它可能会更具预测性 人类神经毒性与目前临床前研究中使用的动物测试系统的对比。第一阶段活动将 开发多路系统，以测试化合物对作用机制(MOA)的毒性影响 包括质膜离子通道、内质网功能和完整性、线粒体功能以及 神经元回路(培养井内神经元的相关活动)。共有30个化合物具有已知的MOA 和人类的影响，将在第一阶段进行测试，这些数据将用于训练人工智能(AI)聚类 识别MOA的算法，特别是对于低剂量的测试代理。这些化合物也将是 对代表APOE-3/4和APOE-4/4基因型的HIPSC神经元/星形胶质细胞/小胶质细胞进行测试，作为 载脂蛋白E-4与神经毒性增加有关。在第二阶段，总共有350种探针/药物(其中275种已知 有临床效果的，预计75人将无效)将进行筛查，结果用于 进一步训练人工智能，提高整体临床预测性。将开发更多版本的化验方法 具有与阿尔茨海默氏症和躁郁症相关的突变/基因类型。除了认可之外 对于神经毒性，该检测系统还将增加我们对神经元回路的了解，神经元回路是有反应的 给大脑倡议计划。神经毒性测试将向制药公司销售 开发针对各种疾病/病痛的治疗药物，包括化疗药物和 神经退行性疾病的潜在疗法。

Neurotoxicity is among the most common reasons for failure of drugs during clinical trials, illustrating that current preclinical test models, which feature whole animals or animal cells are extremely poor at predicting human neurotoxicity. Additionally, many currently FDA-approved therapeutics, such as anti-cancer chemotherapeutics, are neurotoxic, which limits their therapeutic dosage and often results in impaired cognition during and after treatment. FDA “black box” warnings, describing neurological side-effects, are common and currently prescribed therapeutics for AIDs may contribute to HIV-Associated Neurocognitive Disorder (HAND). Accordingly, Vala Sciences Inc., proposes a Fast-Track SBIR project to develop an assay system utilizing neurons, astrocytes, and microglia, derived from human induced pluripotent stem cells (hiPSC-neurons/astrocytes/microglia) to test candidate therapeutics for neurotoxicity. The cells will be cultured in 384-well dishes to enable high throughput multiplexed assay of compound effects on neuronal activity (via Kinetic Image Cytometry [KIC] – acquisition/analysis of videos of cells labeled with calcium and voltage indicators), and synapses, mitochondria, endoplasmic reticulum, and cell viability. Since the assay will feature human cells, it will likely be more predictive of human neurotoxicity vs. animal test systems currently used in preclinical research. Phase I activities will develop the multiplexed system, to enable testing of compounds for toxic effects on mechanisms of action (MOA) including alteration of plasma membrane ion channels, ER function and integrity, mitochondrial function, and neuronal circuitry (correlated activity of neurons within a culture well). A total of 30 compounds with known MOA and effects of humans, will be tested in Phase I, these data will be used to train artificial intelligence (AI) clustering algorithms to recognize the MoAs, particularly for low doses of the test agents. The compounds will also be tested on hiPSC-neurons/astrocytes/microglia representing APOE-3/4 and APOE-4/4 genotypes, as the APOE-4 is associated increased neurotoxicity. In Phase II a total of 350 probes/drugs (of which 275 are known to have clinical effects, and 75 will be expected to have no effect) will be screened, with the results used to further train the AI and improve the overall clinical predictivity. Additional versions of the assay will be developed featuring mutations/genotypes relevant to Alzheimer’s Disease and Bipolar Disease. In addition to recognition of neurotoxicity, the assay system also will increase our understanding of neuronal circuits, which is responsive to the BRAIN Initiative program. The neurotoxicity assay will be marketed to pharmaceutical companies developing therapeutics across a broad spectrum of diseases/afflictions, including chemotherapeutics and potential therapeutics for neurodegenerative diseases.