Computational modeling of stem cells predicts clinical trial outcomes for hypoplastic left heart syndrome

项目总结/摘要 先天性心脏病每年影响美国每1000名新生儿中估计有8名。一种复杂的 先天性心脏缺陷，左心发育不良综合征（HLHS）是缓解了一系列的三个手术 需要右心室维持体循环尽管结局有所改善，但HLHS死亡率 由于右心室功能障碍/衰竭，移植仍然是唯一的治疗选择。 考虑到对移植可用性和排斥的担忧，触发内源性修复的干细胞 机制已成为治疗HLHS的有吸引力的候选者。目前，我们的团队参与了两个 在美国进行的三项HLHS干细胞临床试验中， 间充质干细胞（MSC）和心脏ckit+祖细胞（CPC）。然而，尽管一些 在临床前和临床试验中取得成功并证明安全性，干细胞群变异较大 患者的治疗效果仍然是一个关键问题。此外，缺乏定量研究 调查这些差异在这个建议中，我们将采取系统生物学的方法来理解 干细胞及其旁分泌的修复作用背后的生物分子或信号 信号外泌体（含有不同货物的30 - 150 nm囊泡）。我们之前已经证明了（1） 用CPC及其外来体治疗在体外和体内产生促血管生成和抗纤维化反应， 体内，和（2）这些反应可以通过模拟细胞和外泌体表达模式来预测 使用偏最小二乘回归（PLSR）。考虑到我们参与了HLHS干细胞临床试验， 我们将扩大我们以前的努力，建立一个干细胞含量的计算模型，能够预测 患者改善。我们将用CPC和CPC外泌体测序和质量来训练我们的模型。 来自我们实验室的44个CPC系的库的质谱（MS）数据（先前从心脏活检中分离， 先天性心脏病患者）。然后，我们将测序并执行MSC、CPC和MSC的MS。 他们的外泌体来自两个临床试验，并将这些数据输入到体外训练的模型中。我们预计 我们的模型来预测患者在临床试验中的改善。总的来说，我们的模型不仅会产生 一种预测性的临床工具，但也可以识别与这些修复反应直接相关的共变信号 以作进一步调查。在创建一个强大的，可推广的模型，我们将获得心脏的机制洞察力， 修复并为儿科干细胞试验提供有价值的临床工具。最终，这项工作将有助于 为患有先天性心脏病的儿童提供最佳治疗策略，如HLHS。

PROJECT SUMMARY/ABSTRACT Congenital heart defects affect an estimated 8 in 1000 births in the US annually. One complex form of congenital heart defects, hypoplastic left heart syndrome (HLHS) is palliated by a series of three surgeries which demands the right ventricle sustain systemic circulation. Despite improved outcomes, HLHS mortality remains high due to right ventricular dysfunction/failure and transplant remains the only curative option. Considering concerns over transplant availability and rejection, stem cells which trigger endogenous repair mechanisms have become an attractive candidate for treating HLHS. Currently, our group is involved in two of the three stem cell clinical trials for HLHS in the US, investigating the use of bone marrow derived- mesenchymal stem cells (MSCs) and cardiac ckit+ progenitor cells (CPCs). However, despite some successes and demonstrated safety in preclinical and clinical trials, large variation in stem cell populations and patient outcomes remains a critical problem. Furthermore, there is a lack of quantitative studies investigating these discrepancies. In this proposal, we will take a system-biology approach to understand the biological molecules, or signals, underlying the reparative effects of stem cells and their paracrine signaling exosomes (30-150nm vesicles containing diverse cargo). We have shown previously that (1) treatment with CPCs and their exosomes produce pro-angiogenic and anti-fibrotic responses in vitro and in vivo, and (2) these responses can be predicted by modeling cellular and exosomal expression patterns using partial least squares regression (PLSR). Considering our involvement in HLHS stem cell clinical trials, we will expand our previous efforts to build a computational model of stem cell content, capable of predicting patient improvements. We will train our model with CPC and CPC exosome sequencing and mass spectrometry (MS) data from our lab’s bank of 44 CPC lines (previously isolated from cardiac biopsies of patients with congenital heart defects). Then, we will sequence and perform MS of the MSCs, CPCs, and their exosomes from the two clinical trials and input these data into the in vitro trained model. We expect our model to predict patient improvements from the clinical trials. Overall, our model will not only generate a predictive, clinical tool, but also identify co-varying signals directly related to these reparative responses for further investigation. In creating a robust, generalizable model, we will gain mechanistic insight of cardiac repair and provide a valuable clinical tool for pediatric stem cell trials. Ultimately, this work will be helpful in providing the best treatment strategies for children with congenital heart defects, like HLHS.