Computational Design of Antibody-Drug-Excipient Nanoparticles

摘要 纳米颗粒能够将治疗药物输送到所需的组织，从而提高疗效和安全性。 然而，只有大约30种纳米颗粒疗法获得了FDA的批准，而这30种药物中没有一种 纳米颗粒使用先进的靶向功能。阻碍更广泛纳米颗粒的关键挑战 部署的是复杂的纳米颗粒合成方案，载药量通常在以下 10%，以及一刀切的材料优化方法。新型药物-辅料共聚体(Reker等， NAT Nanotechnol 2021)通过简单的合成、高达95%的载药量以及 通过机器学习对新型纳米粒子进行合理的设计和优化。然而， 这些用于主动靶向给药的新材料的功能化尚未建立，限制了它们的 仅部署到一组狭窄的组织和适应症。这里提出的研究将解决 实现药物-辅料共聚体功能化的新技术的需求尚未得到满足 纳米粒子。具体来说，我们将开发新的实验(目标1)和计算(目标2)协议来 用抗体功能化药物辅料纳米粒并验证其体外和体内靶向能力 活着。该项目将(1)用于定向纳米颗粒开发的原型机器学习，(2)首次 功能化药物辅料纳米粒定性提高高载药的靶向能力 纳米粒，以及(3)产生一组新颖的、仔细表征的具有潜力的治疗性纳米粒 用于进一步的临床开发。通过快速综合和机器学习指导的设计，这里 提出的平台可以快速扩展纳米医学工具箱并简化纳米颗粒的开发， 评估和制造。通过我们的模块化方法来“混合和匹配”纳米颗粒组件， 我们期望抗体、药物和赋形剂的合理选择能够使设计精确度 用于个性化药物输送的纳米颗粒。

ABSTRACT Nanoparticles enable the delivery of therapeutics to the desired tissue and thereby improve efficacy and safety. However, only about 30 nanoparticle therapeutics have been FDA approved, and none of these 30 nanoparticles use advanced targeting functionality. Key challenges that impede broader nanoparticle deployment are the complexity of nanoparticle synthesis protocols, a drug loading capacity commonly below 10%, and a one-size-fits-all approach in material optimization. Novel drug-excipient co-aggregates (Reker et al, Nat Nanotechnol 2021) address these shortcomings through facile synthesis, drug loading of up to 95%, and by using machine learning for the rational design and optimization of new nanoparticles. However, the functionalization of these novel materials for actively targeted drug delivery is not yet established, limiting their deployment to only a narrow set of tissues and indications. The here presented research will address the unmet need for novel technologies to enable the functionalization of drug-excipient co-aggregate nanoparticles. Specifically, we will develop novel experimental (aim 1) and computational (aim 2) protocols to functionalize drug-excipient nanoparticles with antibodies and validate their targeting capabilities in vitro and in vivo. This project will (1) prototype machine learning for targeted nanoparticle development, (2) for the first time functionalize drug-excipient nanoparticles to qualitatively enhance the targeting capabilities of highly loaded nanoparticles, and (3) generate a set of novel, carefully characterized therapeutic nanoparticles with potential for further clinical development. Through rapid synthesis and machine learning-guided design, the here proposed platform can rapidly expand the nanomedicine toolbox and streamline nanoparticle development, evaluation, and manufacturing. Through our modular approach to “mix-and-match” nanoparticle components, we expect the rational selection of antibodies, drugs, and excipients to enable the design of precision nanoparticles for personalized drug delivery.