Healthy humans can be a source of antibodies countering COVID-19

Here, we describe the isolation of 18 unique anti SARS-CoV-2 human single-chain antibodies from an antibody library derived from healthy donors. The selection used a combination of phage and yeast display technologies and included counter-selection strategies meant to direct the selection of the receptor-binding motif (RBM) of SARS-CoV-2 spike protein’s receptor binding domain (RBD2). Selected antibodies were characterized in various formats including IgG, using flow cytometry, ELISA, high throughput SPR, and fluorescence microscopy. We report antibodies’ RBD2 recognition specificity, binding affinity, and epitope diversity, as well as ability to block RBD2 binding to the human receptor angiotensin-converting enzyme 2 (ACE2) and to neutralize authentic SARS-CoV-2 virus infection in vitro. We present evidence supporting that: 1) most of our antibodies (16 out of 18) selectively recognize RBD2; 2) the best performing 8 antibodies target eight different epitopes of RBD2; 3) one of the pairs tested in sandwich assays detects RBD2 with sub-picomolar sensitivity; and 4) two antibody pairs inhibit SARS-CoV-2 infection at low nanomolar half neutralization titers. Based on these results, we conclude that our antibodies have high potential for therapeutic and diagnostic applications. Importantly, our results indicate that readily available non immune (naïve) antibody libraries obtained from healthy donors can be used to select high-quality monoclonal antibodies, bypassing the need for blood of infected patients, and offering a widely accessible and low-cost alternative to more sophisticated and expensive antibody selection approaches (e.g. single B cell analysis and natural evolution in humanized mice).

本研究描述了从健康供体来源的抗体文库中，分离得到18株独特的抗严重急性呼吸综合征冠状病毒2型（SARS-CoV-2）人源单链抗体（single-chain antibody）。筛选过程采用噬菌体展示技术（phage display technology）与酵母展示技术（yeast display technology）联用方案，并引入反向筛选策略（counter-selection strategy），以定向靶向SARS-CoV-2刺突蛋白（spike protein）受体结合域（receptor binding domain, RBD2）中的受体结合基序（receptor-binding motif, RBM）。我们通过多种技术平台对筛选获得的抗体进行了表征，包括以免疫球蛋白G（immunoglobulin G, IgG）形式开展的流式细胞术（flow cytometry）、酶联免疫吸附实验（enzyme-linked immunosorbent assay, ELISA）、高通量表面等离子体共振（high-throughput surface plasmon resonance, SPR）以及荧光显微镜成像（fluorescence microscopy）。本研究报道了这些抗体对RBD2的识别特异性、结合亲和力（binding affinity）以及表位多样性（epitope diversity），同时评估了其阻断RBD2与人类受体血管紧张素转换酶2（angiotensin-converting enzyme 2, ACE2）结合的能力，以及在体外（in vitro）中和活SARS-CoV-2病毒感染的效能。本研究提供的实证支持以下结论：1）18株抗体中有16株可特异性识别RBD2；2）性能最优的8株抗体靶向RBD2上8种不同的表位；3）经夹心实验（sandwich assay）验证的其中一组抗体对可实现亚皮摩尔级的RBD2检测灵敏度；4）两组抗体对可在低纳摩尔级的半中和滴度（half neutralization titer）下抑制SARS-CoV-2感染。基于上述结果，本研究认为所获得的抗体具备极高的治疗与诊断应用潜力。值得注意的是，本研究结果显示，从健康供体中获取的现成未免疫天然抗体文库（naïve antibody library），可用于筛选高质量单克隆抗体（monoclonal antibody），无需使用感染患者的血液，为更为复杂且成本高昂的抗体筛选方法（如单B细胞分析（single B cell analysis）、人源化小鼠（humanized mouse）体内自然进化法）提供了一种可广泛获取且低成本的替代方案。