Fully human antibodies have minimal immunogenicity and safety profiles. At present, most potential antibody drugs in clinical trials are humanized or fully human. Human antibodies are mostly generated using the phage display method (in vitro) or by transgenic mice (in vivo); other methods include B lymphocyte immortalization, human–human hybridoma, and single-cell polymerase chain reaction. Here, we describe a structure-based computer-aided de novo design technology for human antibody generation. Based on the complex structure of human epidermal growth factor receptor 2 (HER2)/Herceptin, we first designed six short peptides targeting the potential epitope of HER2 recognized by Herceptin. Next, these peptides were set as complementarity determining regions in a suitable immunoglobulin frame, giving birth to a novel anti-HER2 antibody named “HF,” which possessed higher affinity and more effective anti-tumor activity than Herceptin. Our work offers a useful tool for the quick design and selection of novel human antibodies for basic mechanical research as well as for imaging and clinical applications in immune-related diseases, such as cancer and infectious diseases.

完全人源抗体具有最小的免疫原性和安全性。目前，临床试验中的大多数潜在抗体药物都是人源化或完全人源化的。人源抗体大多使用噬菌体展示法（体外）或转基因小鼠（体内）产生；其他方法包括 B 淋巴细胞永生化、人-人杂交瘤和单细胞聚合酶链反应。在这里，我们描述了一种基于结构的计算机辅助从头人类抗体生成的设计技术。基于人表皮生长因子受体2（HER2）/赫赛汀的复杂结构，我们首先设计了六个短肽，靶向赫赛汀识别的HER2潜在表位。接下来，将这些肽段设置为合适的免疫球蛋白框架中的互补决定区，诞生了一种名为“HF”的新型抗HER2抗体，该抗体比赫赛汀具有更高的亲和力和更有效的抗肿瘤活性。我们的工作为快速设计和选择用于基础机械研究以及免疫相关疾病（如癌症和传染病）的成像和临床应用的新型人类抗体提供了有用的工具。