High-Throughput, Highly Multiplexed In Situ Proteomic Imaging of Human Tissues

点击翻译按钮获取中文摘要

Summary High multiplexing capability is indispensable for high-content mapping. However, spectral overlap and lack of orthogonal labeling create severe limitations for conventional fluorescence imaging. For protein imaging, these limitations have been circumvented through multiplexed detection via unconventional probes and specialized instruments, iterative sequential antibody labeling and imaging, or sequential detection through DNA-barcoding. Although all of these techniques can theoretically achieve high multiplexing, they come at a cost of limited throughput. Utilizing in situ signal amplification would substantially reduce exposure - and thus imaging - time per frame, allowing for high throughput as well as improved sensitivity. Although amplification methods exist, they have not been robustly multiplexed beyond 5-8 spatially overlapping targets. A critical unmet technical need for fully realizing the HuBMAP vision is a highly multiplexed signal amplification technique that can simultaneously amplify tens of distinct targets by tens- to hundreds-fold, thereby enabling highly multiplexed, high throughput, in situ imaging of proteins in human tissues. We propose such an in situ signal amplification method, based on a novel molecular mechanism that we recently published. In the new method, staining with multiple probes (DNA-barcoded primary antibodies) will be performed simultaneously, and then all barcodes will be simultaneously extended into long concatemers in situ. Finally, mapping of concatemers will be sequentially performed through rapid exchange and imaging cycles, improving throughput by 10-fold while enabling detection of rare targets in tissues. Beyond proteins, the method will be applicable to RNA and DNA (chromosome) targets. We will integrate the method with commercially available, automated staining and imaging systems, and apply it to image diverse human tissues via broad collaboration within and beyond HuBMAP community.