FAO-fueled OXPHOS and NRF2-mediated stress resilience in MICs drive lymph node metastasis

The lymph nodes (LN) are the most frequent metastatic sites of esophageal squamous cell carcinoma (ESCC), but the associated mechanisms are poorly understood. Tumor cells lodge and survive at the metastatic site during the early adaptation step, which is the bottleneck for successful metastasis. However, the phenotypic and molecular characteristics of tumor cells that contribute to this key process remain elusive. Here, we established a robust human xenograft spontaneous LN metastasis model to capture individual cancer cells and used single-cell RNA sequencing to profile transcriptomic changes in cancer cells during different steps of metastasis. Overall design: To establish an LN metastasis mouse model, 1-3×10^6 tumor cells tagged with GFP in 20 µl PBS were injected into the right hind foot pad of nude mice. Metastases were evaluated frequently through in vivo imaging. Mice were harvested between 2 weeks and 2 months post-injection. Primary tumors at the injection site and right-side popliteal lymph nodes (draining LNs) were excised for subsequent examination. The absence or presence of LN metastases was confirmed by FACS or examination of 100-µm LN sections for the presence of a GFP signal in the LNs. The single tumor cells were rinsed with DPBS on ice before loading into the Gel Bead & Multiplex Kit and Chip Kit (10× Genomics, Pleasanton, CA, USA) for the construction of a chromium single-cell 3'-library according to the manufacturer's instructions. Libraries were sequenced on an HiSeq X Ten system and mapped to the human reference genome (build GRCh38) using CellRanger-4.0.0 (10× Genomics).