CXCL13的分泌和固定在次级和三级淋巴结构(TLS)内显示出不同的地理分布

The purpose of this study is to investigate the formation and maintenance of germinal centers in TLS by defining mechanisms and spatial patterns of CXCL13 chemokine gradients in tumor tissue. In secondary lymphoid tissue, it is recognized that CXCL13 is immobilized by follicular dendritic cells (FDCs), but the mechanisms of immobilization remain unclear. Furthermore, spatial patterns and mechanisms of CXCL13 immobilization in TLS are unknown. We developed a CXCL13 Immunohistochemistry /in situ hybridization (ICH/ISH) dual stain to determine which cells are responsible for CXCL13 production and display. We stained 3 formalin-fixed paraffin-embedded tonsil specimens and found that CXCL13 protein and mRNA staining have distinct localizations. Specifically, spatial analyses were performed on 10 individual follicular structures from each tonsil to compare the percent surface area of either germinal centers or mantle zones (n=30) demonstrating staining by IHC or ISH. This analysis revealed that CXCL13 protein was highly expressed in germinal centers and expressed relatively lower in mantle zones (31% vs 12% surface area, p<0.0001). However, CXCL13 ISH staining was the inverse: mRNA expression was lower in germinal centers, and higher in mantle zones (20% vs 37%, p<0.0001). Flow cytometry analysis of fresh tonsil tissue revealed stromal cells displaying CXCL13, (PDPN + CD31 - CXCL13 + ), supporting the CXCL13 IHC findings. Similarly, spatial transcriptomics showed similar findings to the CXCL13 ISH, providing additional validation of the spatial expression pattern. Together, these results support a model where CXCL13 is primarily secreted from mantle zones, then immobilized in germinal centers. To investigate if these findings extend to TLS in tumor tissues, we applied our CXCL13 IHC/ISH dual stain to lung cancer tissue from the definitive resection specimen from patients treated with neoadjuvant anti-PD-1 immunotherapy. Spatial analyses revealed that similar to what was observed in tonsil, the density of CXCL13 mRNA is highly correlated with CXCL13 protein staining (n=3, Spearman's r: 0.9, p<9.311e-19), though these do not immediately overlap spatially, with an average distance of 30 μm between mRNA and protein expression, suggesting capture of the diffusing CXCL13. We also observed that stromal cells in mature TLS immobilize CXCL13, but stromal cells in immature TLS do not. In addition, we found CXCL13 immobilized on elastin fibers near mRNA staining. These findings show that CXCL13 immobilization may contribute to the formation and maintenance of TLS. New therapies enhancing the formation and maturation of TLS by modulating CXCL13 may warrant exploration. Ongoing single cell RNA sequencing and in vitro studies are focused on determining the mechanisms responsible for the immobilization of CXCL13 in lymphoid and tumor tissue.