Distinct CXCL13 expression pattern in NPs with and without eLTs
Although both CXCL12/CXCR4 and CXCL13/CXCR5 axis are vital for B cell attraction and aggregation, our previous study only found an elevation of CXCL13, but not CXCL12, in NPs with eLTs.8 In this study, we first confirmed increase of CXCL13 production in NPs with eLTs in comparison to NPs without eLTs and control tissues at both mRNA and protein level (Fig 1, A and B). Compared to control tissues, NPs without eLTs also demonstrated a mild increase of CXCL13 production (Fig 1, A and B). We therefore explored whether there is a difference in cellular sources of CXCL13 in NPs with and without eLTs. Immunofluorescence study showed that CXCL13 was mainly expressed by CD20+ B cells and vimentin+ stromal cells in NPs with and without eLTs (Fig 1, C). Given to the difficulty in counting the number of CXCL13 positive cells in eLTs in immunofluorescence study and the interference of membrane-bound CXCL13 on quantification of cytoplasmic CXCL13 in B cells by flow cytometry, we detected CXCL13 mRNA levels in different portions of cells in NPs. We depleted stromal cells from total polyp cells and found that CXCL13 mRNA expression was remarkably reduced in NPs without eLTs (mean, 66% reduction), while mildly reduced in NPs with eLTs (mean, 36% reduction) (Fig 1, D). On the contrary, after B cell depletion, there was about 31% and 60% (mean) reduction of CXCL13 mRNA expression in relation to the expression of total cells from NPs without and with eLTs, respectively (Fig 1, E). These results suggest that CXCL13 is mainly produced by stromal cells in NPs without eLTs, whereas by B cells in NPs with eLTs.