The high incidence of unusual mannose-rich glycans on the expressed surface immunoglobulins of follicular lymphoma (FL) suggests a pathogenic role for this modification. While lectin-mediated signaling has been proposed to provide trophic support to FL cells, the localization of the lectins has not been extensively characterized. Dendritic Cell-Specific Intercellular adhesion molecule-3-Grabbing Non-integrin (DC-SIGN) is a mannose-selective lectin named for its expression on classic marrow-derived dendritic cells. In vitro, recombinant DC-SIGN induces signaling in primary FL cells. The nodules of Follicular lymphoma (nFL) are the neoplastic counterparts of normal germinal centers (GC). Both GC and nFL contain variable numbers of centrocytic and/or centroblastic B cells, T-cells, and follicular dendritic cells (FDCs), but macrophages are typically found only in GC. Although termed “follicular dendritic cells”, FDCs are thought to be derived from non-blood stem-cells, potentially pericytes. Therefore, nFL lack classic dendritic cells and macrophages so it is unclear what cells in FL could express DC-SIGN. Therefore, we systematically characterized FDC and DC- SIGN localization in sixteen FL and nine Follicular hyperplasia (FH) lymph nodes. Using flow cytometry of disaggregated FL nodes, we were not able to identify a specific cell population expressing DC-SIGN. However, using immunohistochemistry and Immunofluorescence methods, we identified nine cases of FH and five cases of FL (n=5/16) with DC-SIGN and FDC co-localization. Immunohistochemical stains of the FDCs using CD23 and CD21 showed that the follicular dendritic structures were disrupted in almost all the FL cases assessed showing either a peripheral crescentic staining or broken FDC meshwork within the follicle. We identified five cases of FL with an average of more than four well visualized and complete FDC structures staining per low power field (4X) while the remaining cases had an average of less than two complete FDC network staining. There was near complete absence of CD21 positive FDCs meshwork identified in 50% of the FL cases compared to only one case lacking CD23 positive FDCs structure staining. In contrast, FH cases (n=9) showed complete staining of the FDC meshwork within the follicles with both CD21/CD23. For FH, the average number of complete FDC staining with CD23 was 18 while an average of 10 complete FDC meshworks stained with CD21. Furthermore, immunofluorescence imaging analysis showed different pattern of localization of CD23 positive FDC with DC-SIGN with between the two conditions. In FH, DC-SIGN and CD23 colocalization was identified mainly in the light zone. For FL, we observed that DC-SIGN overlap with CD23 localized in the intra-follicular areas and co-localized with CD163 (M2 macrophages) in the inter-follicular zones. Confocal imaging confirmed colocalization of CD23 and DC-SIGN on FDCs with close association to MUM1-positive B cells in both GC and nFL. Additional analysis also showed that 80% (n=5) of the cases which had more than four complete FDC structures staining with CD23 co-localized with DC-SIGN in the intra-follicular areas and the cases with complete FDC network disruption did not. There was no significant relationship identified between the extent of FDC networks and the fraction of non-neoplastic B cells in the FL cases. In summary, our data show that DC-SIGN is expressed on FDCs in FL and FH even though FDCs are not the marrow-derived DCs for which DC-SIGN is named. Furthermore, localization of DC-SIGN to the follicle suggests a potential role for the microenvironment in the pathogenesis of FL and possible therapeutic target.
Disclosures
No relevant conflicts of interest to declare.
