FDC Dendrite Activation

FDC Purification requires the use of collagenase and selection columns and leaves FDCs without detectable dendrites when examined by light microscopy. We have found that in vitro, FDCs reattach to the collagen matrix, and extend their dendrites and form FDC networks. When tested, 80-90% of FDCs attached to collagen type I, laminin, biglycan and hyaluronan in 12 hours but not to plastic. Within 30 days the dendrites formed networks through interdigitation with neighboring FDCs.

This figure shows the regeneration of dendritic processes and the formation of networks among several FDCs 30 days after plating on collagen type I. Note that the dendrites typically emanate from the pole of the FDC that tends to be toward the substrate matrix (figures a,b, and c). For additional illustrations go to the publication.
Next, to determine the phenotypic identity of the plated cells, that is that they are FDCs or not, we have tested the cultures for FDC-M1 labeling which is specific for FDCs and also tested the cells' function for immune complex retention. FDCs typically bind immune complexes on their surfaces. To do this 30 day old cultures of cells were incubated with the monoclonal antibody FDC-M1. The HRP conjugated FDC-M1 was developed and the cells and the networks formed by their dendrites were positive for FDC-M1 indicating that the cells were the correct phenotype. To ascertain that these cells also functioned like FDCs in situ the cells were incubated with HRP conjugated immune complexes. After development of the peroxidatic activity the cells showed positivity for immune complexes, confirming immune complex binding, the typical function of FDCs.

Figure a, shows the FDC-M1 labeling (brown) of the cells and networks. Figure b, is the similarly treated negative control of a culture of peritoneal macrophages. Figure c, is the immune complex labeled culture showing positivity for immune complexes (brown). Figure d, is the negative control culture of peritoneal macrophages (no labeling).


Connections and attachment of collagen fibers to developing FDCs is not only an in vitro phenomenon. During the development of the follicle and FDC networks there is a close interaction between FDCs and reticular fibers. To substantiate this in vivo, adjacent sections from popliteal lymph nodes were labeled for collagen type I (a,b) and FDC-M1 (c,d) 5 hrs (a,b) and 2 days (c,d) after challenge of OVA immune mice with OVA, using the biotin streptavidin system, collagen and FDC labeling was developed (see paper). Fibers labeled with anti-collagen were colocalized at 5hrs with the FDCs labeled with FDC-M1 (arrows). After 2 days the reticular fibers were restricted to the periphery of the follicles surrounding the FDCs. Essentially no contact between FDCs and reticular fibers were apparent at day 2. The day 1 observations attest to the necessity of FDC collagen type I contact for the development of dendrites and networks.
Following overnight incubation of FDCs on plastic the FDCs could easily be washed off. Similarly, after incubation of lymphocytes on collagen type I, laminin, biglycan, and hyaluronic acid the cells could be readily washed off. In contrast 80% of the cells in FDC preparations bound to these substrate matrices and remained adherent and the FDCs developed dendrites and by interdigitating with neighboring FDC dendrites formed distinct FDC networks. This study also confirms the need for a direct contact with collagen type I for the development of FDC dendrites and networks.
We reasoned that a receptor-mediated mechanism was probably involved in attaching FDCs to collagen and in signaling the formation of dendrites and FDC networks. Flow cytometric analysis indicated that FDCs expressed CD29, the B1 integrin subunits shared by collagen (a1, 2,3,9,10,11B1), laminin (a6B1), and fibronectin (a4B1) receptors, and the hyaluronic acid receptor CD44. However, the blocking of these molecules did NOT prevent FDCs from binding to collagen type I suggesting that additional receptors were involved.