Abstract 3696

Background:

The stromal microenvironment is thought to play a critical role in supporting the survival and growth of follicular lymphoma (FL) tumor cells. However, the key stromal factors involved are not well characterized. We hypothesized that FL tumor cells hijack the survival and growth mechanisms that are usually used by normal B cells in the lymph node microenvironment. For example, follicular dendritic cells (FDC) residing in germinal centers of lymph nodes play a vital role in supporting normal B cell survival and proliferation through adhesion molecules and by secreting chemokines and cytokines. In normal lymph nodes, lymphotoxin-alpha (LTA, TNFSF1) secreted by B cells activates FDC through LTbR and induces secretion of BAFF (TNFSF13B) and CXCL13. CXCL13 recruits follicular helper T cells (Tfh) into the B cell zone and CD40L (expressed by Tfh) together with BAFF prevents apoptosis and promotes proliferation of B cells. Here, we determined whether a similar cross-talk exists between FL tumor B cells, FDC, and Tfh.

Methods:

FL tumor samples were processed into single-cell suspension and cryopreserved in aliquots for later use. Primary tumor B cells were isolated by magnetic cell separation from single-cell suspensions of FL. The FDC-like cell line HK derived from human tonsil was used in place of primary FDC. Levels of BAFF, CXCL13, and LTA in culture supernatants were determined by ELISA. Apoptosis of tumor cells was determined by Annexin V assay and proliferation by tritiated thymidine incorporation assay. Activation of NFkB was determined by the TransAM NFkB p52 ELISA kit.

Results:

We developed a co-culture model to mimic the FL tumor microenvironment to study the interactions between tumor cells, FDC and Tfh. We observed that FL tumor cells spontaneously produced low levels of LTA but its production was significantly enhanced when tumor cells were cultured with either recombinant BAFF or CD40L. Dose response curves from FL tumor cells treated with both BAFF and CD40L recombinant proteins showed synergistic increase in LTA secretion. Recombinant BAFF and CD40L also synergistically enhanced the survival and proliferation of FL tumor cells. HK cells spontaneously produced low levels of BAFF and CXCL13 but their production was significantly enhanced by recombinant LTA in a dose-dependent manner.

Consistent with the above studies, co-culturing FL tumor cells with HK cells significantly enhanced LTA secretion, possibly due to the effects of BAFF secreted by HK cells. In addition, in this co-culture, we also observed enhanced production of BAFF and CXCL13 possibly due to the effects of LTA secreted by the tumor cells. The production of BAFF and CXCL13 in this co-culture could be inhibited by an LTA neutralizing antibody in a dose-dependent manner suggesting the role of LTA in this mechanism. Co-culturing FL tumor cells with HK cells also significantly enhanced the survival and proliferation of FL tumor cells greater than two-fold compared with FL tumor cells alone. This effect also could be inhibited by LTA neutralizing antibody but not isotype control antibody. The survival benefit for tumor cells was further enhanced upon co-culture with autologous Tfh cells FACSorted from the same tumor sample. We also studied the downstream NFkB signaling mediated by BAFF on tumor B cells and by LTA on HK cells in the co-culture. We found that both tumor B cells and HK cells had significantly higher NFkB activity upon co-culture compared to cells alone.

Conclusions:

Our results suggest a model where LTA secreted by tumor B cells forms the afferent axis and induces production of BAFF and CXCL13 by FDC. CXCL13 preferentially recruits CXCR5-expressing Tfh into the tumor microenvironment and CD40L expressed by Tfh and BAFF secreted by FDC form the efferent axis and promote further production of LTA by tumor cells. In addition, CD40L and BAFF synergistically enhance the survival and proliferation of FL tumor cells via activation of NFkB. Our results suggest that because of their redundant functions, blocking BAFF-BAFF-R pathway alone or CD40-CD40L pathway alone with monoclonal antibodies as is being done in ongoing clinical trials may not be optimal. In contrast, blocking both pathways simultaneously is likely to significantly enhance the clinical responses. Our results also suggest that neutralizing LTA or blocking its receptor may be another novel therapeutic strategy for management of FL.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution