Although the requirement of the NF-kB pathway in B-cell lymphomas is clear, the role of the individual REL (c-rel, relA, and relB) proteins is not understood. This is particularly interesting for several reasons. The first is the observation that c-rel is often amplified in several lymphomas: 50% of Hodgkin's (HL) and 20-30% of diffuse large B-cell lymphomas (DLBCL). However, the specific role of c-rel amplification, as opposed to relA and rel B is not known. Secondly, the lack of similarities in the primary sequences of the three RELs is considerable. This disparity does not lie within the DNA binding and dimerization Rel Homology Domain (RHD), but rather in the C-terminal transcriptional transactivation domains (TAD) of the three. There is only 10% similarity between c-rel and relA, for example. We have therefore constructed the hypothesis that the three REL proteins are functionally distinct and unique. The differences in the TADs suggest that each REL protein has different promoter targets and unique transcriptional biochemistry. Furthermore, the specificity of the c-rel amplification implies that the REL proteins each will have distinct, nonoverlapping functions in lymphoma tumorigenesis. We are therefore addressing the functions of the REL family in B-cell lymphomas in vivo and in vitro. We are addressing REL family functions in several ways. We have begun an in vitro analysis of the protein factors that are necessary for the transcriptional activity of each REL family member. As stated above, we expect these factors to be different for each REL protein. Secondly, we have embarked on an extensive analysis of the genomic distribution of the REL proteins and the promoters they regulate. We are analyzing several different human B-cell lines that represent both Hodgkin's and non-Hodgkin's lymphomas. Finally, we are using cell viability assays and shRNAs that target each REL member to assess their relative importance in the maintenance of lymphoma cell viability. The genomic distribution of the REL factors appears to be unique for each type of lymphoma indicating that each REL regulates a unique transcriptional regulatory network and which probably serve as the defining feature of these lymphoma classes. We unexpectedly found that although all germinal center lymphomas we tested required relA and c-rel, only Hodgkin's lymphoma (HL) was additionally dependent on relB. This underscores the unique functions of relB in HL and argues that the defining characteristic of HL is relB expression. The expression of the activated relB pathway offers an additional method of identifying HL cells in patients. Indeed, we found that the relB pathway (characterized by NIK expression) was active in primary HL tumors. Furthermore, we found that relB controls BCL2 expression but that loss of BCL2 via relB shRNA depletion does not kill the HL cells by apoptosis as was widely expected. Rather, the HL cell lines viability losses are due to autophagic cell death. This is the first observation of non-apoptotic cell death mechanisms being employed by B-cells. Indeed, it suggests this might be an additional mechanism by which B-cells containing non-productive IgG rearrangements are removed from the germinal center. In any case, these data point to defining HL and NHL in terms of their REL transcriptional regulatory networks.
