Searching for Second Events in Follicular Lymphoma: A New Target for Therapy

Although the pathognomonic t(14;18) translocation has long been characterized as necessary but not sufficient for malignant transformation in follicular lymphoma (FL), there remains a good deal of uncertainty about the nature of the other molecular events that result in the full malignant phenotype, and more still about the processes that underlie transformation to diffuse large B-cell lymphoma. Classical cytogenetics have revealed the common occurrence of abnormalities on the long arm of chromosome 6, but no one consistent region of gain or loss has been implicated, and most of the changes are hemizygous, leaving a large number of potential gene candidates to be examined.

The group at Memorial Sloan-Kettering Cancer Center has taken a combined approach to this problem, using the results of array-comparative genomic hybridization analysis to define the main regions of loss on 6q and constructing a library of short hairpin RNAs (sh-RNAs) designed to knock down expression of the genes from the most commonly deleted regions. By transfecting the shRNAs into a B-cell line dependent on IL-3, they were able to identify genes whose inhibition rendered the cells relatively immune to IL-3 deprivation, indicating that the inhibited genes functioned as tumor suppressors. Among the genes identified by this screening procedure were A20, already known to be the target of deletions in B-cell lymphoma, and the ephrin receptor A7 (EPHA7), whose silencing had previously been found to be associated both with aberrant methylation in lymphoma and with epithelial malignancies such as those of the stomach, colon, and prostate.¹  Suppression of EPHA7 by shRNA knockdown in a spontaneous mouse-lymphoma model resulted in accelerated lymphoma development, at a rate equivalent to that observed with knockdown of p53. Further studies confirmed that translation of EPHA7 was lower in most FL than in normal germinal-center B cells, and that 72 percent of FL tested by immunohistochemistry had absent or weak staining for EPHA7 protein. In contrast, control studies showed that EPHA7 was abundant in the cytoplasm of normal tonsillar B cells. Much of the loss of expression in EPHA7 in FL appears to be the result of epigenetic silencing through promoter methylation, an effect that was reversible in B-cell lymphoma lines using the methylation inhibitor, 5-azacytidine.

Ephrin receptors mediate cell-cell interactions, embryonic development, neural development, and angiogenesis signaling through several different kinase pathways. A truncated form of EPHA7 protein, EPHA7TR, is expressed in normal B cells and is found in lymphocyte-conditioned media and in normal human serum where it may act as a growth inhibitor. The group showed that a tagged EPHA7 ectodomain protein could bind to the homologous EPHA2 receptor on B cells thereby blocking receptor-mediated oncogenic signaling as evidenced by inhibition of downstream phosphorylation of ERK, STAT3, and a variety of SRC family kinases. Studies in B-lymphoma xenograft models showed that soluble EPHA7TR protein had a suppressive effect on tumor growth, especially when administered locally at the injection site. Importantly, a fusion protein of anti-CD20 antibody and EPHA7TR showed a growth inhibitory effect upon Raji xenografts when given systemically, with the fusion protein having significantly greater potency than EPHA7TR alone. These experiments suggest a mechanism for targeting EPHA7TR to B cells through binding to CD20.