Somatic hypermutation and B-cell lymphoma

During the B-cell response to T-cell-dependent antigens, the B cells underg o a rapid proliferative phase in the germinal centre. This is accompanied b y the introduction of mutations into the immunoglobulin (Ig) variable regio n (V) genes. The B cells are then selected according to the affinity of the encoded immunoglobulin for antigen, resulting in affinity maturation of th e response. Analysis of mutations in IgV genes has given insight into the h istory of individual B cells and their malignancies. In most cases, analysis of mutations confirms classifications of B-cell lin eage designated by studies of cellular morphology and surface antigen expre ssion. However, of particular interest is the subdivision of groups of mali gnancies by analysis of somatic hypermutation. It is now apparent that ther e are two subsets of chronic lymphocytic leukaemia (CLL), one with a low lo ad of mutations and poor prognosis, and one with a heavy load of mutations with a much more favourable prognosis. In addition, in Burkitt's lymphoma, sporadic and endemic subtypes are now considered possibly to have a differe nt pathogenesis, reflected in differences in the numbers of mutations. Hodgkin's disease, which was a mystery for many years, has now been shown t o be a B-cell tumour. Although in many cases the Ig genes are crippled by s omatic hypermutation, it is thought that failure to express Ig is more like ly to be associated with problems of transcription. It has been proposed that the distribution of mutations in a B-cell lymphom a can be used to determine whether a lymphoma is selected. We have investig ated the load and distribution of mutations in one group of lymphomas-margi nal zone B-cell lymphomas of mucosa-associated lymphoid tissues (MALT-type lymphoma), which are dependent on Helicobacter pylori for disease progressi on, to investigate the limits of information that can be derived from such studies. Comparison of the load of mutations demonstrates that these tumour s have approximately the same load of mutations as normal mucosal marginal zone B cells from the Peyer's patches and mucosal plasma cells. This is con sistent with the origin of these cells from mucosal marginal zone B cells w ith plasma cell differentiation. To investigate selection in MALT lymphomas we compared a region of the fram ework region three in ten MALT lymphomas which use the V-H4 family, with th e same codons in groups of V-H4 genes that are out of frame between V and J . The latter accumulate mutations but are not used and are not selected. A group of V-H4 genes are in-frame between V and J were also included for com parison. There were no obvious differences in the distribution of mutations between the groups of genes; the same hot spots and cold spots were appare nt in each. In the MALT lymphomas, selection was apparent in the framework regions only and the tendency was to conserve. We therefore feel that there is selection to conserve antibody structure and that this does not reflect selection for antigen. We do not believe that antigen selection can be ded uced reliably from sequence information alone. It is possible that somatic hypermutation could be a cause of malignancy si nce it has been shown that the process may generate DNA strand breaks and i s known to be able to generate insertions and deletions. Such events may me diate the translocation of genes-a process that is pivotal in the evolution of many lymphomas.