Immunogenetic analysis of aggressive B-cell malignancies

Abstract

Normal B cells undergo a series of recombinatorial and mutational changes during differentiation that lead to unique sequences in the immunoglobulin (Ig) heavy and light chain variable (V) region genes. The defining event is the rearrangement of the heavy and light chain genes in the bone marrow (BM), with further diversification following interaction with antigen in the secondary lymphoid organs by means of somatic mutation and class switching. In the event of neoplastic transformation, these sequence modifications are preserved and feature in every cell of the tumour clone. Since B-cell malignancies typically have functional Ig gene rearrangements and express surface Ig, analysis of the Ig V genes can provide an independent marker for a clonal B-cell population and may help to establish the developmental stage at which neoplastic transformation occurred. Therefore, Ig V gene analysis can contribute to our understanding of the differentiation and development of malignant B cells. In this study, we focus on the Ig V genes of two Non- Hodgkin's lymphomas that together encompass ~80% of all B-cell malignancies: follicular lymphoma (FL) and diffuse large B-cell lymphoma (DLBCL). An unusually high incidence of N-glycosylation motifs, introduced by somatic mutation, has been reported to feature in the VH genes of FL and other germinal centre (GC)-associated malignancies, and contrasts that found for normal B cells. Our primary aim was to investigate the introduction of yV-glycosylation sites in both the VH and VL genes of FL, in order to gain a complete view of the B- cell receptor. Here, we confirm that the most dramatic distinction between the Ig V genes of normal B cells and those of FL lies in the accumulation of motifs available for addition of oligosaccharides in the latter, where 94% of tumours acquired at least one novel N-glycosylation site. Furthermore, there appears to be a drive towards the acquisition of such sites, with 21% tumours continuing to accumulate N-glycosylation sites as a result of ongoing mutation. Using recombinant scFv proteins derived from FL, we next confirm that these sites are functional and reveal that the added glycans are commonly oligomannose. We further show that the oligomannose glycans confer the ability to bind specifically to recombinant mannose receptor. Our data support the concept that an interaction with a mannose-binding lectin expressed by stromal cells in the GC may play a role in lymphoma pathogenesis, possibly by providing specific signals for the growth and/or survival of these tumours within the GC microenvironment. de novo DLBCLs represent a highly heterogeneous group of lymphoid neoplasms, and for which N-gylcosylation sites in the Ig V genes have been reported in ~40% of cases. Our second aim was to determine whether the introduction of yV-glycosylation sites is restricted to the Ig V genes of tumours displaying a GCB-like phenotype. Here, we report that DLBCLs classified by cDNA/tissue microarray and iFISH analysis into t(14;18)+-GCB-like, t(14;18)-GCB-like and ABC-like subgroups show an incidence of A^-glycosylation sites of 50%, 56% and 50%, respectively. Therefore, our findings do not support a link between t(14;18)+-GCB-like DLBCL and FL and argue against a common pathway for pathogenesis. Primary CNS lymphomas (PCNSL) represent highly malignant DLBCLs, which arise within and are confined to the CNS in the absence of systemic disease. Little is known about the site of origin of the malignant lymphocyte that gives rise to these tumours. It has been suggested that clonal proliferation and transformation may take place in the periphery, followed by the specific homing of tumour cells to the CNS. The third aim of our study was to investigate the existence of tumour cells at peripheral sites in PCNSL patients at the time of diagnosis. Using the unique tumour- derived Ig V gene sequences as a marker we could track tumour cells in the BM and/or blood of 3 of 3 patients and further show the existence of tumour-related subclones unique to extracerebral sites. We hypothesise that, at some stage, a proportion of tumour cells in the periphery may have migrated to the brain, with the transformation event occurring either in the periphery or after entry into the CNS.

More information

Identifiers

Catalogue record

Export record

ASCII CitationAtomBibTeXData Cite XMLDublin CoreDublin CoreEP3 XMLEndNoteHTML CitationHTML CitationHTML ListJSONMETSMODSMPEG-21 DIDLOpenURL ContextObjectOpenURL ContextObject in SpanRDF+N-TriplesRDF+N3RDF+XMLRIOXX2 XMLReferReference ManagerSimple Metadata