Diffuse large B-cell lymphomas (DLBCL) are a heterogeneous diagnostic entity of B-cell tumors whose behavior is variably influenced by genetic changes and environmental stimuli. They are usually divided into two major subgroups, the germinal center B-cell-like DLBCL (GCB-DLBCL) and the activated B-cell-like DLBCL (ABC-DLBCL), with different cells of origin and distinct clinical behavior.
From a previous analysis of a small number of patients, we found that a fraction of DLBCL acquires N-glycosylation sites by somatic hypermutation of the tumor surface immunoglobulin (sIg) variable region, suggesting a connection with follicular lymphoma (FL). In FL, this leads to addition of mannosylated glycans in the antigen binding site (sIg-Mann), which allow interaction with microenvironmental lectins including dendritic cell-specific intercellular adhesion molecule 3 grabbing non-integrin (DC-SIGN). However, the distribution and the consequences of acquired N-glycosylation motifs in DLBCL was unknown.
In this study, we investigated frequency, pattern and function of the mannosylated sIg in DLBCLs. DLBCL cell lines and primary samples were analyzed for the acquisition of the N-glycosylation motifs Asparagine-X-Serine/Threonine (N-x-S/T, where x ≠ Proline) in the tumor IGHV-D-J rearranged transcripts, and for binding to DC-SIGN by flow cytometry. HILIC-UPLC and crystallography were used to define structure of the glycans located in the sIg variable region. Interaction of sIg-Mann with DC-SIGN expressing cells was measured by flow cytometry and imaged by inverted fluorescence microscopy. Intracellular signaling was measured by Phosflow.
Analysis of GCB-DLBCL and ABC-DLBCL cell lines and primary samples revealed that acquired N-glycosylation sites (AGS) were common in a subset of GCB-DLBCL (51%), especially in cases with a t(14;18) translocation (88%). Remarkably, the motifs were selectively acquired in the complementary-determining-regions (CDRs) of the tumor Ig (93%). In contrast, sites were infrequent in primary ABC-DLBCL (19%) and preferentially acquired in the framework regions (51% in all ABC-DLBCL cases, 88% in IGHV4-34 ABC-DLBCL). DLBCL cell lines with AGS which bound DC-SIGN had a t(14;18) translocation and were enriched with EZH2 and KMT2D mutations, while those without AGS, and unable to bind to DC-SIGN, were not.
The sites acquired in the CDRs were permissive for addition of glycans terminating at high-mannose, as revealed by immunoblotting following EndoH treatment (that digests only glycans terminating at high-mannose) of the tumor sIg and by binding to soluble DC-SIGN. This was also confirmed by HILIC-UPLC and crystal structure of a sIg-Mann+ve lymphoma-derived recombinant Fab.
Binding of DC-SIGN to sIg-Mann mediated an antigen-independent signal of lower levels than that mediated by anti-Ig, as measured by increased SYK phosphorylation in the tumor B cells. The sIg-Mann+ve GCB-DLBCL, but not sIg-Mann-ve DLBCL, formed clusters round DC-SIGN expressing cells. These interactions were inhibitable or disrupted by antibodies specifically targeting the DC-SIGN carbohydrate-recognition domain.
Our results refine the phenotypic and functional characteristics of a GCB-DLBCL subset, in which the cell of origin has been selected to carry glycans terminating at high-mannose in the antigen-binding region. The acquisition of sites particularly in tumors harboring the t(14;18) translocation and mutations of epigenetic modifiers suggest a cell of origin common to FL, where these features occur early at transformation. Therefore, our data suggest the presence of a tumor cell ancestor with sIg glycans and genetic features common to FL and DLBCL. These results also document that those mannoses placed in the sIg variable region are functional and engage with DC-SIGN, to receive low level signals reminiscent of those protecting B cells from apoptosis. The possibility of inhibiting this antigen-independent interaction with anti-DC-SIGN antibodies in vitro suggests a potentially exploitable way for new therapeutic intervention.
Forconi:Menarini: Consultancy; Novartis: Honoraria; Janssen-Cilag: Consultancy, Honoraria, Other: Travel, Accommodations, Expenses, Speakers Bureau; Roche: Honoraria; Gilead Sciences: Research Funding; Abbvie: Consultancy, Honoraria, Other: Travel, Accommodations, Expenses, Speakers Bureau.
Author notes
Asterisk with author names denotes non-ASH members.