Unique Versus Common: Disease-Biased Immunoglobulin Gene Repertoires Along with Public Antigen Receptor Stereotypes in Marginal Zone B-Cell Lymphoproliferations

B cells residing the marginal zone (MZ) provide a first line of defense against blood borne pathogens, producing the greater part of circulating natural antibodies conferring protection against infection. Dysregulated homeostasis and function of MZ B cells has been implicated in a wide range of B lymphoproliferations, encompassing the distinct MZ lymphomas recognized by the WHO Classification, the related provisional entities and even chronic lymphocytic leukemia (CLL), for which a MZ derivation has been proposed. Here, taking advantage of a large multi-institutional series, we aimed at obtaining insight into the ontogenetic relationship of MZ lymphoproliferations, related entities and CLL through cross-comparison of their B cell receptor immunoglobulin (BcR IG) gene repertoires. Our sequence dataset included 3660 unique IGHV-IGHD-IGHJ gene rearrangement sequences from our collaborative centers and/or public databases derived from: (1) MZ lymphomas: splenic (SMZL), n=379; nodal (NMZL), n=37; extranodal (ENMZL), n=95; (2) provisional entities of postulated MZ origin, including splenic diffuse red pulp lymphoma (SDRL, n=16) and clonal B cell lymphocytosis of MZ origin (CBL-MZ, n=60); (3) persistent polyclonal B cell lymphocytosis (PPBL), n=286 (from 2 cases); (4) MZ cells isolated from six spleen specimens free of neoplastic cells at histological inspection (non-malignant MZ), obtained at surgery for cancer, n=489; (5) autoimmune conditions, n=1243; (6) various types of normal B cells, n=1055. The most pronounced IG gene repertoire skewing was observed in SMZL with the IGHV1-2*04 gene accounting for 26% of cases. Restrictions, though less striking, were also identified in the other MZ lymphomas as well: (i) the IGHV4-34 gene predominated in NMZL (14.3%); and, (ii) the IGHV1-69 gene predominated in ENMZL (14.6%), albeit with significantly different distribution depending on the primary site of involvement, ranging from 38% in salivary ENMZL to 11% in gastric ENMZL to 4% in ocular adnexa ENMZL (p<0.01). The vast majority of MZL cases showed at least some impact of somatic hypermutation (SHM), with the proportion of cases lacking any SHM ranging from 0% in salivary ENMZ to only 13% in SMZL. Following established bioinformatics approaches, we searched for stereotyped BcR IG sequences i.e. IGHV-IGHD-IGHJ gene rearrangements with restricted antigen-binding site sequence motifs. For the purposes of this analysis, the present sequence dataset was cross-compared to a large dataset of 20451 IGHV-IGHD-IGHJ gene rearrangement sequences from CLL patients from the IMGT/CLL-DB. Overall, 6437 different clusters with stereotyped BcR IG sequences were identified in the merged dataset, including from only 2 to more than 350 sequences. Two categories of clusters with stereotyped BcR IG were identified: disease-specific (n=4813) and 'mixed' (n=1624) i.e. comprised of cases with different diagnosis. The great majority of clusters in the former category concerned exclusively CLL and corresponded to well-established CLL stereotyped subsets, while only a small minority concerned exclusively MZ lymphomas, all with a diagnosis of SMZL. Mixed clusters were relatively small in size, with only 4 populated by more than 10 cases; of these, 2 utilized the IGHV1-69 gene, while the remaining 2 utilized the IGHV3-7 and IGHV4-59 gene, respectively. They comprised rearrangements from various entities, including SMZL, ENMZL (gastric, salivary gland, ocular adnexa), CLL, hepatitis C virus-associated diffuse large B cell lymphoma (DLBCL), but also rheumatoid factors and non-malignant spleen MZ cells. Notably, shared (recurrent) amino acid changes introduced by SHM (i.e. the same amino acid replacement at the same position) were identified in each mixed cluster. In conclusion, we document different immunogenetic signatures for MZ lymphomas, with limited overlap both amongst the various distinct and provisional WHO entities but also versus CLL. These findings indicate distinct antigen exposure histories and/or different (micro)environments underlying the ontogeny of MZ lymphomas. That said, the existence of rare public stereotypes raises the intriguing possibility that common, pathogen-triggered, immune-mediated mechanisms, may result in diverse B lymphoproliferations due to targeting versatile progenitor B cells and/or operating in particular microenvironments.