Lymphoma B Cell Receptors in Rheumatoid Arthritis Bind Modified Proteins

Diffuse large B cell lymphomas arising in patients with rheumatoid arthritis carry B cell receptors that frequently bind inflammation-associated protein modifications, and in several cases show strikingly broad polyreactivity, according to a study published in Arthritis Research & Therapy. The findings, based on sequencing tumour tissue from a unique Swedish cohort and re-expressing the lymphoma receptors as recombinant antibodies, suggest a possible mechanistic link between chronic inflammatory burden and lymphoma development in this autoimmune disease.

Rheumatoid arthritis patients are known to have a two- to threefold increased risk of B cell lymphoma compared with the general population, and high cumulative disease activity is considered the strongest risk factor. The importance of inflammation is underscored by recent evidence that the introduction of effective biological disease-modifying antirheumatic drugs may have reduced lymphoma risk in these patients. Most rheumatoid arthritis patients who develop lymphoma have diffuse large B cell lymphoma, which is often of the activated B cell-like or non-germinal centre subtype, a category whose cell lines have been shown to depend on tonic B cell receptor signalling for survival.

In total, 22 patients with diffuse large B cell lymphoma were included, 45 percent of them male, with a median age of 70 years. All fulfilled the 2010 ACR/EULAR classification criteria for rheumatoid arthritis and had established disease of more than two years duration, with a median of 21 years. Lymphomas were diagnosed between 2010 and 2021.

Because the tumour tissue had been preserved as formalin-fixed paraffin-embedded blocks, DNA quality varied considerably. Only nine of the 22 tumour samples yielded DNA of sufficient quality for B cell receptor sequencing, and the nine sequenced patients had a median age of 72 years at lymphoma diagnosis. Quantitative polymerase chain reaction testing detected Epstein-Barr virus in the tumour tissue of three of the 22 patients, or 14 percent, a figure slightly higher than the 5 to 9 percent reported for diffuse large B cell lymphoma in general but in line with studies of lymphoma in elderly populations. Among the nine sequenced patients, seven had anti-CCP2 antibodies or rheumatoid factor in their serum.

Using an in-house multiplex polymerase chain reaction protocol followed by Illumina sequencing, the team amplified full-length recombined heavy and light chain variable regions directly from the fixed tumour sections. Heavy chain sequencing succeeded for all nine patients and light chain sequencing for seven. For six patients, a dominant heavy chain clonotype could be identified, accounting for 66 to 99 percent of the heavy chain sequencing data. Within these dominant lineages, the most frequent unique sequence represented 14 to 71 percent of the total reads. No bias in variable gene usage was detected: five of the lymphoma heavy chains belonged to the VH3 family and one patient carried a dominant VH1-2 gene, and notably none of the tumour receptors were encoded by VH4-34, a gene that other studies have found enriched in diffuse large B cell lymphoma.

The sequencing also revealed evidence of ongoing evolution within the tumours. Somatic hypermutation patterns in three patients showed amino acid replacements concentrated in both the complementarity-determining regions and the framework regions, a distribution the authors interpret as indicative of antigen-driven selection. In three patients, multiple prevalent amino acid variants occurred at frequencies of 15 to 35 percent within the clonotype, with recurrent conversions at specific framework residues, providing evidence of intra-clonal evolution. The authors caution that the multiplex polymerase chain reaction and sequencing approach introduces errors that can mimic diversity, and that true intra-clonal variation can only be definitively established by single-cell methods. Still, previous work has associated ongoing B cell receptor mutation within lymphomas with poorer prognosis.

The pivotal step came when the dominant paired heavy and light chain sequences from six patients were synthesized and expressed as recombinant human IgG1 monoclonal antibodies, allowing the actual binding properties of the lymphoma receptors to be measured. The antibodies were screened in an in-house polyreactivity assay using a soluble membrane protein fraction from HEK293 cells, a panel of enzyme-linked immunosorbent assays covering modified and unmodified antigens, a commercial anti-CCP2 assay, a rheumatoid factor assay, Hep-2 cell immunofluorescence for anti-nuclear reactivity, and a custom solid-phase microarray containing more than 400 full-length human proteins.

The results showed a consistent pattern of autoreactivity, but not of the classical rheumatoid arthritis type. All six lymphoma antibodies were negative for anti-CCP2 reactivity, and none showed significant anti-nuclear staining. Five of the six, however, bound proteins modified by malondialdehyde or malondialdehyde acetaldehyde, adducts generated during oxidative stress and lipid peroxidation. The antibodies from patients RAL01, RAL03, RAL09 and RAL10 bound malondialdehyde-modified low-density lipoprotein as well as malondialdehyde- and malondialdehyde acetaldehyde-modified albumin. When the researchers blocked formation of the cyclic malondialdehyde acetaldehyde adducts with cyanoborohydride, binding by the RAL03 and RAL09 antibodies dropped significantly, indicating preferential recognition of the acetaldehyde-derived adducts, consistent with previously characterized anti-malondialdehyde antibodies from rheumatoid synovium. Two clones stood out for their specificity: the RAL03 antibody bound malondialdehyde-modified antigens in a pattern comparable to a well-characterized control antibody, while the RAL04 antibody bound almost exclusively to methylated protein.

The remaining antibodies were strikingly promiscuous. The two lymphoma B cell receptors from patient RAL10, which differed by a single amino acid at position 37 in framework region 2, where a valine in the germline version was replaced by leucine, both showed broad polyreactivity, with the leucine variant binding 71 percent of the antigens on the array and the valine variant 27 percent, compared with 86 percent for established polyreactive control antibodies and under 2 percent for negative controls and the non-polyreactive lymphoma clones. The RAL01 and RAL09 antibodies also displayed strong polyreactivity on the array, binding 27 to 70 percent of antigens, and the RAL01 antibody additionally bound rabbit IgG and unmodified bovine serum albumin. Notably, the leucine variant was more polyreactive than its valine counterpart despite differing by only one residue of similar biochemical character, and both variants were present among expanded sequences within the same tumour, which the authors suggest may indicate an evolutionary drive related to polyreactive binding. No clear association between Epstein-Barr virus positivity and receptor reactivity could be established, since only one of the highly polyreactive clones came from virus-positive tissue, and the material was too limited for conclusions.

The authors propose that high and chronic inflammation in rheumatoid arthritis may increase the burden of modified self-antigens such as malondialdehyde adducts, driving chronic activation of polyreactive and anti-malondialdehyde B cells that, with disease-associated reduced tolerance, could enter germinal centres and, under persistent stimulation, be at higher risk of malignant transformation into diffuse large B cell lymphoma. They emphasize that IgG antibodies against malondialdehyde-modified proteins are elevated in rheumatoid arthritis patients and correlate with disease activity and inflammatory markers, and that a subset of chronic lymphocytic leukemia has similarly been shown to carry receptors with anti-malondialdehyde acetaldehyde reactivity. The team also acknowledges the study’s central limitation: only nine patients could be sequenced and six receptors expressed, a small sample in a rare and unique cohort. Within that constraint, the demonstration that lymphoma receptors in these patients are polyreactive or directed against oxidation-associated protein modifications, while remaining distinct from classical rheumatoid autoantibodies, offers a concrete molecular thread connecting long-standing inflammation to the biology of lymphoma cells surviving in an inflammatory environment.

Subject of Research: B cell receptor autoreactivity and polyreactivity in diffuse large B cell lymphoma arising in rheumatoid arthritis

Article Title: Diffuse large B cell lymphoma in rheumatoid arthritis express polyreactive and autoreactive BCRs binding protein modifications

Article References: Euler, N., Gonzalez-Sanchez, A. C., Clark, E., Mildner, H., Galesic, M., Kemp, A., Jangard, M., Hellbacher, E., Landegren, N., Diaz-Gallo, L. M., Klareskog, L., Malmström, V., Baecklund, E., & Grönwall, C. (2026). Diffuse large B cell lymphoma in rheumatoid arthritis express polyreactive and autoreactive BCRs binding protein modifications. Arthritis Research & Therapy, 28(1), Article 188. https://doi.org/10.1186/s13075-026-03937-0

Image Credits: AI Generated

DOI: 10.1186/s13075-026-03937-0

Keywords: rheumatoid arthritis, diffuse large B cell lymphoma, B cell receptor, polyreactivity, malondialdehyde, autoantibodies, somatic hypermutation, Epstein-Barr virus, inflammation, AUTO-LYMPHOMA cohort, recombinant monoclonal antibodies, lymphomagenesis