Frequent somatic mutations of <i><scp>KMT</scp>2D</i> (<i><scp>MLL</scp>2</i>) and <i><scp>CARD</scp>11</i> genes in primary cold agglutinin disease

Primary chronic cold agglutinin disease (CAD) is a rare type of haemolytic anaemia mediated by monoclonal IgM anti-I autoantibodies. The antibodies bind to erythrocytes at low temperatures, mostly in acral parts of the body, causing agglutination and complement activation (Berentsen et al, 2015). CAD represents 15% of all cases of autoimmune haemolytic anaemia and has an incidence of 1 per million per year. Although CAD is characterized by monoclonal gammopathy, patients typically display no signs or symptoms of lymphoma or myeloma at diagnosis. Cold-agglutinins detected in patients with lymphoma, such as lymphoplasmacytic lymphoma, are not included in the definition of CAD. CAD does not progress to myeloma or lymphoma (Berentsen et al, 2015). However, clonal B-cells are demonstrated in the bone marrow of most patients. More recent studies, including flow cytometry analysis, demonstrated a homogeneous histology and immunophenotype of the bone marrow B-cell lymphoproliferative disease (Figure S1) (Randen et al, 2014). Furthermore, the MYD88 L265P mutation, typical of lymphoplasmacytic lymphoma, is not present in CAD (Randen et al, 2014; Malecka et al, 2016). Together, the clinical and pathology findings suggest that CAD patients have a unique low-grade clonal B-cell lymphoproliferative disease of the bone marrow. The current study was undertaken to find molecular changes in CAD-associated lymphoproliferative disease that could be potentially useful for diagnosis or targeted treatment. Clonal B-cells and normal T cells, used as control, were purified from bone marrow samples of 16 CAD patients (Table SI) using fluorescence-activated cell sorting (Figure S2, Table SII). Exome sequencing was performed in six cases and results were validated using targeted sequencing of 10 cases (Tables 1 and SIII). The total number of mutations detected and the coverage are given in Tables SIV and SV. All mutations were verified by Sanger sequencing (Figures S3 and S4). Materials and Methods are available online as Supplementary Data. Recurrent somatic mutations of KMT2D and CARD11, but not other genes, were demonstrated. Somatic KMT2D mutations were detected in 11 out of 16 patients (69%). Out-of-frame deletions or insertions, gained stop codons or splice donor variants were demonstrated in seven patients (Table 1, Fig 1A). The mutations occurred N-terminally from the KMT2D SET domain, with lysine-(K) specific methyltransferase (KMT) activity. The mutations result in an enzymatically inactive truncated protein. KMT2D SET domain-inactivating mutations are not unique to CAD, but are frequent in follicular lymphoma, nodal marginal zone lymphoma and diffuse large B-cell lymphoma (DLBCL). In two patients, missense mutations located in the C-terminal domain at Arg5048 or Cys5481 residues were identified (Table 1), which may also impair the activity of the SET domain or result in KMT loss-of-function. Similar mutations are found in Kabuki syndrome (Ng et al, 2010; Makrythanasis et al, 2013). Kabuki syndrome is a congenital autosomal dominant disorder caused by KMT2D malfunction (Ng et al, 2010). Additionally, two patients showed KMT2D mutations in splice regions. The variant allele frequency range of KMT2D was approximately 40–60% (Figures S3 and S4), indicating that only one allele was mutated, as previously demonstrated in lymphoma and in Kabuki syndrome, demonstrating that haplo-insufficiency affects normal KMT2D function. Loss of KMT2D function increases B cell proliferation and impedes class switch recombination (Ortega-Molina et al, 2015; Zhang et al, 2015). Loss of KMT2D function might act in concert with B cell survival signals induced by stimulation of the auto-reactive IGHV4-34-encoded immunoglobulin receptor typically expressed on the surface of CAD B cells. A growth advantage through auto-antigen stimulation of the IGHV4-34-encoded B-cell receptor was recently demonstrated for a subset of DLBCL. CARD11 was somatically mutated in 5 out of 16 patients (31%). All mutations were classified as of moderate impact by the SnpEff program (http://snpeff.sourceforge.net/). Three samples had a missense mutation, one an in-frame deletion and one an in-frame insertion. The missense mutations are classified as pathogenic by Cosmic (http://cancer.sanger.ac.uk/cosmic) (Table 1). The five CARD11 mutations were located within a 20 bp stretch of exon 6, coding for the BAR domain of the coiled-coil region of CARD11 (Fig 1B). Similar mutations were previously demonstrated in DLBCL and result in constitutive nuclear factor-κB activation (Lenz et al, 2008). The variant allele frequency of CARD11 was approximately 40–60% for the five patients (Figures S3 and S4). Mono-allelic CARD11 coiled-coil domain mutations are not oncogenic per se in mice, but result in B-cell proliferation and auto-antibody production. Four of the five CARD11 mutations were detected in patients with a concurrent KMT2D mutation. The two mutated genes might act in concert. The demonstration of recurrent KMT2D and CARD11 mutations in CAD suggests that targeted treatment might be attempted. Histone deacetylase (HDAC) inhibitors, counteracting diminished KMT2D function, have also been used for treatment of lymphoma and myeloma (Imai et al, 2016). Also, HDAC inhibitors have recently been tested in models of Kabuki syndrome. Targeted therapy to counteract the effect of CARD11 gain-of-function mutations is currently also being evaluated for use in DLBCL (Young & Staudt, 2012). KMT2D and CARD11 mutations, combined with previously described pathology findings that are distinct from other lymphoproliferative diseases, including lymphoplasmacytic lymphoma, with which it is most often confounded (Randen et al, 2014; Malecka et al, 2016), establishes CAD-associated B-cell lymphoproliferative disease as a unique disease. Lymphoplasmacytic lymphoma displays the MYD88 L265P mutation in more than 90% of cases, but does not exhibit KMT2D or CARD11 mutations. MYD88 L265P mutation is not found in CAD-associated lymphoproliferative disease (Malecka et al, 2016). Our results suggest therefore that KMT2D and CARD11 mutations may help to diagnose CAD and better distinguish it from lymphoplasmacytic lymphoma. However, recurrent KMT2D and CARD11 mutations need to be verified in larger CAD case series. It also remains to be demonstrated whether additional genes, such as CXCR4, mutated in one case in this series and previously demonstrated to be frequently mutated in lymphoplasmacytic lymphoma in addition to MYD88, are recurrently mutated. CAD-associated lymphoproliferative disease is an indolent disease (Berentsen et al, 2015), that does not progress to systemic lymphoproliferative disease. Therefore, the term CAD-associated B-cell lymphoproliferative disease instead of CAD-associated B-cell lymphoma is suggested. This study was supported by South-Eastern Norway Regional Health Authority and by the Norwegian Cancer Society. AM, GT, AT, SB, GET and JD designed the study. AM, GT, IØ, JM and JW performed the analyses. AT, SB, GET and JD supervised the study. AT, UR and JD reviewed the diagnostic patient samples, GET and SB collected the clinical data. AM, GT, JM and JD prepared the manuscript. All authors have critically read the manuscript. The authors have no competing interests. Table SI. Clinical characteristics of CAD patients. Table SII. Flow cytometry analysis of bone marrow and blood samples from CAD. Table SIII. Non-synonymous mutations in monotypic B cells from CAD patients detected by exome sequencing. Table SIV. Summary of somatic mutations in monotypic B cells from CAD patients detected by exome sequencing. Table SV. Coverage of exome sequenced samples from CAD patiens. Fig S1. The figure illustrates the typical pathology of CAD-associated B-cell lymphoproliferative disease. Fig S2. Sorting strategy for isolation of monoclonal B cells from bone marrow by flow cytometry. Fig S3. Examples of mutations detected by exome or targeted sequencing of clonal B cells. Fig S4. Examples of NGS analysis and Sanger sequencing verification of somatic mutations in CARD11 and KMT2D genes. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.