Somatic deficiency of the human E3 ubiquitin ligase CBL in leukocytes impairs B cell but not T cell development and function

Abstract

The E3 ubiquitin ligase Casitas B-lineage lymphoma (CBL) promotes positive selection and antigen responses in mouse T lymphocytes by ubiquitinating ZAP70. Conversely, mouse CBL and CBL-B mutually redundantly regulate SYK ubiquitination and B cell receptor signaling. Here we studied individuals with somatically homozygous CBL loss-of-function variants in leukocytes. Human CBL is largely redundant for the development and function of human T cells. Conversely, B cell development is altered at the immature stage, with a tenfold increase in transitional cells, enhanced survival of autoreactive clones and impaired tolerance manifested by autoantibody production. B cell maturation is intrinsically impaired by reduced apoptosis and dysregulated B cell receptor signaling. CBL deficiency impairs humoral immunity by limiting memory B cell formation and reducing class switching and somatic hypermutation. Consequently, antigen-specific B cell generation and adaptive immune memory are disrupted, predisposing individuals to infection. Human CBL is critical for B cell development and function but redundant for T cell biology.

Fig. 1. Cohort of 11 individuals with CBL-LOH with leukemia, autoinflammation and infections.

a, Family pedigrees of individuals with CBL-LOH. The participants are shown in dotted black, indicating somatic mosaicism; black vertical line, asymptomatic heterozygotes; the question mark (?) indicates unknown genotype; M, mutant. b, Allelic ratio of variants on chromosome 11 in participant P1 as determined by WES from whole-blood genomic DNA (gDNA). The CBL p.Y371C variant is marked as a red dot. c, Schematic illustration of the position of UPD on chromosome 11 in individuals for whom raw WES data were available. d,e, Quantitative genotyping by amplicon sequencing of patient tissues (d) and peripheral leukocyte subsets (e) targeting the relevant CBL variants. JMML, juvenile myelomonocytic leukemia; AML, acute myeloid leukemia; CMV, cytomegalovirus; BCG, Bacillus Calmette–Guerin; Mbp, mega-base pairs; mut, mutant; PMNs, polymorphonuclear cells. Source data

Fig. 2. Intact T cell development and function in individuals with CBL-LOH.

a, sjTREC quantification in HDs (black dots), heterozygous donors (gray dots) and individuals with CBL-LOH (colored dots), as determined by qPCR of whole-blood DNA; WBCs, white blood cells. b, Recent thymic emigrant CD4⁺ and CD8⁺ T cells quantified in peripheral fresh blood by mass cytometry and gating of CD31⁺ cells among naive T cells; data are shown as mean ± s.d. The statistical significance of differences was assessed by multiple two-sided Mann–Whitney tests, with correction for multiple testing; *P < 0.05 and **P < 0.005. c, Quantification of the indicated T cell subsets in the peripheral blood of HDs, heterozygous HDs and individuals with CBL-LOH of the indicated ages as determined by mass cytometry; data are shown as mean ± s.d. The statistical significance of differences was assessed by multiple two-sided Mann–Whitney tests, with correction for multiple testing; *P < 0.05 and **P < 0.005. In b and c, controls 0–3 years old (n = 2), controls 4–15 years old (n = 9), controls 16–100 years old (n = 28), pediatric participants (LOH) (n = 5), adult participants (LOH) (n = 2) and heterozygous individuals (n = 3). d, Percentage of dead cells in cultures of activated fresh PBMCs from HDs, the heterozygous father and participants (P1–P3) after 5 days of TCR stimulation, as determined by dead cell marker staining and flow cytometry; n = 3 HDs and n = 3 patient; data are shown as mean ± s.d. e,f, Cell division index of CD4⁺ (e) and CD8⁺ (f) T cells of HDs, the heterozygous father and participants (P1–P3) after 5 days of the indicated TCR stimulation, as determined by dilution of CFSE; n = 3 HDs and n = 3 patients; data are shown as mean ± s.d. g, Cytokine response by STAT5 phosphorylation (left) and cytokine production (right) by T cell blasts that are homozygous (red), heterozygous (gray) and homozygous WT for CBL Ub^LOF variants following the indicated stimuli. The bars show the mean of the displayed data points (one for each T blast line); NS, not significant; TEMRA, terminally differentiated effector memory T cells; MAIT, mucosal-associated invariant T cells; MFI, median fluorescence intensity; PHA, phytohemagglutinin; PMA, phorbol 12-myristate 13-acetate. Source data

Fig. 3. Dysregulated B cell development in individuals with CBL-LOH.

a, Counts of peripheral B cells in individuals with CBL-LOH over time compared with the healthy range. Data from eight participants are shown. Healthy ranges are from pediatric clinical recommendations. b, IgG levels in individuals with CBL-LOH (P1–P8) over time compared with the healthy range. Data from eight participants are shown. Healthy ranges are from pediatric clinical recommendations. c, Quantification of the indicated B cell subsets in the peripheral cryopreserved mononuclear cell blood of HDs, heterozygous HDs and individuals with CBL-LOH of the indicated ages as determined by mass cytometry; controls 0–3 years old (n = 2), controls 4–15 years old (n = 9), controls 16–100 years old (n = 28), pediatric participants (LOH) (n = 5), adult participants (LOH) (n = 2), heterozygous individuals (n = 3). Data are shown as mean ± s.d. The statistical significance of differences was assessed in multiple two-sided Mann–Whitney tests, with correction for multiple testing; **P < 0.005 and ***P < 0.0005. y.o., years old. d, Frequency of B cell subsets in cryopreserved PBMCs from HDs and individuals with CBL-LOH as determined by flow cytometry; HDs (n = 13), individuals with CBL-LOH (n = 5). Data are shown as mean ± s.d. The statistical significance of differences was assessed using multiple two-sided Mann–Whitney tests, with correction for multiple testing; *P < 0.05 and **P < 0.005. e, CD5, CD9, CD21 and CD38 expression on transitional B cells of HDs (black) and individuals with CBL-LOH (red) as determined by flow cytometry; HDs (n = 19), individuals with CBL-LOH (n = 6). Data are shown as mean ± s.d. The statistical significance of differences was assessed by multiple two-sided Mann–Whitney tests, with correction for multiple testing; *P < 0.05, **P < 0.005 and ****P < 0.00005. f, Percentage of CD21^hi cells among transitional B cells of HDs (black) and individuals with CBL-LOH (red) as determined by flow cytometry; HDs (n = 9), individuals with CBL-LOH (n = 5). Data are shown as mean ± s.d. The statistical significance of differences was assessed by Mann–Whitney test; ***P < 0.0005; gMFI, geometric mean fluorescence intensity. Source data

Fig. 4. Cell-autonomous defect in B cell maturation in CBL Ub^LOF cells.

a,b, Defective B cell maturation in the BM of individuals with CBL-LOH. Flow cytometry staining of cryopreserved BM mononuclear cells of HDs and participants P1, P2 and P3 with CBL-LOH. a, Representative flow staining of CD20 versus CD10 expression levels on CD34⁻CD19⁺ cells in BM samples. b, Quantification of these subsets for HDs (n = 8), individuals with CBL-LOH (n = 3) and individuals with PIK3CD^GOF (n = 3). The line shows the mean of the data points. Statistical significance was assessed using multiple two-sided Mann–Whitney tests corrected for multiple testing; *P < 0.05. c,d, In vitro differentiation of control AAVS1-edited and CBL-edited CD34⁺ HSPCs toward B cell identity. c, Flow cytometry staining for CD10 and CD20 among CD19⁺ cells in differentiation cultures after 3 weeks of coculture. d, Quantification of B cell ‘subsets’ based on flow cytometry marker expression (pre-BI, CD10⁺CD20⁻; pre-BII, CD10⁺CD20⁺; immature B, CD10⁺CD20⁺⁺; mature B, CD10⁻CD20⁺⁺) in this culture in control and two CBL-edited reactions. The lines show the means of three biological replicates, except for AAVS1 single guide RNA (sgRNA) 1, where two replicates are shown. Statistical significance was assessed using multiple two-sided Mann–Whitney tests corrected for multiple testing; **P < 0.005. e, Transcriptional overlap between CBL-edited and PI3K^GOF HSPC-derived B cell progenitors. Gene set enrichment analysis for PI3K^GOF gene signatures in CBL Ub^LOF samples is shown; NES, normalized enrichment score. No correction for multiple testing was performed for the two binomial tests. f, Quantitative genotyping by amplicon sequencing of B cell subsets and monocytes in individuals with CBL-LOH and HDs, as well as parents of P1–P3; HDs (n = 3), individuals with CBL-LOH (n = 5). Data are shown as mean ± s.d. g, CD38 staining intensity of primary B cell subsets of pediatric individuals with CBL-LOH compared with age-matched HDs; HDs (n = 9), individuals with CBL-LOH (n = 5). Data are shown as mean ± s.d. The statistical significance of differences was assessed by multiple two-sided Mann–Whitney tests, with correction for multiple testing; *P < 0.05. h, Rate of apoptosis following stimulation with daratumumab of CD19⁺ cells from in vitro differentiation cultures of control and CBL-edited HSPCs. Data show the mean of three technical replicates. The experiment is representative of three biological replicates. i, Rate of apoptosis following stimulation with daratumumab of control and CBL^Y371C KI REH cells. Each dot represents one biological replicate; n = 7. The statistical significance of differences was assessed using multiple paired, two-sided t-tests, with correction for multiple testing. j, Western blot of control and CBL^Y371C KI REH cells following stimulation with monoclonal anti-CD38 (daratumumab) for the indicated times (min). Source data

Fig. 5. Cell-autonomous defect in B cell function of mature CBL Ub^LOF B cells.

a, Ig production by sorted primary B cell subsets from HDs and individuals with CBL-LOH from cryopreserved PBMCs after the indicated stimulations. Supernatants were collected after 5–7 days, and Ig levels were measured by enzyme-linked immunosorbent assay (ELISA). The line shows the mean of the displayed data points (one per individual); HDs (n = 13), participants with CBL-LOH (n = 5). The statistical significance of differences was assessed using multiple two-sided unpaired t-tests, with correction for multiple testing; *P < 0.05, **P < 0.005 and ***P < 0.0005. b,c, Ig production by plasma cells sorted from cryopreserved BM mononuclear cells of participants P1 and P2 with CBL-LOH and HDs (n = 2; b) and IgM production of WT, CBL-KO and rescue BJAB cell lines within 24 h of culture (c); data are shown as mean ± s.d. of n = 5 independent biological replicates. d, Western blot of WT and CBL^Y371C KI BJAB cells following BCR stimulation with monoclonal anti-IgM for the indicated time periods. Data are representative of three biological replicates. e, Primary monocytes from HDs (n = 15) and participants with CBL-LOH (n = 6) were sorted from fresh blood samples. After 24 h of nonstimulated culture, supernatants were collected, and B cells from HDs were stimulated with the supernatants for 24 h. Ig production was assessed by ELISA. Data are shown as mean ± s.d. Statistical significance was assessed using multiple two-sided Mann–Whitney tests adjusted for multiple testing; **P < 0.005 and ***P < 0.0005. f, Frequency of IgA⁺ and IgG⁺ B cells among memory B cells in patients (n = 3) and HDs (n = 13), as determined by flow cytometry. Statistical significance was assessed by unpaired, two-sided t-tests; **P < 0.005 and ***P < 0.0005. g, Frequency of spike⁺ B cells in patients (n = 4) and HDs (n = 10), as determined by flow cytometry with tetramer staining. Data are shown as mean ± s.d. Statistical significance was assessed by unpaired, two-sided t-tests; *P < 0.05 and **P < 0.005. h, Frequency of IgG⁺, IgA⁺ and IgG⁻IgA⁻ B cells among spike⁺ B cells in patients (n = 4) and HDs (n = 6). Error bars indicate s.e.m. Statistical significance was assessed by unpaired, two-sided t-tests; *P < 0.05; EV, empty vector; Vinc, vinculin; wk, weeks; mo, months; yrs, years. Source data

Fig. 6. BCR repertoire of individuals deficient in CBL reveals a defect in immunoglobulin V gene usage and somatic hypermutation.

a,b, Usage of the top three IGHV IGHV4-34, IGHV4-59 and IGHV3-23 gene elements (a) and Ig JH4 and JH6 elements (b) in transitional, naive and memory B cells isolated from HDs and the indicated individuals deficient in CBL. c,d, CDR3 lengths in transitional, naive and memory cells isolated from HDs and individuals deficient in CBL (c) or CD10⁺CD20^dim HSPC-derived B cell progenitors edited at the AAVS1 or CBL locus (d). aa, amino acids; Trans, transitional. e, Frequency of Ig somatic hypermutations (SHM) in memory B cells defined by the expression of distinct class-switched Ig isotypes. f, Frequency of clones with different levels of somatic hypermutation within IgM⁺ memory B cells. g, CDR replacement:silent (R:S) ratios in IgM, IgG and IgA memory B cells. Statistical significance was assessed with a Wilcoxon test with Bonferroni correction for multiple testing (if needed; b–d and g) or Dunn’s test for multiple comparisons (e and f); *P < 0.05. Data shown were generated from BCR sequencing of HDs (n = 5) and individuals with CBL-LOH (n = 5). Boxes and whiskers indicate the median (center line), quartiles (box) and data range within 1.5× interquartile range (whiskers), and dots show data values beyond 1.5× interquartile range. All individual data points are shown. Source data

Fig. 7. Autoimmunity in individuals with CBL-LOH.

a,b, 9G4 staining of primary B cells from cryopreserved PBMCs from HDs and individuals with CBL-LOH (P1–P5) by flow cytometry. a, Median fluorescence intensity of the indicated B cell subsets in HDs (n = 9) and individuals with CBL-LOH (n = 5). Data are shown as mean ± s.d. Statistical significance was assessed using multiple two-sided Mann–Whitney tests adjusted for multiple testing; **P < 0.005. b, Frequency of 9G4^lo, 9G4^int and 9G4^hi cells among transitional (T), naive (N) and memory (M) B cells in HDs (n = 9) and individuals with CBL-LOH (n = 5). Data are shown as mean ± s.d. Statistical significance was assessed with multiple two-sided Mann–Whitney tests adjusted for multiple testing. c,d, Human protein microarray autoantibody detection. c, Protein microarray fluorescence intensity. The ratio of values for plasma from HDs (n = 8) to those from samples from individuals with CBL-LOH (n = 8) is shown. Data are shown as mean values from pairs of duplicates. One protein microarray was used per individual, and the results have been normalized to account for interexperiment variation. Fluorescence intensity is expressed in arbitrary units (a.u.). d, Reactivities for the indicated autoantigens common to P1–P8 and absent from eight age-matched control individuals. Serum samples from participants and blood donor controls were screened for IgG reactivity to 20,000 full-length human proteins on microarrays (HuProt). Red dots indicate autoantibodies previously shown to be associated with a clinical condition. Source data

Extended Data Fig. 1. Clinical Imaging of participants’ torsos.

(a) CT scan during pneumonia of P1 at 5 years of age. (b) CT scans of parenchymal lesions of P1, P2 and P3 at 9 years of age. (c) Pneumonia in X-ray imagine of P7 at 15 months old. (d) X-ray image of P10 at 7 months old with lymphadenopathy affecting the right axillary lymph node.

Extended Data Fig. 2. Humoral immunity of CBL deficient individuals and healthy donors.

(a) Serology for the indicated microbes of CBL-LOH patients, their relatives and healthy donors. The serology is indicated as the log₂(fold-change) over the positivity limit. (b) Virome-wide serological profiling (virscan) of CBL-LOH patients and healthy donors. Signals indicate the adjusted species score calculated from all peptide counts of the indicated species. Source data

Extended Data Fig. 3. Characterization and functional study of patients NK and T cells.

(a) CytoF immunophenotyping of patients total CD3⁺ T cells, NK cells and indicated NK cell subsets. Controls 0–3 y.o. (n = 2), Controls 4–15 y.o. (n = 9), Controls 16–100 y.o. (n = 28), pediatric patients (LOH) (n = 5), adult patients (LOH) (n = 2), heterozygous individuals (n = 3), mean ± s.d. Statistical significance was assessed with multiple Mann Whitney tests adjusted for multiple testing. *p<0.05. (b) Assessment of T cell proliferation of patients P1, P2 and P3. (top) gating strategy, (bottom) CFSE dilution plots. (c) Intracellular TNF production of CD4⁺ and CD8⁺ T cell blasts in homozygous (n = 4), heterozygous (n = 2) and wildtype (n = 5) state of CBL Ub^LOF variants. Line indicates the mean of the displayed datapoints. (d,e) Intracellular and extracellular cytokine production by patients T_fh cells stimulated with CD2/CD3/CD28 under T_h0 conditions for 5 days. (d) Intracellular staining of the indicated cytokines. Healthy donors (n = 17), CBL-LOH patients (n = 4). Mann-Whitney testing with correction for multiple testing did not reveal significant (p< 0.05) differences between healthy donors and CBL-LOH patients. Line indicates the mean of the displayed datapoints. (e) Extracellular detection of cytokines by ELISA. Healthy donors (n = 6), CBL-LOH patients (n = 3). Mann-Whitney testing with correction for multiple testing did not reveal significant (p< 0.05) differences between healthy donors and CBL-LOH patients. Line indicates the mean of the displayed datapoints. Source data

Extended Data Fig. 4. Characterization of B cell biology in CBL deficient participants and cells.

(a) Quantification of B-cell subsets in cryopreserved PBMCs from healthy donors, heterozygous carriers, and CBL-LOH patients of the indicated ages by mass cytometry. Controls 0–3 y.o. (n = 2); controls 4–15 y.o. (n = 9); controls 16–100 y.o. (n = 28); pediatric CBL-LOH (n = 5); adult CBL-LOH (n = 2); heterozygous individuals (n = 3). Mean ± s.d. Statistical significance was assessed using two-sided Mann–Whitney tests with correction for multiple testing. (b) Frequency of B-cell subsets in healthy donors (n = 10) and CBL-LOH patients (n = 4) by flow cytometry. Mean ± s.d. Significance was evaluated with two-sided Mann–Whitney tests adjusted for multiple comparisons. **p < 0.005. (c) MFI of B-cell markers on CD21^lo and transitional B cells from healthy donors (n = 6) and patients P1–3 (n = 3). Lines show means. (d–f) Modeling the CBL ΔExon8 variant in primary human CD34⁺ HSPCs. (d) Agarose gel electrophoresis of PCR products from AAVS1 or CBL loci 72 h after nucleofection; representative of >5 biological replicates. (e) Editing efficiency using sgRNA pair 1+2 by NGS, showing ~80% of the ~400 bp exon-8 deletion. (f) Western blot of CBL protein levels after CBL or AAVS1 editing. (g) Quantification of three biological replicates from (f). Mean ± s.d. (h) NGS quantification of exon-8 deletions for all three guide pairs (n = 2 biological replicates). Bars show means. (i) Editing efficiencies at days 8 and 21 in differentiation cultures by NGS. Mean ± s.d. from three biological replicates. (j–m) Bulk RNA-seq of AAVS1- and CBL-edited CD19⁺CD10⁺CD20^low HSPC-derived B-cell progenitors. (j) Gene-set enrichment analysis showing significantly enriched (red) or depleted (blue) pathways (NES: normalized enrichment score). (k,l) Differential expression of leading-edge genes in the (k) Hallmark G2–M checkpoint and (l) Hallmark mTORC1 signaling gene sets (Z-transformed normalized counts). (m) Transcriptional overlap between CBL-edited and PI3K^GOF progenitors; Venn diagrams show shared significantly up- or downregulated genes, with overlap significance by binomial test. Source data

Extended Data Fig. 5. Intact cell extrinsic determinants of B cell development and function.

(a) Levels of sCD40L, APRIL and BAFF in the plasma of CBL-LOH patients (n = 6) and healthy donors (n = 14). Line shows the mean of the displayed datapoints (one point per individual assessed). (b) Quantification of the production of sCD40L, APRIL and BAFF by PBMCs of CBL-LOH patients (n = 3) and healthy donors (n = 5). Line shows the mean of the displayed datapoints (one point per individual assessed). (c-f) Impact of sCD40L levels on B cell differentiation in vitro using CD34+ HSPCs. (c,d) B cell output at day 21 of co-culture of CD34+ HSPCs from two healthy donors. The MS5 co-culture was supplement with IL-7 (20 ng/mL) and the indicated doses of sCD40L. Mean of dots that represent technical replicates. (e,f) Quantification of B cell subsets based on CD10 and CD20 marker expression. Cells were treated as in (c,d). Mean ± s.d. (g) IL-21 production by sorted CD4⁺ naïve and memory T cells upon the indicated stimuli. Mean ± s.d. (h) Surface staining of CD38 expressed on CD19+ HSPC-derived B cell progenitors edited with the indicated sgRNAs. Mean ± s.d. of three biological replicates. (i) Surface staining of CD38 expressed on HEK293T, REH and BJAB cells as determined by flow cytometry. Mean ± s.d. of three biological replicates. (j) CBL Y371C KI BJAB (n = 9) cells are not more resistant than WT (n = 5) cells to BCR-induced apoptosis. Mean ± s.d. of clones over three independent experiments. Source data

Extended Data Fig. 6. Assessment of signalling pathway activation in CBL deficient cells.

(a-f) Western blots of wildtype and CBL Y371C KI REH cells upon CD38 crosslinking with daratumumab for the indicated time periods (min/h). (b,c, e,f) Shows quantifications of three biological replicates. Mean ± s.d. Statistical significance was assessed with Mann Whitney tests. *p<0.05. (g,h) ERK phosphorylation upon CD38 crosslinking in gene-edited HSPC-derived CD19+ B progenitors. Total CD19+ (g) or CD19+CD20high (h) B progenitors were sorted from co-cultures edited at the AAVS1 or CBL locus. Cells were stimulated for 15 min with daratumumab, followed by fixation, permeabilization and intracellular staining for phosphorylated ERK (pERK). n = 2 biological replicates. Bars show the mean. Source data

Extended Data Fig. 7. Functional phenotyping of mature CBL deficient B cells.

(a,b) Transitional B cells (n = 4 independent experiments) and naïve B cells (n = 12 independent experiments) were sort-purified from PBMCs of healthy donors and then cultured in vitro. Levels of IgM, IgG and IgA were measured in supernatants from transitional and naïve B cells after 7 days following stimulation with: (a) CD40L alone or in combination with CpG and/or BCR crosslinking, or (b) CD40L and IL-21. Box and whiskers indicate median (central line), quartiles (box) and deciles (whiskers). (c) Western blot of CBL KO and stable lentiviral overexpression on BJAB cells. (d) Sanger genotyping of BJAB CBL Y371C KI and control wildtype cells. (e) Ig production of control wildtype and CBL Y371C KI BJAB cell lines within 24 h of culture unstimulated (n = 4 biological replicates) and upon IL-4 stimulation (n = 3 biological replicates). Mean ± s.d. The statistical significance of differences was assessed in multiple two-sided Mann-Whitney tests, with correction for multiple testing. *p<0.05. (f) Quantification of three biological replicates of the western blot shown in Fig. 5e. Mean ± s.d. Statistical significance was assessed with Mann Whitney tests. *p<0.05. (g) Pathway enrichment analysis of bulk RNA-sequencing of healthy donor and CBL LOH patients’ primary naïve B cells directly after sorting from cryopreserved PBMCs or after 24 h of non-stimulated culture. NES: normalized enrichment ratio. p values were adjusted for multiple testing. (h) Immunoglobulin production by sorted primary B cell subsets of healthy donors (n = 10) and CBL-LOH patients (n = 5) from fresh blood samples without stimulation. Supernatants were collected after 24 h and Igs were measured by ELISA. Line indicates the mean of the displayed datapoints (one point per individual tested). Statistical significance was assessed using Mann Whitney tests and correction for multiple testing. *p < 0.05, ***p< 0.0005. Source data

Extended Data Fig. 8. Profiling of humoral immunity and autoimmunity.

(a) Flow cytometry for Spike tetramers on primary B cells of healthy donors and patient P2. (b) Cell painting of healthy donor transitional and memory B cells by the indicated plasma samples. (c) Validation of presence of anti-TXLNB autoantibodies in three CBL-deficient patients by multiplex bead assay. Line indicates the mean of the displayed datapoints (one point per individual tested). Source data