UBR5 HECT domain mutations identified in mantle cell lymphoma control maturation of B cells

Mantle cell lymphoma (MCL) is a rare, aggressive form of non-Hodgkin lymphoma (NHL).¹  Although MCL represents only ∼6% of NHL lymphoma cases, it has one of the highest mortality rates of all lymphomas with only a 50% 5-year survival.²  Given the high mortality rate and propensity for recurrence, understanding mutations found in MCL and disease development in B cells will open avenues for identifying new therapies. Recently, monoallelic mutations in the ubiquitin protein ligase E3 component n-recognin 5 (UBR5) were found in ∼18% of patients with MCL.^3,4  Approximately 60% of mutations identified in UBR5 were frame shift mutations within its HECT domain, which can accept and transfer ubiquitin molecules to the substrate, leading to a premature stop codon before the cysteine residue associated with ubiquitin transfer.

UBR5 is a large ∼300 kDa protein HECT E3 ligase with a conserved carboxyl terminal HECT domain. In HECT E3 ligases, the N-terminal portion (N-lobe) of the enzyme interacts with E2 ubiquitin-conjugating enzymes and determines substrate specificity, whereas the C-terminal HECT domain (C-lobe) contains a catalytic cysteine residue that binds ubiquitin.⁵  The 2 lobes are connected by a flexible linker that allows for shifting orientation between N- and C-lobes during ubiquitin transfer to allow for efficient movement of ubiquitin from the E3 ligase to the substrate protein. UBR5 regulates a number of cellular processes including metabolism, apoptosis, angiogenesis, gene expression, and genome integrity.^6-11  Overexpression of UBR5 has been found in a number of cancers including ovarian, breast, hepatocellular, squamous cell carcinoma, and melanoma.^12-15 

Determining if, and at what stage (transcriptional, translational, or proteomic) UBR5 controls maturation of B cells is important for fully understanding B-cell development and lymphoma transformation. To elucidate the role of UBR5 in B-cell maturation and activation, we generated a conditional mutant disrupting the C-terminal HECT domain mimicking mutations found in MCL. Loss of the HECT domain leads to a block in maturation of B cells with follicular B cells that are phenotypically abnormal with low expression of immunoglobulin D (IgD) and high expression of IgM. Upon immune stimulation, B cells lacking the HECT domain show decreased germinal center (GC) formation and reduced antibody producing plasma cells, suggesting functional defects. Proteomic studies reveal upregulation of proteins associated with messenger RNA (mRNA) splicing via the spliceosome and indicates that UBR5 interacts with splicing factors (SF3B3, PRPF8, DHX15, SNRNP200, and EFTUD2). Our studies reveal a novel role of UBR5 in B-cell maturation and suggest UBR5 mutations in MCL contribute to disease initiation.

Ubr5 HECT mutant (Ubr5^ΔHECT) mice were developed in C57BL/6 mice using Easi-CRISPR as previously published¹⁶  and crossed with E2A^CRE mice (Jackson Laboratory, Bar Harbor, ME) or Mb1^CRE mice (kind gift of Dr. Michael Reth) on C57BL/6 background. Eμ-CyclinD1 mice were a kind gift from Dr. Samuel Katz.¹⁷  Targeted alleles were validated by polymerase chain reaction (PCR). For immune stimulation, mice were injected intraperitoneally with 1 × 10⁸ sheep red blood cells (SRBC) (Innovative Research, Novi, MI). All mice were housed in a pathogen-free facility and procedures were approved by Institutional Animal Care and Use Committee of University of Nebraska Medical Center in accordance with National Institutes of Health guidelines.

Total bone marrow (BM) and splenocytes were isolated from 6-week-old mice. B220⁺ or CD23⁺ cells were isolated using MojoSort Streptavidin Nanobeads (BioLegend, San Diego, CA) following manufacturer’s protocol. B cells were cultured in RPMI1640 (Hyclone/GE Healthcare, Chicago, IL), 10% fetal bovine serum, 2 mM l-glutamine (Corning, Corning, NY), 50 µM 2-mercaptoethanol (Corning), 20 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (Hyclone/GE Healthcare), 1X penicillin/streptomycin, and 10 µg/mL lipopolysaccharide (MilliporeSigma, Burlington, MA). Cells were treated with 200 µg/mL cycloheximide (MilliporeSigma) and/or 20 µg/mL MG132 (Cell Signaling Technologies, Danvers, MA) for 4 hours. For irradiation experiments, CD23⁺ splenocytes were given 10 Gy X-ray and collected after 1-hour recovery.

For flow cytometry, cells were stained for 1 hour in 3% fetal bovine serum in phosphate-buffered saline (PBS). For cell-cycle analysis, cells were fixed and permeabilized following BioLegend intracellular staining protocol and stained with Ki67 and DAPI (4′,6-diamidino-2-phenylindole).

Mice were bled on days 0 and 8 following immune stimulation. For enzyme-linked immunosorbent assay (ELISA), serum was collected following Abcam ELISA sample preparation guide, diluted 1:10 with PBS. SRBC was diluted 1:10 as a control. ELISA was performed with positive reference antigen mixture and PBS as a negative control according to manufacture (BD Pharmingen, San Jose, CA). ELISpot was performed using MabTech Kit for mouse IgG ELISpot^BASIC (ALP; Cincinnati, OH) following manufacturer’s protocol. A total of 1.76 × 10⁶ SRBC and 75 000 splenocytes were plated on polyvinyl fluoride plates. BCIPNBT substrate was from Promega (Madison, WI).

Spleen sections were stained with UBR5 antibody and Ki67. For GC analysis, spleen sections from immunized mice were stained with biotinylated peanut agglutinin (PNA) antibody.

Nuclei from 293Ts (ATCC, Manassas, VA) were collected as described previously.¹⁸  Nuclei were lysed in 10 mM Tris-HCl pH7.4, 0.2 mM MgCl₂, and 1% Triton-X 100 containing protease and phosphatase inhibitors for 15 minutes at 4°C. Continuous 10% to 30% glycerol gradients were prepared as previously described.¹⁹  Samples were spun at 28 000g for 13 hours at 4°C.

Total RNA was harvested using QIAGEN RNeasy Kit (QIAGEN, Hilden, Germany). Complementary DNA (cDNA) was synthesized using High Capacity RNA-to-cDNA Kit (Thermo Fisher Scientific, Waltham, MA), followed by quantitative real-time PCR (qRT-PCR) using iTaq Universal SYBR Green (BioRad, Hercules, CA).

For global proteome quantification, B220⁺ splenocytes were isolated from 3 mice per genotype. Samples were prepared and tandem mass tag (TMT) labeled (TMT10plex Mass Tag Labeling Kits; Thermo Fisher Scientific) as previously described.²⁰  Data are available via ProteomeXchange with identifier PXD014307.

For immunoprecipitation, cells were lysed in 20 mM Tris pH 7.5, 150 mM NaCl, 1 mM EDTA, and protease and phosphatase inhibitors. Immunoprecipitation was performed overnight at 4°C using anti-UBR5 antibody (Abcam, Cambridge, MA) or rabbit IgG control (Cell Signaling Technologies, Danvers, MA) and using protein A agarose beads (Cell Signaling Technologies) for IgG pulldown.

All experiments were performed in triplicate unless noted and statistical analyses were performed using an unpaired, 2-tailed Student t test. *P < .05, **P < .01, ***P < .001, ****P < .0001.

Antibodies, primers, and additional methods are provided in supplemental Data, available on the Blood Web site.

Meissner et al originally identified UBR5 mutations in ∼18% of MCL patients.⁴  In recent cross-sectional genomic profiling of multiple lymphoma subtypes, we identified UBR5 as 1 of 8 genes that had a significantly higher frequency of mutation in MCL compared with other lymphoma subtypes. These monoallelic mutations were observed in 34 of 196 MCL tumors (15.8%) and 15 of 559 tumors (2.7%) of other subtypes. However, mutations within the HECT domain were found only in MCL tumors (Figure 1A).³  Patients with UBR5 mutations had between 4 and 12 additional mutations, but did not significantly cooccur with other mutations (Figure 1B).³  These findings suggest that HECT domain mutations of UBR5 are a disease-specific genetic feature of MCL.

Because the role of Ubr5 in lymphopoiesis is unknown, we evaluated the expression of Ubr5 in B-cell subgroups during development by purifying pro-, pre-, and immature B cells from BM of C57/BL6 wild-type (WT) mice. Additionally, transitional, follicular, and marginal zone B-cell populations were purified from spleens. The pro B-cell population showed lowest expression of Ubr5 at both the RNA and protein level, whereas the highest expression of Ubr5 was found in mature splenic populations (follicular and marginal) (Figure 1C-D). These studies imply a role for UBR5 during B-cell development.

To understand the role of UBR5 mutations in B-cell development, we generated a conditional allele targeting exon 58 that would lead to a truncated protein lacking the cysteine required for ubiquitin conjugation in the HECT domain and mimicking mutations found in MCL patients (Figure 1E).²¹  We first crossed our Ubr5^ΔHECT mice to E2A^CRE mice that delete in early embryogenesis.^22,23 E2A^CRE;Ubr5^ΔHECT/WT mice were fertile and showed no physical abnormalities. In contrast to Ubr5-null mice that are not viable past E10.5,²²  3 of 49 pups found at birth were E2A^CRE;Ubr5^{ΔHECT/ΔHECT}; however, the pups were not viable, indicating the HECT domain is not required for embryogenesis and fetal hematopoiesis (supplemental Figure 1A-B).

Because E2A^CRE;Ubr5^{ΔHECT/ΔHECT} mice die before or at birth, to study B-cell development we crossed Ubr5^{ΔHECT/ΔHECT} mice to Mb1^WT/CRE, which truncates UBR5 in early B-lymphocytes (supplemental Figure 1C).²⁴  qRT-PCR showed no decrease in Ubr5 expression, but significant decrease in Ubr5 HECT domain expression (Figure 1F). Because of the size of UBR5 protein (∼300 kDa) we are unable to determine changes in protein size of the HECT mutant (∼290 kDa); therefore, we isolated RNA from splenic B cells from Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice and WT littermates and amplified cDNA followed by Sanger sequencing. Sequencing confirmed loss of exon 58 along with expected frame shift and early stop codon (Figure 1G).

Early B-cell development occurs within the BM. In Mb1^WT/CRE;Ubr5^ΔHECT/WT and Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice, the number of total BM cells and frequency of B220⁺ B cells showed no significant difference compared with WT littermates (Figure 2A-C). Further analysis of specific subtypes of B cells revealed a striking decrease in IgD⁺ mature B-cell populations in both Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT}and Mb1^WT/CRE;Ubr5^ΔHECT/WT mice (Figure 2D-F). Decreases in mature B cells were compensated in BM by a slight increase in pro/pre B-cell population, specifically in pro-B cells in the Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} cohort (Figure 2G-I). These studies demonstrate an impact to mature cells within BM, as well as changes to composition of pro-B cells following loss of HECT domain.

Following development in BM, B cells migrate to the spleen where they undergo maturation and activation.²⁵ Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice have smaller spleens and a reduction in total splenocytes (Figure 3A; supplemental Figure 2A). Although mice had a reduction of splenocytes, frequency of B220⁺ cells in all genotypes was ∼45% (Figure 3B), and splenic architecture was unaltered following loss of the HECT domain (supplemental Figure 2B). Transitional B-cell populations in the spleen had no significant differences (supplemental Figure 2C-E). However, there was significant impact on mature B1 and B2 subsets within the spleen. We found in the B2 subset, marginal zone B cells were significantly reduced from approximately 10% of B220⁺ splenocytes to 2% in both Mb1^WT/CRE;Ubr5^ΔHECT/WT and Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice whereas the follicular B-cell compartment frequency was slightly increased, despite a reduction in absolute number of follicular B cells in Mb1^WT/CRE;Ubr5^ΔHECT/WT and Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice (Figure 3C-E).

The B1 population plays a key role innate immunity. The splenic B1 population was reduced by approximately twofold in Mb1^WT/CRE;Ubr5^ΔHECT/WT and Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice and the reduction was exclusively in the B1a subpopulation (Figure 3F-H). Similarly, B1 populations in the peritoneal cavity was ∼75% reduced (supplemental Figure 2F-H). These findings demonstrate a significant loss of populations required for innate immunity with loss of B1 and marginal zone B cells following deletion of the HECT domain of Ubr5.

Evaluating cell surface markers revealed Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} follicular B cells had abnormal expression with low IgD and high IgM compared with their WT littermates (Figure 4A-B). Follicular B cells also had high CD23 expression, but normal expression of CD5 and CD1d (Figure 4A-B). To further define alterations in the B-cell compartment, we analyzed cell cycle. Although follicular B cells are typically in the resting state, Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} cells are more quiescent with increased cells in G₀ (supplemental Figure 3A). Additionally, staining with proliferation marker Ki67 showed a reduction of Ki67 staining, specifically in white pulp of spleens in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice (supplemental Figure 3B). These studies demonstrated alterations in mature spleen cells with both phenotypic and cell-cycle alterations.

Activation of follicular B cells by T-dependent antigens leads to formation of GC in secondary lymphoid tissues and generation of antibody producing plasma cells. To determine if Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} follicular B cells have normal function, we immunized mice with SRBC. Eight days following immunization, we performed immunohistochemical (IHC) analysis on spleens with PNA, a GC marker. Staining revealed significant decrease in the proportion of GC to white pulp in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice (Figure 4C-D). Quantification of GC B cells by flow cytometry showed a trend of decrease in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice; however, it was not statistically significant (Figure 4E). We evaluated the ability of follicular B cells to terminally differentiate into plasma cells within the peripheral blood and found a ∼50% reduction of plasma cells in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice, whereas Mb1^WT/CRE;Ubr5^ΔHECT/WT mice had ∼25% reduction in plasma cells (Figure 4F-H). Correlating with the decreased number of plasma cells, B cells isolated from Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice showed a decrease in antibody-producing cells (Figure 4I). Although we had a significant decrease in CD138⁺ antibody producing cells, levels of antibodies IgG, IgM, and IgA within sera were unaltered implying that Ubr5 HECT domain deletion does not affect basal antibody levels (Figure 4J). These studies indicate an important role of Ubr5 in the function and activation of B cells.

MCL is characterized by t(11;14)(q13;q32) translocation that leads to aberrant expression of CyclinD1 (CCND1).¹  To determine whether Ubr5^{ΔHECT/ΔHECT} drives lymphoma transformation, we crossed our Mb1^CRE/WT;Ubr5^{ΔHECT/ΔHECT} mice with CCND1 B-cell specific overexpression model (Eµ^CCND1) to obtain Mb1^CRE/WT;Ubr5^{ΔHECT/ΔHECT};Eµ^CCND1 mice.¹⁷  We immune stimulated mice with SRBC biweekly for 4 months, and aged 12 months. Similar to Mb1^CRE/WT;Ubr5^{ΔHECT/ΔHECT} mice, Mb1^CRE/WT;Ubr5^{ΔHECT/ΔHECT};Eµ^CCND1 mice had decreased mature recirculating cells and plasma cells in the peripheral blood (supplemental Figure 4A-C). Interestingly, in the spleen, after the addition of Eµ^CCND1, there was no difference in weight of the spleens or frequency of B220⁺ cells (supplemental Figure 4D-F). Strikingly, the addition of Eµ^CCND1 rescued the loss of both marginal zone and B1 B cells (supplemental Figure 4G-H). Although mice did not show signs of disease at 12 months, spleen cells maintained the IgM^hiIgD^lo phenotype similar to phenotype commonly seen in MCL patients (supplemental Figure 4I).^26,27  MCL patients are generally CD23⁻; however, a small subset are CD23⁺.²⁸ Mb1^CRE/WT;Ubr5^{ΔHECT/ΔHECT};Eµ^CCND1 mice displayed varied CD23 expression compared with Mb1^CRE/WT;Ubr5^{ΔHECT/ΔHECT} with some mice expressing little to no CD23. UBR5 mutations affect the pre-GC B cells, the tumor population in MCL, but these findings propose additional mutations may be required for lymphoma transformation.

The HECT domain of UBR5 is thought to be required for its ubiquitination activity, implicating deletion could lead to increased accumulation of UBR5 substrates targeted for degradation by the proteasome. To quantitatively determine protein differences in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} vs WT littermates, we labeled splenic B220⁺ cells with TMT, combined samples in equal concentrations, and analyzed by mass spectrometry (MS) (Figure 5A). Proteomic analysis identified 15 584 unique peptides, 2797 quantifiable proteins, and 1675 proteins with ≥3 unique peptides. A total of 153 proteins were significantly changed in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} splenocytes (Figure 5B; supplemental Table 2). Principal component analysis plot showed genotypes clustered together, and proteins identified were distributed throughout cellular compartments (Figure 5C-D). Of the differentially expressed proteins, 104 ≥1.3-fold significantly overexpressed were enriched for proteins associated with mRNA processing, and mRNA splicing via the spliceosome, whereas the 49 ≤0.70-fold significantly decreased from Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} were proteins associated with immune system process and protein transport (Figure 5E-G). As revealed by flowcytometry, IgD protein was the highest downregulated protein following deletion of Ubr5 HECT domain (Figure 5E). CD22, which is associated with B-cell activation, was also downregulated in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice (Figure 5E). Intriguingly, the second highest expressed protein in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} was UBR5 (Figure 5E-F). Proteomic profiling revealed an increase in UBR5 and proteins associated with mRNA splicing in B cells lacking HECT domain of UBR5, suggesting a novel role of UBR5 in mRNA splicing.

To identify interacting partners of UBR5 in MCL, we performed immunoprecipitation followed by MS using human MCL patient-derived cell lines, JEKO1 and MINO (Figure 6A).^29,30  Of note, sequencing of MINO and JEKO1 revealed no mutations in UBR5. Using MCL cell lines allowed for a homogeneous B-cell population and circumvented limitations in protein quantity in primary cells for immunoprecipitation. We identified 115 proteins with ≥3 unique peptides and, similar to our TMT analysis, gene ontology analysis showed proteins enriched in mRNA splicing via the spliceosome (Figure 6B; supplementary Table 3). Comparing MS datasets revealed 89 proteins overlapping and 6 proteins were upregulated in the TMT study (Figure 6 C-E). Intriguingly, all 6 of the other identified proteins are associated with mRNA splicing (SF3B3, SMC2, PRPF8, DHX15, SNRNP200, and EFTUD2). These proteins are classified as core spliceosome components including U2 (SF3B3) and associated U5 small nuclear ribonucleoprotein (snRNP) complex (EFTUD2, SNRNP200, and PRPF8) (Figure 6E). Of the identified proteins, none have previously been characterized as UBR5 interactors or substrates, providing a novel pathway of UBR5 regulation.

Flow cytometry validated cell-surface markers associated with activation and maturation, CD22 and IgD, were significantly decreased in splenocytes (Figure 6F). UBR5 was the second highest expressed protein in MS; however, none of the identified peptides had coverage within the HECT domain. To confirm overexpression of UBR5 protein, we performed IHC and western blot analysis and found higher protein expression of UBR5 in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} spleens (Figures 6G and7A). To determine if loss of HECT domain leads to stabilization of UBR5, we performed half-life analysis with cycloheximide. Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} CD23⁺ splenocytes had an increased protein half-life (Figure 7B). Stabilization of UBR5 could be due to loss of self-ubiquitination. To determine UBR5 ubiquitination, UBR5 was immunoprecipitated in Mb1^WT/CRE;Ubr5^WT/WT and Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} CD23⁺ splenocytes; however, changes in ubiquitin were not seen (supplemental Figure 5B). Further investigation is required to determine cause of UBR5 stabilization. Phosphorylation sites have been identified by MS on the C-terminus that are lost in the HECT mutant and may play a role in UBR5 stabilization (supplemental Figure 5C-D).³¹  These studies confirm that loss of the HECT domain leads to overexpression and stabilization of UBR5; however, the mechanism of UBR5 stabilization remains unknown.

To evaluate protein overexpression of RNA splicing components, we performed western blot analysis on CD23⁺ splenocytes. Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} and Mb1^WT/CRE;Ubr5^ΔHECT/WT had increased protein expression of EFTUD2, SNRNP200, PRPF8, and DHX15 (Figure 7A). Of note, Mb1^WT/CRE;Ubr5^ΔHECT/WT spleens had increased UBR5 expression as well (Figure 7A). These findings are of significance because of MCL patient having monoallelic mutations of UBR5. Interestingly, MCL cell lines JEKO1 and MINO also express high levels of UBR5, EFTUD2, SNRNP200, PRPF8, and DHX15 protein compared with normal B cells, providing evidence that high expression is indicative of MCL (supplemental Figure 5A). Because 3 of the identified proteins (EFTUD2, SNRNP200, and PRPF8) are part of the U5 snRNP complex, we evaluated whether UBR5 elutes with the U5 complex. We performed a glycerol density gradient and found that EFTUD2, SNRNP200, PRPF8, and UBR5 eluted in fractions 15-19 alluding to components associating with 1 another (Figure 7C). Similar to Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} leading to increased half-life of UBR5, we determined that loss of the HECT domain leads to increased protein half-life of EFTUD2, SNRNP200, PRPF8, and DHX15 (Figure 7B). These combined findings propose that UBR5 interacts with the U5 complex and that loss of the HECT domain stabilizes the U5 complex.

To determine whether EFTUD2, SNRNP200, PRPF8, and DHX15 are ubiquitination targets of UBR5, we treated CD23⁺ splenocytes with MG132, a proteasome inhibitor. Upon MG132 treatment, accumulation of EFTUD2, SNRNP200, PRPF8, and DHX15 was not observed in WT nor Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} B cells (Figure 7D). Treatment with MG132 also did not increase the protein half-life of U5 components (supplemental Figure 5E). Immunoprecipitation of polyubiquitin using TR-TUBE agarose beads demonstrated decrease in polyubiquitinated SNRPN200 Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} B cells; however, it is unclear whether ubiquitin chain in K48 that would target the protein for proteasomal degradation (supplemental Figure 5F). The lack of accumulation of proteins following MG132 treatment or changes in protein half-life suggests UBR5 does not target the proteins for proteasomal degradation via K48 ubiquitin chains; however, other conjugation of K33, K63, K27, or K11 ubiquitin chains could potentially lead to stabilization of the protein.

Although the C-terminal cysteine of UBR5 in the HECT domain is lost in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} mice, it is unclear whether catalytic activity is required for all UBR5 functions. It is known that UBR5 polyubiquitinates RNF168 in conjunction with E3 ligase TRIP12 following irradiation in the DNA damage response pathway.⁷  To determine whether mutant UBR5 is required for RNF168 ubiquitination, we irradiated CD23⁺ splenocytes and evaluated RNF168 expression. In Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} B cells, high UBR5 expression led to lower RNF168 expression that was further attenuated upon irradiation treatment (Figure 7E). Additionally, RNF168 half-life was significantly reduced in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} B cells overexpressing UBR5 following irradiation and Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} B cells had increased in RNF168 polyubiquitination (Figure 7F; supplemental 5F). These studies demonstrate that UBR5 overexpression yields a similar phenotype to inhibition of RNF168 deubiquitinating enzyme USP34, supporting the conclusion that the catalytic cysteine in the UBR5 HECT domain is not required for RNF168 ubiquitin mediated degradation.³²  Taken together, this suggests mutant UBR5 activity and function is substrate/complex dependent.

EFTUD2, SNRNP200, PRPF8, and DHX15 are expressed in early stages of B-cell development, predominantly in the pro/pre B cell stage, whereas in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} B cells, they are more abundant in the immature and mature populations predicting alterations in splicing could affect B cells early in development (Figure 7G). Expression of IgM and IgD is regulated by alternative splicing of the Ig heavy chain locus. As shown in the follicular B-cell population, IgM and IgD have aberrant expression, suggesting defects in splicing. To confirm altered splicing, we isolated follicular B cells from Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} and WT littermates for analysis of transcript levels. PCR confirmed increased IgM and decreased IgD transcripts in Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT} (Figure 7H). It was previously shown that mRNA of Zinc Finger Protein 318 (Zfp318) has 2 known splice variants.³³  We found decreased expression of transcript spliced from exons 7-11, and decreased transcript containing exons 7-9, suggesting expression of Zfp318 splice variants may also be contributing to B-cell defects (Figure 7H). These studies demonstrate that over-expression of spliceosome components in UBR5 mutant B cells promote aberrant splicing.

In this report, we demonstrate that ubiquitin E3 ligase UBR5 plays a key role in B-cell maturation and activation. Although UBR5 is present in all B-cell populations, deletion of Ubr5 HECT domain impacted specifically maturation of B cells within the spleen. The follicular B-cell population is phenotypically abnormal and has a reduced capacity to generate plasma cells. The block in differentiation of pre-GC B cells corresponds to the MCL tumor population. Although expression of Eµ^CCND1 in the UBR5 mutant mice did not lead to lymphoma transformation, MCL patients with UBR5 mutations had between 4 and 12 additional mutations indicating additional mutations may be required for transformation. In addition, expression of Eµ^CCND1 alone does not lead to lymphoma. Previous studies demonstrated Bim deletion in Eµ^{CCND1 mice} resulted in an MCL-like disease. However, disease was found at a low frequency in mice >12 months old, further supporting additional mutations are required to drive disease.¹⁷  MCL is a pre-GC lymphoma and patients can vary in phenotype but are commonly IgM^hiIgD^loCD5⁺CD23⁻.^26,27  Interestingly, pre-GC B cells express high levels of IgD, whereas MCL is IgD^lo suggesting our mouse model, which is IgM^hiIgD^lo, has phenotypical characteristics of MCL. Of note, when we cross our mouse with Eµ^CCND1 mice, a number of mice lose CD23 expression, similar to MCL, and the B1 population is rescued.

Interrogation of global proteomic changes in the splenic B-cell population suggests a novel role of UBR5 in mRNA splicing via the spliceosome. The spliceosome plays an important role in providing genetic diversity. In immune cells, alternative splicing is suggested to play a key role in B-cell differentiation, activation, and survival.³⁴  In human B cells, ∼90% of genes with multiple exons undergo alternative splicing.³⁵  Recurrent splicing mutations have been found in chronic lymphocytic leukemia (CLL) patients, another pre-GC B-cell malignancy where approximately 10% to 15% of patients have mutations to the U2 spliceosome component SF3B1. Interestingly, SF3B1 mutations lead to defects in B-cell development, suggesting that the U2 complex plays a key role in proper B-cell development.^36,37  CLL patients have mutations at low frequency in other splicing components including PRPF8 and EFTUD2.^38,39  The striking phenotype of IgD in HECT domain mutants supports defects in splicing since IgM and IgD are generated from alternative splicing of the heavy chain gene.⁴⁰  It is well established that IgM and IgD play a key role in immunity and are required for B-cell receptor signaling and immune response, but little is known regarding how alternative splicing of IgD and IgM occur.

UBR5 has been mainly associated with ubiquitin dependent functions, but also plays a role in ubiquitin independent functions. Molecular pathways regulated by UBR5 include DNA damage repair, microRNA silencing, transcription, and translation.^6,7,11,41-43  Our findings suggest mutant UBR5 interacts with the U5 spliceosome complex and mutations lead to stabilization of EFTUD2, SNRNP200, PRPF8, and DHX15. Although our data allude to interaction of UBR5 with the U5 spliceosome, we do not find alterations in ubiquitination of U5 in mutant UBR5; however, this does not exclude a direct ubiquitin dependent or independent effect of UBR5 on U5 spliceosome components. Further investigation is required to decipher the mechanism promoting stabilization of U5 and the role of UBR5. UBR5 is found in complex with additional E3 ligases and although the ubiquitin conjugating cysteine in the HECT is lost, we found UBR5 does not always require its catalytic cysteine for substrate ubiquitination; and in the case of RNF168, overexpression of mutant UBR5 leads to increased ubiquitination where the ubiquitin is potentially transfered from another component of the complex.^7,44  This suggests the gain of function vs loss of function of mutant UBR5 is substrate/complex dependent.

Patient mutations in UBR5 are monoallelic indicating the Mb1^WT/CRE;Ubr5^ΔHECT/WT mouse phenotype would be more representative to patient mutations. As with Mb1^WT/CRE;Ubr5^{ΔHECT/ΔHECT}, Mb1^WT/CRE;Ubr5^ΔHECT/WT mice had defects in B-cell maturation and overexpressed spliceosome components indicating the mutant protein acts as a dominant negative over the WT protein. UBR5 is commonly overexpressed in solid tumors that correlates with data presented herein where HECT mutations lead to protein stabilization and overexpression indicating Ubr5^{ΔHECT/ΔHECT} mouse model could provide additional clues to the role of UBR5 in cancer.^12,14,15  Overall, our findings reveal a novel mechanism of regulation by UBR5 in B-cell maturation and provides a key understanding of the role of UBR5 mutations in MCL.

The authors thank M. Reth for the use of the Mb1^CRE mice, S. Korolov for helpful discussions and providing Mb1^CRE mice, S. Katz for Eµ^CCND1 mice, and P. Swanson (Creighton University, Omaha, NE), for helpful discussions. The authors also thank the UNMC Flowcytometry Research Facility, UNMC Mouse Genome Engineering Core Facility, and UNMC Mass Spectrometry and Proteomics Core Facility for expert assistance.

The core facilities are administrated through the Office of the Vice Chancellor for Research and supported by state funds from the Nebraska Research Initiative and The Fred and Pamela Buffett Cancer Center's National Cancer Institute Cancer Support Grant. This work was supported by the National Institutes of Health, National Institute of General Medical Sciences (P20GM121316) (S.M.B., N.T.W.), an American Cancer Society institutional grant, the Frances E. Lageschulte & Evelyn B. Weese New Frontiers Medical Research Fund, a UNMC National Institutes of Health National, Cancer Institute training grant (5T32CA009476-23) (W.H.S.), and the Fred & Pamela Buffett Cancer Center Support Grant from the National Institutes of Health, National Cancer Institute (P30 CA036727).

Contribution: S.A.S., T.J.G., and S.M.B. conceived and designed the experiments; S.A.S., T.J.G., H.V., J.H.G., M.R.G., and S.M.B. performed experiments and analysis; S.A.S., T.J.G., R.W.H.-S., and S.M.B. wrote the manuscript; H.C.-H.L., C.A., and N.T.W. provided technical and material support; and all authors reviewed the manuscript before submission.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Shannon M. Buckley, Department of Genetics, Cell Biology, and Anatomy, University of Nebraska Medical Center, 985805 Nebraska Medical Center, Omaha, NE 68198; e-mail: shannon.buckley@unmc.edu.