The present study utilized 10 female NOD/SCID mice (age, 6-8 weeks; Henan Skobes Biotechnology Co., Ltd.) raised in an independent ventilation system with a constant temperature of 22˚C and a light/dark cycle of 12/12 h. These mice (weight, 18-22 g) were maintained in specific pathogen-free conditions at Xinxiang Medical University (Xinxiang, China). A mouse model of human DLBCL was established by subcutaneously injecting 2x10⁷ OCI-LY10 cells into the right flank of the NOD/SCID mice (14). Tumor volume was quantified using the formula: Volume=(length x width²)/2. Once the average tumor volume reached 80-100 mm³, treatment with MG132 commenced. The NOD/SCID mice bearing OCI-LY10 tumors were then allocated into a control group (n=5) or a treatment group (n=5). Mice in the treatment group received intraperitoneal injections of MG132 at a concentration of 50 mg/kg (200 µl), while the control group was administered an equal volume of the solvent (4% DMSO + 30% PEG300 + 20% propylene glycol + ddH₂O) via the same route (14). Tumor growth was monitored every two days. All mice were sacrificed 45 days following the initiation of MG132 therapy. Euthanasia was performed using an overdose of sodium pentobarbital (100 mg/kg), administered intravenously to ensure rapid and painless loss of consciousness. The tumors in the control and treatment groups were excised and weighed. The present study was approved by The Ethics Committee of Xinxiang Medical University (Xinxiang, China; approval no. XYLL-2022001253), and all the animal procedures were performed in accordance with the ‘Guide for the Care and Use of Laboratory Animals’ published by the National Institutes of Health (22).

The pCas9-AID constructs were successfully generated through the ligation of gRNA targeting AID into the pL-CRISPR.EFS.PAC plasmids (cat. no. 57828; Addgene, Inc.) (12,13). The specific target sequences are detailed in Table SI (NCBI accession no. NM_020661.4). These gRNAs were designed using online software E-CRISPR (http://www.e-crisp.org/) to generate AID knock out. Additionally, the primer sequences utilized for amplifying the AID cDNA to produce the pWPI-AID-GFP lentivirus constructs are provided in Table SII.

The OCI-LY10 ABC-DLBCL cell line was acquired from the BeNa Culture Collection (Beijing Beina Chunglian Institute of Biotechnology; cat. no. BNCC337742). The 293T cells were maintained in the cell bank of Xinxiang Key Laboratory of Tumor Vaccine and Immunotherapy (Xinxiang, China). The AIDKO OCI-LY10 ABC-DLBCL cell line was generated by the AIDKO lentivirus, and the lentivirus was obtained from the supernatant of 293T cells co-transfected by pCas9-AID constructs (or pL-CRISPR.EFS.PAC plasmids as control) and virus generation plasmids including ΔR9 and pVSVG. Both cell lines were cultured in a humidified incubator at 37˚C and 5% CO₂, using either Iscove's modified Dulbecco's medium or Dulbecco's modified Eagle's medium (Hyclone; Cytiva) as the base medium, supplemented with 10% fetal bovine serum (MilliporeSigma), non-essential amino acids and a 1% penicillin-streptomycin mixture.

To detect AID ubiquitination, pWPI-AID-GFP and Ub-HA constructs were co-transfected into 293T cells and cultured for 24, 48 and 72 h at 37˚C, respectively. Following these time points, cells were harvested for total protein extraction and subsequent immunoprecipitation analysis.

To assess MG132-induced cell death, cells were incubated with MG132 (10 µM; Selleck Chemicals; cat. no. S2619) for 8 h at 37˚C. The MG132 powder was prepared using DMSO as a solvent. As a control, an equivalent volume of DMSO was used in the treatment of the control cells.

Total RNA was extracted from OCI-LY10 ABC-DLBCL cell pellets using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 15596026) in accordance with the manufacturer's protocol. Subsequently, cDNA was synthesized utilizing the PrimeScript™ RT Reagent Kit (Takara Bio, Inc.; cat. no. RR037A) using reverse transcription for 15 min at 37˚C and terminated for 5 sec at 85˚C. qPCR was conducted on the QuantStudio™ 5 System (Applied Biosystems; Thermo Fisher Scientific, Inc.). The thermocycling conditions were as follows: 15 sec at 95˚C, 30 sec at 60˚C and 30 sec at 72˚C, for a total of 40 cycles. The relative mRNA levels were determined using the 2^-ΔΔCq method (23) calculated using Microsoft Office 2016, with the expression of β-actin taken as the internal control. The primer sequences employed for the quantification of AID and β-actin transcription are provided in Table SIII.

To assess the apoptosis rate of OCI-LY10 ABC-DLBCL cells induced by MG132, the treated cells were harvested and twice washed with 1X PBS at 4˚C. The cells were then incubated with anti-PI and anti-Annexin V (BD Biosciences; cat. no. 556547) for 15 min at room temperature. Following incubation, the cells were resuspended in flow cytometry buffer and subjected to analysis via flow cytometry. All data were acquired and processed using a CytoFLEX Flow Cytometer (with BECKMAN CytoExpert version 2.4) (Beckman Coulter, Inc.).

For protein extraction, the OCI-LY10 cell pellet was lysed in RIPA buffer (cat. no. P0013B; Beyotime Institute of Biotechnology) for 30 min. The RIPA preparation included 50 mM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate and 0.1% SDS. The lysates were subsequently centrifuged at 13,000 x g for 20 min at 4˚C, and the protein-containing supernatant was collected. The quality and concentration of the total protein were determined by BCA protein assay Kit (Beyotime Institute of Biotechnology). Protein samples (20-100 µg per lane) were then loaded onto 10% (w/v) Tris-HCl SDS-PAGE gels for electrophoresis, transferred to a PVDF membrane (MilliporeSigma), and subjected to blotting. The membrane was blocked with 5% (w/v) skimmed milk, then probed with an anti-AID antibody (1:1,000; cat. no. 4959; CST Biological Reagents Co., Ltd.) and anti-GAPDH (1:5,000; cat. no. ab9485; Abcam) served as the loading control. The primary antibody incubation was performed at 4˚C overnight. The signal was detected using a goat anti-rabbit secondary antibody conjugated with horseradish peroxidase (1:1,000; cat. no. 31460; Thermo Fisher Scientific, Inc.). The secondary anti-rabbit antibody was incubated with the membrane for 1 h at room temperature. The band signals were captured using the FUSION FX7 system (Vilber Lourmat).

For protein extraction, the OCI-LY10 cell pellet was lysed in RIPA buffer (Beyotime Institute of Biotechnology) for 30 min. The lysates were subsequently centrifuged at 13,000 x g for 20 min at 4˚C, and the protein-containing supernatant was collected. The quality and concentration of the total protein were determined by BCA protein assay Kit (Beyotime Institute of Biotechnology). Following pre-clearance of the chromatin using Dynabeads Protein G beads (2X; 25 µl for each IP reaction; cat. no. 10003D; Invitrogen; Thermo Fisher Scientific, Inc.), a portion of the aliquot was set aside as the input sample. Subsequently, proteins from 5x10⁶ cells were incubated with 5 µg of a specific antibody or normal goat IgG (Santa Cruz Biotechnology, Inc.; cat. no. sc2346) overnight at 4˚C. Anti-GFP (Abcam; cat. no. ab13970) immune complexes were then captured by incubating with Dynabeads Protein G beads (2X; 25 µl for each IP reaction) (2X; Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 10003D) for 3 h. The beads were washed at 4˚C using RIPA buffer containing varying concentrations of NaCl. The proteins from the pull down were denatured at 100˚C and loaded onto SDS-PAGE gels for subsequent immunoblotting.

The AID^-/- mice were a gift from Professor Ji of Xi'an Jiaotong University (Xi'an, China) (14,15), and were bred in the Laboratory Animal Centre of Xinxiang Medical University (Xinxiang, China). The WT and AID^-/- C57 mice were confirmed to carry the homozygous disruption of the AID gene according to the generation of AID^-/- mice reported by Honjo T (24). C57 and AID^-/- mice were administered a tail vein injection of MG132 (10 µg/kg/day) for 24 h (25). Splenic resting B cells were purified by negative selection using anti-CD43-conjugated microbeads (Miltenyi Biotec GmbH) (26). A total of ~3-5x10⁷ splenic resting B cells were obtained from one mouse spleen. Purified B cells were cultured at 5x10⁵ cells/ml in RPMI medium supplemented with 15% fetal bovine serum (MilliporeSigma), non-essential amino acids, penicillin-streptomycin (1%) and β-mercaptoethanol (50 µM). Cells were cultured in a humidified incubator at 37˚C and 5% CO₂. Cells were cultured in the presence of LPS and IL-4 for 5 days: i) LPS (50 µg/ml); and ii) LPS (3 µg/ml) plus IL-4 (50 ng/ml), to induce class switching to IgG1, IgG3 and IgE. MG132 (5 µm) treatment was sustained throughout the stimulation process. The antibody class switch was detected using RT-qPCR. The relative mRNA levels were determined using the 2^-ΔΔCq method (23) calculated using Microsoft Office 2016, with the expression of β-actin taken as the internal control. The primers for RT-qPCR performed as aforementioned are listed in Table SIV (NCBI accession no. MF430613.1, V00818.1 and KU613484.1).

The expression levels of AID in clinical specimens were accomplished through interrogating the online databases GENT2 (http://gent2.appex.kr/gent2/) by searching tissue, blood disease and gene symbols. The significance cut-off level was defined as P<0.05 and |fold-change|>2.

Data analysis were performed using GraphPad Prism 8.0 software (Dotmatics). Unpaired t-tests were conducted when comparing datasets of two groups. One-way ANOVA and Dunnett's multiple comparisons test were used to perform the comparison of multiple groups. Data shown are representative of 3 independent experiments and represented as the mean ± SD. P<0.05 was considered to indicate a statistically significant difference.

The association between the expression of AID and several lymphoma subtypes including Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL) was examined using the GENT2 database, which showed that AID expression was significantly higher in the NHL subtype compared with the HL subtype (Fig. 1A). The association between the expression of AID and the subtypes of NHL including follicular lymphoma, mantle cell lymphoma, DLBCL and chronic lymphocytic leukemia were evaluated using the GENT2 database. AID showed a significantly higher expression in DLBCL in comparison with the other NHL subtypes (Fig. 1B). These data indicated that the upregulation of AID is associated with DLBCL progression.

To investigate therapeutic strategies to treat the refractory ABC-DLBCL subtype, MG132 was introduced on the basis of previous studies (14,15). ABC-DLBCL cell xenograft tumor-bearing mice were established to explore the MG132 treatment effect in vivo. Tumor growth was monitored for 45 days, while MG132 treatment was initiated on day 14, when tumor volume had reached 80-100 mm³. The results showed that tumor volume was decreased in the MG132 treated tumor-bearing mice in comparison with the mice in the DMSO-treated group (Fig. 1C; Fig. S1), and the tumor weight showed the same trend (Fig. 1D). These results indicate the inhibitory effect of MG132 on tumor growth in ABC-DLBCL xenograft tumors. Collectively, these results imply an association between MG132 treatment and AID in ABC-DLBCL.

The post-transcription regulation of AID is achieved through the proteasome pathway (21). Considering the significant effect of AID in DLBCL (Fig. 1) and the inhibition of MG132 to ubiquitination through the proteasome pathway (19-21), AID was taken as a key factor to link the mechanism of MG132 induced anti-lymphoma to ABC-DLBCL. The OCI-LY10 ABC-DLBCL cells were transduced using CRISPR/Cas9 with three gRNAs for AID to generate the AID knockout OCI-LY10 (10AIDKO) cell lines. The levels of mRNA and protein for AID were significantly depleted in 10AIDKO cells compared with their wild-type (WT) counterparts and the negative control (EV; Fig. 2A and B). Specifically, two OCI-LY10 cell lines (10AIDKO2 and 10AIDKO3) with efficient AID deficiency were identified.

To determine whether AID loss had an impact on cellular function, the apoptosis rate of 10AIDKO2 cells was examined using Annexin V and PI staining in presence of DMSO. The AIDKO2 cells presented significantly less Annexin V⁺PI⁺ populations compared with the WT cells (Fig. 2C), indicating reduced apoptosis rate of ABC-DLBCL cells after AID deficiency. By contrast, these results showed an anti-lymphoma effect of AID to ABC-DLBCL cells.

To assess the effect of MG132 on AID degradation through proteasomes, the anti-lymphoma effect of MG132 on ABC-DLBCL was confirmed through inhibiting AID degradation. The ubiquitination of AID after MG132 treatment was detected, with the anti-GFP purified proteins indicating the ubiquitination of AID with MG132 treatment. The results showed that AID underwent degradation without MG132 administration (Fig. S2; Fig. 3A, lanes 1 and 2; Fig. 3B), while MG132 administration resulted in almost a complete disappearance of ubiquitination and a single AID band in the immunoblots (Fig. 3A, lane 3 and lane 4; Fig. 3B), suggesting that MG132 resulted in the inhibition of proteasome degradation of AID. Therefore, the results indicate that MG132 induces AID accumulation by inhibiting ubiquitination.

To further explore the mechanism of MG132 treating ABC-DLBCL, the inhibitory effects of MG132 on AID and the important role of AID in CSR was assessed. Therefore, class switch in WT and AID^-/- mice treated with or without MG132 in vivo was detected (Fig. 4A). AID accumulation was observed after MG132 treatment (WT + MG132 group; Fig. 4B), suggesting that MG132 induced AID accumulation, which may be associated with the anti-lymphoma effect of MG132 on ABC-DLBCL.

After confirmation of MG132 inhibition of AID ubiquitination, the AID-induced class switch in the presence or absence of MG132 was also assessed in vivo. By stimulating with LPS or/and LPS plus IL-4, the purified spleen resting B cells underwent class switch to IgG1, IgG3 and IgE (Fig. 4C). Through stimulation of LPS (3 µg/ml) plus IL-4 (50 ng/ml), a notable elevated class switch to IgG1 was detected in mice with complete AID expression, while no class switch to IgG1 was observed in spleen resting B cells from AID^-/- mice treated with MG132 (Fig. 4D). Similar trends were also observed in the detection of class switch to IgG3 and IgE with stimulation by LPS (50 µg/ml) only and LPS (3 µg/ml) plus IL-4 (50 ng/ml) for 5 days (Fig. 4D). As hypothesized, notable class switching to IgE and IgG3 was detected by RT-qPCR in spleen resting B cells of WT mice, while AID^-/- mice had no class switching to IgE and IgG3 due to AID deficiency (Fig. 4D). Collectively, MG132 inhibits AID ubiquitination to elevate AID levels, which thus induces apparent CSR.

To evaluate the impact of MG132 on ABC-DLBCL cells, ABC-DLBCL OCI-LY10 cells were treated with MG132 (5 µM) in the presence or absence of AID. The ubiquitination of MG132 treatment at different times (24, 48 and 72 h) was detected by immunoblotting (Fig. 5A). The results showed that MG132-treated WT OCI-LY10 cells, specifically in the 24 h treatment group, showed increased AID accumulation compared with the untreated counterparts (Fig. 5A, lanes 2-4 compared with lane 1). In addition, as time progressed, AID accumulation reduced, which could be attributed to the declined drug efficiency along with increased time (Fig. 5A, lane 2-4).

In order to further detect MG132 induced CSR in ABC-DLBCL, 10WT and 10AIDKO cells were stimulated by LPS or/and LPS plus IL-4 (Fig. 5B). A significant increase in class switching to IgG1, IgG3 and IgE was detected in WT ABC-DLBCL cells stimulated by LPS (3 µg/ml) alone or LPS plus IL-4 (50 ng/ml), but not in AIDKO cells (Fig. 5C). Thus, the data indicate that MG132 inhibits AID ubiquitination to elevate AID levels, thus inducing apparent CSR in ABC-DLBCL cells.