Introduction
Multiple myeloma (MM) is a hematologic malignancy characterized by clonal plasma cell proliferation and excessive monoclonal immunoglobulin production, leading to multi-organ dysfunction [ 1 ]. Predominantly affecting older adults [ 2 ], MM presents with significant heterogeneity, typically including osteolytic bone lesions, anemia, renal impairment, and immunodeficiency [ 3 – 5 ]. Although immune checkpoint inhibitors (ICIs) targeting the programmed cell death protein 1/programmed cell death-ligand 1 (PD-1/PD-L1) axis have reshaped cancer immunotherapy by reinvigorating exhausted T cells [ 6 , 7 ], their clinical efficacy in MM remains constrained by a highly immunosuppressive tumor microenvironment (TME) [ 8 ]. This therapeutic challenge underscores the necessity of developing innovative and effective combination therapies to achieve better clinical outcomes.
Tumor-associated macrophages (TAMs) are critical regulators of TME remodeling and tumor progression [ 9 ]. Microenvironmental stimuli frequently drive TAMs toward an immunosuppressive M2-like phenotype, which accelerates disease progression by inducing T-cell exhaustion and suppressing cytotoxic CD8 + T-cell function [ 10 , 11 ]. Conversely, repolarizing TAMs toward a pro-inflammatory M1-like phenotype converts immunosuppressive “cold” tumors into inflamed “hot” environments [ 10 – 12 ]. Emerging evidence indicates that the reprogramming of TAMs toward an M1-like phenotype can substantially enhance the efficacy of immune checkpoint blockade. Combining the M1-like repolarization of TAMs with anti-PD-L1 therapy reduces primary tumor burden and increases survival in mouse models, correlating with increased CD8 + T-cell infiltration and decreased T-cell exhaustion [ 13 – 15 ]. Therefore, driving TAM repolarization, particularly through epigenetic or cytokine modulation, represents a promising immunotherapeutic strategy to release immunosuppressive brakes and overcome ICI resistance [ 16 ].
Suppressor of cytokine signaling 3 (SOCS3) is a multifaceted regulator of tumor pathogenesis, influencing metastasis, inflammatory cascades, and immune cell infiltration [ 17 , 18 ]. SOCS3 exhibits context-dependent effects across various tumor types, functioning as an oncogene in colon and lung cancers [ 19 , 20 ], in contrast to a tumor suppressor in breast and gastric cancers [ 21 , 22 ]. SOCS3 regulates tumor immunology primarily through the inhibition of the Janus kinase (JAK)/signal transducer and activator of transcription 3 (STAT3) pathway [ 23 ]. SOCS3 acts as a critical immunomodulator by restricting M2-like macrophage polarization and coordinating T-cell responses [ 24 , 25 ]. In the specific context of MM, circulating SOCS3 levels inversely correlate with disease risk [ 26 ]. Its expression is frequently silenced by promoter methylation [ 27 ]. Functionally, Casitas B lineage lymphoma (c-Cbl)-mediated degradation of SOCS3 has been shown to promote MM cell survival and proliferation [ 28 ], whereas restoring SOCS3 expression effectively induces MM cell apoptosis via STAT3 suppression [ 29 ]. Furthermore, SOCS3 mediates negative feedback against inflammatory cytokines in myeloma cells, such as tumor necrosis factor-alpha (TNF-α) and interleukin 6 (IL-6) [ 30 ]. However, the broader impact of SOCS3 on MM progression and microenvironmental immune reprogramming remains poorly understood.
N 6 -methyladenosine (m6A) RNA modification critically contributes to MM pathogenesis by regulating transcript stability. For instance, upregulated methyltransferase-like 3 (METTL3) drives malignant MM cell proliferation through m6A-dependent post-transcriptional regulation of target oncogenes [ 31 ]. Moreover, inhibition of demethylase alkylation repair homologue protein 5 (ALKBH5) suppresses MM angiogenesis and metastasis through downregulating oncogene via m6A methylation [ 32 ]. In other disease models, METTL3 and its co-factor methyltransferase-like 14 (METTL14) suppress SOCS3 expression via m6A modification [ 33 , 34 ]. However, whether this METTL3/m6A/SOCS3 axis operates in MM to drive disease progression and microenvironmental immunosuppression requires systematic investigation.
This study investigated the roles of SOCS3 in immune regulation during MM progression, particularly focusing on macrophage polarization and cytotoxic T-cell activation within the TME. We aimed to explore a novel immunotherapeutic strategy to restore the antitumor efficacy of anti-PD-1 therapy.
Materials and methods
Clinical patients
Bone marrow aspirates were obtained from 32 newly diagnosed MM patients (18 males and 14 females, aged 55–74 years) and 18 healthy controls (11 males and 7 females, aged 52–71 years) at the Quzhou People’s Hospital (Approval No.: 2022-041). Patients with concurrent malignancies or severe immune diseases were excluded. This study protocol was approved by the Ethics Committee of Quzhou People’s Hospital (Approval No.: 2022-041), and all participants provided written informed consent. Primary CD138 + plasma cells were isolated using CD138 magnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany).
Gene Expression Omnibus (GEO) database analysis
The GSE6477 dataset ( GPL96 platform; 17 normal samples and 73 newly diagnosed MM samples), GSE24870 dataset ( GPL571 platform; 20 normal samples and 23 MM samples), and GSE16558 dataset ( GPL6244 platform; 5 normal samples and 60 MM samples) were downloaded from the GEO database. Datasets were analyzed using the GEO2R tool to identify differentially expressed genes (DEGs). We screened the downregulated genes according to the criteria ( P < 0.05 and log 2 (FC) < −1). The overlapping genes across the three datasets were identified and visualized using a Venn diagram.
Cell culture, transfection, and lentiviral transduction
MM cell lines (U266, NCI-H929, RPMI-8226, and MM.1 S) were obtained from American Type Culture Collection (ATCC; Manassas, USA), and the murine MM cell line 5TGM1 was obtained from YaJi Biological Co., Ltd (Shanghai, China). All cell lines were cultured in RPMI-1640 medium (Gibco, USA) containing 10% fetal bovine serum (FBS) and 1% antibiotics at 37 °C in 5% CO 2 . Primary normal plasma cells (nPCs) were isolated from bone marrow mononuclear cells of healthy donors by immunomagnetic positive selection using the EasySep™ Human CD138 Positive Selection Kit (STEMCELL Technologies, Canada). For in vitro assays, cell transfection was conducted using Lipofectamine 3000 reagent (Invitrogen, USA) according to the manufacturer’s instructions. Overexpression plasmids targeting METTL3 (oe-METTL3) and SOCS3 (oe-SOCS3), and small interfering RNA (siRNA) targeting METTL3 (si-METTL3) and SOCS3 (si-SOCS3) were provided by GenePharma (Shanghai, China). For in vivo animal studies, 5TGM1 cells were stably transduced with a lentiviral SOCS3 overexpression vector (GenePharma) or a negative control vector. Briefly, 5TGM1 cells were seeded in 6-well plates and infected with the respective lentivirus in the presence of 8 µg/mL polybrene (Sigma-Aldrich, USA) to enhance transduction efficiency. After 24 h of incubation, the viral medium was replaced with fresh complete medium. Stably transduced cells were subsequently selected by culturing in complete medium containing 2 µg/mL puromycin for two weeks before being utilized for subcutaneous inoculation. The transduction efficiency was evaluated by measuring SOCS3 mRNA expression, which was shown in Supplementary Fig. S1 .
Co-culture of MM cells and macrophages
Macrophage differentiation of THP-1 cells (ATCC) was induced by 24 h treatment with 100 nM phorbol myristate acetate (PMA). Co-culture of NCI-H929 cells and M0 macrophages was conducted using a co-culture Transwell system (0.4 μm pore size, polycarbonate membrane; Corning, USA) following a published method [ 35 ]. Briefly, THP‑1‑derived macrophages (1 × 10 6 cells/well) were seeded in the lower chamber, and transfected NCI-H929 cells (5 × 10 5 cells/well) were placed in the upper chambers, yielding an MM cell-to-macrophage ratio of 1:2. The two cell types were co-cultured in RPMI-1640 medium supplemented with 10% FBS for 72 h at 37 °C in 5% CO 2 . After co-culture, the conditioned media (CM) were collected, centrifuged at 1,500 rpm for 10 min to remove cell debris, and either used immediately for downstream assays or aliquoted and stored at − 80 °C for subsequent analysis.
Cell Counting Kit-8 (CCK-8) assay
NCI-H929 cells were seeded in 96-well plates (1 × 10 3 cells/well) and cultured for designated time periods (24, 48, and 72 h). Then, 10 µL of CCK-8 reagent (Nanjing Jiancheng, China) was dispensed into each well at each time interval, followed by an additional 2 h of incubation. Absorbance at 450 nm was subsequently assessed using a microplate reader (Bio-Rad, USA).
Colony formation assay
NCI-H929 cells were seeded at 500 cells per well in 6-well plates and cultured for 14 days to form colonies. The colonies were then fixed with 4% paraformaldehyde and stained with 0.1% crystal violet for 20 min. Colonies were counted using ImageJ software.
Transwell migration assay
Cells resuspended in serum-free medium were seeded into the upper chambers of Transwell inserts (8 μm pore size; Corning, USA), while complete medium was added to the lower compartment. After 24 h of incubation, the migrated cells on the lower surface of the membrane were stained with 0.5% crystal violet, imaged with microscope, and quantified using ImageJ software.
Cell apoptosis
Harvested cells were dual-stained with 5 µL of Annexin V-Fluorescein isothiocyanate (FITC) and 5 µL of propidium iodide (PI) using an Apoptosis Detection Kit (Solarbio, Beijing, China) for 20 min in the dark. Apoptosis was analyzed using a flow cytometer (BD Bioscience, USA).
Reverse transcription quantitative polymerase chain reaction (RT-qPCR)
Total RNA was extracted from cellular samples using TRIzol reagent (Beyotime, Dalian, China). Complementary DNA (cDNA) was subsequently synthesized using the PrimeScript™ RT Master Mix (Takara, Dalian, China) for cDNA synthesis. Quantitative analysis of gene expression was performed using the SYBR Green PCR Kit (Applied Biosystems, USA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as the internal reference for data normalization. Relative mRNA expression was quantified through the 2 −ΔΔCT method. Primer sequences are listed in Supplementary Table S1 .
Western blot analysis
Experimental samples were lysed in radioimmunoprecipitation assay (RIPA) buffer (Beyotime, Dalian, China) to extract total proteins. Equal amounts of protein lysates (30 µg per lane) were separated via 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, USA). The membranes were blocked with 5% non-fat milk for 1 h at room temperature, followed by overnight incubation at 4 °C with primary antibodies. After washing, the membranes were incubated with a secondary antibody (1:2000, Abcam, ab6721) for 2 h at 37 °C. Protein bands were visualized using an enhanced chemiluminescence (ECL) reagent (Pierce, IL, USA) and quantified via densitometric analysis using ImageJ software. For data normalization, the densitometric values of all target proteins were normalized to the internal loading control GAPDH. Subsequently, JAK2/STAT3 pathway activation was determined by calculating the ratio of phosphorylated to total proteins (p-JAK2/JAK2 and p-STAT3/STAT3). The following primary antibodies (Abcam, USA) were used: SOCS3 (1:1000, ab280884), METTL3 (1:1000, ab195352), JAK2 (1:5000, ab108596), p-JAK2 (1:1000, ab32101), STAT3 (1:5000, ab68153), p-STAT3 (1:2000, ab76315), and GAPDH (1:2500, ab9485).
RNA immunoprecipitation (RIP) assay
To confirm the direct binding of METTL3 and YTH N 6 -methyladenosine RNA binding protein 2 (YTHDF2) to the SOCS3 transcript, RIP assays were performed using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, USA). Briefly, NCI-H929 cells were transfected with siRNAs targeting METTL3 or YTHDF2 for 48 h. Harvested cells were lysed in complete RIP lysis buffer, and cell lysates were then incubated overnight at 4 °C with magnetic beads pre-coated with highly specific antibodies against METTL3 (ab195352, Abcam), YTHDF2 (ab220613, Abcam), or a normal rabbit immunoglobulin G (IgG) control (ab172730, Abcam). After extensive washing, the immunoprecipitated RNA-protein complexes were treated with Proteinase K to digest proteins. The enriched RNA was subsequently extracted, purified, and subjected to reverse transcription. The relative enrichment of SOCS3 transcripts in the METTL3-IP and YTHDF2-IP groups compared with the IgG control was quantified via RT-qPCR.
Methylated RNA immunoprecipitation-qPCR
Total RNA was isolated from NCI-H929 and RPMI-8226 cells, and RNA fragments (approximately 100–200 nt in length) were prepared using the Magnesium RNA Fragmentation Module (New England Biolabs, USA). These fragments were immunoprecipitated with magnetic beads (Millipore, Billerica, MA, USA) pre-coated with either an anti-m6A antibody (ab151230, Abcam, 5 µg per IP reaction) or an isotype-matched control normal rabbit IgG control (ab172730, Abcam; 5 µg per IP reaction). Antibody specificity for enriching m6A-modified RNA was pre-validated by the manufacturer. Following extensive washing, the enriched m6A-modified RNA was eluted, and analyzed by RT-qPCR. The relative m6A enrichment of SOCS3 was calculated as a percentage of the input RNA (% Input) and compared with the IgG control.
Single-base elongation- and ligation-based qPCR (SELECT-qPCR)
To validate the predicted high-confidence m6A motifs (GAACT; Probability > 0.99) at positions 1427 and 2625 on the SOCS3 transcript (GenBank accession number: NM_001378932.1 ) predicted by the sequence-based RNA adenosine methylation site predictor (SRAMP) database, a SELECT-qPCR assay was employed. NCI-H929 cells were first transfected with METTL3 siRNA or a negative control for 48 h. Total RNA was extracted and mixed with site-specific Up and Down probes flanking the predicted m6A sites (1427 and 2625) of SOCS3. The mixture was heated at 90 °C for 1 min and gradually cooled to 40 °C for optimal annealing. Subsequently, a reaction mixture containing Bst 2.0 DNA polymerase, SplintR ligase (New England Biolabs, USA), ATP, and dTTP was added to the annealed products, followed by incubation at 40 °C for 20 min to facilitate elongation and ligation. The ligated products were then amplified and quantified using standard qPCR. Given that m6A modification sterically hinders the elongation and ligation efficiency of Bst 2.0 and SplintR, the relative m6A abundance at specific sites was inversely calculated based on the qPCR cycle threshold (Ct) values.
Dual-luciferase reporter assay
To verify whether the predicted sites (1427 and 2625) functionally mediate METTL3-regulated SOCS3 expression, a dual-luciferase reporter assay was conducted. The cDNA fragments corresponding to the wild-type (WT) sequence of the SOCS3 transcript, encompassing the high-confidence m6A motifs (GAACT), and its corresponding mutant (Mut) sequences were chemically synthesized and cloned into the pmirGLO Dual-Luciferase reporter vector (Promega) by GenScript (Nanjing, China). For the Mut constructs, the consensus adenosine (A) residues at positions 1427 and 2625 were site-specifically mutated to cytosine (C) to abolish the m6A modification upon transcription (5’-GAACT-3’ to 5’-GCACT-3’; as shown in Supplementary Fig. S2 ), yielding three mutant constructs: a single mutation at 1427, a single mutation at 2625, and a double mutation at both sites. NCI-H929 cells were co-transfected with the respective reporter plasmids (WT or Mut) and METTL3 overexpression plasmid (or empty vector) using Lipofectamine 3000 (Invitrogen, USA). After 48 h of transfection, cell lysates were collected, and both firefly and Renilla luciferase activities were sequentially measured using the Dual-Luciferase Reporter Assay System (Promega, USA).
mRNA stability assay
To evaluate the stability of SOCS3 mRNA, NCI-H929 and RPMI-8226 cells (transfected with vector control or METTL3-overexpression plasmids) were treated with the transcriptional inhibitor actinomycin D (5 µg/mL; Sigma-Aldrich) to block de novo RNA synthesis. Total RNA was harvested at designated time points (0, 3, 6, and 9 h) post-treatment using TRIzol reagent (Invitrogen, USA). The mRNA expression levels of SOCS3 at each time point were quantified by RT-qPCR and normalized to the internal control GAPDH using the 2 −ΔΔCT method. The relative remaining mRNA level at each time point was expressed as a percentage of the level at time 0. The mRNA half-life (t 1/2 ) was calculated by fitting the degradation data to a one-phase exponential decay model using GraphPad Prism 8.0 software.
Flow cytometric analysis
For in vivo tumor samples, fresh tissues underwent enzymatic disaggregation at 37 °C for 1 h using 10 U/mL collagenase I (Thermo Fisher, USA) combined with 30 U/mL DNase I (Thermo Fisher, USA) in RPMI-1640 medium, followed by filtration through a 40-µm cell strainer. Erythrocyte depletion was achieved using ACK lysing buffer (Gibco, USA) prior to centrifugation to harvest the final single-cell suspension. For in vitro assays, THP-1-derived macrophages were directly harvested using cell scrapers to obtain single-cell suspensions. Following cell collection, macrophages or tissue-derived cells (2 × 10 6 cells/mL) were prepared in ice-cold fluorescence-activated cell sorting (FACS) buffer. Fc gamma receptors were blocked with a CD16/CD32 monoclonal antibody (1:50, BioLegend) at 4 °C for 15 min to reduce non-specific binding. To validate marker specificity and establish accurate gating boundaries for the four-quadrant scatter plots, appropriate isotype-matched antibodies or fluorescence-minus-one (FMO) controls were utilized. Prior to specific subpopulation analysis, a stringent sequential gating strategy was applied to all samples. First, the main cell population was identified and cellular debris was excluded using forward and side scatter characteristics (FSC-A vs. SSC-A), followed by doublet exclusion (FSC-A vs. FSC-H) to ensure single-cell analysis. Dead cells were strictly excluded utilizing a Fixable Viability Dye. The detailed, representative step-by-step gating strategies for all analyzed immune cell populations are provided in Supplementary Fig. S3 . From the gated viable singlets, specific cellular immunostaining was performed according to the experimental design: For in vitro macrophage polarization analysis, human THP-1-derived macrophages were first stained with anti-human CD86-PE (#374205, BioLegend) and CD206-FITC (#321103, BioLegend). Then, cells were fixed and permeabilized using the Intracellular Fixation & Permeabilization Wash Buffer Kit (BioLegend) and subsequently stained with anti-human CD68-APC (#333809, BioLegend). Quadrant analysis was then performed within the CD68 + gate to distinguish M1-like (CD68 + CD86 + ) and M2-like (CD68 + CD206 + ) phenotypes. For in vivo macrophage polarization analysis, total tumor-infiltrating leukocytes were first identified by staining with an anti-mouse CD45-BV510 antibody (#103138, BioLegend). Within the gated viable CD45⁺ leukocyte population, macrophages were identified using anti-mouse F4/80-APC (#123115, BioLegend), followed by quadrant analysis of CD86-PE (#159203, BioLegend) and CD206-FITC (#141703, BioLegend) within the F4/80⁺ population to distinguish M1-like and M2-like phenotypes. For T-cell phenotyping, the identical initial gating strategy was applied to exclude cellular debris, doublets, and dead cells. For the evaluation of the CD4⁺/CD8⁺ T-cell ratio, these CD45 + tumor-infiltrating lymphocytes were stained with anti-mouse CD3-PerCP-Cy5.5(#100207, BioLegend), CD8a-PE (#100707, BioLegend), and CD4-FITC (#100405, BioLegend). For the evaluation of surface exhaustion markers PD-1 and T-cell immunoglobulin and mucin receptor 3 (TIM-3) on CD8⁺ T cells, cells were stained with anti-mouse CD3-PerCP-Cy5.5 (#100207, BioLegend), CD8a-PE (#100707, BioLegend), PD-1-APC (#135209, BioLegend), and TIM-3-BV421 (#134019, BioLegend). This surface staining process was conducted in the dark at 4 °C for 1 h. For the evaluation of functional intracellular proteins interferon-gamma (IFN-γ) and Granzyme B in CD8⁺ T cells, the obtained single-cell suspensions were first stimulated ex vivo with Cell Activation Cocktail (containing PMA, ionomycin, and Brefeldin A) (BioLegend, USA) at 37 °C for 4–6 h to trigger cytokine production and prevent their secretion. Following this stimulation, cells from the viable singlet gate were first stained with the surface markers anti-mouse CD45-BV510 (#103138, BioLegend), CD3-PerCP-Cy5.5 (#100207, BioLegend) and CD8a-PE (#100707, BioLegend). After washing off unbound antibodies, cells underwent fixation and permeabilization utilizing the Intracellular Fixation & Permeabilization Wash Buffer Kit (BioLegend, USA). Subsequently, the permeabilized cells were incubated with anti-mouse IFN-γ-APC (#505809, BioLegend) and Granzyme B-BV421 (#396413, BioLegend) antibodies in the dark. Finally, all stained samples were analyzed using a flow cytometer (BD Bioscience, USA), and the data and gating strategies were comprehensively analyzed utilizing FlowJo software (Tree Star, USA).
Enzyme-linked immunosorbent assay (ELISA)
The levels of TNF-α, IL-6, interleukin 1β (IL-1β), interleukin 10 (IL-10), C-C motif chemokine ligand 2 (CCL2), and C-X-C motif chemokine ligand 10 (CXCL10) in conditioned medium (CM) and tissues were quantified using the corresponding ELISA kits (R&D Systems, USA) according to the manufacturer’s instructions.
Immunofluorescence assay
Cells in different groups were harvested, fixed with 4% paraformaldehyde for 15 min at room temperature, and subsequently permeabilized using 0.1% Triton X-100. After undergoing standard blocking procedures, cells were treated with primary antibodies (4 °C, overnight) and fluorophore-conjugated secondary antibody (2 h, room temperature). Nuclear visualization was achieved through 4’,6-diamidino-2-phenylindole (DAPI) counterstaining. Cell images were examined using a confocal microscope. The primary antibodies used were as follows: arginase 1 (Arg1) (ab96183, Abcam, 1:500 dilution), and inducible nitric oxide synthase (iNOS) (ab178945, Abcam, 1:500 dilution).
Immunohistochemistry
The tissue specimens were fixed in 10% formaldehyde, paraffin-embedded, and sectioned into 4-µm slices using a microtome. After routine deparaffinization and rehydration procedures, antigen retrieval was conducted using microwave-mediated heating in 0.01 M sodium citrate buffer. Next, immunohistochemistry was conducted by incubation with primary antibodies (4 °C, overnight) and a horseradish peroxidase (HRP)-conjugated secondary antibody (1 h, room temperature). A diaminobenzidine (DAB) chromogenic substrate was employed for target protein visualization. Finally, image evaluation was performed using a light microscope and ImageJ software. The primary antibodies used were as follows: SOCS3 (ab280884, Abcam, 1:1000 dilution) Ki67 (ab16667, Abcam, 1:200 dilution).
Animal experiments
C57BL/KaLwRij mice (6–8 weeks) were obtained from YILAIBO Biotechnology (Shanghai, China), and housed in a controlled laboratory environment (22–25 °C temperature, 12-h light/dark cycle). Animal experimental procedures were approved by the Animal Ethics Committee of Quzhou People’s Hospital. Mice were divided into four groups ( n = 8 per group): oe-NC + IgG group, oe-SOCS3 + IgG group, oe-NC+anti-PD-1 group, and oe-SOCS3 + anti-PD-1 group. This sample size was predetermined to ensure adequate statistical power (> 80%, alpha = 0.05) to detect significant differences in tumor progression, while adhering to the “Reduction” principle of animal welfare. Murine 5TGM1 MM cells (1 × 10 7 cells/mL) stably transduced with a lentiviral SOCS3 overexpression vector or its negative control were injected subcutaneously into the left posterior flank area of the mice. For anti-PD-1 therapy, mice were intraperitoneally injected with 200 µg of an anti-PD-1 monoclonal antibody (BE0146, BioXCell, USA) or IgG isotype control (BE0089, BioXCell, USA) every 3 days, starting 7 days post-cell transplantation [ 36 ]. In our study design, tumor volume and final tumor weight were pre-defined as the primary in vivo endpoints to evaluate therapeutic efficacy, with overall survival serving as a secondary observational readout. Accordingly, tumor volume was monitored at 7-day intervals following inoculation. The experiment was concluded at the pre-defined endpoint on day 28, at which point the survival rate was 100% across all groups. All mice were subsequently euthanized, and tumor tissues were harvested for further examination.
Statistical analysis
Experimental data are presented as mean ± standard deviation (SD). The sample sizes for independent biological replicates were standardized to n = 6 for in vitro cellular experiments and n = 8 for in vivo animal studies. The clinical cohort for specimen analysis comprised 32 multiple myeloma (MM) patients and 18 normal controls. Statistical analyses were performed using SPSS 22.0 software. The normality of data distribution was assessed using the Shapiro‑Wilk test, and homogeneity of variances was evaluated using Levene’s test. For normally distributed data with equal variance, group differences were assessed via parametric tests: Student’s t -test was used for comparisons between two groups, and one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test was applied for multi-group analyses. If data failed to meet the assumptions of normality or equal variance, non-parametric tests (Mann-Whitney U test or Kruskal-Wallis test) or Welch’s t-test were utilized, respectively. Statistical significance was defined as P < 0.05 for all quantitative assessments.
Results
SOCS3 is downregulated in clinical MM specimens
Through integrative analysis of three MM-related GEO datasets ( GSE6477 , GSE24870 and GSE16558 ), we identified consensus downregulated genes as potential therapeutic targets. Volcano plot visualization (Fig. 1 A) and Venn diagram confirmation (Fig. 1 B) revealed 5 overlapping candidates, including neuron-derived neurotrophic factor (NDNF), Dickkopf-1 (DKK1), SOCS3, MORC family CW-type zinc finger 1 (MORC1), and interferon-induced transmembrane protein 1 (IFITM1). Subsequently, mRNA expression examination in clinical specimens demonstrated that SOCS3 was the most significantly downregulated gene among them in MM specimens (Fig. 1 C). Consistently, SOCS3 protein expression was downregulated in CD138 + plasma cells (Fig. 1 D and E) and bone marrow aspirates (Fig. 1 F) from MM patients. This downregulation pattern was further validated across multiple MM cell lines (Fig. 1 G and H), suggesting a potential tumor-suppressive role of SOCS3 in MM pathogenesis.
SOCS3 exerts tumor-suppressive effects on MM cells
Functional examination of SOCS3 was conducted in NCI-H929 cells through gain- and loss-of-function approaches. Western blot analysis confirmed efficient SOCS3 overexpression and silencing in NCI-H929 cells via oe-SOCS3 and si-SOCS3 transfection (Fig. 2 A and B). CCK-8 results demonstrated significant proliferation inhibition in SOCS3-overexpressing cells, contrasting with enhanced proliferation upon SOCS3 silencing (Fig. 2 C). The anti-proliferative effect was corroborated by colony formation assays, showing decreased colony numbers in the SOCS3 overexpression group and increased clone numbers in the SOCS3 silencing group (Fig. 2 D and E). Furthermore, SOCS3 overexpression inhibited cell migration (Fig. 2 F and G) and enhanced apoptosis (Fig. 2 H and I) in NCI-H929 cells. In contrast, SOCS3 silencing facilitated cell migration and reduced apoptosis.
Co-culture with SOCS3-overexpressing MM cells promotes macrophage M1-like polarization, accompanied by JAK2/STAT3 pathway suppression
To elucidate the role of SOCS3 in tumor microenvironment underlying MM pathogenesis, we established a co-culture system of SOCS3-modified NCI-H929 cells (overexpression/silencing) and macrophages. Functional analysis illustrated that co-culture with SOCS3-overexpressing NCI-H929 cells promoted THP-1-derived macrophage migration, whereas the SOCS3 silencing group impaired this process (Fig. 3 A and B). Moreover, flow cytometry analysis confirmed that co-culture with SOCS3-overexpressing NCI-H929 cells increased the M1-like macrophage proportion (Fig. 3 C and D) and decreased the M2-like proportion (Fig. 3 C and E). Conversely, SOCS3 silencing yielded the opposite results. To further validate this phenotypic shift beyond surface markers, we evaluated a comprehensive panel of intracellular enzymes and cytokines. Consistently, M1-like macrophage markers including iNOS (Fig. 3 F and H), TNF-α (Fig. 3 I), IL-6 (Fig. 3 J), and IL-1β (Fig. 3 K) were increased and M2-like macrophage markers including Arg1 (Fig. 3 G and H) and IL-10 (Fig. 3 L) were decreased in the SOCS3 overexpression group. In contrast, SOCS3 silencing exerted opposing effects. Since SOCS3 is an intracellular protein, we hypothesized that SOCS3-modified MM cells regulate macrophage polarization via a paracrine mechanism. ELISA of conditioned media from SOCS3-modified MM cells suggested that SOCS3 overexpression significantly decreased the secretion of canonical JAK/STAT activators (IL-6 and IL-10) and M2-recruiting chemokine (CCL2), while enhancing the secretion of M1-promoting chemokine (CXCL10) (Supplementary Fig. S4 ). In contrast, SOCS3 silencing yielded the opposite secretory profile. This altered secretome effectively deprived neighboring macrophages of immunosuppressive activating stimuli. We then determined whether the observed changes in JAK2/STAT3 phosphorylation occurred in macrophages. Western blot results showed that co-culture with SOCS3‑overexpressing MM cells decreased p-JAK2 and p-STAT3 levels in THP-1-derived macrophages (Fig. 3 M and N), while SOCS3 silencing conversely enhanced these phosphorylation events. Given that IL‑6 and IL‑10 are canonical activators of JAK2/STAT3 signaling, the reduced secretion of these cytokines by SOCS3‑overexpressing MM cells provides a paracrine explanation for the suppressed JAK2/STAT3 pathway in macrophages. Collectively, our findings demonstrate that co-culture with SOCS3-overexpressing MM cells promotes macrophage M1-like polarization, a phenotypic reprogramming that is closely associated with paracrine-mediated suppression of the macrophage-intrinsic JAK2/STAT3 signaling pathway in macrophages.
METTL3 inhibits SOCS3 expression through m6A modification in MM cells
SRAMP database showed multiple m6A methylation sites in the SOCS3 transcript (Fig. 4 A), indicating that SOCS3 is regulated by m6A modification. Notably, SOCS3 m6A level was upregulated in MM samples (Fig. 4 B). To identify the key m6A regulator in MM, we profiled the expression of six major enzymes in clinical MM and normal samples (8 cases each). RT-qPCR analysis revealed METTL3 as the most markedly upregulated gene in MM samples (Fig. 4 C). Our subsequent experiments confirmed the upregulated METTL3 expression in all MM samples (Fig. 4 D), and METTL3 expression was positively correlated with SOCS3 m6A level in MM samples (Fig. 4 E). Then, efficient METTL3 overexpression and silencing were confirmed in NCI-H929 and RPMI-8226 cell lines via Western blot analysis (Fig. 4 F and G). To investigate the direct regulatory mechanism, RIP-qPCR assay was performed, revealing a significant enrichment of SOCS3 transcripts in the METTL3-IP group compared to the IgG control, confirming that METTL3 strongly binds to the SOCS3 transcript (Fig. 4 H). MeRIP-qPCR results suggested that METTL3 overexpression increased SOCS3 m6A level in NCI-H929 cells and RPMI-8226 cells, while METTL3 silencing reduced it (Fig. 4 I). To validate the specific m6A sites, we focused on two high‑confidence m6A motifs (GAACT; Probability > 0.99) predicted at positions 1427 and 2625 within the 3’ untranslated region (3’ UTR) of the human SOCS3 transcript (GenBank accession number: NM_001378932.1 ) using the SRAMP database. SELECT-qPCR analysis revealed that silencing METTL3 significantly reduced m6A abundance at both positions 1427 and 2625 (Fig. 4 J). Furthermore, we constructed luciferase reporter plasmids containing either the wild-type (WT) or mutant (Mut) SOCS3 sequences (where the consensus adenosine at 1427 and 2625 was mutated to cytosine), as schematically illustrated in Supplementary Fig. S2 . Dual-luciferase reporter assay demonstrated that METTL3 overexpression significantly suppressed the luciferase activity of the wild-type (WT) reporter. However, mutation of either site alone partially restored luciferase activity, whereas simultaneous mutation of both sites almost abolished METTL3‑mediated suppression (Fig. 4 K). This site-directed mutagenesis structurally and functionally confirms that METTL3 mediates specific m6A modification at positions 1427 and 2625 on the SOCS3 transcript. Furthermore, METTL3 overexpression inhibited SOCS3 expression, and METTL3 silencing conversely upregulated SOCS3 expression consistently across both NCI-H929 and RPMI-8226 cell lines (Fig. 4 L and N). As expected, METTL3 overexpression reduced SOCS3 mRNA stability in both NCI-H929 and RPMI-8226 cell lines (Fig. 4 O and P). Finally, to elucidate the downstream reader protein mediating the degradation of m6A-modified SOCS3, we investigated YTHDF2, a well-known m6A reader that significantly upregulated in MM samples. After confirming the efficient knockdown of YTHDF2 in NCI-H929 cells (Supplementary Fig. S5 A), RIP-qPCR confirmed that YTHDF2 directly bound to SOCS3 mRNA (Supplementary Fig. S5 B). Crucially, YTHDF2 knockdown rescued the SOCS3 mRNA and protein expression levels that were suppressed by METTL3 overexpression (Supplementary Fig. S5 C-S5E). Collectively, these findings demonstrate that METTL3 methylates SOCS3 at specific sites, which are subsequently recognized by YTHDF2 to accelerate transcript degradation.
METTL3 silencing suppresses MM cell progression by enhancing SOCS3 expression
To elucidate the functions of METTL3/SOCS3 axis in MM, genetic rescue experiments were conducted in NCI-H929 cells. Western blot analysis confirmed that METTL3 silencing increased SOCS3 protein expression, while si-SOCS3 transfection inhibited its expression (Fig. 5 A and B). Functional studies demonstrated that METTL3 silencing restrained cellular proliferation (Fig. 5 C) and colony formation potential (Fig. 5 D and E) which were effectively reversed by SOCS3 silencing. Furthermore, the inhibitory effects of METTL3 silencing on migration (Fig. 5 F and G) were also abolished upon SOCS3 silencing. The enhancement of apoptosis induced by METTL3 silencing was counteracted by SOCS3 silencing (Fig. 5 H and I).
Co-culture with METTL3-silenced MM cells promotes macrophage M1-like polarization by enhancing SOCS3 expression
We then illustrated that co-culture with METTL3-silenced NCI-H929 cells enhanced macrophage migration, while SOCS3 silencing counteracted this enhancement (Fig. 6 A and B). Furthermore, co-culture with METTL3-silenced NCI-H929 cells promoted a phenotypic shift in macrophages from an M2-like to an M1-like state (Fig. 6 C and E), accompanied by the upregulated expression of M1-like markers including iNOS (Fig. 6 F and H), TNF-α, IL-6, and IL-1β (Fig. 6 J and L) and the downregulated expression of M2-like signatures Arg1 (Fig. 6 G and I) and IL-10 (Fig. 6 M). However, these effects were abolished upon SOCS3 silencing. Additionally, METTL3 silencing inhibited JAK2/STAT3 pathway activation in co-cultured macrophages (Fig. 6 N and P), which was rescued by SOCS3 silencing. Co-culture with METTL3-silenced MM cells promoted macrophage M1-like polarization by enhancing SOCS3 expression. Collectively, our results suggest that METTL3-mediated m6A modulation of SOCS3 regulates macrophage M1-like polarization in MM progression.
SOCS3 overexpression potentiates the antitumor efficacy of anti-PD-1 therapy in mice
Given the limited therapeutic efficacy of anti-PD-1 therapy in MM, the antitumor efficacy of SOCS3 overexpression and anti-PD-1 combination therapy was explored in a syngeneic murine subcutaneous tumor mouse model. C57BL/KaLwRij mice were subcutaneously injected with SOCS3-overexpressing or control 5TGM1 cells, followed by treatment with an anti-PD-1 monoclonal antibody or an IgG isotype control starting 7 days post-inoculation. We showed that SOCS3 overexpression and anti-PD-1 treatment individually reduced tumor volume and weight (Fig. 7 B and C). Furthermore, the combination group exhibited the most pronounced inhibition of tumor volume and weight. Notably, at the predefined experimental endpoint on day 28, the survival rate was 100% across all groups because the tumor burden remained well below the institutional ethical humane endpoint. Immunohistochemistry assays confirmed the upregulation of SOCS3 protein levels in the SOCS3 overexpression group and the combination group (Fig. 7 D and E). Moreover, Ki67 expression was significantly decreased in the combination group compared with two monotherapy groups (Fig. 7 F and G). Our collective results revealed that SOCS3 overexpression potentiated the antitumor efficacy of anti-PD-1 therapy in mice.
SOCS3 overexpression combined with anti-PD-1 therapy enhances antitumor immune response in MM mice
Flow cytometry analysis illustrated that the combination of SOCS3 overexpression and anti-PD-1 therapy remarkably augmented the M1-like macrophage proportion and decreased the M2-like macrophage proportion (Fig. 8 A and C) compared with the two monotherapies. Consistently, immunohistochemistry assays showed that iNOS level was increased and Arg1 level was decreased in the combination group compared with the monotherapy groups (Fig. 8 D and F). Furthermore, immunohistochemical staining for PD-L1 revealed that SOCS3 overexpression significantly reduced PD-L1 expression on tumor cells regardless of anti-PD-1 treatment, whereas anti-PD-1 monotherapy alone did not alter PD-L1 levels (Fig. 8 G and H). Additionally, combination therapy promoted a pro-inflammatory tumor microenvironment by increasing cytotoxic CD8 + T-cell infiltration while reducing CD4 + T-cell infiltration (Fig. 8 I and K). This shift resulted in a higher CD8⁺/CD4⁺ T-cell ratio (Fig. 8 L), indicating a potentiated antitumor immune response in mice. The effector molecules and exhaustion markers of CD8⁺ T cells were then assessed by flow cytometry. Compared with the two monotherapy groups, the combination treatment significantly increased the percentages of IFN-γ⁺ CD8⁺ T cells and Granzyme B⁺ CD8⁺ T cells, while decreasing the frequencies of PD‑1⁺ CD8⁺ T cells and TIM‑3⁺ CD8⁺ T cells (Fig. 8 M and O), indicating alleviated T‑cell exhaustion. Taken together, these findings suggest that the combination of SOCS3 overexpression and anti-PD-1 therapy enhanced antitumor immune response in MM mice.
Discussion
This study elucidates the crucial immunomodulatory effects of SOCS3 in MM by promoting macrophage M1-like polarization and cytotoxic T-cell activation. SOCS3 downregulation in MM is driven by METTL3-mediated m6A modulation. Functionally, SOCS3 overexpression inhibits MM cell proliferation and migration while promoting apoptosis. Notably, SOCS3 promotes macrophage repolarization towards the M1-like phenotype via a paracrine mechanism, a transition closely associated with suppression of the JAK2/STAT3 signaling pathway in macrophages. Additionally, our in vivo data indicate that SOCS3 overexpression effectively reprogrammes authentic TAMs within the complex TME, thereby enhancing the infiltration and activation of cytotoxic CD8 + T cells. Importantly, combining SOCS3 overexpression with anti-PD-1 therapy enhances anti-tumor immunity in vivo, highlighting its translational potential in overcoming ICI resistance in MM.
Bioinformatics analysis previously revealed an inverse correlation between circulating SOCS3 levels and MM risk [ 26 ], supporting its potential tumor-suppressive role. Corroborating this, we observed consistent SOCS3 downregulation across clinical MM specimens and cell lines. Functional experiments further suggest that SOCS3 overexpression effectively suppresses in vitro MM cell malignant behaviors and in vivo tumor growth. Given that the aberrant JAK/STAT3 signaling is a critical driver of MM development and drug resistance [ 37 ], the direct antitumor effects of SOCS3 are largely attributable to its suppression of this oncogenic cascade. Importantly, our study expands the functional roles of SOCS3 beyond cell-intrinsic tumor suppression, revealing its capacity to reprogram the immunosuppressive TME. Both in vitro co-culture assays and in vivo subcutaneous models demonstrate that SOCS3-overexpressing MM cells drive macrophage polarization towards the antitumor M1-like phenotype. Within the in vivo model, this microenvironmental remodeling potentiated the antitumor efficacy of anti-PD-1 therapy by promoting cytotoxic CD8 + T-cell activation and alleviating T-cell exhaustion. Mechanistically, our findings suggest a paracrine communication network whereby tumor-cell SOCS3 alters the microenvironmental secretome by reducing canonical immunosuppressive cytokines (IL-6, IL-10) and modifying chemokine gradients (decreased CCL2, increased CXCL10). This secretome shift potentially deprives neighboring macrophages of immunosuppressive activating stimuli. Subsequent mechanistic analysis reveals that this M1-like phenotypic reprogramming is closely associated with paracrine-mediated suppression of the macrophage-intrinsic JAK2/STAT3 signaling pathway, a pathway widely implicated in TAM polarization and immune evasion across various malignancies [ 36 , 38 , 39 ]. These findings are consistent with emerging evidence linking SOCS3 to macrophage repolarization, immune cell infiltration, and T-cell exclusion within the TME [ 23 – 25 , 40 ]. This dual capacity to directly inhibit malignant progression while promoting the paracrine reprogramming of TAMs highlights SOCS3 as a promising multifaceted therapeutic target in MM.
Accumulating evidence establishes METTL3 overexpression as a hallmark feature of MM, where its upregulation in clinical specimens correlates with disease progression and drug resistance [ 31 , 41 – 43 ]. Consistently, we demonstrate that METTL3 silencing significantly suppressed MM cell proliferation and migration. Notably, identifying METTL3 as an upstream epigenetic inhibitor of SOCS3 through m6A modulation clarifies the mechanism driving SOCS3 downregulation in MM. METTL3-dependent SOCS3 mRNA destabilization via m6A modification has been documented in other pathological contexts [ 33 , 34 ]. Our study provides direct mechanistic evidence for this regulatory cascade in MM. Specifically, we pinpointed two high-confidence m6A motifs (positions 1427 and 2625) within the SOCS3 transcript and verified their functional necessity through site-directed mutagenesis reporter assays. Furthermore, we identified YTHDF2 as the specific reader protein that recognizes m6A-modified SOCS3 mRNA to accelerate its decay, providing evidence for a conserved METTL3-YTHDF2-SOCS3 regulatory axis across multiple MM cell lines. Moreover, functional rescue experiments revealed that SOCS3 silencing effectively counteracts the tumor-suppressive and M1-polarizing effects of METTL3 depletion. While these mutagenesis and rescue data substantiate SOCS3 as a critical mediator of METTL3-driven tumor progression and immune reprogramming, we acknowledge that SOCS3 is likely not the exclusive downstream target. As a master RNA methyltransferase, METTL3 undoubtedly regulates a broader epitranscriptomic network in MM, potentially modulating other oncogenic or immune-related transcripts within the TME. Future high-throughput MeRIP-seq analyses will be necessary to map its comprehensive immunomodulatory landscape. Nevertheless, these findings substantially extend the biological significance of m6A modifications to microenvironmental immune modulation, underscoring the interplay between epitranscriptomic mechanisms and immune regulation in MM progression.
The ability of SOCS3 overexpression to sensitize MM tumors to anti-PD-1 therapy in vivo highlights the translational significance of this study. ICIs have shown limited efficacy in MM due to T-cell exhaustion and an immunosuppressive TME [ 8 ]. Our findings demonstrate that SOCS3 restoration enhances anti-PD-1 efficacy by driving macrophage M1-like polarization and promoting cytotoxic CD8 + T-cell infiltration and activation. Notably, combination therapy significantly increased the frequencies of IFN‑γ + and Granzyme B + CD8 + T cells while reducing PD‑1 + and TIM‑3 + exhausted CD8 + T cells, indicating that SOCS3 restoration not only promotes CD8 + T cell recruitment but also reinvigorates their effector functions. Mechanistically, we propose that this synergy operates through a potential two-pronged network initiated by tumor-intrinsic SOCS3. First, SOCS3 overexpression downregulates tumor PD-L1, attenuating baseline immune evasion against PD-1 + T cells. Given that aberrant JAK/STAT3 signaling is a canonical transcriptional driver of PD-L1 in cancer [ 44 ], intrinsic SOCS3 restoration likely contributes to this PD-L1 downregulation through JAK/STAT3 pathway blockade. Second, tumor-cell SOCS3 remodels the paracrine secretome to suppress macrophage-intrinsic JAK2/STAT3 signaling, thereby driving M1-like polarization. Our findings suggest that these M1-like macrophages may act as immunostimulatory amplifiers that secrete chemokines, likely contributing to the recruitment and priming of CD8 + T cells. Combining this dual pre-sensitization (tumor PD-L1 reduction and M1-driven T-cell recruitment) with anti-PD-1 therapy potentially mitigates PD-1/PD-L1-mediated resistance. This approach may help alleviate T-cell exhaustion and restore robust cytotoxic functions, thereby facilitating the conversion of an immunologically “cold” myeloma microenvironment into a responsive “hot” tumor [ 45 ]. Our study provides a compelling preclinical rationale for restoring SOCS3 to potentiate ICI efficacy in MM.
Several limitations should be noted. First, the clinical relevance of our primary human sample analysis is currently limited by a small cohort size and the lack of long-term follow-up data. Consequently, the present study is unable to definitively validate the independent prognostic value of the METTL3/m6A/SOCS3 axis in MM patients. Future large-scale, multi-center clinical studies incorporating longitudinal survival cohorts are required. Second, our in vitro co-culture cell model and in vivo subcutaneous 5TGM1 tumor model may not fully recapitulate human TME complexity. Given that MM is predominantly a bone marrow-resident disease, the subcutaneous model lacks the specialized orthotopic niche (e.g., bone marrow stromal cells, specific extracellular matrix, and the characteristic local hypoxia) that profoundly dictates tumor-immune interactions and therapy resistance. Although the subcutaneous 5TGM1 model provides proof‑of‑concept evidence for macrophage reprogramming and antitumor immunity and may partially reflect features of extramedullary myeloma, we explicitly acknowledge that it does not reproduce this hypoxic orthotopic microenvironment. Therefore, future studies employing systemic or orthotopic bone marrow-homing models are necessary to validate these findings within a physiologically relevant niche. Third, although our study provides robust functional evidence of T-cell reinvigoration, the lack of long-term animal survival analysis in our mouse model represents a limitation requiring further investigation. Fourth, given the high heterogeneity and a continuous phenotypic spectrum of in vivo TAMs within the microenvironment, our simplified M1-like/M2-like classification based on conventional surface markers represents a limitation, and future single-cell transcriptomic or high-dimensional spatial analyses are needed to capture macrophage diversity. Fifth, while we focused on TAMs and CD8 + T cells, the regulatory roles of SOCS3 in other immunosuppressive or effector immune cells within the TME, such as regulatory T cells (Tregs), natural killer (NK) cells, and myeloid-derived suppressor cells (MDSCs), remain unexplored. Sixth, our in vitro tumor-macrophage co-culture and downstream functional rescue assays were predominantly characterized in the NCI-H929 cell line. Cross-verification across a broader panel of MM cell lines and primary patient-derived bone marrow macrophages will be critical to validate translational generalizability. Seventh, while our data establish the METTL3-YTHDF2 axis as a primary driver of SOCS3 mRNA decay, we did not perform parallel knockdown screening for other canonical m6A readers (e.g., YTHDF1, YTHDC1, and YTHDC2) or alternative writers/erasers. Consequently, we cannot rule out cooperative or redundant involvement of alternative reader homologues. Future high-throughput profiling and paralog-screening studies are warranted to fully delineate this regulatory network in MM. Eighth, in our in vivo models, SOCS3 was constitutively overexpressed in 5TGM1 tumor cells via lentiviral transduction. We acknowledge that such exogenous constitutive overexpression may yield supraphysiological SOCS3 levels, which cannot fully recapitulate the physiological restoration or reactivation of the endogenous SOCS3 locus. Finally, although SOCS3 restoration significantly inhibits JAK2/STAT3 phosphorylation, we did not conduct specific pharmacological blocking or genetic rescue experiments. Since SOCS3 is a canonical endogenous inhibitor of this pathway, its suppression of JAK2/STAT3 strongly correlates with the observed M1-like polarization. However, we cannot definitively claim that macrophage reprogramming strictly depends on this specific signaling axis. Future investigations utilizing targeted JAK2/STAT3 inhibitors or clustered regularly interspaced short palindromic repeats (CRISPR)-based rescue assays are required to establish this direct signaling dependency.
Beyond these experimental limitations, our translational conclusions must be interpreted with caution, as clinical application remains exploratory. While restoring SOCS3 function demonstrates potent antitumor efficacy in our models, systemic upregulation could induce side effects such as severe immunodeficiency or hematopoiesis abnormalities due to widespread JAK/STAT inhibition. Furthermore, because direct gene delivery via plasmids or lentiviruses faces clinical hurdles, future strategies should explore and meticulously differentiate between intratumoral local delivery and systemic targeted nanoparticles. Specifically, intratumoral injections of nano-formulations into extramedullary plasmacytomas could maximize local macrophage reprogramming while minimizing the risk of widespread JAK/STAT inhibition [ 46 ]. Conversely, for disseminated medullary myeloma, systemic targeted nanocarriers, such as bone marrow-homing lipid nanoparticles functionalized with myeloma-specific ligands, are essential to deliver genetic or pharmacological payloads directly to the tumor microenvironment without off-target toxicity [ 47 ]. Additionally, pharmacological strategies to restore SOCS3 function could primarily rely on specific METTL3 small-molecule inhibitors to translate our mechanistic findings. Existing preclinical literature has validated potent METTL3 inhibitors, including STM2457, UZH1a, and the clinical-stage STC-15, which demonstrate robust in vivo efficacy in myeloid leukemia and solid tumors [ 48 – 50 ]. To formulate a concrete translational roadmap, our subsequent studies will execute a phased preclinical validation strategy. First, building upon the established reference dosage of STM2457 (typically 50 mg/kg/day via systemic administration [ 48 ]), we propose encapsulating this inhibitor within B-cell maturation antigen-targeted lipid nanoparticles. This targeted design will specifically direct the pharmacological payload to myeloma cells and enhance drug accumulation within the bone marrow niche. To strictly mitigate potential on-target/off-tumor hematopoietic toxicity, our specific dosing regimen will employ a rationally designed intermittent schedule, such as 30–50 mg/kg administered every other day (QOD) or through a “5-days-on/2-days-off” cycle. This intermittent approach aligns with current dose-optimization paradigms in targeted oncology, which emphasize rational dosing schedules to mitigate toxicity and provide therapeutic windows for hematopoietic recovery [ 51 ]. Finally, this targeted nano-formulation, in combination with anti-PD-1 therapy, will undergo rigorous validation in an orthotopic myeloma model (e.g., via intravenous engraftment of luciferase-tagged 5TGM1 cells) [ 52 ]. This pipeline will allow us to evaluate systemic pharmacokinetics, monitor local macrophage-reprogramming efficacy, and conclusively confirm their immunomodulatory synergy prior to any clinical advancement.
In conclusion, our findings demonstrate that the downregulation of SOCS3 in MM is driven by METTL3-mediated m6A modification. We highlight the dual ability of SOCS3 in directly inhibiting tumor cell malignant progression while promoting macrophage M1-like polarization and cytotoxic T-cell activation in MM. The enhanced therapeutic efficacy achieved by combining SOCS3 restoration with anti-PD-1 therapy highlights a promising biological rationale for overcoming immunotherapy resistance in MM. Fully realizing this translational potential will require developing tumor-targeted delivery mechanisms to minimize systemic toxicity and advancing clinically translatable pharmacological approaches.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1 (DOCX 1.33 MB)
Acknowledgements
None.
Abbreviations
Multiple myeloma
Suppressor of cytokine signaling 3
tumor-associated macrophage
tumor microenvironment
immune checkpoint inhibitors
methyltransferase like 3
Janus kinase 2
signal transducer and activator of transcription 3
Authors’ contributions
Gang Wang: Methodology, Investigation, original draft. Fangyu Zhou: Methodology, Investigation, Data curation. Xuefen Yan: review & editing. Jiaheng Wang: Idea, Supervision, review & editing. Mengni Yan: Investigation, Data curation. Jinle Liu: Supervision, review & editing. Li Yu: Data curation, original draft.
Funding
This work was supported by grants from the Zhejiang Provincial Medical and Health Science and Technology Program (2023KY389).
Data availability
The datasets used during the present study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
This study involving human participants was approved by the Medical Ethics Committee of Quzhou People’s Hospital (Approval No.: 2022-041). All procedures were performed in accordance with the 1964 Helsinki Declaration and its later amendments. All participants provided the written informed consent. Animal experiments were granted approval by the Animal Ethics Committee of Quzhou People’s Hospital. All methods were carried out in accordance with relevant guidelines and regulations. This study was carried out in compliance with the ARRIVE guidelines.
Consent for publication
Not applicable.
Conflict of interest
The authors have no conflicts of interest to declare.
Clinical trial number
Not applicable.
Footnotes
Gang Wang and Fangyu Zhou are the co-first authors.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Gang Wang, Email: gumpwang@163.com.
Li Yu, Email: 15505708530@163.com.
References
1. van de Donk N, Pawlyn C, Yong KL (2021) Multiple myeloma. Lancet 397:410–427. 10.1016/s0140-6736(21)00135-5
2. George LL, Malik MN, Miller EJ, Hicks K, Khanam R, Saterehaseman A, McNichol M, Anwer F (2021) Special considerations for supportive care and management of complications in elderly patients with multiple myeloma. Clin Lymphoma Myeloma Leuk 21:812–822. 10.1016/j.clml.2021.07.013
3. Rasch S, Lund T, Asmussen JT, Lerberg Nielsen A, Faebo Larsen R, Østerheden Andersen M, Abildgaard N (2020) Multiple myeloma associated bone disease. Cancers (Basel) 12:2113. 10.3390/cancers12082113
4. Ohara F, Suzuki A, Suzuki N, Kanematsu T, Okamoto S, Tamura S, Kiyoi H, Matsushita T, Katsumi A (2024) Newly diagnosed multiple myeloma with bleeding and coagulation abnormalities caused by a thrombin-inhibiting substance. Int J Hematol 120:743–749. 10.1007/s12185-024-03849-2
5. Dimopoulos MA, Merlini G, Bridoux F, Leung N, Mikhael J, Harrison SJ, Kastritis E, Garderet L, Gozzetti A, van de Donk N, Weisel KC, Badros AZ, Beksac M, Hillengass J, Mohty M, Ho PJ, Ntanasis-Stathopoulos I, Mateos MV, Richardson P, Blade J, Moreau P, San-Miguel J, Munshi N, Rajkumar SV, Durie BGM, Ludwig H, Terpos E (2023) Management of multiple myeloma-related renal impairment: recommendations from the International Myeloma Working Group. Lancet Oncol 24:e293–e311. 10.1016/s1470-2045(23)00223-1
6. Liu Z, Xu X, Liu H, Zhao X, Yang C, Fu R (2023) Immune checkpoint inhibitors for multiple myeloma immunotherapy. Exp Hematol Oncol 12:99. 10.1186/s40164-023-00456-5
7. Le Calvez B, Moreau P, Touzeau C (2021) Immune checkpoint inhibitors for the treatment of myeloma: novel investigational options. Expert Opin Investig Drugs 30:965–973. 10.1080/13543784.2021.1955103
8. Alkharabsheh O, Trisel Z, Badami S, Aljama MA, Sidiqi MH (2021) Checkpoint inhibitors in multiple myeloma: intriguing potential and unfulfilled promises. Cancers (Basel) 14:113. 10.3390/cancers14010113
9. Li C, Xu X, Wei S, Jiang P, Xue L, Wang J (2021) Tumor-associated macrophages: potential therapeutic strategies and future prospects in cancer. J Immunother Cancer 9:e001341. 10.1136/jitc-2020-001341
10. Basak U, Sarkar T, Mukherjee S, Chakraborty S, Dutta A, Dutta S, Nayak D, Kaushik S, Das T, Sa G (2023) Tumor-associated macrophages: an effective player of the tumor microenvironment. Front Immunol 14:1295257. 10.3389/fimmu.2023.1295257
11. Gao J, Liang Y, Wang L (2022) Shaping polarization of tumor-associated macrophages in cancer immunotherapy. Front Immunol 13:888713. 10.3389/fimmu.2022.888713
12. Xu J, Ding L, Mei J, Hu Y, Kong X, Dai S, Bu T, Xiao Q, Ding K (2025) Dual roles and therapeutic targeting of tumor-associated macrophages in tumor microenvironments. Signal Transduct Target Ther 10:268. 10.1038/s41392-025-02325-5
13. Jordanides PP, Jagadeesha S, Upadhaya P, Ryan NM, Anderson K, Lamenza FF, Shrestha S, Siddiqui A, Bopp AR, Pracha SH, Roth P, Kehres R, Mo X, Oghumu S (2026) PI3Kγ inhibition drives M1 macrophage differentiation and synergizes with PD-L1 blockade to improve survival in poorly immunogenic head and neck squamous cell carcinoma. Cancer Biol Ther 27:2600701. 10.1080/15384047.2025.2600701
14. Wang C, Han L, Zhang J, Ji Q, Guo X, Li Y, Li S, He J, He F, Dai W, Li M, Yu J, Chen D, Gong Y, Xie C (2025) Radiotherapy in combination with PD-1 and TIGIT blockade mediate antitumor abscopal effects and immune memory via CD8 + T cells. Cancer Lett 631:217935. 10.1016/j.canlet.2025.217935
15. Luo F, Li H, Ma W, Cao J, Chen Q, Lu F, Qiu M, Zhou P, Xia Z, Zeng K, Zhan J, Zhou T, Luo Q, Pan W, Zhang L, Lin C, Huang Y, Zhang L, Yang D, Zhao H (2024) The BCL-2 inhibitor APG-2575 resets tumor-associated macrophages toward the M1 phenotype, promoting a favorable response to anti-PD-1 therapy via NLRP3 activation. Cell Mol Immunol 21:60–79. 10.1038/s41423-023-01112-y
16. Pu Y, Ji Q (2022) Tumor-associated macrophages regulate PD-1/PD-L1 immunosuppression. Front Immunol 13:874589. 10.3389/fimmu.2022.874589
17. Zheng X, Shao J, Wei S, Gu Y, Qian J (2021) Prognostic significance of SOCS3 in patients with solid tumors: a meta-analysis. Front Surg 8:802143. 10.3389/fsurg.2021.802143
18. Dai L, Li Z, Tao Y, Liang W, Hu W, Zhou S, Fu X, Wang X (2021) Emerging roles of suppressor of cytokine signaling 3 in human cancers. Biomed Pharmacother 144:112262. 10.1016/j.biopha.2021.112262
19. Li X, Yang Z, Chen B, Gu L, Tian G, Sui X (2023) SOCS3 as a potential driver of lung metastasis in colon cancer patients. Front Immunol 14:1088542. 10.3389/fimmu.2023.1088542
20. Jeong IH, Yun JK, Jin JO, Hong JH, Lee JY, Lee GD, Lee PC (2024) E3 ligase SOCS3 regulates NOD2 expression by ubiquitin proteasome system in lung cancer progression. Cell Oncol (Dordr) 47:819–832. 10.1007/s13402-023-00896-5
21. Li X, Peng B, Li J, Tian M, He L (2023) Unleashing breast cancer progression: miR-455-5p’s targeting of SOCS3 drives proliferation, migration, and invasion. Protein Pept Lett 30:992–1000. 10.2174/0109298665245603231106050224
22. Al Saqri A, Malgundkar SH, Al Kindi F, Gupta I, Al Moundhri M, Tamimi Y (2022) SOCS3 gene silencing does not occur through methylation and mutations in gastric cancer. Hum Cell 35:1114–1125. 10.1007/s13577-022-00715-3
23. Jafarzadeh A, Zandvakili R, Jafarzadeh Z, Nemati M (2024) Dysregulated expression of the suppressors of cytokine signaling (SOCS) contributes to the development of prostate cancer. Pathol Res Pract 262:155558. 10.1016/j.prp.2024.155558
24. Chikuma S, Kanamori M, Mise-Omata S, Yoshimura A (2017) Suppressors of cytokine signaling: Potential immune checkpoint molecules for cancer immunotherapy. Cancer Sci 108:574–580. 10.1111/cas.13194
25. Ni J, Xi X, Xiao S, Xiao X (2024) Tumor cell-derived exosomal miR-191-5p activates M2-subtype macrophages through SOCS3 to facilitate breast cancer. Mol Biotechnol 66:1314–1325. 10.1007/s12033-023-01034-0
26. Wang Q, Shi Q, Wang Z, Lu J, Hou J (2023) Integrating plasma proteomes with genome-wide association data for causal protein identification in multiple myeloma. BMC Med 21:377. 10.1186/s12916-023-03086-0
27. Wilop S, van Gemmeren TB, Lentjes MH, van Engeland M, Herman JG, Brümmendorf TH, Jost E, Galm O (2011) Methylation-associated dysregulation of the suppressor of cytokine signaling-3 gene in multiple myeloma. Epigenetics 6:1047–1052. 10.4161/epi.6.8.16167
28. Lin L, Cao L, Liu Y, Wang K, Zhang X, Qin X, Zhao D, Hao J, Chang Y, Huang X, Liu B, Zhang J, Lu J, Ge Q (2019) B7-H3 promotes multiple myeloma cell survival and proliferation by ROS-dependent activation of Src/STAT3 and c-Cbl-mediated degradation of SOCS3. Leukemia 33:1475–1486. 10.1038/s41375-018-0331-6
29. Liang P, Cheng SH, Cheng CK, Lau KM, Lin SY, Chow EY, Chan NP, Ip RK, Wong RS, Ng MH (2013) Platelet factor 4 induces cell apoptosis by inhibition of STAT3 via up-regulation of SOCS3 expression in multiple myeloma. Haematologica 98:288–295. 10.3324/haematol.2012.065607
30. Usui E, Nishii K, Katayama N, Lorenzo VF, Chen F, Monma F, Otsuki T, Shiku H (2004) Upregulated production of IL-6, but not IL-10, by interferon-alpha induces SOCS3 expression and attenuates STAT1 phosphorylation in myeloma cells. Hematol J 5:505–512. 10.1038/sj.thj.6200565
31. Huang X, Yang Z, Li Y, Long X (2023) m6A methyltransferase METTL3 facilitates multiple myeloma cell growth through the m6A modification of BZW2. Ann Hematol 102:1801–1810. 10.1007/s00277-023-05283-6
32. Yu T, Yao L, Yin H, Teng Y, Hong M, Wu Q (2022) ALKBH5 Promotes Multiple Myeloma Tumorigenicity through inducing m6A-demethylation of SAV1 mRNA and myeloma stem cell phenotype. Int J Biol Sci 18:2235–2248. 10.7150/ijbs.64943
33. Zhang Y, Wang L, Yan F, Yang M, Gao H, Zeng Y (2023) METTL3 mediated m6A methylation involved in epithelial-mesenchymal transition by targeting SOCS3/STAT3/SNAI1 in cigarette smoking-induced COPD. Int J Chron Obstruct Pulmon Dis 18:1007–1017. 10.2147/copd.s398289
34. Chen Y, Miao X, Xiang Y, Kuai L, Ding X, Ma T, Li B, Fan B (2023) Qinzhu Liangxue inhibits IL-6-induced hyperproliferation and inflammation in HaCaT cells by regulating METTL14/SOCS3/STAT3 axis. J Ethnopharmacol 317:116809. 10.1016/j.jep.2023.116809
35. Yang H, He Y, Qu F, Zhu J, Deng L, Jiang F, Wu X, Chen Y, Kashif A, Wang X (2025) Maraviroc enhances Bortezomib sensitivity in multiple myeloma by inhibiting M2 macrophage polarization via PI3K/AKT/RhoA signaling pathway in macrophages. Cell Div 20:5. 10.1186/s13008-025-00145-1
36. Thuya WL, Cao Y, Ho PC, Wong AL, Wang L, Zhou J, Nicot C, Goh BC (2025) Insights into IL-6/JAK/STAT3 signaling in the tumor microenvironment: Implications for cancer therapy. Cytokine Growth Factor Rev 85:26–42. 10.1016/j.cytogfr.2025.01.003
37. Jung YY, Um JY, Nasif O, Alharbi SA, Sethi G, Ahn KS (2021) Blockage of the JAK/STAT3 signaling pathway in multiple myeloma by leelamine. Phytomedicine 87:153574. 10.1016/j.phymed.2021.153574
38. Xu X, Xu Z, Cai Y, Chen X, Huang C (2025) CKIP-1 inhibits M2 macrophage polarization to suppress the progression of gastric cancer by inactivating JAK/STAT3 signaling. Cell Biochem Biophys 83:1289–1298. 10.1007/s12013-024-01562-9
39. Qi Y, Li R, Han M (2024) Tumor-associated macrophages induce epithelial-mesenchymal transition and promote lung metastasis in breast cancer by activating the IL-6/STAT3/TGM2 axis. Int Immunopharmacol 143:113387. 10.1016/j.intimp.2024.113387
40. Tang B, Zhu J, Wang Y, Chen W, Fang S, Mao W, Xu Z, Yang Y, Weng Q, Zhao Z, Chen M, Ji J (2023) Targeted xCT-mediated ferroptosis and protumoral polarization of macrophages is effective against HCC and enhances the efficacy of the anti-PD-1/L1 response. Adv Sci (Weinh) 10:e2203973. 10.1002/advs.202203973
41. Che F, Ye X, Wang Y, Wang X, Ma S, Tan Y, Mao Y, Luo Z (2023) METTL3 facilitates multiple myeloma tumorigenesis by enhancing YY1 stability and pri-microRNA-27 maturation in m6A-dependent manner. Cell Biol Toxicol 39:2033–2050. 10.1007/s10565-021-09690-1
42. Bao J, Xu T, Wang W, Xu H, Chen X, Xia R (2024) N6-methyladenosine-induced miR-182-5p promotes multiple myeloma tumorigenesis by regulating CAMK2N1. Mol Cell Biochem 479:3077–3089. 10.1007/s11010-023-04906-w
43. Wang G, Wu W, He D, Wang J, Kong H, Wu W (2025) N6-methyladenosine-mediated upregulation of H19 promotes resistance to bortezomib by modulating the miR-184/CARM1 axis in multiple myeloma. Clin Exp Med 25:102. 10.1007/s10238-025-01624-z
44. Yao Y, Tao G, Xie D, Xu Y, Li L (2025) Effect of JAK/STAT signaling on anti-PD-L1 immunotherapy efficacy in lung cancer. Crit Rev Oncol Hematol 214:104861. 10.1016/j.critrevonc.2025.104861
45. Yamamoto L, Amodio N, Gulla A, Anderson KC (2020) Harnessing the immune system against multiple myeloma: challenges and opportunities. Front Oncol 10:606368. 10.3389/fonc.2020.606368
46. Yun WS, Kim J, Lim DK, Kim DH, Jeon SI, Kim K (2023) Recent studies and progress in the intratumoral administration of nano-sized drug delivery systems. Nanomaterials (Basel) 13:2225. 10.3390/nano13152225
47. Alhallak K, Sun J, Wasden K, Guenthner N, O’Neal J, Muz B, King J, Kohnen D, Vij R, Achilefu S, DiPersio JF, Azab AK (2021) Nanoparticle T-cell engagers as a modular platform for cancer immunotherapy. Leukemia 35:2346–2357. 10.1038/s41375-021-01127-2
48. Yankova E, Blackaby W, Albertella M, Rak J, De Braekeleer E, Tsagkogeorga G, Pilka ES, Aspris D, Leggate D, Hendrick AG, Webster NA, Andrews B, Fosbeary R, Guest P, Irigoyen N, Eleftheriou M, Gozdecka M, Dias JML, Bannister AJ, Vick B, Jeremias I, Vassiliou GS, Rausch O, Tzelepis K, Kouzarides T (2021) Small-molecule inhibition of METTL3 as a strategy against myeloid leukaemia. Nature 593:597–601. 10.1038/s41586-021-03536-w
49. Lin T, Chen X, Huang H, Zhang J, Xu Y, Lin J (2026) Pre-clinical anti-tumor activity and preliminary safety of METTL3 inhibitors STM2457 and UZH1a in neuroblastoma. Transl Cancer Res 15:186. 10.21037/tcr-2025-aw-2392
50. Tang H, Zhang R, Zhang A (2025) Small-molecule inhibitors targeting RNA m6A modifiers for cancer therapeutics: latest advances and future perspectives. J Med Chem 68:18114–18142. 10.1021/acs.jmedchem.5c01570
51. Shah M, Rahman A, Theoret MR, Pazdur R (2021) The drug-dosing conundrum in oncology: when less is more. N Engl J Med 385:1445–1447. 10.1056/nejmp2109826
52. Gowda PS, Wildman BJ, Trotter TN, Xu X, Hao X, Hassan MQ, Yang Y (2018) Runx2 suppression by miR-342 and miR-363 inhibits multiple myeloma progression. Mol Cancer Res 16:1138–1148. 10.1158/1541-7786.mcr-17-0606
Associated Data
Supplementary Materials
Supplementary Material 1 (DOCX 1.33 MB)
Data Availability Statement
The datasets used during the present study are available from the corresponding author upon reasonable request.