BM samples derived from healthy donors (HD; n=5) and newly diagnosed patients with MM (n=9) were obtained from Xiangya Hospital, the Second Xiangya Hospital of Central South University (Changsha, China), from January 2020 to June 2023. CD138⁺ plasma cells were isolated by using anti-human CD138 magnetic beads (Miltenyi Biotec GmbH) and incubated in 4°C for 15 min with monocytes isolated from BM samples using lymphocyte separation medium (cat. no. LTS1077; TBD; https://www.tbdscience.com/). The patients with MM enrolled in the present study were newly diagnosed. International Myeloma Working Group criteria (19) were processed by hematologists for the diagnosis of MM. Patients with monoclonal gammopathy of undetermined significance (MGUS) and smoldering MM were excluded, together with patients with MM combined with other diseases. The clinical information of the enrolled patients with MM in the present study is included in Table SI. The present study was approved by Cancer Research Institute Review Board of Central South University (Changsha, China).

The human MM cell lines ARP1 (20), KMS28-BM (21), KMS28-pleural effusion (PE) (20) and OCI-My5 (20,22), were utilized to explore the function of LGR4 in MM. Cell lines were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (FBS; cat. no. FSP500; Shanghai ExCell Biology, Inc.) and 1% penicillin and streptomycin (P/S; Gibco; Thermo Fisher Scientific, Inc.). The cell lines have been used in previous studies (20,22). The human MM cell line OCI-T3^rd-luc (derived from OCI-My5-luc) was established by our group using tail vein injection into NCG mice with three rounds of homing transplantation and enriched from BM. KMS28-BM (21) and KMS28-PE cells are the paired cell lines originated from BM and PE of MM patients with immortalization. The human BM stromal cell line HS5 (a gift from Dr. Jiaxi Zhou, Institute of Hematology, Chinese Academy of Medical Sciences) was maintained in DMEM low glucose supplemented with 10% FBS and 1% P/S.

Reagents included QNZ (cat. no. EVP4593; Selleck Chemicals) and doxycycline (DOX; cat. no. 24390-14-5; MilliporeSigma). The antibodies were as follows: Anti-LGR4 (1:1,000; cat. no. A12657; for western blotting), anti-β-actin (1:5,000; cat. no. AC004), anti-nuclear factor kappa B (NF-κB) 2 (1:1,000; cat. no. A19605), anti-P-NF-κB2-S866 (1:1,000; cat. no. Ap0418), anti-IκBα (1:1,000; cat. no. A19714), anti-P-IκBα-S36 (1:1,000; cat. no. A191714), anti-Snail (1:1,000; cat. no. A5243) and anti-TNFRSF1B (1:1,000; cat. no. A13556) were all obtained from ABclonal Biotech Co., Ltd. Anti-GAPDH (1:5,000; cat. no. 10494-1-AP), anti-MUC2 (1:1,000; cat. no. 27675-1-AP), anti-Caspase (1:1,000; cat. no. 319677-1-AP) and anti-Zinc Finger E-Box Binding Homeobox 1 (ZEB1; 1:1,000; cat. no. 21544-1-AP) were all obtained from Proteintech Group, Inc. Anti-poly-(ADP-ribose) polymerase (PARP) antibody (1:1,000; cat. no. 9532), anti-cleaved caspase 3 antibody (1:1,000; cat. no. 9664) and anti-Vimentin (1:1,000; cat. no. D21H3) were obtained from Cell Signaling Technology, Inc. Anti-N-cadherin (1:800; cat. no. WL011047) was obtained from Wanleibio Co., Ltd. PE anti-human CD138 (cat. no. 352306) was obtained from BioLegend, Inc. and CXCR12 (cat. no. 350-NS) from R&D Systems, Inc.

LGR4-overexpression (LGR4-OE) constructs were constructed by cloning LGR4 cDNA into a pSIN-EF2-Puro (23) lentiviral vector using EcoRI (cat. no. R3101S; New England Biolabs) and BamHI (cat. no. R3136S; New England Biolabs). LGR4-knockdown constructs using two pairs of short hairpin RNA sequences (shRNA1 and shRNA2) were ligated into a pLKO-tet-on lentiviral vector. Aim Lentiviruses (5 μg) were packaged in 293T cells (a gift from Dr. Rong Chang, Kunming Institute of Zoology, the Chinese Academy of Sciences) using pMD2G (1.25 μg) and psPAX2 (3.75 μg) helper vectors and polybrene (8 μg/ml)-mediated transduction (cat. no. H9268-5G; MilliporeSigma). After 60 h the 10 ml virus was collected and 1ml virus was used to transfect 1×10⁶ ARP1 or OCI-My5 cell lines in 1 ml medium. A total of 48-72 h after the transfection, puromycin (1 μg/ml) was added to screen the positive cells. The final concentration of siRNA transfection was 50 nM. Transient transfection was performed using a Nefect DNA Transfection reagent (cat. no. TF20121201; Neofect; http://www.neofect.cn/) according to the specification. All primer and siRNA sequences are listed in Tables SII and SIII.

Western blot analysis was performed as previously described (20). Proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a 0.45-μm polyvinylidene fluoride membrane. The blots were then probed with specific primary antibodies overnight at 4°C, followed by HRP-conjugated secondary antibodies incubation for 1 h at room temperature (RT). Protein signals were developed with SuperSignalTM West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific, Inc.).

To detect mRNA expression in MM cells, RT-qPCR was performed as previously described (22). Total RNA was extracted using TRIzol^® (cat. no. 15596026; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Total RNA was reverse transcribed using the SuperScript™ II Reverse Transcriptase kit (cat. no. 18064071; Thermo Fisher Scientific, Inc.). qPCRs were performed by using ABsolute qPCR SYBR Green Mixes (cat. no. AB1163A; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. All primer sequences are listed in Table SIV.

A total of 4×10⁴ CD138⁺ cells were spun down on glass slides and then fixed with methanol for 15 min at 20°C. Diluted LGR4 antibodies (1:200) and ZEB1 antibodies (1:200) were placed on glass slides and incubated overnight at 4°C. Then, the cells were incubated with secondary antibodies (1:1,000) conjugated with Goat anti-Rabbit Alexa Fluor™ 488 (cat. no. A-11008; Thermo Fisher Scientific, Inc.) or Goat anti-Mouse Alexa Fluor™ 594 (cat. no. A-11012; Thermo Fisher Scientific, Inc.) for 2 h at RT (protected from light). Nuclei were labeled with 4′,6-diamidino-2-phenylindole (DAPI, 1 mg/ml; 1:5,000) (cat. no. S2110; Beijing Solarbio Science & Technology Co., Ltd.). Fluorescence was observed under a Leica fluorescence microscope. The fluorescence intensity was quantified using ImageJ 1.54 software (National Institutes of Health).

To determine cell proliferation, MM cells were plated in 24-well plates with 5,000 cells per well by counting alive cells after trypan blue exclusion. Cell numbers were counted using a cell counting chamber for six days. To determine cell viability, MM cells were plated in 96-well plates at a density of 5,000 cells per well. The cells were treated with QNZ for 48 h and counted using Cell Counting Kit-8 (cat. no. B34302; Bimake). For each well, 10 μl of reagent was added, followed by incubation at 37°C for 2-3 h. The optical density (OD) was measured at 450 nm. Each test was repeated three times. IC₅₀ was calculated using GraphPad Prism 9 software.

Soft agar colony formation assay was performed as previously described (22). The colonies were treated with RPMI-1640 complete medium in the presence or absence of DOX twice every week. One colony was defined if >50 cells were observed. Plates were imaged, and colonies were enumerated using ImageJ software.

For BrdU assay, cells were starved with 2% FBS-RPMI1-640 for 12 h and recovered in 10% FB-RPMI1640 for 24 h for cell synchronization. Cells were labeled with BrdU in culture medium for 1 h. All procedures followed the standard protocol with the allophycocyanin (APC) BrdU Flow Kit (cat. no. 552598; BD Biosciences). Cell cycle progression was determined by propidium iodide (PI) staining. Cells were fixed in 75% ethanol at −20°C overnight and incubated with PI/RNase Staining Buffer (cat. no. 550825; BD Biosciences) for 15-20 min at RT (protecting the cells from light). For apoptosis assay, cells were labeled by APC-conjugated Annexin V and PI/7-aminoactinomycin D (cat. no. A7313020; Shanghai Yeasen Biotechnology Co., Ltd.) according to the manufacturer's protocol. Stained cells were analyzed by flow cytometry (DxP Athena B4-R2; Cytek Biosciences) and analyzed with FlowJo 10.0 software (FlowJo LLC).

The Transwell assay was performed using the Transwell chambers (Corning, Inc.) with a filter membrane (aperture 8 μm). For the migration assay, the chemoattractant was 10% FBS. The invasion experiment required the addition of Matrigel to the upper chamber and incubation at 37°C for 2 h, followed by the same steps as the migration assay. The chemoattractant of the cell homing assay was CXCR12 (40 ng/ml). Cells were starved with 2% FBS-RPMI1640 for 12 h. Then 5×10⁵ cells/400 μl serum-free medium/well were inoculated into the upper chamber, and 600 μl/well medium with chemoattractant was added into the lower chamber. After incubation in 37°C for 24 h, the migratory cells in the lower chamber were harvested and counted through flow cytometry. After removing the cells in the upper chambers, chambers were stained with 0.2% crystal violet (CV) for 20 min. The filter membrane was cut and placed on the glass slide for counting using a Leica light microscope.

A co-culture system was used to detect the direct adhesion ability of MM cells. 96-well plates were incubated at 37°C with fibronectin (FN; 50 μg/ml; 100 μl/well; cat. no. 354008; Corning, Inc.) or seeded HS5 cells (2×10⁴ cells/well) overnight. LGR4-OE or LGR4-knockdown ARP1 and OCI-My5 were harvested and resuspended with serum-free RPMI-1640 medium, seeded with 1×10⁵ MM cell/100 μl into 96-well plate. To detect the adhesion ability of MM cells after being treated with QNZ for 48 h, MM cells were harvested and seeded into the 96-well pre-coated with FN or HS5 cells. After co-culture for 4 h, non-adherent MM cells were removed. Adherent cells were stained with 0.2% CV for 2 h at RT. Superfluous CV was washed off with distilled water, and the plates were dried overnight at RT. The dye was dissolved with 2% SDS for 2 h on the shaking platform, and the plates were measured at 570 nm using a microplate reader. The optical densities (ODs) from HS5s cultured alone were tested as background absorption.

All animal experiments were performed in accordance with the guidelines of the Institutional Animal Care and local Veterinary Office and Ethics Committee of the Animal Center of Hunan Normal University School of Medicine (approval no. D2021013; Changsha, China). The SPF housing conditions were maintained at 20-26°C, with relative humidity at 40-70% and a 12/12-h light/dark circadian rhythm. OCI-My5 cells (1×10⁶ cells in 200 μl PBS) were injected by tail vain intravenously into 8 weeks-old female NCG mice (n=22; weight, 19-24 g) (NOD/ShiLtJGptPrkdc^em26Cd52Il2rg^em26Cd22/Gpt, GemPharmatech). MM progression in the mice was monitored by measuring the tumor burden through live imaging. Homing assay of injected MM cells to the BM was measured by flow cytometry. Mice were euthanized using sodium pentobarbital at a dose of 100 mg/kg via intraperitoneal injection, and living imaging of leg bones was performed at week 6. Homing MM cells were flushed out of the BM with 1X PBS. After lysis of the erythrocyte lysate (ACK lysis Buffer; cat. no. SL1070; Coolaber science & technology Co., Ltd.), the cells were labeled with PE-anti-human CD138 (1:100; cat. no. 352306; BioLegend, Inc.). Flow cytometry was used to analyze the proportion of the homing cells to BM.

Micro-CT scanning was performed as previously described (22). The mouse tibia was fixed in 4% paraformaldehyde (PFA) in 4°C for 48 h. Tibia scans were performed by High-resolution Micro-CT (Skyscan 1176; Bruker, https://www.bruker.com/zh.html?ao=1) at a resolution of 6.5 μm per pixel. Measuring the bone parameters, including trabecular bone volume fraction (Tb. BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th) and trabecular separation (Tb.Sp). These bone parameters were analyzed using DataViewer 1.4.1.9 (CTAn version 1.11) and (μCTVol version 2.2; (both from Bruker Corporation).

The mouse tibia was fixed in 5 ml of 4% PFA solution at 4°C overnight and then decalcified the tibia in 10 ml of 0.5 M EDTA at 4°C for 24 h. Paraffin sections (6 μm) were stained with (TRAP; cat. no. 387a-1KT; MilliporeSigma). Images of TRAP were obtained through a light microscope (Keyence Corporation).

The experimental mice tibia was fixed with PFA in 4°C, embedded in paraffin following gradient ethanol dehydration, and sliced into 3 μm for IHC. The slides were dewaxed with xylene, rehydrated, and subjected to antigen retrieval treatment using an IHC kit (cat. no. KIT-9720; MXB Biotechnologies; http://www.maxim.com.cn/sitecn/myzhjcxthsjh/981.html). Subsequently, the slides were incubated with anti-CD138 antibody at a 1:400 dilution overnight at 4°C. Next, the slides were incubated in 25°C with HRP-conjugated antibody (Reagent 3 in IHC kit) and stained with 3,30-diaminobenzidine tetrahydrochloride hydrate (DAB) for 3 min. Finally, cell nuclei were counterstained with hematoxylin.

Total RNA was extracted from MM cells using TRIzol (cat. no. 15596018CN; Invitrogen; Thermo Fisher Scientific, Inc.), and its quality and quantity were assessed using a Fragment Analyzer (Agilent Technologies, Inc.). Library preparation was performed using Optimal Dual-mode mRNA Library Prep Kit (BGI; https://www.bgi.com/). The loading concentration of the final library is 300g/lane quantified by Qubit (Thermo Fisher Scientific, Inc.). Sequencing was performed on T7 platform (BGI) using paired-end 150-base reads are generated. MM cell lines, including OCI-EV, OCI-LGR4-OE, OCI-Ctrl and OCI-LGR4-shRNA1, were used for RNA-seq. Gene set enrichment analysis (GSEA) enrichment from differentially expressed genes between OCI-EV and OCI-LGR4-OE was performed using clusterProfiler_4.2.2 (24) function of R language, and the threshold was set as P<0.05.

Data were analyzed and represented using GraphPad Prism software (version 9; Dotmatics). All data are presented as the mean ± standard deviation (SD). The statistical significance of the data was determined using the two-tailed unpaired Student's t-test, one-way or two-way ANOVA with Tukey or Dunnett post-hoc test, or Kruskal-Wallis test. Overall survival was measured using the Kaplan-Meier method, and the log-rank test was used for group comparison based on GraphPad Prism 9 software. The statistical significance of the data in the Table SVI was determined through the Fisher-Freeman-Hanlton test. Each experiment was performed three times. The analysis of the Gene Expression Programming database utilized dataset GSE2658 and GSE24080, accessed from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). *P<0.05 was considered to indicate a statistically significant difference.

To identify the role of LGR4 in MM, the dataset GSE2658 (25) was analyzed, which contains data from 22 HD, 44 patients with MGUS and 351 newly diagnosed patients with MM. The analysis revealed a gradual increase in LGR4 expression (Fig. S1A). Higher expression of adhesion-associated molecules was observed in patients with MM compared with both HD and MGUS on the dataset GSE2658, same as LGR4 (Fig. 1A). These results were further validated using immunofluorescence in CD138⁺ cells derived from HD (n=5) and patients with MM (n=9) (Fig. S1B). The results confirmed a higher expression of LGR4 in MM, consistent with previous studies (16,17).

The focal proliferation of MM cells in the BM is a hallmark of MM, where increased adhesion enhances cell homing (9). To investigate whether LGR4 influences adhesion and homing, the association of LGR4 and adhesion-associated molecules was examined. Pearson's correlation heatmap analysis exhibited a positive correlation between LGR4 and adhesion-associated molecules (Fig. 1B). Subsequently, RNA-seq data of OCI-T3^rd-luc cells revealed significant upregulation of LGR4 and adhesion genes compared with OCI-My5-luc cells (Fig. 1C). Additionally, the high expression of LGR4 along with adhesion genes was verified using RNA-seq data from paired KMS28-BM and KMS28-PE cell lines (Fig. 1D). Among the positive correlated gene, it was found that ZEB1, a transcription factor associated cell adhesion (26), is associated with worse overall survival; as MM patients with LGR4^highZEB1^high (n=85) exhibited significantly worse overall survival (P=0.068) (Figs. 1F and S1F). Furthermore, ZEB1 expression was significantly increased in KMS28-BM compared with KMS28-PE cells, at both mRNA (Fig. S1D) and protein levels (Figs. 1E and S1C-E). These results suggested that LGR4 is associated with cell adhesion and promoted MM cell homing to the BM. In summary, high expression of LGR4 is associated with increased cell adhesion and is associated with poor prognosis in MM.

To explore whether high expression of LGR4 promotes MM progression, LGR4 was the BrdU incorporation assay (Figs. 2B and S2C) and a marked increase in colony formation in the soft-agar colony formation assay (Fig. 2C) compared with the EV group. Additionally, cell cycle assays indicated that LGR4 overexpression increased the percentages of both the S phase and G2/M phase in ARP1 and OCI-My5 cells (Fig. S2D). The aforementioned results confirmed that LGR4 overexpression promotes the proliferation of MM cells.

Next, the effect of LGR4 was examined on the migration, invasion and homing abilities of MM cells. A Transwell assay using different chemo-attractants was performed to assess cell migration (Fig. 2D). The results indicated that LGR4 overexpression significantly promoted cell migration and invasion, which was quantified by counting migratory cells (Fig. 2E and F). Additionally, a cell homing assay using CXCR12, a chemokine known to induce immune cell homing to BM (27), demonstrated that more cells homed to the BM in the LGR4-OE group (Fig. 2G). As previously reported, FN acts as a connection between cells and matrix (26), and the HS5 cell line mimics the bone marrow stromal cells (BMSCs), promoting MM cell proliferation and adhesion (28). To determine whether LGR4 improves the interaction between MM cells and the BMME, a cell adhesion co-culture assay was performed using FN and HS5 cells (29)(Fig. 2H). The OD value at 570 nm indicated a direct increase in adhesion to both FN and HS5 in the LGR4-OE group (Fig. 2I). Even though in ARP1-LGR4-OE cells was observed increase tend of adhesion (P=0.06), which may cause by the complex genetic characteristics such as TP53^del (30), the aforementioned results demonstrated that LGR4-OE enhances the cell homing and adhesion ability in MM cells.

To further investigate the function of LGR4 on adhesion, migration and homing of MM cells, two shRNA sequences (shRNA1 and shRNA2) targeting LGR4 were designed. A DOX-inducible lentiviral expression system expressing LGR4 shRNA was used to knock down LGR4 in MM cell lines. LGR4 knockdown was confirmed at both mRNA and protein levels (Figs. 3A, S3A and C). Among these, LGR4-shRNA1 silencing was confirmed to be more effective. Growth curves indicated that LGR4 knockdown significantly inhibited proliferation in MM cells following DOX induction (Fig. S3B). Additionally, the proportion of BrdU-positive cells in the LGR4-shRNA1 groups was significantly lower than in the control group (Fig. 3B and C). The colony formation of LGR4-shRNA1 cells exhibited a significant inhibition (Fig. 3D). Cell cycle assays revealed that LGR4 silencing decreased the proportion of cells in the S and G2/M phases in both ARP1 and OCI-My5 (Fig. S3D). Furthermore, the proportion of Annexin-V-positive cells was significantly increased (Fig. S3E), and cleaved caspase 3 and PARP were significantly upregulated in the LGR4-knockdown group (Fig. S3F and G), indicating that LGR4-knockdown induced apoptosis in MM cells. These results suggested that LGR4-knockdown inhibited the proliferation and induced apoptosis in MM cells.

Next, Transwell migration and invasion assays revealed that LGR4-knockdown reduced both migration and invasion, with a corresponding decrease in the number of migratory cells (Fig. 3E and F). Additionally, LGR4-knockdown decreased the homing of MM cells induced by the BM chemokine CXCR12, as confirmed by statistical analysis (Fig. 3G). Using cell-adhesion co-cultured assay, the absorbance at 570 nm exhibited that LGR4-knockdown suppressed the adhesion ability of MM cells to FN and BMSCs (Fig. 3H and I). The aforementioned results confirmed that LGR4-mediated interaction between malignant plasma cells and BMME is crucial for cell adhesion and homing to BM niches.

To further explore the role of LGR4 in MM cell homing in vivo, OCI-My5 cells with LGR4-OE were generated and injected through the tail vein into NCG mice. MM cells are typically home to BM, where they proliferate and cause symptoms, including hindlimb paralysis (31). Tumor burden was monitored through whole-animal live imaging, evaluating the proportion of human MM cells in BM and their homing efficiency (Fig. 4A). Compared with the control mice, the LGR4-OE mice significantly demonstrated an increased tumor-associated luminescence intensity at weeks 4 and 6 (Fig. 4B and C). LGR4-OE mice exhibited 60% of paralysis, while control mice had no expression at week 6 (Fig. S4). Due to reaching the humane endpoint, the mice were euthanized at week 6. Flow cytometric analysis revealed that LGR4 overexpression significantly increased the proportion of homing MM cells in the BM. As a result, LGR4-OE significantly increased the proportion of human MM cells (AVG 55%) in the BM compared with the control mice (AVG 23.8%) (Fig. 4D).

Since MM cell homing and proliferation within the BM are key drivers of bone disease (21), bone health was further assessed by evaluating bone fractures and CD138 immunostaining. Micro-CT scanning was used to detect bone damage, finding that LGR4-OE had markedly more severe trabecular bone loss compared with the control mice (Fig. 4E). Quantitative bone microstructure parameters exhibited that LGR4-OE in mice had a lower BV/TV and Tb.N, while Tb.Sp was markedly higher compared with the controls (Fig. 4F). Consistent with these results, TRAP staining indicated the increased positive osteoclast number in the LGR4-OE mice femora compared with control mice (Fig. 4G and I). Moreover, plasma cell morphology was clearly visible at ×80 magnification and exhibited enrichment in LGR4-OE mouse bones (Fig. 4G). CD138 immunostaining revealed the increased plasma cell number in the LGR4-OE mice compared with the control (Fig. 4H and J). Additionally, the potential association between LGR4 expression and clinical characteristics was assessed, using the GSE24080 dataset. It was found that LGR4 expression correlated strongly with the percentage of plasma cells in the BM (P=0.037), and the number of magnetic resonance imaging (MRI)-defined focal lesions (P=0.019) (Table SVI). These findings suggest that high expression of LGR4 is linked to myeloma cell homing, promotes bone destruction and contributes to malignant progression in patients with MM. In summary, these results demonstrated that overexpression of LGR4 enhances MM cell homing to the BM and accelerates disease progression.

To further understand the signaling pathways regulated by LGR4 in MM, RNA-seq was performed on OCI-LGR4-OE and OCI-LGR4-shRNA1, along with the control cells. Moreover, GSEA exhibited major types of gene signatures in LGR4-OE cells that were enriched in the regulation of cell migration and cell adhesion (Figs. 5A-C, S5A and B). Then, the changes in the expression of key cell-adhesion genes were verified at both Mrna and protein levels, including N-Cadherin, Snail, Vimentin, MUC2, ZEB1, TNFRSF1B in LGR4-OE, LGR4-shRNA1 in ARP1 and OCI-My5 cells (Figs. 5D and E and S5C). The results indicated that LGR4-OE significantly increased the cell-adhesion molecules while LGR4-knockdown resulted in a significant decrease (Figs. 5F and G and S5D). A recent study has suggested that LGR4 regulates intestinal epithelial cell proliferation and development through C-terminal activation of NF-κB signaling (32). R-spondin signals drive NF-kB activity through LGR4 and stimulate the proliferation of stem cells (33). It has been reported that activated NF-κB signaling enhances the ability of hematopoietic stem cell homing (9). Besides, NF-κB signaling can activate endothelial cell adhesion molecules (34). Therefore, it was hypothesized that LGR4 promotes MM cell homing by activating NF-κB signaling. To investigate whether LGR4 influences MM cell adhesion through NF-κB activation, the protein level of key NF-κB genes was determined, including p65, phosphorylated (p-)p65, IκBα and p-IκBα. The results indicated that LGR4-OE significantly activated NF-κB signaling, while LGR4-knockdown decreased as western blotting illustrated (Figs. 5H and I, S5E and F). The aforementioned data indicate that LGR4 promotes cell homing to BM through activating NF-κB signaling.

It was further investigated whether QNZ, an NF-κB signaling pathway inhibitor, could suppress cell proliferation and homing in MM. The IC₅₀ of QNZ was significantly lower in LGR4-OE ARP1 and OCI-My5 cells compared with the control group (Fig. 6A). Growth curves indicated that the proliferation was inhibited in the LGR4-OE cells treated with QNZ (Fig. S6C and E). Cell cycle assays exhibited that QNZ alleviated the increased proportion of S-phase cells caused by the overexpression of LGR4 in ARP1 and OCI-My5 cells (Fig. 6B). Subsequently, to investigate whether QNZ inhibits LGR4-induced cell adhesion and homing, Transwell assay of cell homing was performed. The results revealed that cell homing ability induced by LGR4-OE was inhibited after 48 h of QNZ treatment (Figs. 6C and D). Additionally, the Transwell migration and Matrigel invasion assays confirmed that QNZ reduced cell migration and invasion of cells, with the quantification of migratory cells supporting these results (Figs. 6F and G, and S6A and B). Consistently, the results of the cell co-culture adhesion assay exhibited that MM cell adhesion ability was reduced after 48 h of QNZ treatment in LGR4-OE ARP1 and OCI-My5 cells (Fig. 6H and I). Furthermore, western blot analysis revealed that QNZ treatment significantly inhibited NF-κB signaling in LGR4-OE cells, with a corresponding decrease in the expression of adhesion-associated molecules ZEB1 (Figs. 6J, S6D and F). As a critical function subunit of NF-κB signal, RELA (p65) contains transcriptional activation domains of gene transcription and facilitates the binding of p50 to DNA (35,36). Subsequently, to verify interaction molecules of NF-κB signaling activation by LGR4, siRNA was used to knock down RELA in LGR4-OE MM cells. Protein levels (Fig. S7A and B) confirmed a significant knockdown of RELA in LGR4-OE ARP1 and OCI-My5 cells, compared with the same cell lines transduced with SiNC serving as controls. The results indicated that siRNA1 and siRNA3 were more pronounced. Knockdown of RELA reversed the LGR4-induced proliferation and homing effects, as demonstrated by growth curve analysis (Fig. S7C), cell cycle assays (Fig. S7D), Transwell analysis about cell homing (Fig. S7E and F) and migration (Fig. S7G and H). By activating the NF-κB pathway, LGR4 facilitates the entry of the p50/RELA dimer into the nucleus to initiate gene transcription. The aforementioned results confirmed that inhibition of the NF-κB pathway alleviated the promoting effect of LGR4-OE on adhesion ability and MM cell homing to BM.

To explore whether QNZ has similar effects in vivo as observed in vitro, 1×10⁶ OCI-ctrl and OCI-LGR4-OE cells were injected through the tail vein into NCG mice to establish a xenografts mouse model. A total of 10 days post-transplantation, QNZ was injected intraperitoneally every two days (Fig. 7A). As compared with the solvent-treated mice (Fig. 7B), the mice treated with QNZ exhibited a reduction in tumor-associated luminescence intensity in the LGR4-OE group at weeks 3 and 6 (Fig. 7C). Due to reaching the humane endpoint at week 6, one mouse from each in the solvent group was observed to exhibit paralysis, then succumbed unexpectedly the following day, therefore the mice were euthanized. To verify whether QNZ affected cell homing, femur and tibia were dissected for image of bone tissue. The fluorescence intensity in the bone of the LGR4-OE group was higher, whereas QNZ treatment significantly reduced this intensity, particularly in the LGR4-OE group (Fig. 7D and E). Flow cytometric analysis confirmed that QNZ treatment significantly decreased the proportion of human MM cells in the BM of LGR4-OE mice compared with the solvent group (Fig. 7F and G).

Next, micro-CT scanning was used to detect bone damage in the tibia, revealing that QNZ treatment rescued severe trabecular bone loss caused by LGR4-OE, compared with the solvent-treated mice (Figs. 7J). Quantitative analysis of bone microstructure parameters exhibited that both Tb.BV/TV and Tb.N were improved (Fig. 7H). TRAP staining revealed that the number of osteoclasts was reduced in QNZ-treated LGR4-OE mice in contrast to solvent mice (Fig. 7K). Quantitative analysis of TRAP-positive osteoclast cells confirmed this reduction (Fig. 7I). In summary, these results demonstrated that inhibition of the NF-κB pathway relieves the effects of LGR4 overexpression on MM cell proliferation and cell homing in vivo.