O-GlcNAcylation homeostasis controlled by calcium influx channels regulates multiple myeloma dissemination

Human MM-derived cell lines RPMI8226 and National Cancer Institute (NCI)-H929 were obtained from the American Type Culture Collection (Manassas, VA, USA). Cell lines were cultured in RPMI1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Invitrogen, Waltham, MA, USA) at 37 °C and 5% CO₂. Mycoplasma contamination was regularly checked using a commercial test kit (MycoAlert™ PLUS, Lonza, Cologne, Germany) and was found to be absent in all cell lines tested.

mRNA expression analysis of clinical samples was conducted using Oncomine™ public database (https://​www.​oncomine.​org/​resource/​main.​html). Analyzed datasets included Zhan myeloma [24], Zhan myeloma 3 [25], and Agnelli myeloma 3 [26].

Intracellular Ca²⁺ measurements were performed using Fura-2 AM (Abcam, Cambridge, UK) as described previously with minor modifications [21]. After specific treatments, 5 × 10⁴ cells were suspended in HBSS buffer containing 4 μM Fura-2 AM and loaded onto a 96-well black plate for 1 h at 37 °C. Following an internal Ca²⁺ depletion by 2 μM thapsigargin for 10 min, 2 μM extracellular Ca²⁺ was added and Fura-2 AM signals were measured using a fluorescence microplate reader (Synergy H1, BioTek, Winooski, VT, USA) at 340/510 or 380/510 nm.

All-in-one pLentiCRISPR v2 plasmids carrying a single guide RNA (sgRNA) against specific target gene, spCas9 and puromycin resistance were obtained from GenScript (Piscataway, NJ, USA), while LentiCas9-blasticidin and TRPM7 sgRNA plasmids were a kind gift from Profs. Zhang, Doench, and Root (Addgene #52962 and #76111). The oligos sequences of all sgRNAs were listed in Additional file 1: Supplementary Table S1. Lentiviral particles were generated in HEK293T cells using pCMV.dR8.2 dvpr packaging and pCMV-VSV-G envelope vectors (Addgene #8454 and 8455). Cells were transduced with the virus in the presence of hexadimethrine bromide (8 μg/mL), selected with puromycin (8 μg/mL) or blasticindin (10 μg/mL), and assessed for gene knockdown efficiency by Western blotting.

Cells were transfected with ITGA4 or ITGB7 plasmid (Genscript) using Lipofectamine 3000 Transfection Kit (Life Technologies, Carlsbad, CA). The transfected cells were allowed to recover for 48 h and protein level was determined by Western blotting before each experiment.

Cell migration was assessed in a 24-well plate Transwell system (Corning, Kennebunk, ME, USA) using a cell culture insert with a pore size of 5.0 μm, while cell invasion was assessed using an insert coated with 0.5 mg/mL Matrigel (BD Bioscience, San Jose, CA, USA) [27, 28]. After starvation for 4 h, cells (1 × 10⁵ cells/well) were loaded onto the insert chamber containing low-serum (1% FBS) medium and complete medium was added to the lower chamber as a chemoattractant. Migrating/invading cells in the lower chamber were collected at 24–48 h, stained with Hoechst 33342, and visualized under an inverted fluorescence microscope (Eclipse Ti-U, Nikon, Tokyo, Japan).

All animal studies were performed in accordance with the protocol approved by Institutional Animal Care and Use Committee of West Virginia University (WVU) (#1702005551). Cells were labeled with UBC-RFP-T2A-Luciferase dual reporter for live cell tracking. Male NSG (NOD.Cg-Prkdc^scid. Il2rg^tm1Wjl/SzJ) mice (WVU Transgenic Animal Core Facility, Morgantown, WV, USA) at 6–8 weeks of age were intravenously injected with 1.5 × 10⁷ luciferase-labeled cells and tumor burden was monitored on an IVIS Lumina II in Vivo Imaging system (PerkinElmer, Waltham, MA, USA). Mice were euthanized 4 weeks after the infusion or as recommended by a veterinarian. Organs were collected and imaged ex vivo to evaluate a disseminated lesion. After which, isolated organs were formalin fixed, paraffin embedded, and cut into 5-μm sections. Engraftment was confirmed by immunohistochemistry (IHC) using CD138 antibody (Invitrogen, #36–2900).

Protein was extracted using commercial protein lysis buffer (Cell Signaling Technology) supplemented with a protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany). A total protein of 30–50 μg was subjected to SDS-PAGE and transferred to PVDF membranes. Membranes were blocked with 5% fat-free milk and incubated with indicated primary antibodies overnight at 4 °C and with secondary antibodies conjugated with HRP for 2 h at room temperature. The immunoreactive proteins were captured and analyzed with ECL reagent (Millipore, Billerica, MA) using a digital imaging system (ImageQuant LAS, GE Healthcare, Pittsburgh, PA). For IP, cell lysates (100–200 μg) were incubated with anti-integrin α4 and or anti-integrin β7 antibody at 4 °C overnight, followed by incubation on ice with agarose beads for another 2 h to precipitated the protein-antibody complex. After extensive washing with ice-cold IP lysis buffer, samples were resuspended in protein loading buffer, boiled at 95 °C for 5 min, and subjected to immunoblotting.

Data are presented as mean ± SD from three or more independent experiments. Unpaired, two-tailed Student t-test or one-way ANOVA with Tukey’s multiple comparison tests were used to determine statistical significance at the significance level of P < 0.05 (GraphPad Prism, San Diego, CA, USA).

mRNA expression analysis of critical Ca²⁺ influx channels was first performed in human clinical specimens available on Oncomine™ bioinformatics database. We observed that TRPM7, ORAI1, and STIM1 expression was significantly higher in smoldering MM (SMM), a precancerous form of MM, when compared to NPCs (Fig. 1A–C). An upregulation of STIM1 expression in symptomatic MM was observed in two other datasets (Fig. 1C), while information on TRPM7 and ORAI1 expression in symptomatic MM was not available. These data suggest that aberrant Ca²⁺ influx channels may be involved in MM pathogenesis and progression and support the previous study demonstrating a close relation between elevated SOC influx channels and clinical outcome of MM patients [29].

To investigate the potential role of TRPM7, ORAI1, and STIM1 in MM cell motility, we first used various small-molecule inhibitors (SMIs) of the channels, including 2-APB, AnCoA4, and SKF96365, as schematically depicted in Additional file 2: Fig. S1. To ensure that their effect on cell motility was not due to their adverse effects on cell viability and proliferation, we used sub-cytotoxic concentrations of the SMIs, i.e. up to 50 μM for 2-APB and up to 40 μM for AnCoA4 and SKF96365 in MM RPMI8226 and NCI-H929 cells (see Additional file 2: Fig. S2 for MTT assay and Fig. S3 for cell cycle analysis and Annexin V-PE/7-AAD apoptosis assay) in our study. Free intracellular Ca²⁺ in response to 2-APB treatment was evaluated using Fura-2 AM as a fluorescent probe. Figure 1D and E shows that 2-APB, AnCoA4, and SKF96365 dose-dependently inhibited Ca²⁺ influx in both RPMI8226 and NCI-H929 cells, indicating that Ca²⁺ influx channels are functional in these MM cells and that they are responsive to the SMI treatments. Phenotypically, we observed that the cells treated with 2-APB, AnCoA4, and SKF96365 were less migratory and invasive by 30–95% depending on the dose, as evaluated by Transwell assays (Fig. 1F and G; Additional file 2: Fig. S4). These data suggest that Ca²⁺ influx channels, particularly TRPM7, ORAI1, and STIM1, are critical for MM cell migration and invasion.

To specifically determine the role of TRPM7, ORAI1, and STIM1 influx channels in MM cell motility, CRISPR/Cas9 system was used to repress the specific gene expression in MM cells. The cells were transduced with lentiviral particles comprising sgRNA against TRPM7 (sgTRPM7), ORAI1 (sgORAI1), and STIM1 (sgSTIM1) and Cas9, and their effect on protein expression was determined by Western blotting (Fig. 2A–C; Additional file 2: Fig. S5). Figure 2D–F shows that sgTRPM7, sgORAI1, and sgSTIM1-transduced cells exhibited significantly lower migratory and invasive activities than vector control (wild-type, WT) cells. We also verified that the sgTRPM7, sgORAI1 and sgSTIM1 cells exhibited a similar proliferative activity, cell cycle profile, and apoptosis rate as their respective WT control cells (Additional file 2: Fig. S6), thus ruling out their anti-proliferative effect on the observed cell motility. Together, the results from our CRISPR/Cas9 gene depletion and SMI studies strongly indicate that TRPM7, ORAI1, and STIM1 are key regulators of MM cell motility.

Recent studies have demonstrated a plausible crosstalk between Ca²⁺ influx and cellular O-GlcNAcylation in excitable cells [30]. To test this possibility in MM cells, cellular O-GlcNAc level was manipulated in MM RPMI8226 cells and its effect on Ca²⁺ influx channels was determined by Western blotting, and vice versa. Additional file 2: Fig. S7 shows that the levels of TRPM7, ORAI1, and STIM1 were relatively unchanged in cells treated with OGA inhibitor PugNAc (10–50 μM) or thiamet G (2.5–10 μM), which caused an increase in O-GlcNAc level. In contrast, cells treated with Ca²⁺ influx channel inhibitors, including 2-APB (10–50 μM), AnCoA4 (10–40 μM), and SKF96365 (10–40 μM), exhibited a dose-dependent increase in cellular O-GlcNAc level as compared to non-treated control (Fig. 2G and H). The increase in cellular O-GlcNAc level upon inhibition of TRPM7, ORAI1, and STIM1 was validated in sgTRPM7, sgORAI1, and sgSTIM1 cells, which displayed elevated O-GlcNAc levels as compared to WT cells (Fig. 2I–K). These data indicate that O-GlcNAcylation is regulated by Ca²⁺ influx via TRPM7, ORAI1, and STIM1 channels and that O-GlcNAcylation may regulate MM cell motility.

Although O-GlcNAcylation has been associated with metastasis of various solid tumors [31], its role in MM cell migration and invasion has not been demonstrated. In this study, we used OGT and OGA inhibitors, as schematically depicted in Fig. 3A, to evaluate the dose-response effect of O-GlcNAcylation on cell motility. To ascertain the clinical significance of O-GlcNAcylation in MM, we also analyzed OGT and MGEA5 (encoding OGA) expression in MM patients using the Oncomine™ database. We found that MGEA5 expression was significantly higher in MM patients as compared to NPCs, whereas the expression of OGT was not significantly different in the two groups (Fig. 3B). These results suggest a reduced O-GlcNAcylation in MM. Treatment of MM cells with PugNAc (10–50 μM) and thiamet G (2.5–10 μM) decreased cell migration and invasion in a dose-dependent manner as compared to non-treated control, in an inverse relationship with the O-GlcNAc level (Fig. 3C and D; Additional file 2: Fig. S8 and S9). Consistently, genetic inhibition of MGEA5 by CRISPR/Cas9 (sgMGEA5), which caused an increase in O-GlcNAcylation, reduced the migration and invasion of MM cells (Fig. 3E and F; Additional file 2: Fig. S10). It should be noted that inhibition of OGA by both SMIs and CRISPR/Cas9 did not affect cell proliferation, cell cycle or apoptosis (Additional file 2: Fig. S11 and S12), indicating that the observed inhibitory effects were indeed due to impaired cell motility. Similar results were obtained in human MM-derived NCI-H929 cells (Additional file 2: Fig. S13). These findings are in good agreement with the TRPM7, ORAI1, and STIM1 inhibition data showing their effect on hyper-O-GlcNAcylation, and with the clinical data linking hypo-O-GlcNAcylation with aggressive MM.

Based on the findings that inhibition of TRPM7, ORAI1, and STIM1 elevated cellular O-GlcNAcylation level, and that inhibition of such channels and O-GlcNAcylation reduced MM cell migration and invasion, we postulated that TRPM7, ORAI1, and STIM1 may regulate MM cell motility through O-GlcNAcylation. Hence, hypo-O-GlcNAcylation should be able to rescue the effects on cell motility caused by inhibition of the channels. To test this possibility, sgTRPM7, sgORAI1, and sgSTIM1 cells were treated with OGT inhibitor OSMI-1 (3 μM), which reduced O-GlcNAcylation (Fig. 3G and H), and its effects on cell migration and invasion were determined by Transwell assays. Figure 3I–K shows that OSMI-1 rescued the inhibitory effects in sgTRPM7, sgORAI1, and sgSTIM1 cells, substantiating that TRPM7, ORAI1, and STIM1 regulate MM cell motility through O-GlcNAcylation (Additional file 2: Fig. S14). These results demonstrated a novel role of Ca²⁺ influx/O-GlcNAc via TRPM7, ORAI1, and STIM1 channels in regulating MM cell motility.

Having demonstrated that Ca²⁺ influx acts upstream of O-GlcNAcylation in regulating MM cell motility, we experimentally verified the involvement of TRPM7, which showed the most pronounced effect on cell motility, and O-GlcNAcylation in MM dissemination in a xenograft mouse model [32, 33]. sgTRPM7, sgMGEA5 and WT cells were genetically labeled with luciferase and intravenously injected into NSG mice via tail vein to induce systemic disease, followed by a weekly bioluminescence imaging for quantitative analysis of tumor progression. Bioluminescence analyses revealed that both sgTRPM7 and sgMGEA5 cells yielded substantially lower whole-body signals at 1, 2, 3 and 4 weeks post-injection comparing to WT cells (Fig. 4A, B, F, and G), suggesting less tumor burden and experimental dissemination of MM cells. For better comparison, the luminescence signals in each group were normalized to their initial signals at week 0 and relative to WT control. Figure 4C and H confirmed a lower myeloma burden in mice bearing sgTRPM7 and sgMGEA5 cells at week 1 to 4 post-injection. To support the finding of MM dissemination, mice were euthanatized at the end of the experiments and various organs, including brain, lung, kidney, bone, heart, spleen, and liver, were collected for ex vivo bioluminescence imaging and histochemistry. Figure 4D and I demonstrated substantially lower tumor signals in the various organs in mice bearing sgTRPM7 and sgMGEA5 cells as compared to those in WT mice. These results are consistent with the IHC finding showing a decreased number of infiltrated MM cells in the various organs, including bone (femur) and liver, and the flow cytometric analysis of human CD138+ cells in the BM of femur bone (Fig. 4E and J; see also Additional file 2: Fig. S15–18). Together, these results support the in vitro findings that Ca²⁺ influx and hypo-O-GlcNAcylation are critical regulators of MM cell motility and dissemination.

O-GlcNAcylation has been long known to modify proteins and alter their functions and cellular pathways [34]. To gain a more in-depth understanding of the underlying mechanisms, we used a literature mining approach to identify cell adhesion molecules that are involved in MM progression, including the integrin family of proteins such as integrin β1 (ITGB1), β3 (ITGB3), β7 (ITGB7), α4 (ITGA4), α5 (ITGA5) and αV (ITGAV), E-cadherin (E-cad), N-cadherin (N-cad), and β-catenin (B-cat) [35‐38]. The expression level of these proteins in response to thiamet G (2.5–10 μM) treatment was examined in MM cells by Western blotting. Figure 5A and B shows that thiamet G minimally affected ITGB1, ITGAV, ITGA5, N-cad, and B-cat expression, but significantly downregulated ITGB3, ITGA4, ITGB7, and E-cad in a dose-dependent manner (see also Additional file 2: Fig. S19A). Bioinformatics analysis of the adhesion molecules in MM patients using Oncomine™ database revealed that ITGA4 and ITGB7 were significantly upregulated in MM tissues as compared to NPCs, while ITGB3 and CDH1 expression was not different (Fig. 5C; see also Additional file 2: Fig. S19B), suggesting the role of ITGA4 and ITGB7 in MM cell adhesion. Genetic inhibition of ITGA4 and ITGB7 by CRISPR/Cas9 in MM cells, designated as sgITGA4 and sgITGB7 cells (Fig. 6A and B), also showed a substantial reduction in MM cell migration and invasion when compared to WT control (Fig. 6C and D), while having minimal effect on cell proliferation, cell cycle and apoptosis (Additional file 2: Fig. S20). The change in ITGA4 and ITGB7 expression following O-GlcNAcylation was verified in sgMGEA5 cells (Fig. 5 D and E), thus confirming that ITGA4 and ITGB7 are key regulators of MM cell motility under O-GlcNAcylation.

To further validate that ITGA4 and ITGB7 are downstream of Ca²⁺ influx under the Ca²⁺ influx/O-GlcNAc axis, rescue experiments were conducted in which ITGA4 and ITGB7 plasmid were ectopically overexpressed in various Ca²⁺ influx-inhibited cells, including sgTRPM7, sgORAI1, and sgSTIM1 RPMI8226 cells, and their effects on cell motility and protein level were determined. Figure 7A–C showed the reduced ITGA4 and ITGB7 levels in sgTRPM7, sgORAI1, and sgSTIM1 cells, indicating that Ca²⁺ influx via TRPM7, ORAI1, and STIM1 channels regulated ITGA4 and ITGB7 abundance. Overexpression of ITGA4 and ITGB7 did restore the reduced migration and invasion in sgTRPM7, sgORAI1, and sgSTIM1 cells (Fig. 7D–F), in agreement with the increased ITGA4 and ITGB7 levels, strengthening that the ITGA4 and ITGB7 levels and their functional contributions were indeed regulated by the Ca²⁺ influx/O-GlcNAc axis.

Notably, we also monitored the pan matrix metalloproteinase (MMP) activity using MMP Activity Fluorometric Assay Kit (Abcam) in sgTRPM7, sgORAI1, sgSTIM1 cells as well as in sgMGEA5 cells and WT control cells since MMPs are known to be involved in cancer cell motility of various cancers and have been reported to promote MM dissemination [39, 40]. However, our results showed no significant difference in total MMP activity in sgTRPM7, sgORAI1, sgSTIM1, and sgMGEA5 cells, indicating that MMPs may not play a significant role in the MM dissemination under the experimental Ca²⁺ influx/O-GlcNAc conditions (Additional file 2: Fig. S21).

Proteasomal degradation is known to play a major role in cellular proteostasis via the regulation of protein stability and degradation [41]. We tested whether O-GlcNAcylation regulates ITGA4 and ITGB7 stability by proteasomal degradation by first performing a rescue experiment using an established proteasome inhibitor, MG-132 (50 μM). Figure 8A shows that ITGA4 and ITGB7 downregulation by thiamet G was rescued by MG-132 treatment, indicating that O-GlcNAcylation regulates ITGA4 and ITGB7 via a proteasome-dependent pathway. Since O-GlcNAcylation and ubiquitination are PTMs that similarly occur at Ser/Thr residues [42], we next determined the interplay between the two on ITGA4 and ITGB7 proteins using co-immunoprecipitation. Figure 8B reveals physical interaction between ITGA4 and ITGB7 and the association between O-GlcNAcylation and ubiquitination of the two proteins. Elevated ITGA4 and ITGB7 O-GlcNAcylation by thiamet G caused an increase in ITGA4 and ITGB7 ubiquitination. These results indicate that O-GlcNAcylation of ITGA4 and ITGB7 promotes their ubiquitin-proteasomal degradation, leading to reduced protein expression and attenuated MM cell motility.