Targeting Caveolin-1 in Multiple Myeloma Cells Enhances Chemotherapy and Natural Killer Cell-Mediated Immunotherapy

Abstract

The cell membrane transport capacity and surface targets of multiple myeloma (MM) cells heavily influence chemotherapy and immunotherapy. Here, it is found that caveolin-1 (CAV1), a primary component of membrane lipid rafts and caveolae, is highly expressed in MM cells and is associated with MM progression and drug resistance. CAV1 knockdown decreases MM cell adhesion to stromal cells and attenuates cell adhesion-mediated drug resistance to bortezomib. CAV1 inhibition in MM cells enhances natural killer cell-mediated cytotoxicity through increasing CXCL10, SLAMF7, and CD112. CAV1 suppression reduces mitochondrial membrane potential, increases reactive oxygen species, and inhibits autophagosome-lysosome fusion, resulting in the disruption of redox homeostasis. Additionally, CAV1 knockdown enhances glutamine addiction by increasing ASCT2 and LAT1 and dysregulates glutathione metabolism. As a result of CAV1 inhibition, MM cells are more sensitive to starvation, glutamine depletion, and glutamine transporter inhibition, and grow more slowly in vivo in a mouse model treated with bortezomib. The observation that CAV1 inhibition modulated by 6-mercaptopurine, daidzin, and statins enhances the efficacy of bortezomib in vitro and in vivo highlights the translational significance of these FDA-approved drugs in improving MM outcomes. These data demonstrate that CAV1 serves as a potent therapeutic target for enhancing chemotherapy and immunotherapy for MM.

Keywords: caveolin‐1; glutathione metabolism; immunotherapy; multiple myeloma; natural killer cell; redox homeostasis.

Figure 1

CAV1 knockdown attenuates cell adhesion to BMSCs and adhesion‐mediated drug resistance to bortezomib. A) CAV1 expression (GSE2113) in CD138⁺ plasma cells from patients with monoclonal gammopathy of undetermined significance (MGUS, n = 7), multiple myeloma (MM, n = 39), and plasma cell leukemia (PCL, n = 6). The vertical axis represents the normalized signal expression value. B) Overall survival according to high and low CAV1 expression in plasma cells was determined in MM patients (GSE4581) using the Log‐rank (Mantel‐Cox) test (optimal cut‐off points). C) mRNA expression of CAV1 in MM cells transduced with lentivirus expressing shRNAs against NC (LV‐shNC) or CAV1 (LV‐shCAV1) was determined using qRT‐PCR. D and E) Protein expression of CAV1 in MM cells expressing shNC or shCAV1 was detected using a western blot. GAPDH was included as a loading control. F) Volcano plot showing the differential mRNA expression between RPMI 8226 expressing shNC and shCAV1. Red dots represent up‐regulated genes and blue dots represent down‐regulated genes. G) Bubble plots showing the top five GO functional enrichment of genes significantly regulated by CAV1 in RPMI 8226 cells. H) Bubble plots showing the top 15 KEGG pathways of genes significantly regulated by CAV1 in RPMI 8226 cells. I) The surface level of CD325 (CDH2) in MM cells expressing shNC or shCAV1 was determined using flow cytometry. J and K) Adhesion assays of (J) RPMI 8226 or (K) LP‐1 cells expressing shNC or shCAV1 in co‐culture with BMSCs derived from MM patients. L) RPMI 8226 or (M) LP‐1 cells expressing shNC or shCAV1 were cultured with BMSCs derived from MM patients in the presence of bortezomib (Btz) at the indicated concentration for 48 h and the cell viability was measured. p‐values were calculated with an unpaired two‐sided Student t‐test. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01; ^***, p < 0.001; ^****, p < 0.0001.

Figure 2

CAV1 regulates the expression of SLAMF7, Nectin‐2, and CXCL10, and modulates NK cell‐mediated cytotoxicity. A) A circular heatmap representing the mRNA expression of NK‐related inhibitory or stimulatory molecules in MM cells expressing shNC or shCAV1. B and C) The surface level of (B) SLAMF7 and (C) Nectin‐2 (CD112) in MM cells expressing shNC or shCAV1 was determined using flow cytometry. D and E) NK‐92 cells were co‐cultured with MM cells expressing shNC or shCAV1 at the indicated ratio for 12 h, and the apoptotic (Annexin V⁺) cells were determined using flow cytometry. F and G) Human cytokine array of conditioned medium obtained from RPMI 8226 cells expressing shNC or shCAV1. H) CXCL10 expression (GSE5900) in CD138⁺ plasma cells from HD (n = 22), MGUS (n = 44), and smoldering MM (SMM, n = 12). I) Overall survival according to high and low CXCL10 expression on MM cells was determined in MM patients (GSE9782). J and K) NK92 cells were co‐cultured with RPMI 8226 cells in the presence of CXCL10 at the indicated concentrations for 12 h. The apoptotic (Annexin V⁺) cells were determined using flow cytometry. p‐values were calculated with unpaired two‐sided Student t‐test or one‐way ANOVA test. Apop, apoptotic. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01; ^***, p < 0.001; ^****, p < 0.0001.

Figure 3

Targeting CAV1 enhances NK cell‐mediated cytotoxicity and improves daratumumab efficiency. A and B) In vitro BMMCs (iBMMCs, CD45⁺) were co‐culture with RPMI 8226 (CD45⁻) cells expressing shNC and shCAV1 at the ratio of 4:1 for six hours and the apoptosis of MM cells was determined using flow cytometry. C) viSNE maps colored by the normalized expression of the indicated markers in iBMMCs. D) viSNE map colored by eight cell populations after clustering. R‐cells, rest of cells. E) A heatmap showing the normalized median expression of five indicated markers in seven identified cell subsets. F) A heatmap showing the normalized mean expression of the CD107a in seven cell subsets. G) Representative contour plots for CD107a in CD8 T cells after co‐cultured with MM cells expressing shNC or shCAV1. H) Representative contour plots for CD107a in CD16⁺ or CD16⁻ NK cells after co‐cultured with MM cells expressing shNC or shCAV1. I) Bar plots showing the frequencies of the CD107a⁺ cells in NK cell subsets. J and K) Purified NK (pNK) cells were co‐culture with RPMI 8226 cells expressing shNC and shCAV1 at the ratio of 2:1 for six hours and the apoptosis of MM cells was determined using flow cytometry. L and M) In the presence of 10 µg mL⁻¹ daratumumab (Dara) or isotype IgG1, pNK cells were co‐culture with RPMI 8226 cells expressing shNC and shCAV1 at the ratio of 2:1 for six hours and the apoptosis of MM cells were determined using flow cytometry. p‐values were calculated with unpaired two‐sided Student t‐test or one‐way ANOVA test. Apop, apoptotic. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01; ^***, p < 0.001; ^****, p < 0.0001.

Figure 4

CAV1 knockdown inhibits MM cell growth in vivo and modulates ROS and the autophagy process. A) B‐NDG mice (n = 4) were subcutaneously inoculated with MM1S‐GFP‐Luc cells expressing shNC or shCAV1 in the left or right flank, respectively. Representative bioluminescence images of mice inoculated with MM cells. B–D) B‐NDG mice (n = 6) were subcutaneously inoculated with MM1S cells expressing shNC or shCAV1 in the left or right flank, respectively. (B) Tumor size of MM1S cells expressing shNC or shCAV1 at the indicated days were recorded. Error bar, mean ± SEM. (C and D) Tumor weight and images were obtained from B‐NDG mice subcutaneously inoculated with MM1S cells expressing shNC or shCAV1 for 31 days. E) Bubble plots showing the top five GO functional enrichment of genes significantly regulated by CAV1 in MM1S cells. F) Bubble plots showing the top 15 KEGG pathways of genes significantly regulated by CAV1 in MM1S cells. G) Representative confocal microscope images showing the levels of TMRE (red) in MM1S cells expressing shNC or shCAV1. DAPI (blue) was used to stain the cell nuclei. H) TMRE level in MM1S cells expressing shNC or shCAV1 was measured using flow cytometry. I) MM1S cells expressing shNC or shCAV1 were incubated with CellROX Deep Red for 15 min and the ROS level was determined using flow cytometry. J) MM1S cells expressing shNC or shCAV1 were treated with bafilomycin A1 (Baf‐A1) at the indicated concentrations for 48 h and the cell viability was measured. K) MM1S cells expressing shNC or shCAV1 were treated with Baf‐A1 (5 nM) for 48 h and the protein level of p62 and LC3‐I/II was detected using western blot. GAPDH was included as a loading control. L) MM1S cells expressing shNC or shCAV1 were cultured with or without serum for 24 h and cell viability was measured. M) MM1S cells expressing shNC or shCAV1 were cultured with or without serum for 24 h and incubated with CellROX Deep Red for 15 min. The ROS level was then determined using flow cytometry. N) MM1S cells expressing shNC or shCAV1 were cultured with or without serum for 24 h and the protein level of p62 and LC3‐1/II was detected using western blot. p‐values were calculated with unpaired two‐sided Student t‐test or Mann–Whitney U test. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01; ^***, p < 0.001; ^****, p < 0.0001.

Figure 5

CAV1 inhibition reprograms the cell metabolism in MM cells. A–E) Multiple metabolism‐related proteins in MM1S cells expressing shNC or shCAV1 were analyzed using mass cytometry. (A) viSNE maps colored by the normalized expression of the indicated markers in MM1S cells. (B and C) A heatmap showing the normalized mean expression of the indicated markers in 15 FlowSOM clusters of (B) MM1S or (C) RPMI 8226 cells. (D and E) Bar plots showing the frequencies of the indicated clusters in (D) MM1S or (E) RPMI 8226 cells expressing shNC or shCAV1. F) Bubble plots showing the top 15 KEGG pathways of metabolites significantly regulated by CAV1 in MM1S cells. G) Chordal graph showing the relationships between KEGG pathways and metabolites significantly regulated by CAV1 in MM1S cells. H) Lollipop Chart showing the top 10 import metabolites under POS and NEG model for distinguishing MM1S cells expressing shNC and shCAV. p‐values were calculated with an unpaired two‐sided Student t‐test. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01.

Figure 6

Targeting CAV1 regulated glutamine metabolism and enhanced the therapeutic efficiency of bortezomib. A) The relative level of L‐Glutamine and L‐Cysteine in MM1S cells expressing shNC or shCAV1. B) Relative GSH level in MM1S or RPMI 8226 cells expressing shNC or shCAV1. C) Relative GSH/GSSG ratio level in MM1S or RPMI 8226 cells expressing shNC or shCAV1. D and E) Relative total antioxidant capacity (T‐AOC) and malondialdehyde (MDA) levels in MM1S or RPMI 8226 cells expressing shNC or shCAV1. F) Relative GPX activity in MM1S or RPMI 8226 cells expressing shNC or shCAV1. G) A heatmap showing the expression of cell metabolism‐related genes in MM cells expressing shNC or shCAV1. H) mRNA expression of SLC1A5 and SLC7A5 in MM1S cells expressing shNC or shCAV1 was determined using qRT‐PCR. I) Protein expression of SLC1A5 and SLC7A5 in MM1S cells expressing shNC or shCAV1 was determined using a western blot. J) The total expression of SLC1A5 and SLC7A5 in MM1S cells was determined using flow cytometry. K) MM1S cells expressing shNC or shCAV1 were cultured with or without L‐Glutamine (Gln, 4 mM) for 24 h, and cell viability was measured. L) MM1S cells expressing shNC or shCAV1 were treated with V9302 at the indicated concentrations for 48 h and the cell viability was measured. M) In the absence of L‐Glutamine, MM1S cells expressing shNC or shCAV1 were treated with bortezomib at the indicated concentration for 48 h and the cell viability was measured. N and O) B‐NDG mice were intravenously inoculated with MM1S‐GFP‐Luc cells expressing shNC (n = 6) or shCAV1 (n = 6) and after 24 days, bortezomib (0.5 mg kg⁻¹) was intraperitoneally injected three times a week. N) The distribution of MM cells in these mice was measured at the indicated days after inoculation using a living imaging system. O)The total flux in each mouse after inoculation for the indicated days was determined (Error bar, mean ± SEM.). p‐values were calculated with unpaired two‐sided Student t‐test or Mann–Whitney U test. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01; ^***, p < 0.001; ^****, p < 0.0001.

Figure 7

6‐MP and daidzin improve MM treatment with bortezomib. A) MM1S cells were treated with 1087 FDA‐approved drugs and the expression of CAV1 (green) was determined using in‐cell western. DRAQ5 (red) was used to stain DNA/total cells for normalization of cell numbers. B) MM1S cells were treated with 6‐MP or daidzin at the indicated concentrations for 48 h and the expression of CAV1 was detected using western blot. C) The signal density of CAV1 was quantified and normalized to GAPDH. D and E) MM1S cells were treated with bortezomib (1.25 nM) in combination with 6‐MP or daidzin for 48 h and cell apoptosis was determined using flow cytometry. Apop, apoptotic. F–H) B‐NDG mice were inoculated with MM1S‐Luc cells and randomly divided into six groups (n = 5–7). After 11 days of inoculation, these mice were intraperitoneally injected with 6‐MP (10–20 mg kg⁻¹), daidzin (30 mg kg⁻¹), or vehicle daily until day 67. After 15 days of inoculation, mice were intraperitoneally injected with bortezomib (0.5 mg kg⁻¹) three times a week until day 67. F) The distribution of MM1S‐Luc cells in these mice was measured at the indicated days after inoculation using a living imaging system. G) The total flux in each mouse after inoculation for the 40 days was determined. Error bar, mean ± SEM. H) Kaplan–Meier curve showing mice survival. A logarithmic‐rank (Mantel–Cox) test was used to determine P values. p‐values were calculated with unpaired two‐sided Student t‐test or one‐way ANOVA test. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01; ^***, p < 0.001; ^****, p < 0.0001.

Figure 8

Statin‐mediated downregulation of CAV1 enhances the anti‐MM efficiency of bortezomib. A) MM1S cells were treated with lovastatin or simvastatin at the indicated concentration for 48 h and the expression of CAV1 was detected using western blot. B) The signal density of CAV1 was quantified and normalized to GAPDH. C,D) MM1S cells were treated with bortezomib (1.25 nM) in combination with lovastatin or simvastatin for 48 h and cell apoptosis was determined using flow cytometry. Apop, apoptotic. E–G) B‐NDG mice were inoculated with MM1S‐Luc cells and randomly divided into two groups. After 7 days of inoculation, these mice were intraperitoneally injected with simvastatin (20 mg kg⁻¹, n = 7) or vehicle (n = 7) daily until day 60. After day 15 after inoculation, all these mice were intraperitoneally injected with bortezomib (0.5 mg kg⁻¹) three times a week until day 60. (E) The distribution of MM1S‐Luc cells in these mice was measured at the indicated days after inoculation using a living imaging system. (F) The total flux in each mouse after inoculation for the indicated days was determined. Error bar, mean ± SEM. (G) Kaplan–Meier curve showing mice survival. A logarithmic‐rank (Mantel–Cox) test was used to determine P values. H) Schematic diagram showing how targeting CAV1 facilitates MM treatment through regulating cell adhesion, NK cell activation, autophagy, redox homeostasis, and glutamine transport. The figure was created with BioRender.com/v03k780. Gln, glutamine; Glu, glutamate, Cys, cysteine; Gly, glycine. p‐values were calculated with unpaired two‐sided Student t‐test or one‐way ANOVA test. Error bar, mean ± SD. ^*, p < 0.05; ^**, p < 0.01; ^***, p < 0.001; ^****, p < 0.0001.