Potassium/sodium cation carriers robustly upregulate CD20 antigen by targeting MYC, and synergize with anti- CD20 immunotherapies to eliminate malignant B cells

Abstract

Our investigation uncovers that nanomolar concentrations of salinomycin, monensin, nigericin, and narasin (a group of potassium/ sodium cation carriers) robustly enhance surface expression of CD20 antigen in B-cell-derived tumor cells, including primary malignant cells of chronic lymphocytic leukemia and diffuse large B-cell lymphoma. Experiments in vitro, ex vivo, and animal model reveal a novel approach of combining salinomycin or monensin with therapeutic anti-CD20 monoclonal antibodies or anti-CD20 chimeric antigen receptor T cells, significantly improving non-Hodgkin lymphoma therapy. The results of RNA sequencing, genetic editing, and chemical inhibition delineate the molecular mechanism of CD20 upregulation, at least partially, to the downregulation of MYC, the transcriptional repressor of the MS4A1 gene encoding CD20. Our findings propose the cation carriers as compounds targeting MYC oncogene, which can be combined with anti-CD20 antibodies or adoptive cellular therapies to treat non-Hodgkin lymphoma and mitigate resistance, which frequently depends on the CD20 antigen loss, offering new solutions to improve patient outcomes.

Figure 1.

Monovalent cation carriers markedly increase CD20 on the surface of lymphoma and leukemia cells. (A) The chemical structures of selected cation carriers, namely salinomycin (SAL; left), monensin (MON; middle), nigericin (NIG; middle), and narasin (NAR; right), were visualized using ChemDraw Ultra 12.0. (B) Flow cytometry analysis of surface levels of CD20 antigen (quantification of mean fluorescence intensity [MFI] values) and (C) viability of propidium iodide-stained Raji cells, pretreated for 48 hours (h) with corresponding cation carriers (concentration range 0.01-2.5 uM). (D, E) Flow cytometry analysis of cell surface levels of CD20 antigen in 3 cohorts of primary tumor cells (CD19-positive). In the cohorts of chronic lymphocytic leukemia (CLL) samples (D), the cells were pretreated ex vivo with increasing concentrations of either SAL (left graph, N=24) or MON (right graph, N=13), while in the cohort of 10 samples of diffuse large B-cell lymphoma (DLBCL) (E), the cells were pretreated ex vivo with SAL for 48 h. MFI values of anti-CD20-FITC staining in cation-carrier-treated cells were normalized to the MFI values in vehicle-treated cells and presented as fold change.

Figure 2.

Salinomycin and monensin enhance the complement-dependent and natual killer cell-dependent cytotoxicity of rituximab and the anti-CD20 CAR T-cell cytotoxic activity. (A) Complement-dependent cytotoxicity (CDC) assays showing the improved killing of Raji cells pretreated with either salinomycin (SAL) (0.25-0.5 µM; left graph) or monensin (MON) (0.05-0.1 µM; right graph) for 48 hours (h) followed by treatment with rituximab (RTX; 1-100 μg/mL) for 1 h, in the presence of human serum as a source of complement. The viability of cells was assessed with propidium iodide (PI) staining followed by flow cytometry analysis. The results were presented as a percentage of alive control cells (untreated with RTX). (B) Antibody-dependent cellular cytotoxicity (ADCC) assays showing improved cytotoxicity of natural killer (NK) cells towards Raji cells pretreated with either SAL (0.25 µM; left panel) or MON (0.05 µM; right panel) for 48 h, followed by staining of Raji cells with carboxyfluorescein succinimidyl ester (CFSE) and co-incubation with unstained donor-derived NK cells for 3 h, in the absence or presence of RTX (0.03 µg/mL). The survival of CFSE-positive Raji cells was assessed with PI, as above, and presented as a percentage of control cells (Raji not incubated with NK cells). Graphs show data from 3 experiments (NK cells isolated from 3 donors). (C) The heat map presents the log2 fold change in the levels of surface proteins in Raji cells treated with either SAL (0.25 µM) or vehicle for 48 h (6 samples of each treatment were analyzed). The list includes only the proteins potentially implicated in regulating NK cell activity. The changes in CD20 were also included to serve as a positive control. (D) Anti-CD20 CAR T-mediated killing assays showing improved cytotoxicity of CAR T cells towards CFSE-stained Raji cells pretreated with either SAL (left panel) or MON (right panel) for 48 h. For these cytotoxicity assays, Raji cells were co-incubated with the unstained effector cells, either T cells or CAR T cells, for 24 h. Survival of Raji cells was assessed with PI staining followed by flow cytometry analysis of CFSE-positive cells. Results were presented as a percentage of control cells (Raji not incubated with T cells). Graphs summarize data from 3 experiments.

Figure 3.

Salinomycin and monensin increase CD20 levels and rituximab efficacy in a panel of B-cell tumor cell lines. (A, B) The Left panels present flow cytometry analysis of surface levels of CD20 antigen (mean fluorescence intensity [MFI] of anti-CD20-FITC antibody) in diffuse large B-cell lymphoma (DLBCL) cell lines pretreated with either salinomycin (SAL) (0.1-0.5 uM), monensin (MON) (0.01-0.1 uM) or corresponding vehicles for 48 hours (h). The right panels present their response to rituximab in complement-dependent cytotoxicity (CDC) assays upon preincubation with either SAL or vehicle (at 2 selected concentrations); (A) germinal center B-cell (GCB) subtype (OCI-Ly1 and OCI-Ly7 cells), (B) activated B-cell (ABC) subtype (HBL-1 and U2932 cells). (C) CDC assays show the improved killing of Raji cells pretreated with either SAL, MON, or vehicle for 48 h, followed by treatment with ofatumumab (OFA; 1-100 ug/mL) for 1 h, in the presence of human serum as a source of complement. The survival of cells was determined with flow cytometry and was presented as a percentage of alive control cells (untreated with OFA). (D) Assays showing improved cytotoxicity of natural killer (NK) cells towards B-cell tumor cell lines, such as CA46 (left panel), Ramos (middle panel), OCI-Ly1 (middle panel), and HBL-1 (right panel) pretreated with SAL (0.1-0.25 uM) for 48 h, followed by staining of these cells with CFSE and co-incubation with unstained donor-derived NK cells for 3 h. The survival of carboxyfluorescein succinimidyl ester-positive tumor cells was assessed with PI and presented as a percentage of control cells (not incubated with NK cells). Graphs show data from at least 3 experiments (NK cells isolated from 3 donors).

Figure 4.

Salinomycin and monensin potentiate the antitumor activity of rituximab in vivo. SCID mice (CB17/Icr-Prkdc^scid/IcrIcoCrl) were inoculated subcutaneously (sc.) with Raji cells. Mice were then injected intraperitoneally (ip.) with either salinomycin (SAL) (2.5 mg/kg), monensin (MON) (1 mg/kg), or vehicle on days 5 and 7. The ip. administration of rituximab (RTX) (10 mg/kg) started on day 9 and has been applied every second day, together with injections of SAL, MON, or vehicle. (A) The graph presents the volume of tumors measured every 3-4 days. (B) The tumors were post-mortally isolated on day 30 and weighed (N=6-11 tumors/ group). (C) The SCID mice were inoculated sc. with a mix of 3 clones (in proportion 1:1:1) of Raji cells transduced with either empty vector (v.) or sgMS4A1. Mice were then injected ip. with either SAL or vehicle, followed by injections of RTX, as described above. The photos (left panel) and the weight (right panel) of tumors (isolated post-mortally on day 32) were documented (N=5-6 tumors/group).

Figure 5.

Salinomycin and monensin regulate the MS4A1 gene transcription. (A) Analysis of mRNA levels of MS4A1 (estimated by quantitative real-time polymerase chain reaction [qRT-PCR]; 2 minus δ CT method) in Raji cells treated with either salinomycin (SAL) (0.5 µM), monensin (MON) (0.1 µM), or corresponding vehicles (Veh.) for 12-18 hours (h). (B) The analysis was performed like in (A), using Raji cells pretreated with transcription inhibitor actinomycin D (ActD; 5 µg/mL) for 2 h. SAL (0.5 µM) or Veh. was then added for the next 18 h (estimation of MS4A1 mRNA level; left panel) or 24 h (analysis of the surface level of CD20; right panel). (C) The differential expression heatmap (left panel) shows the comparison of gene expression profiles (estimated by RNA sequencing [RNA-seq]; log2 fold change; q value cutoff <0.05) in Raji cells treated with either Veh., SAL (0.5 µM), or MON (0.1 µM) for 18 h (in 2 biological replicates). The volcano plot (right panel) depicts the number of significantly upregulated (red dots) and downregulated (green dots) mRNA upon the treatment of Raji cells with SAL. (D) The Venn diagram represents the prediction of transcription factor/miRNA binding sites in the regulatory elements of the differentially expressed target genes (analyzed with GSEA/MSigDB website v6.3). SAL and MON affected 6 (the small dark blue circle) and 101 (the big bright blue circle) transcription factors/microRNA (miRNA), respectively. The activity of 4 transcription factors/miRNA (FOXO, MYC, NF-Y, and mir181) appeared to be commonly affected by both SAL and MON.

Figure 6.

The cation carriers affect FOXO signaling pathways. (A) Quantitative real-time polymerase chain reaction (qRT-PCR) (2 minus δ CT method) validation of changes in mRNA levels of selected genes, namely SGK1 (left panel) and IL7R (right panel), upon treatment with either salinomycin (SAL) or monensin (MON) for 12-18 hours (h). (B) Western blotting analysis of both phosphorylated and total protein levels of AKT and FOXO1, as well as the total level of SGK1 kinase in Raji cells, pretreated with either SAL (0.25 and 0.5 µM), MON (0.05 and 0.1 µM) or vehicle (-) for 18 h. The level of GAPDH was used as a loading control. Western blotting images are presented in the top panel, while the quantification of phosphorylated AKT (p-AKT; Ser473 residue), phosphorylated FOXO1 (p-FOXO1; Ser256 residue) and SGK1 from 3-4 images are presented in the bottom panel.

Figure 7.

The cation carriers affect MYC levels. (A) Downregulation of MYC mRNA upon 12 hours (h) of treatment with either salinomycin (SAL) or monensin (MON), estimated by quantitative real-time polymerase chain reaction (qRT-PCR) (2 minus 5 CT method). (B) Estimation of MYC protein levels by western blotting analysis in both Raji cells and P493-6 lymphoblastoid cell line, pretreated with either SAL, MON, or vehicle (-) for 18 h. The level of GAPDH was used as a loading control. (C) The differential expression heatmap with RNA-sequencing (RNA-seq) data listing the genes regulated by MYC (according to gene set enrichment analysis [GSEA]; hallmarks - MYC targets v1 and v2) and presents the fold change of their expression in Raji cells treated for 18 h with either SAL (0.5 uM) or MON (0.1 uM). The treatment with MON downregulated all the listed genes with statistical significance (q value <0.05), while the treatment with SAL decreased their expression, with the statistical significance reached for the RPS10 gene only. (D) qRT-PCR validation of changes in the expression of known MYC target genes, namely PLK1 (left panel), TNFAIP3 (middle panel), and PTPN6 (right panel) upon 18 h of treatment with either SAL, MON, or corresponding vehicle (Veh.). (E) Flow cytometry analysis of either intracellular level of MYC (left panel) or surface level of CD20 (right panel) in P493-6 cells expressing MYC under the control of Tet-off system. P493-6 cells were treated with either SAL, doxycycline (DOX, 0.1 ug/mL), or both for either 24 or 48 h (left and right panel, respectively). Additionally, the number of CD20 molecules per cell was quantified by comparison of anti-CD20-PE mean fluorescence intensity (MFI) to the fluorescence intensity of quantibrite beads labeled with the known number of PE molecules. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001

Figure 8.

Derivatives of salinomycin unable to downregulate MYC are also unable to upregulate CD20. (A) Comparison of the changes in surface CD20 induced by increasing concentrations of either salinomycin (SAL) or selected SAL derivatives (derivatives 1 and 5). Raji cells were treated with either vehicle (Veh.) at low (L), medium (M), or high (H) concentrations or SAL derivatives (concentration range, 0.01-2.5 µM). The mean fluorescence intensity (MFI) values of anti-CD20-PE detected in the case of SAL derivative-treated cells were normalized to the MFI values corresponding to vehicle-treated cells and were presented as fold change. The chemical structures of SAL derivatives 1 and 5 are shown in the corresponding graphs, with modifications of the SAL structure marked in red and blue, respectively. (B) Western blotting analysis of MYC and SGK1 in Raji cells, pretreated with either SAL, its derivatives, or vehicle (-) for 18 h. Levels of MYC (graph in the middle) and SGK1 (right graph) were quantified from 3 experiments and presented as a fold change of protein levels in SAL- or derivatives-treated cells versus levels in vehicle-treated cells. (C) Western blotting analysis of MYC and SGK1 (left panel) in Raji cells nucleofected with RNP complexes consisting of single guide RNA (sgRNA) targeting either MYC (sgMYC) or SGK1 (sgSGK1) and Cas9 nuclease followed by treatment with MON for 18 hours (h). For the mock nucleofection, the ribonucleoprotein (RNP) lacked sgRNA. Forty-eight hours upon nucleofection, the MFI of CD20-PE (right panel) was quantified in Raji cells nucleofected with RNP consisting of sgRNA targeting either MYC (sgMYC #1 and sgMYC #2) or SGK1 (sgSGK1 #1 and sgSGK1 #2) and was normalized to MFI in mock-nucleofected cells. (D) Comparison of the upregulation of surface CD20 in Raji cells pretreated with either SAL, MYC inhibitor (10058-F4; 100-150 µM) or corresponding vehicles for 48 h. (E) Results of the CUT&RUN protein-DNA binding assay show the binding of MYC to the MS4A1 promoter in either vehicle- or SAL-treated Raji cells (18 h of the treatment). The reverse transcription polymerase chain reaction (RT-PCR)-amplified fragment of the MS4A1 promoter (from -313 to -198). ^*P< 0.05; ^**P< 0.01; ^***P< 0.001; ^****P< 0.0001