C6-ceramide nanoliposomes target the Warburg effect in chronic lymphocytic leukemia

Abstract

Ceramide is a sphingolipid metabolite that induces cancer cell death. When C6-ceramide is encapsulated in a nanoliposome bilayer formulation, cell death is selectively induced in tumor models. However, the mechanism underlying this selectivity is unknown. As most tumors exhibit a preferential switch to glycolysis, as described in the "Warburg effect", we hypothesize that ceramide nanoliposomes selectively target this glycolytic pathway in cancer. We utilize chronic lymphocytic leukemia (CLL) as a cancer model, which has an increased dependency on glycolysis. In CLL cells, we demonstrate that C6-ceramide nanoliposomes, but not control nanoliposomes, induce caspase 3/7-independent necrotic cell death. Nanoliposomal ceramide inhibits both the RNA and protein expression of GAPDH, an enzyme in the glycolytic pathway, which is overexpressed in CLL. To confirm that ceramide targets GAPDH, we demonstrate that downregulation of GAPDH potentiates the decrease in ATP after ceramide treatment and exogenous pyruvate treatment as well as GAPDH overexpression partially rescues ceramide-induced necrosis. Finally, an in vivo murine model of CLL shows that nanoliposomal C6-ceramide treatment elicits tumor regression, concomitant with GAPDH downregulation. We conclude that selective inhibition of the glycolytic pathway in CLL cells with nanoliposomal C6-ceramide could potentially be an effective therapy for leukemia by targeting the Warburg effect.

Figure 1. Nanoliposomal C6-ceramide selectively induces cell death in CLL cells.

JVM3 cells were treated with varying doses of ghost or C6-ceramide or dihydro-C6-ceramide nanoliposomes for 24 hours then A). MTT assay, B). Trypan blue staining was performed. ANOVA statistical test was used to determine dose dependency between various C6-ceramide treatment groups. P < .0001. C). JVM3 cells were treated with different doses of ghost or C6-ceramide nanoliposomes for 24 hours, then cells were stained with annexin V and 7AAD for apoptosis assay. D). Percentage of apoptotic cells was determined via TUNEL analysis after 24 hours. E). PBMC isolated from either CLL patients (n=3) or normal donors (n=3) were treated with 25 µM ghost or C6-ceramide nanoliposome for 2, 16 and 24 hours, then MTT assay was performed. F). Percentage of apoptotic cells was determined in PBMC from normal donors (n=3) via annexin V/7AAD staining. Cells were treated with different doses of ghost or C6-ceramide nanoliposome for 24 hours or treated with 25 µM ghost or C6-ceramide nanoliposome for 2 and 24 hours. * P < 0.05.

Figure 2. Cell death induced by nanoliposomal C6-ceramide occurs through caspase 3/7-independent necrosis.

JVM3 cells were treated with A). different doses of ghost and C6-ceramide nanoliposome for 24 hours, B) 5 µM and 10 µM dasatinib, as well as DMSO vehicle control for 24 hours, then Western Blot analysis was performed for caspase 3 and PARP cleavage. C). Enzymatic activities of caspase 3/7 were measured using caspase 3/7 luminescence kit. * P < 0.05. D). JVM3 cells were treated with varying doses of ghost or C6-ceramide nanoliposomes for 24 hours following a 2 hour pre-treatment or no pre-treatment with zVAD-fmk (15 µM). Cell viability was assessed via MTT assay. * P < 0.05. E). JVM3 cells were treated with 25 µM ghost nanoliposomes for 24 hours or C6-ceramide nanoliposomes for 2, 6 and 24 hours and phase contrast microscopy images were taken. Arrows indicate morphology of necrotic cell death.

Figure 3. C6-ceramide nanoliposomes target GAPDH in CLL.

A). JVM3 cells were treated with (i) 25µM of ghost or C6-ceramide nanoliposomes for varying times, (ii) varying doses of ghost or C6-ceramide nanoliposomes for 24 hours, then Western Blot analysis was performed for GAPDH. Densitometry analysis from replicate experiments (n=4) was performed via ImageJ software. In Figure A(ii) the representative blot demonstrates a decrease in GAPDH at 12.5µM C6-ceramide, a result not observed in the other two distinct replicate experiments. * P < 0.05; ** P <0 .005. B). CLL patient cells (n=3, analyzed individually) were treated with different doses of ghost or C6-ceramide nanoliposomes for 24 hours, then Western Blot analysis was performed for GAPDH. * P < 0.05. The blot represents the effect of C6-ceramide nanoliposomes on one patient sample, however, the graph is an average of the effect on all three patient cells. C). JVM3 cells were treated with different doses of ghost and C6-ceramide nanoliposomes for 24 hours, then qRT-PCR analysis was performed for expression of GAPDH mRNA. Expression was normalized to 18S. * P < 0.05. D). PBMC from normal donors (n=3) were treated with 25 µM ghost or C6-ceramide nanoliposome for 24 hours and then Western Blot analysis was performed for GAPDH. E). Basal protein expression of GAPDH was determined via Western Blot analysis on PBMC isolated from normal donors (n=8) or from CLL patients with either a lower WBC count (n=9) or higher WBC count (n=5). Patients with WBC counts >50,000 cells/µL were identified as patients with high WBC count. * P < 0.05. F). CLL patient cells with either a lower WBC count (n=4) or higher WBC count (n=3) were treated with 25µM ghost or C6-ceramide nanoliposomes for 2, 16 and 24 hours then an MTT assay was performed * P < 0.05.

Figure 4. C6-ceramide targets the glycolytic pathway.

JVM3 cells were treated with varying doses of ghost or C6-ceramide nanoliposomes for 24 hours, then A) lactate production was analyzed, B) ATP production was analyzed; * P < 0.05. The graphs depict average results from three independent experiments. C). GAPDH was effectively knocked down in JVM3 cells via a lentiviral shRNA approach (inset). ATP production was then assessed after 24 hours of treatment with 50µM of ghost or 25µM of C6-ceramide nanoliposomes; * P < 0.05; ** P < 0.005. D). Glucose uptake was assessed after JVM3 cells were treated with ghost and C6-ceramide nanoliposomes for 24 hours. Cytochalasin B was used as a positive control. E). JVM3 cells were treated with ghost or C6-ceramide nanoliposomes for 24 hours, then Western Blot analysis was performed for GLUT1.

Figure 5. Pharmacological and molecular confirmation that nanoliposomal C6-ceramide targets the glycolytic pathway at the level of GAPDH.

A). JVM3 cells were pre-treated for 2 hours with 10mM pyruvate, then treated with ghost or C6-ceramide nanoliposomes for 24 hours. MTT cell viability assay was performed. B). ATP production in these cells was also determined; ** P < 0.005. C). JVM3 cells were transduced with lentiviral particles overexpressing GAPDH or viral particles expressing RFP (control). Post transduction, Western blotting analysis was done to determine levels of GAPDH in experimental and control cells (insert). Cells were treated with ghost or C6-ceramide nanoliposomes for 24 hours. Cell viability was determined using MTT assay and alamarBlue assay; * P < 0.005.

Figure 6. Nanoliposomal C6-ceramide displays anti-leukemic effect in a CLL animal model.

A). Two weeks after ten million JVM3 cells were inoculated in the right flank of female Balb/c Nu/nu mice, animals were treated with 40 mg/kg ghost (n=8) or C6-ceramide (n=8) nanoliposomes via tail vein injection. Treatment regimen was every other day over a three week period of time. Tumor size was assessed every other day; * P < 0.05; ** P < 0.005. B). Leukemic mice following an identical dose regimen were sacrificed on day 8 (n=5), day 14 (n=6) and day 17 (n=6) of C6-ceramide treatment and on day 17 for ghost treatment. Immunoblot analysis for GAPDH protein expression was performed on tumor tissues. A representative blot from day 17 is shown (Figure 6B inset); * P < 0.05.