Pro-apoptotic activity of α-bisabolol in preclinical models of primary human acute leukemia cells

Abstract

Background: We previously demonstrated that the plant-derived agent α-bisabolol enters cells via lipid rafts, binds to the pro-apoptotic Bcl-2 family protein BID, and may induce apoptosis. Here we studied the activity of α-bisabolol in acute leukemia cells.

Methods: We tested ex vivo blasts from 42 acute leukemias (14 Philadelphia-negative and 14 Philadelphia-positive B acute lymphoid leukemias, Ph-/Ph+B-ALL; 14 acute myeloid leukemias, AML) for their sensitivity to α-bisabolol in 24-hour dose-response assays. Concentrations and time were chosen based on CD34+, CD33+my and normal peripheral blood cell sensitivity to increasing α-bisabolol concentrations for up to 120 hours.

Results: A clustering analysis of the sensitivity over 24 hours identified three clusters. Cluster 1 (14 ± 5 μM α-bisabolol IC50) included mainly Ph-B-ALL cells. AML cells were split into cluster 2 and 3 (45 ± 7 and 65 ± 5 μM IC50). Ph+B-ALL cells were scattered, but mainly grouped into cluster 2. All leukemias, including 3 imatinib-resistant cases, were eventually responsive, but a subset of B-ALL cells was fairly sensitive to low α-bisabolol concentrations. α-bisabolol acted as a pro-apoptotic agent via a direct damage to mitochondrial integrity, which was responsible for the decrease in NADH-supported state 3 respiration and the disruption of the mitochondrial membrane potential.

Conclusion: Our study provides the first evidence that α-bisabolol is a pro-apoptotic agent for primary human acute leukemia cells.

Figure 1

α-bisabolol structure and solubilization in the culture medium. (A) α-bisabolol is a small oily sesquiterpene alcohol with a molecular mass of 222.37 Da. (B) 250 μM α-bisabolol was added to culture medium: concentration raised during the first 3 hours, then lowered to around 65% of the initially added α-bisabolol after 24 hours. (C) By this time, the linear function relating added to measured concentrations of α-bisabolol shows that the incremental ratio was 0.65 for 14 evaluations representing a double series of 7 scaled concentrations tested by a RP-HPLC method. Each point is the mean ± SD of 2 measurements.

Figure 2

Cytotoxicity of α-bisabolol in normal hematologic cells. (A) Peripheral blood cells. (B) Bone marrow stem cells. Time- and dose-response curves between 20 and 160 μM α -bisabolol in the 120-hour cytotoxicity assays. Means ± SD of 5 normal donors are depicted.

Figure 3

24-hourcytotoxicity of α-bisabolol in primary blasts from 42 acute leukemias. (A) α-bisabolol activity against blasts ex vivo, here grouped by diagnosis. The corresponding IC₅₀ values and number of cases (n) are shown. The differences between the Ph^-B-ALL sensitivity curves and the other ones were statistically significant (p < 0.05). Each point is the mean ± SD of 14 cases. (B) α-bisabolol sensitivity clustering analysis. The samples were grouped by complete linkage hierarchical clustering algorithm available in MultiExperiment Viewer http://www.tm4.org/mev/. The heat map was obtained by subtracting spontaneous mortality to scaled α-bisabolol 24-hour cytotoxicity expressed as percentage. Three main groups of patients were identified based on their cytotoxicity response. The Ph^-B-ALL (ALL) cases shared the highest sensitivity and were grouped mainly in the first sensitivity cluster, whereas AML cases were split into two groups with intermediate and lower sensitivities. Ph⁺B-ALL (Ph+) cells were scattered among the three groups, although they were mainly clustered in the second group. At the bottom, the 24-hour dose-response curves of the three sensitivity clusters are depicted, and the corresponding IC₅₀ values and number of cases (n) are shown. The differences between the curves were statistically significant (p < 0.05).

Figure 4

24-hour cytotoxicity of α-bisabolol in each individual case. 14 Ph^-B-ALL, 14 Ph⁺B-ALL, and 14 AML cell samples were treated with 20, 40, 80, and 160 μM α-bisabolol over 24 hours. Captions identify the cases in Table 1 and in Figure 3B (clustering analysis).

Figure 5

24-hour cytotoxicity of α-bisabolol in Ph⁺ cells as compared to imatinib mesylate. (A) Scaled α-bisabolol alone (solid line) or in combination with 3 μM imatinib mesylate (dashed line) in 2 representative cases out of 10 (Ph⁺B-ALL #04 and #05 in Table 1, Figure 3B and Figure 4, where α-bisabolol concentrations are represented up to 160 μM). The imatinib mesylate-dependent cytotoxicity is indicated at point 0,0. Cells resistant to imatinib mesylate were sensitive to α-bisabolol. In cells sensitive to imatinib mesylate, α-bisabolol potentiated the effect of the other drug. (B) Analysis of synergism between imatinib and α-bisabolol in the imatinib-sensitive BCR/ABL⁺ human cell line CML-T1. Left side. Effects of 40 μM α-bisabolol and 0.1 μM imatinib, alone and combined, on proliferation of the cell line. Means ± SD of 5 experiments. Right side. Plot showing the corrisponding combination index (CI) vs. the fraction affected (Fa). CI values are <1, indicating that the two drugs are synergistic. Bars represent the variability of effects according to the sequential deletion analysis [16].

Figure 6

BID and NADH-supported state 3 respiration in normal PBMCs and leukemic blasts treated with α-bisabolol. (A) 24-hour α-bisabolol did not induced the cleavage of BID (full length 22 kDa, cleaved 15 kDa) at any concentration. Etoposide-treated Jurkat cells were used as a positive control for tBID. (B) No BID translocation was detected in mitochondrial fraction at different times and solubilized doses of α-bisabolol. α-tubulin and Hsp60 were used as markers for the cytosol and mitochondria fractions, respectively. A representative case is shown. (C) Permeabilized leukemic cells and healthy lymphocytes were incubated for 10 minutes in respiration buffer at 30°C in the presence or in the absence of 3 μM α-bisabolol. In treated leukemic cells, the G/M oxygen consumption was clearly lower than in untreated leukemic controls (p < 0.05). The S/G3P oxygen consumption was not modified by treatment, and the mitochondrial respiration was not stimulated by FCCP addition. This is in line with a direct effect of α-bisabolol on mitochondrial integrity. Healthy lymphocyte respiration was not affected by treatment. G/M: glutamate plus malate; S/G3P: succinate plus glycerol-3-phosphate; FCCP: carbonylcyanide-4-(trifluoromethoxy)-phenyl-hydrazone. Means ± SD of 6 leukemias and 6 normal donors are depicted.

Figure 7

α-bisabolol-induced mitochondrial damage in primary leukemic blasts. Cells were stained with JC-1. In non-damaged cells, JC-1 forms red-emitting aggregates in the mitochondrial matrix. A loss of red fluorescence and an increase in cytoplasmic green-emitting monomers signal the disruption of the mitochondrial transmembrane potential (ΔΨm). (A) The representative case Ph^-B-ALL #01 is shown out of the 6 leukemias tested. Microscopy (magnification, × 400). Whereas untreated leukemic blasts showed well-polarized mitochondria marked by punctated red fluorescent staining, blasts treated with 40 μM α-bisabolol had staining that was quite completely replaced by diffuse green fluorescence, indicating loss of ΔΨm. Flow cytometry. Untreated blasts with well-polarized mitochondria localized in the upper region of the plot (high ΔΨm). Blasts exposed to 40 μM α-bisabolol shifted right and downward (intermediate and low ΔΨm), due to the progressive dislocation of JC-1 from the mitochondria to the cytoplasm, which signaled the disruption of the mitochondrial ΔΨm. (B) Both untreated and α-bisabolol-treated normal lymphocytes used as a negative control maintained well-polarized mitochondria and did not undergo apoptosis. Apoptosis of leukemic blasts was also documented by (C) PARP cleavage and DNA laddering in the same representative case depicted in (A).