Hyperforin inhibits Akt1 kinase activity and promotes caspase-mediated apoptosis involving Bad and Noxa activation in human myeloid tumor cells

Abstract

Background: The natural phloroglucinol hyperforin HF displays anti-inflammatory and anti-tumoral properties of potential pharmacological interest. Acute myeloid leukemia (AML) cells abnormally proliferate and escape apoptosis. Herein, the effects and mechanisms of purified HF on AML cell dysfunction were investigated in AML cell lines defining distinct AML subfamilies and primary AML cells cultured ex vivo.

Methodology and results: HF inhibited in a time- and concentration-dependent manner the growth of AML cell lines (U937, OCI-AML3, NB4, HL-60) by inducing apoptosis as evidenced by accumulation of sub-G1 population, phosphatidylserine externalization and DNA fragmentation. HF also induced apoptosis in primary AML blasts, whereas normal blood cells were not affected. The apoptotic process in U937 cells was accompanied by downregulation of anti-apoptotic Bcl-2, upregulation of pro-apoptotic Noxa, mitochondrial membrane depolarization, activation of procaspases and cleavage of the caspase substrate PARP-1. The general caspase inhibitor Z-VAD-fmk and the caspase-9- and -3-specific inhibitors, but not caspase-8 inhibitor, significantly attenuated apoptosis. HF-mediated apoptosis was associated with dephosphorylation of active Akt1 (at Ser(473)) and Akt1 substrate Bad (at Ser(136)) which activates Bad pro-apoptotic function. HF supppressed the kinase activity of Akt1, and combined treatment with the allosteric Akt1 inhibitor Akt-I-VIII significantly enhanced apoptosis of U937 cells.

Significance: Our data provide new evidence that HF's pro-apoptotic effect in AML cells involved inhibition of Akt1 signaling, mitochondria and Bcl-2 members dysfunctions, and activation of procaspases -9/-3. Combined interruption of mitochondrial and Akt1 pathways by HF may have implications for AML treatment.

Figure 1. Chemical structure of hyperforin HF.

5-hydroxy-6R-methyl-1R,3,7S-tris(3-methyl-2-butenyl)-5S-(2-methyl-1-oxopropyl)-6R-(4-methyl-3-pentenyl)-bicyclo[3.3.1]non-3-ene-2,9 dione; C₃₅H₅₂O₄; 536 g/mol. HF was purified according to the method described (purity≥98%) and yielded less than 1% hypericin. HF was stored frozen in EtOH under conditions preventing its sensitivity to light, oxygen and aqueous solvents.

Figure 2. Effects of HF on U937 cell growth.

U937 cells (10⁵/ml) were treated with HF (A) at the indicated concentrations for 72 h or (B) or with 0.5 and 1.4 µg/ml HF for the indicated times. Control EtOH (vehicle). Cell growth was measured by direct cell counting (in duplicates). Data are the mean ± SD of results from at least 6 independent experiments, each performed in duplicates. (C) U937 cells were incubated with 1.4 µg/ml HF for 72 h. Cells were stained with PI and DNA contents analyzed by flow cytometry. (D) DNA fragmentation in U937 cells treated for 72 h with 1.4 µg/ml HF, EtOH (vehicle) or 100 nM flavopiridol (F).

Figure 3. HF induces apoptosis in AML cell lines.

(A) U937 cells were treated with 1.4 µg/ml HF for 72 h. Detection of apoptotic cells after annexin-V-FITC/propidium iodide staining and flow cytometry. Results are expressed as log PI fluorescence intensity (y-axis) vs log annexin-V-FITC fluorescence intensity (x-axis). L1, necrotic cells; L2, apoptotic + secondary necrotic cells; L3, healthy cells; L4, apoptotic cells. (B) Percent of apoptotic cells (L2+L4 gates) treated at the indicated concentrations for 72 h. Data are the mean ± SD of results from at least 4 independent experiments. (C) Percent of apoptotic cells (L2+L4 gates) treated with 1 or 1.4 µg/ml HF for the indicated times. Data are the mean ± SD of results from at least 4 independent experiments. (D) AML cell lines were treated 1.4 µg/ml HF for 72 h. Cell growth was quantified as % of untreated cells. Percent of apoptotic cells was determined as in (B). Data are the mean ± SD of results from at least 4 independent experiments (n = 4 for HL-60, OCI and NB4, n = 8 for U937). Basal apoptosis was ≤10% for HL-60 and U937, ≤20% for OCI and NB4).

Figure 4. HF treatment induces apoptosis in primary AML cells.

AML cells are characterized by the French American British (FAB) M0/undifferentiated, M1/myeloblastic, M2/myeloblastic with maturation, M3/promyelocytic, M4/myelomonocytic and M5/monoblastic phenotypes. PBMCs (10⁶/ml) from normal donors (n = 4) and AML patients were cultured with or without HF 2 µg/ml for 72 h, and then stained with annexin-V-FITC/PI and analyzed by flow cytometry to assess the percentage of apoptotic cells (L2+L4 gates) in untreated (basal apoptosis) and HF-treated cultures.

Figure 5. HF-induced apoptosis in U937 cells is caspase-dependent.

(A) U937 cells were treated for 14 and 24 h with 1.4 µg/ml HF or left untreated. Caspase-8, -9 and -3 activities were determined using IETD-pNA (caspase-8), LEHD-pNA (caspase-9) DEVD-pNA (caspase-3) respectively. The release of pNA was measured at 405 nm. The data are expressed as fold-increase relative to respective untreated samples (basal values for caspase-8, -9 and -3 activities were respectively 23±0.02, 9±0.01 and 13±0.02 pmol pNA/60 min/µg protein). Data are from three determinations ± SD. (B) Lysates from U937, NB4 and OCI cells treated for 24 h with HF (1, 1.4, 2 µg/ml) were examined for PARP-1 expression by immunoblot. The arrows indicate the cleaved form (cf) of PARP-1. (C) U937 cells were incubated with 1.4 µg/ml HF for 48 h after 1 h-pretreatment with 50 µM Z-VAD-fmk (broad spectrum caspase inhibitor), Z-DEVD-fmk (caspase-3 inhibitor), Z-IETD-fmk (caspase-8 inhibitor) or Ac-LEHD-CHO (caspase-9 inhibitor). The percentage of apoptotic cells was determined after annexin-V-FITC/PI staining and flow cytometry (L2+L4 gates). Data are the mean ± SD of results from 3 independent experiments. *P<0.05 compared with untreated cells or cells pretreated with caspase inhibitors.

Figure 6. HF triggers a dissipation of the mitochondrial transmembrane potential associated with downregulation of Bcl-2 and Bid, and upregulation of Noxa.

(A&B) U937 cells were cultured for 24 h in the absence or presence of 1.4 µg/ml HF or 100 nM flavopiridol (F) (a positive control of apoptosis in U937 cells). Then, cells were incubated for 15 min at 37°C with the fluorescent probe JC-1, subsequently washed and distributed in triplicates in the wells of a microtitration plate. Green and red fluorescences were measured. The loss of mitochondrial membrane potential is characterized by a a significant shift of the red (polarization) fluorescence to the green (depolarization) fluorescence. A) Representative histograms showing flow cytometry analysis of polarization and depolarization. B) Quantification of the polarization to the depolarization ratio. Data are expressed as mean ± SD of three determinations. (C) U937 cells were treated with 1, 1.4 or 2 µg/ml HF or 100 nM F for 24 h, after which lysates were subjected to Western blots using antibodies recognizing Bax, Bid, Bcl-2, Mcl-1, Noxa and actin. Two experiments representative of four are shown.

Figure 7. HF induces the dephosphorylation of Bad and Akt, and inhibits Akt1 kinase activity.

(A) U937 cells were treated with 1, 1.4 or 2 µg/ml HF or 100 nM F for 24 h, after which lysates were subjected to Western blots using antibodies recognizing Ser¹³⁶-p-Bad, Bad, Ser⁴⁷³-p-Akt1, Akt1 and actin. One experiment representative of three is shown. (B) Akt1 kinase activity was determined using a kinase assay kit obtained from Enzolifesciences France using recombinant active Akt1 in the absence or presence of increasing concentrations (0.5–5.5 µM) of Akt1/2 inhibitor Akt-I-VIII, F (flavopiridol), HF or vehicles EtOH (for HF) and DMSO (for F and Akt-I-VIII). Of note, 2.5 µM of HF corresponds to 1.4 µg/ml. Percent of Akt kinase activity was determined. (C) The percentage of apoptotic cells was determined after annexin-V-FITC/PI staining and cytometry, under control conditions, or 1.4 µg/ml HF or Akt-I-VIII (2.5, 5 and 10 µM) and in combination treatment for 48 h. Data are the mean ± SD of results from 2 independent experiments.

Figure 8. HF-induced apoptosis is independent on NF-κB activation.

(A) NB4/GFP cells and NB4/GFP-MAD cells were left untreated or were incubated for 15 min with TNF-α (20 ng/ml), after which lysates were subjected to Western blots using antibodies recognizing Iκ-Bα and actin. (B) cell growth and (C) apoptosis of wild NB4, NB4/GFP and NB4/GFP-MAD cells (10⁵/ml) treated with HF at the indicated concentrations for 72 h. Data are the mean ± SD of results from 4 independent experiments.

Figure 9. Putative model for the convergent roles of Akt1 and mitochondrial caspase-dependent signaling pathways in the induction of apoptosis by HF in AML cells.

HF inactivates Akt1 leading to dephosphorylation of proapoptotic Bad, and downregulates anti-apoptotic Bcl-2 protein. Dephosphorylated Bad sequesters Bcl-2 in the cytoplasm, favoring Bax translocation to the mitochondria. HF induces upregulation of Noxa which exerts its pro-apoptotic function by neutralizing the prosurvival Mcl-1, facilitating activation of Bax. Subsequent oligomerization of Bax inserts into the outer mitochondrial membrane, and triggers cytochrome c release. Cytochrome c forms a complex with procaspase-9, Apaf-1 and dATP (apoptosome). The apoptosome activates caspase-9 which then activates caspase-3 ultimating leading to apoptosis.