Molecular Targets for Therapy (MTT)

Cobalt chloride and low oxygen tension trigger differentiation of acute myeloid leukemic cells: possible mediation of hypoxia-inducible factor-1α


Cellular and systemic O2 concentrations are tightly regulated to maintain delicate oxygen homeostasis. Although the roles of hypoxia in solid tumors have been widely studied, few studies were reported regarding the possible effects of hypoxia on leukemic cells. Here, we showed for the first time that low concentrations of cobalt chloride (CoCl2), a hypoxia-mimicking agent, and 2–3% O2 triggered differentiation of various subtypes of human acute myeloid leukemic (AML) cell lines, including NB4, U937 and Kasumi-1 cells, respectively, from M3, M5 and M2b-type AML, but CoCl2 did not modulate AML subtype-specific fusion proteins promyelocytic leukemia-retinoic acid receptor alpha (PML-RARα) and AML1-ETO. Treatment with CoCl2 also induced primary leukemic cells from some AML patients to undergo differentiation. Similar to what occurs in solid tumor cells, CoCl2-mimicked hypoxia also increased the level of hypoxia-inducible factor (HIF)-1α protein and its DNA-binding activity in leukemic cells. The CoCl2 induction of HIF-1α protein and its DNA-binding activity were inhibited by 3-morpholinosydnonimine, which also blocked CoCl2-induced cell differentiation in leukemic cells. These results provide an insight into a possible link of hypoxia or HIF-1α and leukemic cell differentiation, and are possibly of significance to explore clinical potentials of hypoxia or hypoxia-mimicking agents and novel target-based drugs for differentiation therapy of leukemia


Acute promyelocytic leukemia (APL) is a particular subtype of acute myeloid leukemia (AML) characterized with the specific reciprocal chromosome translocation t(15;17) that results in the expression of leukemia-promoting promyelocytic leukemia-retinoic acid receptor alpha (PML-RARα) chimeric protein.1,2,3 Following the successful clinical application of all-trans retinoic acid (ATRA) in the treatment of most APL patients by differentiation induction,4,5 arsenic trioxide (ATO) is also added recently to the treatment arsenal for APL patients who are refractory to ATRA and conventional chemotherapeutic drugs.6,7 However, the exact mechanism of ATO actions still remains elusive. In vitro studies suggest that high concentrations (1–2 μ M) of ATO induce apoptosis, which is independent of cancer cell types, while low concentrations (0.1–0.5 μ M) of ATO seem to be capable of triggering partial differentiation of APL cells after a prolonged treatment.8,9,10 However, the in vitro differentiation-inducing ability of ATO does not appear to be better than its in vivo activity. Thus, it is reasonable to speculate that some factors in bone marrow (BM) microenvironments could modulate the in vivo activity of ATO.

Cellular and systemic O2 concentrations are tightly regulated via short- and long-acting response pathways that maintain delicate oxygen homeostasis, an important organizing principle for human development and physiology. At the center of these signal pathways is hypoxia-inducible factor 1(HIF-1), a heterodimeric transcription factor consisting of a hypoxia-sensitive α-subunit (HIF-1α) and a constitutively expressing β subunit (HIF-1β, also known as the aryl hydrocarbon receptor nuclear translocator, ARNT).11 It has been well demonstrated that hypoxia is an important selective force in the clonal evolution of solid tumors mainly through HIF-1-drived production of vascular endothelial growth factor (VEGF) and angiogenesis.12,13,14 Although leukemic cells do not form a well-circumscribed ‘mass’ in BM like solid tumors, oxygen levels of BM in AML patients may be decreased as a result of fast growth of leukemic cells. The lack of oxygen in BM is possibly further aggravated with anemia that often accompanies newly diagnosed AML patients. In fact, Jensen et al15 have reported that hypoxic fraction of leukemic cells and normal cells inoculated in the Brown Norwegian rat increase significantly in the BM. On the other hand, leukemic cells are cultured in vitro at ambient oxygen (21%) in most circumstances, while in vivo cells are physiologically exposed under much lower oxygen levels ranging from 16% in pulmonary alveoli to less than 6% in most peripheral organs of the body.11 Based on these understanding, we tried to figure out the possible effect of ATO on APL cells under hypoxia. Unexpectedly, we found that cobalt chloride (CoCl2) mimicked hypoxia or moderate hypoxia (2 and 3% O2) treatment alone could trigger leukemic cells to undergo differentiation to a large extent in an AML subtype-independent manner. This differentiation was accompanied by increased protein level and DNA-binding activity of HIF-1α protein. In contrast, nitric oxide (NO) donor 3-morpholinosydnonimine abrogated HIF-1α accumulation and antagonized CoCl2-induced differentiation of leukemic cells, suggesting a potential link between HIF-1α and leukemic cell differentiation. These novel findings are possibly of significance to explore clinical potentials of hypoxia or hypoxia-mimicking agents and novel target-based drugs for differentiation therapy of leukemia.

Materials and methods


Powdery CoCl2 with purity of 99% was purchased from Shanghai Chemical Agent Company (Shanghai, China) and its 50 mM stock solution as well as 10 mM 3-morpholinosydnonimine (Sigma, St Louis, MI, USA) were freshly prepared in distilled water and phosphate-buffered saline, respectively. Other reagents including ATO and ATRA (Sigma, St Louis, MI, USA) were prepared as described before.8,9,10

Cell lines and cell culture

Three AML cell lines were used for the current studies. NB4, a t(15;17)-positive APL (M3-type) cell line, was kindly provided by Dr M Lanotte's in France;16 Kasumi-1, a t(8;21)-positive M2b-type AML cell line, was obtained from Dr Kamada in Japan17 and U937, an acute monocytic leukemic (M5-type) cell line, was obtained from Cell Bank of Shanghai Institutes of Biologic Sciecnes (Shanghai, China). Fresh (primary) leukemic cells were obtained from the BM of AML patients, which were diagnosed by morphological, immunophenotypic and cytogenetic criteria. The BM from the patients was drawn to a Ficoll's solution and the cells were separated by centrifugation. An informed consent was obtained from all subjects. All cell lines and primary cells were cultured in RPMI-1640 medium (Sigma, St Louis, MI, USA) supplemented with 10% fetal calf serum (Gibco BRL, Gaithersburg, MD, USA). For CoCl2 treatment, cells were seeded into six-well plates at the initial density of 2 × 105 cells/ml and maintained in a 5% CO2–95% air humidified atmosphere at 37°C. To determine the effects of low oxygen tension in leukemic cells, cells were cultured in a specially designed hypoxia incubator (Thermo Electron, Forma, MA, USA) and a specific O2 concentration was generated by flushing an N2/5% CO2 mixture into the incubator. The cell viability was determined by trypan-blue exclusion assays.8

Determination for cell differentiation antigens, cell cycle and apoptosis

Cell morphological features were examined by microscope after Wright's staining of cells that were collected onto slides by cytospin (Shandon, Runcorn, UK). The percentages of mature cells in total cells were estimated in more than 200 cells of at least 20 fields of vision. Cell surface differentiation antigens CD11b, CD11c, CD14, CD15 and CD33 were measured using fluorescein isothyiocyanate (FITC)- or PE-labeled antibodies via flow cytometry (Beckmon-Coulter, Miami, FL, USA).10 Becton Dickinson Simultest Control r1/r2α was used as a negative control. The distribution of nuclear DNA contents was analyzed by flow cytometry as described previously.10 Apoptosis was also evaluated by the annexin-V assay using the ApoAlert Annexin-V kit from Clontech (Palo Alto, CA, USA) following the manufacturer's instruction.

Western blot analysis

Cells were dissolved in the lysis buffer (100 mM Tris-HCl, pH 6.8, 4% SDS, 20% glycerine, 200 mM dithiothreitol) and were quantified for proteins by DC protein assay kit II (Bio-Rad, Hercules, CA, USA). Equal amounts (20–50 μg) of protein extracts were fractionated on a 8–10% SDS-polyacrylamide gel, and the fractionated proteins were then transferred to a nitrocellulose membrane. Ponceau S staining was performed on the membrane to further make sure that equal amounts of proteins were loaded. The blot was blocked in 5% fat-free milk and then incubated with a polyclonal anti-human RARα antibody (a gift from Dr Chambon of France) and an anti-human HIF-1α antibody (H-206, Santa Cruz, CA, USA) in a 5% fat-free milk solution. The immunocomplex was visualized by using a chemiluminescent kit (Amersham, Buckinghamshire, UK).

Immunofluorescent analysis

Immunofluorescence staining of the N-terminal region of PML was performed as described previously.8,10 For immunofluorescent analysis of HIF-1α protein, AML cells were treated with or without 50 μ M CoCl2 for 2 days, cells centrifuged onto slides by cytospin (Shandon, Runcorn, UK) were fixed with cold acetone and stained with the mouse anti-HIF-1α monoclonal antibody (Becton Dickinson Biosciences, Palo Alto, CA, USA) and then with the FITC-conjugated goat anti-mouse IgG (Santa Cruz, CA, USA). Fluorescence signals were examined by fluorescent microscope (Olympus BX-51, Olympus Optical, Japan).

Preparation of nuclear extracts and electrophoretic mobility shift assay (EMSA)

Harvested cells (1 × 107) were resuspended in 500 μl buffer A (10 mM HEPES, pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM dithiothreitol, 0.2% NP-40) and kept on ice for 10 min. After centrifugation (500 g, 2 min), cell pellets were then resuspended in 400 μl buffer A followed by centrifugation at 4°C (2000 g, 1 min). Finally, the pellets were resuspended in 150 μl ice-cold buffer B (200 mM HEPES, pH 7.9, 0.42 M NaCl, 25% glycerol, 0.2 mM EDTA, 0.5 mM dithiothreitol, 100 μg/ml phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin and 10 μg/ml leupeptin) and kept on ice for 20 min. After centrifugation (12 000 g, 10 min), the supernatant was collected for EMSA. Protein concentration was determined by using the DC Protein Assay kit 22 (Bio-Rad, Hercules, CA, USA). The wild-type oligonucleotide, 5′-IndexTermGCCCTACGTGCTGTCTCA-3′, from the 3′-flanking region of human erythropoietin gene encompassing the consensus sequence (coding sequence) for the binding of HIF-1 (hypoxia-responsive elememt, HRE, shown in bold) and a mutant HRE, 5′-IndexTermGCCCTAAAAGCTGTCTCA-3′ (mutated bases are underlined),18 and their equal molar concentrations of complementary oligonucleotides were end labeled with [γ-32P]ATP using T4 polynucleotide kinase (Promega, Madison, WI, USA). The labeled probes were purified using the MicroSpinTM G-25 columns (Amersham, Piscataway, USA). Binding reactions were performed at 4°C for 40 min in a total volume of 20 μl containing 4 μl of a 5 × gel shift binding buffer (Promega, Madison, WI, USA), 10 μg of nuclear protein extracts and 1 μl (0.2 pmol) of the labeled oligonucleotide probe or plus 200-fold molar excess of the unlabeled wild-type HRE oligonucleotide. The DNA–protein complexes were separated in 4.5% nondenaturing polyacrylamide gel by electrophoresis in 0.5 × TBE buffer at room temperature. After electrophoresis, the gel was dried and analyzed on a PhosphorImager (Molecular Dynamics, Sunnyvale, CA, USA).

HIF-1α cDNA sequencing

Total RNAs were extracted from NB4 and U937 cells by TRIzol reagent (Gibco BRL, Gaithersburg, MD, USA) and reverse transcriptions were performed by TaKaRa RNA PCR kit (Takaka, Dalian, China) following the manufacturer's instruction. For cDNA sequencing, we used two pairs of primers to amplify HIF-1α full-length cDNA (GenBank accession no. NM_001530) as follows: primers 5′-IndexTermATTCACCATGGAGGGCGC-3′ (forward) and 5′-TGGGTAGGAGATGGAGATGC-3′ (reverse) amplified 1–1265 cDNA fragments of human HIF-1α, and primers 5′-IndexTermGATGCTTTAACTTTGCTGGC-3′ (forward) and 5′-IndexTermTCAGTTAACTTGATCCAAAGCTC-3′ (reverse) amplified 1173–2481 cDNA fragments of human HIF-1α. PCR amplification was performed for 35 cycles with denaturing at 95°C for 30 s, annealing at 55°C for 40 s and extension at 72°C for 1 min (except for last cycle with extension at 72°C for 10 mins) using a GeneAmp PCR System 9600 (Perkin-Elmer Norwalk, USA). Amplified products were subcloned into pGEM-T Easy vector (Promega, Madison, WI). DNA sequencing was carried out by ABI377 DNA Sequencer (Perkin-Elmer, Boston, MA, USA).


CoCl2 at low concentrations inhibited cell growth without apoptosis induction in NB4 cells

To determine possible effects of CoCl2, a commonly used hypoxia-mimicking agent,19 on cell growth and survival, NB4 cells were treated with various concentrations (12.5, 25, 50, 100 and 200 μ M) of CoCl2 for 1–6 days. After treatment, viable cells and cell viability were quantified with Trypan-blue exclusion assay. As shown in Figure 1a and b, high concentrations (especially 200 μ M) of CoCl2 significantly inhibited NB4 cell growth with reduced viability. Treatment with CoCl2 at 200 μ M for 3 days resulting in the death of almost all the cells. However, at concentrations 12.5–50 μ M, CoCl2 significantly inhibited NB4 cell growth in a concentration- and time-dependent manner without the significant reduction of cell viability (Figure 1b) and apoptosis induction as evidenced by the absence of sub-G1 (Figure 1c)- and apoptosis-specific annexin-V-positive cells (Figure 1d), which lasted for up to 6 days (data not shown). Of note, low concentrations of CoCl2 did not modulate cell-cycle distribution (Figure 1c), although they inhibited cell growth, suggesting that they prolonged cell-doubling time rather than induced cell-cycle arrest.

Figure 1

Effects of CoCl2 on growth, cell-cycle and apoptosis of NB4 cells. NB4 cells were treated by the indicated concentrations of CoCl2. Viable cell number (a) and viability (b), histogramic DNA content-related cell-cycle distribution (c) and annexin-V-positive cells (d) were measured for the indicated days (a, b) or 3 days (c, d). The percentages of cells in the different cell-cycle phases and annexin-V-positive cells were indicated in the corresponding site. Independent experiments were repeated at least 10 times and similar results were obtained. The value indicated means of triplicate with less than 15% of variance.

Low concentrations of CoCl2 triggered morphological and functional differentiation without modulation of PML-RARα fusion protein in NB4 cells

Following the treatment of CoCl2 at 25 μ M and 50 μ M, NB4 cells exhibited morphological changes related to differentiation such as condensed chromatin, a decreased nuclei/cytoplasm ratio with smaller nuclei, although nucleoli remained visible (Figure 2a). More importantly, CoCl2 also induced the expression of differentiation-related antigens CD11 (CD11b and CD11c), but not CD14, CD15 and CD33, as shown in Figure 2b and c. Of note, the enhanced expression of CD11 began to appear at day 2 and appeared greatly significant at day 6 with the treatment of CoCl2 at a concentration of 50 μM, indicating that low concentrations of CoCl2 could modify granulocytic differentiation in NB4 cells.

Figure 2

Nontoxic concentrations of CoCl2 induced differentiation of NB4 cells. (a) NB4 cells were treated with or without 25 and 50 μ M CoCl2 or 50 μ M CoCl2 plus 500 μ M 3-morpholinosydnonimine (SIN-1) for 6 days. Cell morphological features were observed, where values represented the percentage of mature cells from triplicate. (b and c) NB4 cells were treated by CoCl2 for the indicated days (b) or 6 days (c), the percentages of CD11b-positive cells (b) as well as indicated CD-positive cells (c) were measured on flow cytometry. Independent experiments were repeated at least 10 times and similar results were obtained. Each value indicated x±s.d. of triplicate in an independent experiment.

It is well documented that ATRA or ATO can cleave and/or degrade PML-RARα protein, which in turn results in the differentiation and/or apoptosis of APL cells.1,8,9,10 Thus, whether CoCl2 also modulated PML-RARα protein in a manner similar to ATRA and ATO was investigated. Interestingly, unlike ATRA and ATO, CoCl2 at 50 μ M failed to modify the subcellular localization of PML-RARα protein in NB4 cells as analyzed by immunofluorescent analysis (data not shown). Western blots against anti-RARα antibody showed that CoCl2 (50 μ M) treatment for 24–72 h had no effects on PML-RARα protein level, while ATRA (0.1 μ M) and ATO (0.5 μ M) reduced and almost completely decreased the PML-RARα protein, respectively (Figure 3).

Figure 3

Effects of CoCl2 and/or ATRA/ATO on PML-RARα protein in NB4 cells by Western blot analysis. NB4 cells were treated with 50 μ M CoCl2 and/or 0.5 μ M ATO (a)/0.1 μ M ATRA (b) treatment for the indicated hours, Western blots were performed with equal amounts of protein loading.

CoCl2 also triggered cell differentiation in other subtypes of AML cells

The fact that CoCl2 induced NB4 cell differentiation without the alteration of PML-RARα protein prompted us to address whether CoCl2 also induced differentiation of other leukemic cells. To this end, two additional leukemic cell lines, U937 and Kasumi-1, were studied. Consistent with the observation in NB4 cells, treatment with CoCl2 at 25 and 50 μ M did induce differentiation-related morphological changes and the expression of granulocytic mature-related antigens CD11b and CD11c in U937 cells (Figure 4) and in Kasumi-1 cells (Figure 5). CoCl2 also inhibited cell growth without cell apoptosis in these cells (data not shown). Moreover, CoCl2 at 50 μ M did not modulate the expression of M2b-related fusion protein AML1-ETO in Kasumi-1 cells as revealed by Western blot analysis (Figure 5c).

Figure 4

CoCl2 triggered differentiation in U937 cells. (a) Cell morphology of U937 cells treated with 25 and 50 μ M CoCl2 for 6 days. The values indicated x±s.d. of mature cells % of triplicates. (b) U937 cells were treated with the indicated concentrations of CoCl2 for 6 days, CD11b-positive cells were analyzed on flow cytometry. For the CD11b measurement, left curves and right curves, respectively, represented Simultest Control and CD11b-positive cells. (c) U937 cells were treated with 50 μ M CoCl2 for 6 days, the indicated CD-positive cells % were analyzed. Independent experiments were repeated 5–7 times with similar results and values indicated x±s.d. of triplicate from an independent experiment.

Figure 5

CoCl2 triggered differentiation in Kasumi-1 cells without alterations of AML1-ETO protein. (a and b) Kasumi-1 cells were treated with 50 μ M CoCl2. Cell morphology at day 6 (a) and CD11b-positive cells % at the indicated days (b) were measured. Each value indicated x±s.d. of triplicate in one out of 5–7 times of independent experiments with similar results. Of note, whether U937 cells or Kasumi-1 cells kept up to 95% of viability under 25–50 μ M CoCl2. (c) Western blot against anti-ETO antibody was performed on Kasumi-1 cells treated or untreated with 50 μ M CoCl2 for 2 days.

Low oxygen tension induced AML cell differentiation

To determine whether the actual low oxygen tension can induce cell differentiation of leukemic cells, U937 cells were incubated under 2, 3, 5 and 21% O2 concentrations. Compared with cells incubated at 21% O2, cell growth was inhibited at 2–5% O2 environment (Figure 6a, left), while they remained viable (viability >90%) at these O2 concentrations even after 9-day incubation (data not shown). As predicted, incubation under 2 and 3% O2, but not under 5% O2, rendered U937 cells to undergo differentiation as evidenced by cell morphological changes and elevated CD11b expression (Figure 6a, right and 6b). This effect began to appear at day 3 of treatment. After 9-day incubation with 2% O2, U937 cells displayed morphological features of terminal differentiation such as decreased cell size, condensed chromatin, reduced nuclei/cytoplasm ratio and disappearance of nucleoli. Some cells even exhibited horseshoe-shaped nuclei (Figure 6c). The similar differentiation-related changes could also be seen in NB4 and Kasumi-1 cells (Figure 6d and data not shown).

Figure 6

Moderate hypoxia induced differentiation in U937 cells and NB4 cells. U937 cells were incubated under 2, 3, 5 and 21% O2 for 6 days (a) or 6 and 9 days (b), growth inhibition % (a, left) with up to 90% viable and CD11b-positive cells % (a, right and b) were measured. Morphologic alterations of U937 cells (c) and NB4 cells (d) under 2% O2 for the indicated days (c) and 6 days (d) were observed, where values indicated mature cells %. Independent experiments were repeated up to five times with similar results and each value represented x±s.d. of triplicate samples of an independent experiment.

CoCl2 increased HIF-1α protein and its DNA-binding activity in leukemic cells

We next asked whether differentiation of leukemic cells induced by hypoxia-mimicking agent involved HIF-1 protein. After confirming that NB4 and U937 cells carried normal HIF-1α gene by cDNA sequencing (data not shown), we examined the possible effects of nontoxic concentration (50 μ M) of CoCl2 on HIF-1α protein. As shown in Figure 7a, 50 μ M CoCl2 rapidly increased HIF-1α protein levels in NB4, U937 and Kasumi-1 cells without a significant alteration of HIF-1α mRNA, as confirmed with semiquantitative PCR (data not shown). Similarly, immunofluorescent analysis also revealed that a small fraction of untreated NB4 and U937 cells had weak HIF-1α staining mainly in the cytosol. However, NB4 cells treated with 50 μ M CoCl2 showed enhanced HIF-1α signals that were present not only in the cytosol but also in the nucleus (Figure 8a).

Figure 7

CoCl2 increased HIF-1α protein level in leukemic cell lines that was antagonized by SIN-1. (a) U937 cells, Kasumi-1 cells and NB4 cells were treated with or without 50 μ M CoCl2 for the indicated hours, HIF-1α proteins were measured by Western blots. (b) NB4 cells were treated with 50 μ M CoCl2 and/or 500 μ M 3-morpholinosydnonimine (SIN-1) for the indicated days, viable cell numbers were counted. Over 90% of cells kept viable in all the time points. (c) NB4 cells were treated with or without 50 μ M CoCl2 and/or 500 μ M SIN-1 for 48 h, HIF-1α proteins were measured by Western blots. (d) NB4 cells were treated with 50 μ M CoCl2 and/or 500 μ M SIN-1 for the indicated time, CD11b-positive cells % was measured on flow cytometry. For all data, independent experiments were repeated 3–5 times with similar results and each value indicated x±s.d. of triplicate in an independent experiment.

Figure 8

Immunofluorescent staining for HIF-1α protein and EMSA assay for DNA-binding activity of HIF-1α. (a) NB4 cells were treated with or without 50 μ M CoCl2 and/or 500 μ M SIN-1 for 2 days, immunofluorescent staining against HIF-1α antibody was performed. (b) NB4 cells were treated with or without 50 μ M CoCl2 and/or 500 μ M SIN-1 for 48 h, EMSA was performed on nuclear extracts. Lane 1: without nuclear extract with isotope-labeled wild-type HRE; lane 2: nuclear extract from untreated NB4 cells with isotope-labeled wild-type HRE; lane 3: nuclear extract from 50 μ M CoCl2-treated NB4 cells with isotope-labeled wild-type HRE; lane 4: nuclear extract from 500 μ M SIN-1-treated NB4 cells with isotope-labeled wild-type HRE; lane 5: nuclear extract from 50 μ M CoCl2 plus 500 μ M SIN-1-treated NB4 cells with isotope-labeled wild-type HRE; lane 6: nuclear extract from 50 μ M CoCl2-treated NB4 cells with isotope-labeled mutant HRE and lane 7: nuclear extract from 50 μ M CoCl2-treated NB4 cells with isotope-labeled wild-type HRE plus 200-fold cold (unlabeled) HRE. The arrow points to HIF-1α/HRE complex.

To rule out whether 50 μ M CoCl2 increased HRE-binding activity in NB4 cells, EMSAs were performed using their nuclear extracts. We observed that untreated NB4 cells exhibited a weak HRE-binding activity (Figure 8b, lane 2) that was significantly enhanced after 50 μ M CoCl2 treatment for 48 h (Figure 8b, lane 3). No DNA-binding activities were detected in the nuclear extracts of 50 μ M CoCl2-treated NB4 cells when the mutant HRE probe was used (Figure 8b, lane 6). In addition, 200-fold molar excess of unlabeled wild-type HRE probe completely abolished the binding activity in the treated nuclear extracts of NB4 cells (Figure 8b, lane 7), supporting the specificity of HIF-1 binding in this assay.

3-Morpholinosydnonimine inhibited HIF-1α accumulation and HRE-binding activity and abrogated differentiation of leukemic cells induced by CoCl2

NB4 cells were treated with 500 μ M 3-morpholinosydnonimine (SIN-1) that has been reported to modulate HIF-1α in cultured Hep3B cells.20 The results showed that SIN-1 at 500 μ M inhibited cell proliferation to a much greater extent than did CoCl2 at 50 μ M, while SIN-1 treatment alone or in combination with 50 μ M CoCl2 did not induce cell death (Figure 7b). Indeed, 500 μ M SIN-1 almost completely inhibited accumulation of HIF-1α protein in NB4 cells, as confirmed by both Western blotting (Figure 7c) and immunofluorescent staining (Figure 8a). Consistently, this NO donor also antagonized CoCl2-induced HRE-binding activity of HIF-1α in these cells (Figure 8b, lane 5). More intriguingly, SIN-1 treatment alone did not induce, but significantly antagonized, CoCl2-induced leukemic cell differentiation in terms of morphology (Figure 2a) and CD11b expression (Figure 7d). The inhibiting effects of SIN-1 on CoCl2-induced HIF-1α accumulation and differentiation could also be shown in U937 cells (data not shown).

Hypoxia mimicking also induced differentiation of some fresh AML cells

We further addressed the effect of CoCl2-mimicked hypoxia on fresh leukemic cells from four cases of APL (AML-M3), four cases of AML-M2, two cases of AML-M4 and one case of AML-M5 patients. All APL and M2b cases, respectively, carried t(15;17) and t(8;21), while no defective chromosomal abnormalities were observed in other three cases according to conventional cytogenetic analysis. A total of 50 μ M CoCl2 did not accelerate death of these fresh cells under in vitro culture conditions (data not shown). Its treatment alone did not effectively induce differentiation of fresh APL cells of all four cases, although it enhanced ATRA action in two cases (M3-2 and M3-3, Figure 9a). However, 50 μ M CoCl2 did induce differentiation to a large extent in four cases, including three AML-M2 cases (M2-1, M2-2 and M2-4 in Figure 9b) and one AML-M4 case (M4-1 in Figure 9c). Mature-related morphological alterations were also observed in these cells after treatment with 50 μ M CoCl2 for 6 days. Figure 9d is a representative image from case M2-1.

Figure 9

Effects of 50 μ M CoCl2 on the differentiation of primary cells from AML patients. Primary cells from APL/M3-type (a), M2b-type (b), M4- and M5-type AML (c) patients were treated with or without 50 μ M CoCl2 for 6 days, and CD11b-positive cells were measured. Each value was representative of means of triplicate with less than 10% of variance. (d) A representative image for morphological changes of 50 μ M CoCl2-treated fresh AML-M2b cells at day 6.


Previously, Ivanovic et al21 reported that severe hypoxia (0.9–1% O2) favors the self-renewal of murine and human hematopoietic stem cells. Hypoxia has also been reported to modify proliferation and differentiation of CD34+ CML cells,22 although CML-associated oncogene BCR/ABL was shown to induce HIF-1α gene expression. However, to date, there are few reports in literatures that deal with the potential effect of hypoxia on AML cells. In the current study, we have shown for the first time that both hypoxia-mimicking agent CoCl2 and 2 and 3% O2 triggered differentiation of AML cells in a subtype-independent manner. Furthermore, CoCl2 neither induced the degradation of APL-specific PML-RARα fusion protein nor modulated M2b-related AML1-ETO fusion protein. At this moment, we could not rule out the possibility that these fusion proteins may intervene differentiation signals induced by hypoxia due to the following reasons: first, U937 cells having no fusion proteins were the most sensitive, while Kasumi-1 cells with AML1-ETO expression exhibited a weaker response to CoCl2-induced differentiation. Secondly, hypoxia could enhance differentiation-inducing effects of ATRA and ATO in APL cells (unpublished data), both of which are capable of destructing PML-RARα protein.9,10 Finally, inducible overexpression of AML1-ETO gene in U937 cells could over-ride CoCl2-induced differentiation of U937 cells (Jiang Y et al, unpublished data).

As described above, HIF-1 has been regarded as a master regulator for cellular responses to hypoxia.11,23 Under normoxic conditions, von Hippel Lindau tumor suppressor protein recognizes the hydroxylation of the conserved proline residues (mainly proline 564) of HIF-1α through a recently identified oxygen- and iron-dependent mammalian prolyl hydroxylase,24 which tags HIF-1α for rapid degradation by the proteasomal pathway.25,26 Under hypoxia, hydroxylation of HIF-1α is diminished, leading to HIF-1α stabilization and translocation to the nucleus, where it binds to ARNT and activates the transcription of a number of target genes through binding to HRE. As a hypoxia-mimicking agent, CoCl2 has been shown to inhibit proline hydroxylation of HIF-1α protein, thus stabilizing the protein in solid tumor cells.19 Indeed, we have demonstrated that CoCl2 also rapidly increased HIF-1α protein rather than mRNA levels as well as increased HRE-binding activity in leukemic cells, which were inhibited by 3-morpholinosydnonimine. The latter also antagonized hypoxia-induced differentiation of leukemic cells. These combined results imply a potential link between HIF-1α and leukemic cell differentiation, although further studies through overexpression of the gene by transfection or silencing of HIF-1α by RNA interference are needed for clarification of their relationship.

The observation that hypoxia favors leukemic cell differentiation also led to the consideration that the anomaly of hypoxia-related signal molecules possibly contributed to differentiation block of leukemic cells. Indeed, increased levels of leukocyte-associated VEGF and neovascularization of the BM, which might disrupt low oxygen tension, accelerate the progression of AML and adversely affect their prognosis, whereas VEGF antagonist and soluble neuropilin-1, prolongs the survival in a systemic leukemia mice model.27 Furthermore, a t(1;12)(q21;p13) translocation in an AML-M2 results in the expression of a fusion gene between TEL/ETV6 and essentially all of ARNT, a constitutive subunit for HIF-1α activities.28 The breakpoint of ARNT loci is also found in 11 out of 36 patients with various hematopoietic disorders.29 In this work, we also address possible effects of CoCl2 on primary AML cells. Indeed, CoCl2 is capable of inducing apparent differentiation in leukemic cells from some but not all AML patients. Surprisingly, CoCl2-induced differentiation did not occur in primary t(15;17)-carrying APL cells tested in this work, which was inconsistent to that observed in NB4 cells. It remains to be elucidated whether leukemic cells from these patients resistant to hypoxia-induced differentiation presented abnormal hypoxia-related signal molecules.

Finally, most of the current cancer chemotherapies are highly cytotoxic and often nonspecific. Based on the tacit assumption that malignant cells, especially leukemic cells, exhibit reversible defects in their differentiation process, a potentially less cytotoxic therapeutic strategy for cancer known as ‘differentiation therapy’ is being developed. It employs drugs to induce cancer cells to undergo terminal differentiation, thus preventing their further proliferation.30 In the past years, some novel differentiation-inducing agents have been explored such as histone deacetylase inhibitors and novel retinoids.31 However, the successful model of differentiation therapy is still limited to the application of ATRA and maybe ATO in APL. Therefore, it has become increasingly necessary to develop novel mechanism-based differentiation therapy for other leukemia subsets and even solid tumors. Our current studies will help to explore new target-based drugs for differentiation therapy and provide possible clinical potential for hypoxia or hypoxia-mimicking agents in the treatment of AML. Actually, oral CoCl2 intake had been administered for the treatment of anemic patients undergoing long-term hemodialysis 30 years ago.32


  1. 1

    Melnick A, Licht JD . Deconstructing a disease: RARalpha, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia. Blood 1999; 93: 3167–3215.

  2. 2

    Minucci S, Monestiroli S, Giavara S, Ronzoni S, Marchesi F, Insinga A et al. PML-RAR induces promyelocytic leukemias with high efficiency following retroviral gene transfer into purified murine hematopoietic progenitors. Blood 2002; 100: 2985–2989.

  3. 3

    Kogan SC, Brown DE, Shultz DB, Truong BT, Lallemand-Breitenbach V, Guillemin MC et al. BCL-2 cooperates with promyelocytic leukemia retinoic acid receptor alpha chimeric protein (PMLRARalpha) to block neutrophil differentiation and initiate acute leukemia. J Exp Med 2001; 193: 531–543.

  4. 4

    Lo Coco F, Nervi C, Avvisati G, Mandelli F . Acute promyelocytic leukemia: a curable disease. Leukemia 1998; 12: 1866–1880.

  5. 5

    Tallman MS, Nabhan C, Feusner JH, Rowe JM . Acute promyelocytic leukemia: evolving therapeutic strategies. Blood 2002; 9: 759–767.

  6. 6

    Shen ZX, Chen GQ, Ni JH, Li XS, Xiong SM, Qiu QY et al. Use of arsenic trioxide in the treatment of acute promyelocytic leukemia: II. Clinical efficacy and pharmacokinetics in relapsed patients. Blood 1997; 89: 3354–3364.

  7. 7

    Soignet SL, Maslak P, Wang ZG, Jhanwar S, Calleja E, Dardashti LJ et al. Complete remission after treatment of acute promyelocytic leukemia with arsenic trioxide. N Engl J Med 1998; 339: 1341–1348.

  8. 8

    Chen GQ, Shi XG, Tang W, Xiong SM, Zhu J, Cai X et al. Use of arsenic trioxide (As2O3) in the treatment of acute promyelocytic leukemia (APL): I. As2O3 exerts dose-dependent dual effects on APL cells. Blood 1997; 89: 3053–3345.

  9. 9

    Zhu J, Chen Z, Lallemand-Breitenbach V, de The H . How acute promyelocytic leukaemia revived arsenic. Nat Rev Cancer 2002; 2: 705–713.

  10. 10

    Cai X, Shen YL, Zhu Q, Jia PM, Yu Y, Zhou L et al. Arsenic trioxide-induced apoptosis and differentiation are associated respectively with mitochondrial transmembrane potential collapse and retinoic acid signaling pathways in acute promyelocytic leukemia. Leukemia 2000; 14: 262–270.

  11. 11

    Semenza GL . Hypoxia-inducible factor 1: oxygen homeostasis and disease pathophysiology. Trends Mol Med 2001; 7: 345–350.

  12. 12

    Graeber TG, Osmanian C, Jacks T, Housman DE, Koch CJ, Lowe SW et al. Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours. Nature 1996; 379: 88–91.

  13. 13

    Zhong H, De Marzo AM, Laughner E, Lim M, Hilton DA, Zagzag D et al. Overexpression of hypoxia-inducible factor 1alpha in common human cancers and their metastases. Cancer Res 1999; 59: 5830–5835.

  14. 14

    Jiang BH, Agani F, Passaniti A, Semenza GL . V-SRC induces expression of hypoxia-inducible factor 1 (HIF-1) and transcription of genes encoding vascular endothelial growth factor and enolase 1: involvement of HIF-1 in tumor progression. Cancer Res 1997; 57: 5328–5835.

  15. 15

    Jensen PO, Mortensen BT, Hodgkiss RJ, Iversen PO, Christensen IJ, Helledie N et al. Increased cellular hypoxia and reduced proliferation of both normal and leukaemic cells during progression of acute myeloid leukaemia in rats. Cell Prolif 2000; 33: 381–395.

  16. 16

    Roussel MJ, Lanotte M . Maturation sensitive and resistant t(15;17) NB4 cell lines as tools for APL physiopathology: nomenclature of cells and repertory of their known genetic alterations and phenotypes. Oncogene 2001; 20: 7287–7291.

  17. 17

    Asou H, Tashiro S, Hamamoto K, Otsuji A, Kita K, Kamada N . Establishment of a human acute myeloid leukemia cell line (Kasumi-1) with 8;21 chromosome translocation. Blood 1991; 77: 2031–2036.

  18. 18

    Wang GL, Semenza GL . Characterization of hypoxia-inducible factor 1 and regulation of DNA binding activity by hypoxia. J Biol Chem 1993; 268: 21513–21518.

  19. 19

    Piret JP, Mottet D, Raes M, Michiels C . CoCl2, a chemical inducer of hypoxia-inducible factor-1, and hypoxia reduce apoptotic cell death in hepatoma cell line HepG2. Ann NY Acad Sci 2002; 973: 443–447.

  20. 20

    Sogawa K, Numayama-Tsuruta K, Ema M, Abe M, Abe H, Fujii-Kuriyama Y . Inhibition of hypoxia-inducible factor 1 activity by nitric oxide donors in hypoxia. Proc Natl Acad Sci USA 1998; 95: 7368–7373.

  21. 21

    Ivanovic Z, Dello Sbarba P, Trimoreau F, Faucher JL, Praloran V . Primitive human HPCs are better maintained and expanded in vitro at 1 percent oxygen than at 20 percent. Transfusion 2000; 40: 1482–1488.

  22. 22

    Desplat V, Faucher JL, Mahon FX, Dello Sbarba P, Praloran V, Ivanovic Z . Hypoxia modifies proliferation and differentiation of CD34(+) CML cells. Stem Cells 2002; 20: 347–354.

  23. 23

    Harris AL . Hypoxia – a key regulatory factor in tumour growth. Nat Rev Cancer 2002; 2: 38–47.

  24. 24

    Ivan M, Haberberger T, Gervasi DC, Michelson KS, Gunzler V, Kondo K et al. Biochemical purification and pharmacological inhibition of a mammalian prolyl hydroxylase acting on hypoxia-inducible factor. Proc Natl Acad Sci USA 2002; 99: 13459–13464.

  25. 25

    Hon WC, Wilson MI, Harlos K, Claridge TD, Schofield CJ, Pugh CW et al. Structural basis for the recognition of hydroxyproline in HIF-1 alpha by pVHL. Nature 2002; 417: 975–978.

  26. 26

    Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M et al. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 2001; 292: 464–468.

  27. 27

    Schuch G, Machluf M, Bartsch Jr G, Nomi M, Richard H, Atala A et al. In vivo administration of vascular endothelial growth factor (VEGF) and its antagonist, soluble neuropilin-1, predicts a role of VEGF in the progression of acute myeloid leukemia in vivo. Blood 2002; 100: 4622–4628.

  28. 28

    Salomon-Nguyen F, Della-Valle V, Mauchauffe M, Busson-Le Coniat M, Ghysdael J, Berger R et al. The t(1;12)(q21;p13) translocation of human acute myeloblastic leukemia results in a TEL-ARNT fusion. Proc Natl Acad Sci USA 2000; 97: 6757–6762.

  29. 29

    Busson-Le Coniat M, Salomon-Nguyen F, Dastugue N, Maarek O, Lafage-Pochitaloff M, Mozziconacci MJ et al. Fluorescence in situ hybridization analysis of chromosome 1 abnormalities in hematopoietic disorders: rearrangements of DNA satellite II and new recurrent translocations. Leukemia 1999; 13: 1975–1981.

  30. 30

    Leszczyniecka M, Roberts T, Dent P, Grant S, Fisher PB . Differentiation therapy of human cancer: basic science and clinical applications. Pharmacol Ther 2001; 90: 105–156.

  31. 31

    Zwiebel JA . New agents for acute myelogenous leukemia. Leukemia 2000; 14: 488–490.

  32. 32

    Bowie EA, Hurley PJ . Cobalt chloride in the treatment of refractory anaemia in patients undergoing long-term haemodialysis. Aust NZ J Med 1975; 5: 306–314.

Download references


This work was supported in part by National Key Program (973) for Basic Research of China (NO2002CB512806), Key Science and Technology Development Project (CGQ) and Natural Science Foundation (HY) of Shanghai, 100-Talent Program of Chinese Academy of Sciences (CGQ).

Author information

Correspondence to G-Q Chen.

Rights and permissions

Reprints and Permissions

About this article


  • hypoxia
  • cobalt chloride
  • differentiation

Further reading