The ubiquitin-proteasome pathway is the principal mechanism for the degradation of short-lived proteins in eukaryotic cells. We demonstrated that treatment of THP-1 human monocytic leukemia cells with Z-LLL-CHO, a reversible proteasome inhibitor, induced cell death through an apoptotic pathway. Apoptosis in THP-1 cells induced by Z-LLL-CHO involved a cytochrome c-dependent pathway,which included the release of mitochondrial cytochrome c, activation of caspase-9 and -3, and cleavage of Bcl-2 into a shortened 22-kDa fragment. Induction of apoptosis by protease inhibitor also was detected in U937 and TF-1 leukemia cell lines and cells obtained from acute myelogenous leukemia patients but not in normal human blood monocytes. Treatment of human blood monocytes with Z-LLL-CHO did not induce apoptosis or Bcl-2 cleavage in these cells that rarely proliferate. Interestingly, when THP-1 cells were induced to undergo monocytic differentiation by bryostatin 1, a naturally occurring protein kinase C activator, they were no longer susceptible to apoptosis induced by Z-LLL-CHO. Bryostatin 1-induced differentiation of THP-1 cells was associated with growth arrest,acquisition of adherent capacity, and expression of membrane markers characteristic of blood monocytes. Likewise, differentiated THP-1 cells were refractory to Z-LLL-CHO-induced cytochrome crelease, caspase activation, and Bcl-2 cleavage. Resistance to Z-LLL-CHO-induced apoptosis in differentiated THP-1 cells was not due to cell cycle arrest. These findings show that the action of proteasome inhibitors is mediated primarily through a cytochrome c-dependent pathway and induces apoptosis in leukemic cells that are not differentiated.
The ubiquitin-proteasome pathway is the major nonlysosomal tool in eukaryotic cells for the degradation of short-lived intracellular proteins for disposal via an ATP- and ubiquitin-dependent mechanism(1). In this pathway, specific proteins are marked for degradation by conjugation to multiple molecules of ubiquitin, which targets proteins for rapid hydrolysis by the 26S proteasome. The ubiquitin-proteasome pathway was initially regarded as a mechanism of destruction for old and damaged proteins. In recent years, however, it has become clear that proteolysis by the proteasome pathway is a crucial mechanism of regulation of many cellular processes, including cell cycle progression, gene expression, and cell differentiation. Known substrates of this pathway include mitotic and S-phase cyclins(2), p21waf1 (3),cyclin-dependent kinase inhibitor p27 (4), IκBα(5), Bax, Mdm2 (6), and transcriptional factors such as p53 (7, 8), Jun (9), and Fos(10). Alterations of proteasome function have been linked to cellular transformation by oncogenic viruses and immune escape, and correlated to poor prognosis in colon and breast cancer(11).
Furthermore, results from recent studies have suggested that the ubiquitin-proteasome pathway may be involved in the regulation of apoptosis (12). Shinohara et al.(13) showed that inhibition of the proteasome pathway can induce apoptosis in MOLT-4 cells by a p53-dependent mechanism. In contrast, Herrmann et al. (14) found that proteasome inhibitor-induced prostate carcinoma cell death is independent of functional Bcl-2 and p53. Drexler (15)reported that inhibition of proteasome function is associated with apoptosis in HL60 cells, primarily in the G1phase of the cell cycle. Kitagawa et al. (16)established that apoptosis of human glioma cells induced by proteasome inhibitors involves a mitochondria-independent mechanism. More recently, we showed that proteasome inhibitor-induced apoptosis in human M-07e leukemia cells is mediated through a caspase-3-dependent and Bcl-2-sensitive pathway (17). Induction of apoptosis by inhibition of the proteasome pathway appears to be cell cycle independent. There are examples where exposure of quiescent cells to proteasome inhibitors induces apoptosis (18, 19). However,there is evidence that proteasomes may be required for, or are protective against apoptosis under other conditions, such as growth factor withdrawal and ionizing irradiation (19, 20). Thus,the exact role of the ubiquitin-proteasome pathway in regulating apoptosis is far from clear.
During apoptosis, several effector proteases such as caspase-3 mediate the deliberate disassembly of the cell into apoptotic bodies(21). These downstream caspases are activated through proteolytic cleavage by either caspase-8 or caspase-9, two upstream initiator caspases. In the Fas pathway, the activation of caspase-8 involves the formation of a complex with the cytoplasmic death domain of TNF3receptor and its analogous receptors (22). In contrast,the activation of caspase-9 requires the participation of cytochrome c release from the mitochondria. In this pathway, caspase-9 is activated when complexed with extramitochondrial cytochrome c and apoptotic protease activating factor 1(23, 24). Both initiator caspases are responsible for the activation of caspase-3 and other downstream effectors during apoptosis. In addition to caspases, other controllers of apoptosis are the Bcl-2 family proteins, which function upstream of caspases by either promoting or suppressing their protease activities. Several lines of evidences show that Bcl-2 family proteins are involved in controlling the release of cytochrome c from the mitochondria to activate caspase-9 (25, 26, 27). Apoptosis,which is essential for normal cell differentiation and development, is under strict physiological regulations. Deregulation of this process can lead to various defects ranging from embryonic lethality to a high susceptibility of malignant diseases such as leukemia(28). Indeed, leukemia is believed to be caused by impaired apoptosis in hematopoietic cells, resulting in the accumulation of immature nonfunctional cells (29).
We have shown that treatment of THP-1 monocytic leukemia cells with Z-LLL-CHO, a reversible proteasome inhibitor, produced cell death through apoptotic pathways.4 Biochemical analysis showed that apoptosis of THP-1 cells induced by Z-LLL-CHO was associated with the activation of a caspase-3-like protease that cleaved Bcl-2 into a shortened 22-kDa fragment. However, Z-LLL-CHO did not induce apoptosis in normal human blood monocytes, which rarely proliferate. This finding led us to hypothesize that proteasome inhibitors specifically target leukemic cells that are not differentiated. We now report that when THP-1 cells were induced to undergo monocytic differentiation by bryo1(30), a naturally occurring PKC activator, they became refractory to the pro-apoptotic effect of proteasome inhibitors.
MATERIALS AND METHODS
Mouse anti-Bcl-2 monoclonal antibody (SC-509), anti-Bax (B9), and rabbit anti-c-fms were obtained from Santa Cruz Biotechnology, Inc.(Santa Cruz, CA). Monoclonal mouse anti-cytochrome cantibody and polyclonal rabbit anti-caspase-3 and caspase-9 antibodies were purchased from PharMingen Inc. (San Diego, CA). Polyclonal anti-Ron antibody was a gift from Dr. M. H. Wang (Department of Medicine, University of Colorado, Denver, CO). FCS and RPMI 1640 were products of Life Technologies (Grand Island, NY). Z-LLL-CHO, caspase inhibitors and protease inhibitors were obtained from Calbiochem (San Diego, CA). Other reagents were purchased from Sigma Co. (St. Louis, MO). bryo1 was prepared and purified from the marine bryozoan Bugula neritina in Dr. Pettit’s laboratory as described previously (31).
DNA Fragmentation Assay.
DNA was extracted according to the procedure of Miller et al. (32). THP-1 cell (1 × 107) genomic DNA was extracted by adding 3 ml of nuclei lysis buffer [10 mm Tris-Cl, 400 mm NaCl, and 2 mmNa2EDTA, (pH 8.2)], 0.2 ml of 10% SDS, and 0.5 ml proteinase K solution [1 mg proteinase K, 2 mm Na2EDTA (pH 8.2), and 1% SDS] and incubated at 37°C overnight. DNA was precipitated by adding 1 ml of 6 m NaCl and vortexed for 15 s. The supernatant was collected by spinning at 2500 rpm(1300 × g) for 15 min. Two volumes of 95%ethanol were added to the supernatant and gently mixed. DNA precipitate was removed with a plastic spatula and placed in a tube containing 200μl of Tris-EDTA buffer. DNA was allowed to dissolve at 37°C for 2 h; its absorbance was then determined by spectrophotometry. Fractionation of DNA by electrophoresis was performed on 1.2% agarose gel in 1× Tris-borate-EDTA buffer at a constant voltage of 40 V. The agarose gel was stained with ethidium bromide for visualization of DNA.
Acridine Orange Staining.
Cells (1 × 106/ml) were stained with acridine orange (10 μL of a 100 μg/ml solution into 100 μL of cell suspension) for 10 min at 37°C. Thereafter, cells were washed with cold 0.9% NaCl to remove excess stains. The cell pellet was left on ice until fluorescence microscopy was performed. A minimum of 100 cells per sample were counted under a fluorescence microscope. Cells with condensed chromatin and fragmented nuclei were counted as positive, whereas those with a normal chromatin pattern were counted as negative.
Flow Cytometric Analysis.
Phosphatidylserine on the plasma membranes of cells was stained with Annexin V-FITC (Alexis Biochemicals) according to the protocol provided by the manufacturer. Briefly, 2–5 × 105 cells/ml were washed in 1× PBS. The cells were resuspended in 198 μl of binding buffer [10 mmHEPES/NaOH (pH 7.4), 140 mm NaCl, 2.5 mmCaCl2; filtered in 0.2 μm pore filter] and 2μl of FITC-labeled annexin V (annexin V-FITC). The mixtures were then incubated in the dark for 10 min, after which the cells were washed once with PBS and resuspended in 195 μl of binding buffer and 2 μl of 100 μg/ml PI. Apoptotic cells were defined as FITC positive and PI negative. Flow cytometry was analyzed on FACScan (Becton Dickinson),and data analysis was done on PC-LYSYS v1.1.
Whole cell lysates in 2× SDS loading buffer were fractionated by 12%SDS-PAGE at 100 V until the dye front reached the bottom of the gel. The proteins were transferred onto 0.2 μm pore nylon membrane(NYTRAN) at 40 V for 45–60 min. The membranes were blocked with 5% nonfat milk and probed with anti-Bcl-2 monoclonal antibody at a 1:200 dilution, anti-Bax monoclonal antibody at a 1:500 dilution,anti-c-fms polyclonal antibody at a 1:100 dilution, anti-Ron polyclonal antibody at a 1:5000 dilution, anti-caspase-3 polyclonal antibody at a 1:2000 dilution, anti-caspase-9 polyclonal antibody at a 1:1000 dilution, or with anti-cytochrome c monoclonal antibody at 1:500 for 1 h at room temperature. After extensive washes with 1×Tris-borate-EDTA buffer, the blots were incubated with appropriate secondary antibodies conjugated with horseradish peroxidase (1:5000 dilutions) for 1 h at room temperature. The blots were washed three times in 1× Tris-buffered saline, and the protein complexes were detected using enhanced chemiluminescence detection reagents according to the manufacturer’s protocol (Amersham Life Science).
Cytosolic Fraction Isolation.
The procedure for the isolation of the cytosolic fraction was described previously (33). Briefly, cells (1 × 108) were washed in ice-cold PBS and spun at 300 × g for 5 min at 4°C. The cell pellet was resuspended in 1 ml of ice-cold buffer A [250 mm sucrose, 20 mm HEPES-KOH(pH 7.5), 10 mm KCl, 1.5 mmMgCl2, 1 mm sodium EDTA, 1 mm EGTA, 1 mm DTT, and 0.1 mm phenylmethylsulfonyl fluoride] with 1× protease inhibitor cocktail (Boehringer Mannheim). Cells were lysed by 20 strokes with a Dounce glass homogenizer (No. 7726) on ice. The lysate was centrifuged at 750 × g for 10 min at 4°C to remove nuclei and unbroken cells. The supernatant was removed and centrifuged at 15,000 × g for 15 min at 4°C to eliminate mitochondria. The resulting supernatant, the cytosolic fraction, was assayed for the protein concentration by the Bradford protein assay (Bio-Rad), and then boiled in 1× SDS sample loading buffer.
The caspase-9 assay kit was purchased from Medical and Biological Laboratories Co., LTD. (Nagoya, Japan), and the assay was performed exactly according to the manufacturer’s protocol. Briefly, THP-1 cells(5 × 106) were resuspended in 50μL of chilled cell lysis buffer and incubated on ice for 10 min. Cell debris was separated from the supernatant by centrifuging at 10,000 × g for 2 min. Equal volume of 2×reaction buffer and 5 μL of 2 mm LEHD-pNA substrate (100 μm final concentration) were added to the supernatant, and the mixtures were incubated at 37°C for another 2 h. Dilution buffer (500 μL) was added to the sample,and the absorbance at 405 nm was read. Enzyme activity was expressed as pmol/mg protein/min.
Isolation of Human Monocytes and Leukemia Cells.
WBCs from healthy volunteers and AML patients with FAB M3 histological classification were isolated by layering whole blood over Histopaque 1077 (Sigma) at a volume of 1:1, and then centrifuging at 2000 rpm(1200 × g) for 30 min. The white mononuclear layer was carefully removed with a Pasteur pipette, washed once with PBS, and resuspended in cold medium. Human monocytes were isolated by incubating mononuclear cells in RPMI 1640 supplemented with 10% FCS in tissue culture dishes for an additional 3 h at 37°C and 5%CO2, after which the nonadherent cells were removed. Over 90% of the adherent cells were identified as monocytes by morphology criteria. Each immunoblot lane contained 1 × 106 cells. Samples were obtained from the patients after informed consent for this study.
Cell Cycle Arrest.
THP-1 cells were cultured in 0.2% FCS for 48 h, then in serum-free medium for 48 h to arrest the cell cycle at the G1 phase. For G2-M arrest,cells were starved in 0.2% serum for 48 h and then treated with colchicine (2 × 10−5m), a microtubule inhibitor, for an additional 48 h. Growth arrest was confirmed using a[3H]thymidine uptake technique. Briefly, THP-1 cells (1 × 105/100 μl) cultured in 96-well plates in triplicate were labeled with 0.5 μCi of tritiated [3H]thymidine (6.7 Ci/mmol;NEN Life Science Products, Boston, MA) for 16 h. The cells were harvested onto fiberglass filters using an automated PHD cell harvester. The filters were dried and counted in 3 ml of scintillation fluid with a Beckman LS3801 scintillation counter.
Induction of Apoptosis in THP-1 Cells by Z-LLL-CHO.
THP-1 cell line was established from a patient with AML. The cells grew in culture as a suspension with a cell doubling time of ∼16 h. THP-1 cells can be induced to differentiate into monocytes/macrophages but not cells of other hematopoietic lineages and are considered as the leukemic “counterpart” of blood monocytes. Treatment of THP-1 cells with Z-LLL-CHO resulted in growth arrest and cell death through an apoptotic pathway. Apoptosis was confirmed by a DNA fragmentation assay in cells that had been treated with Z-LLL-CHO (Fig. 1,A). Under the light microscope, apoptotic cells,characterized by cytoplasmic vacuolation, membrane blebbing, and apoptotic bodies (34), could be seen at 12 h after Z-LLL-CHO treatment. Acridine orange staining showed nuclear fragmentation and condensed chromatin structures in >25% of the cells(Fig. 1,B). In contrast, <2% of apoptotic cells were detected in control cultures of THP-1 cells. Flow cytometry study with FITC-labeled annexin V also showed a dramatic increase of annexin V-binding activity in Z-LLL-CHO-treated THP-1 cells (Fig. 1 C). At 6 h, 29.14% of the treated cells were apoptotic, staining positive for annexin V and negative for PI,compared with 2.08% in control cultures without treatment. At 24 h, only 9.97% of the cells were identified as apoptotic because of procession of apoptotic cells to death, resulting in positive staining for both PI and annexin V-FITC.
Z-LLL-CHO-induced Bcl-2 Cleavage and Caspase Activation.
To better understand the role of Bcl-2 in the process, we investigated the status of Bcl-2 during Z-LLL-CHO-induced apoptosis in THP-1 cells. Apoptosis of THP-1 cells induced by Z-LLL-CHO was clearly associated with the cleavage of Bcl-2 into a shortened 22-kDa fragment in a time-and dose-dependent manner (Fig. 2 and B). The cleavage of Bcl-2 was detected at 12 h after the addition of the proteasome inhibitor. At the highest dose (80 μm) used in this study,Z-LLL-CHO treatment induced >25% cleavage of total cellular Bcl-2 as estimated from the intensity of the Bcl-2 bands in immunoblots. In contrast, no Bax cleavage was noticed in Z-LLL-CHO-treated cells although the levels of Bax appeared to be significantly reduced during apoptosis (Fig. 2 C). Because of alternative splicing, Bax was detected as a doublet by the anti-Bax antibody obtained from commercial sources (Santa Cruz).
We next asked whether Z-LLL-CHO-induced apoptosis and Bcl-2 cleavage were mediated through a caspase-3-dependent pathway in THP-1 cells. Caspase-3 activation was monitored by cleavage from a 32-kDa precursor to a 17-kDa active fragment, using immunoblot analyses. As shown in Fig. 2,D, the activation of caspase-3 was detected as early as 6 h after Z-LLL-CHO treatment. Caspase-3 activation also was confirmed by a colorimetric assay using Ac-DEVD-pNA as the substrate for caspase-3 (data not shown). To establish that Bcl-2 was cleaved by activated caspase-3, we treated the cells with a highly specific caspase-3 inhibitor, DEVD-CHO, prior to the addition of Z-LLL-CHO. In the presence of DEVD-CHO, the cleavage of Bcl-2 induced by Z-LLL-CHO treatment was inhibited (Fig. 2 E). In contrast, cleavage of Bcl-2 was not inhibited by the caspase-1 inhibitor YVAD-CHO.
Z-LLL-CHO Induced the Release of Mitochondrial Cytochrome c and Caspase-9 Activation in THP-1 Cells.
The preceding experiments showed that Z-LLL-CHO-induced apoptosis in THP-1 cells was associated with Bcl-2 cleavage by activated caspase-3. Because Bcl-2 has been implicated in the regulation of cytochrome c release from the mitochondria to the cytoplasm, we asked whether Z-LLL-CHO-induced apoptosis in THP-1 cells involved a cytochrome c-dependent pathway. THP-1 cells were treated with Z-LLL-CHO for various time periods. The levels of extramitochondrial cytochrome c were determined using an immunoblot analysis with anti-cytochrome c antibody. Treatment with Z-LLL-CHO readily induced the release of cytochrome c into the cytosolic fraction as early as 4 h (Fig. 3,A). The release of cytochrome c was correlated with a transient activation of caspase-9 as indicated by cleavage of procaspase-9 (p48) into a shortened active p37 fragment in Z-LLL-CHO-treated cells (Fig. 3,B). In parallel experiments,the activation of caspase-9 protease activity was detected using a colorimetric assay with a specific substrate, LEHD-pNA (Fig. 3,C). To show that cytochrome c release is related to the activation of caspases, we pretreated THP-1 cells with a general caspase inhibitor, Boc-D-fmk, prior to the addition of Z-LLL-CHO. As shown in Fig. 3 and E, pretreatment of THP-1 cells with the general caspase inhibitor reduced the amount of cell death and cytochrome c release induced by Z-LLL-CHO. Additional support of the activation of caspase-3 and -9 and Bcl-2 cleavage induced by Z-LLL-CHO treatment was shown in two other cell lines, TF-1, a GM-CSF-dependent eurythrocytic leukemia cell line, and U937, a human promonocytic cell line. Fig. 4 shows that when treated with Z-LLL-CHO, Bcl-2 was cleaved in both leukemia cell lines undergoing apoptosis. Likewise, the cleavage of Bcl-2 was associated with the activation of both caspase-3 and caspase-9.
Human Peripheral Blood Monocytes Were Refractory to Z-LLL-CHO-induced Apoptosis.
To further analyze the mechanism whereby Z-LLL-CHO induces apoptosis,we examined the effect of Z-LLL-CHO on normal human blood monocytes and leukemia cells obtained from four AML patients. Compared with THP-1 cells and human leukemia cells, normal monocytes were highly resistant to Z-LLL-CHO-induced apoptosis (Fig. 5) from three normal volunteers. Immunoblot analysis showed that Z-LLL-CHO induced distinct Bcl-2 cleavage in both THP-1 cells and leukemic cells obtained from patients but not normal blood monocytes(Fig. 6). A very slight amount of cleaved Bcl-2 fragment was noticed in blood monocyte samples that had been treated with the highest dose of Z-LLL-CHO (100 μm) for an extended period (3 days).
Z-LLL-CHO Induced Apoptosis in Leukemic Cells but not Differentiated Cells.
The preceding study shows that Z-LLL-CHO did not induce apoptosis in normal blood monocytes, raising the possibility that proteasome inhibitors may target specifically on leukemic cells that are not differentiated. To test this hypothesis, we induced THP-1 cells to undergo monocytic differentiation by bryo1, a naturally occurring PKC activator. Treatment with bryo1 for 12 h resulted in growth arrest of THP-1 cells and induced a major fraction (60%) of them to become adherent with distinct monocyte/macrophage differentiation markers(35), which included the expression of Ron and c-fms. Ron is a receptor for the human macrophage-stimulating protein, MSP, which regulates the motility and shape change of mature macrophages. The product of the c-fms proto-oncogene is the receptor for macrophage colony-stimulating factor (Fig. 7 and B). The expression of these two receptors increases markedly during macrophage differentiation. Bryo1 treatment,however, neither significantly affected the levels of Bcl-2, Bax, and caspase-3 proteins from the same samples (Fig. 7, C–E), nor did it produce DNA fragmentation in THP-1 cells (Fig. 1 A, Lane 4).
Unlike control THP-1 cells, treatment of differentiated THP-1 cells with Z-LLL-CHO did not induce Bcl-2 cleavage (Fig. 8,A). The lack of Bcl-2 cleavage was associated with the failure of Z-LLL-CHO to activate caspase-3 in differentiated THP-1 cells (Fig. 8,B). Because we have shown previously that the action of Z-LLL-CHO was mediated through a cytochrome c-dependent pathway, we asked whether bryo1 treatment affects the release of cytochrome c in Z-LLL-CHO-treated cells. As shown in Fig. 8 and D, the release of cytochrome c and caspase-9 activation triggered by Z-LLL-CHO treatment were inhibited when THP-1 cells were induced to undergo monocytic differentiation by bryo1. Differentiated THP-1 cells became resistant to Z-LLL-CHO-induced apoptosis and also excluded trypan blue. Prolonged treatment (>48 h) of THP-1 cells with bryo1 eventually resulted in death but not apoptosis. Moreover, the induction of apoptosis by Z-LLL-CHO was also dramatically inhibited with prior bryo1 treatment (Fig. 8 E).
To rule out the possibility that cell cycle arrest following bryo1 treatment was responsible for resistance to Z-LLL-CHO-induced apoptosis, we treated THP-1 cells with colchicine, which arrests the cells in G2-M phase. Growth arrest in THP-1 cells was confirmed by reduced thymidine uptake in both cases. As shown in Fig. 9, quiescent cells arrested in G2-M by colchicine were still responsive to Z-LLL-CHO-induced Bcl-2 cleavage and apoptotic cell death. In another experiment, THP-1 cells were cultured in serum-free medium for 48 h, which arrests cells in the G1 phase. Likewise, nonproliferating cells arrested in the G1 phase remained responsive to Z-LLL-CHO-induced apoptosis. These findings excluded the possibility that cells in cell cycle arrest were not responsive for resistance to Z-LLL-CHO-induced apoptosis in differentiated THP-1 cells.
A number of recent reports showed that inhibition of proteasome pathway induced apoptosis in various leukemic and nonhematological tumor cell lines (13, 14, 16, 17, 36). Here, we show that proteasome inhibitor-induced apoptosis in human leukemia THP-1 cells involves a cytochrome c-dependent pathway. The release of cytochrome c was accompanied by a transient activation of both caspase-9 and caspase-3 and apoptosis in Z-LLL-CHO-treated THP-1 cells. The release of cytochrome c and induction of cell death could be inhibited by the addition of a general caspase inhibitor. These results indicated that cytochrome c release is correlated with the activation of caspases and proteasome inhibitor-induced death of THP-1 cells proceeds in a caspase-dependent manner. In contrast to our finding, however, a recent study reported that proteasome inhibitor induces mitochondria-independent apoptosis in human glioma cells (16), raising the possibility that the execution of apoptosis induced by proteasome inhibitors is likely to be mediated, depending on the cell types and cellular factors, through several mechanisms, including both cytochrome c-dependent and -independent pathways.
In addition to THP-1 cells, Z-LLL-CHO induced apoptosis in leukemia cells obtained from AML patients. Two samples of AML were taken from patients who had undergone chemotherapy. Another two samples were taken from newly diagnosed leukemic patients before treatments. Despite the differences in samples, upon Z-LLL-CHO treatment, they invariably underwent apoptosis with distinct Bcl-2 cleavage. Serendipitously, we found that normal monocytes were highly resistant to Z-LLL-CHO-induced apoptosis compared with either THP-1 cells or leukemic cells from the patients. These observations led us to ask whether differentiated cells were less sensitive to proteasome inhibitor-induced apoptosis. To test this hypothesis, we induced THP-1 cells to undergo monocytic differentiation by bryo1, a naturally occurring PKC activator. Induction of monocytic differentiation in THP-1 cells was associated with growth arrest, acquisition of adherent capacity, and expression of some membrane markers characteristic of blood monocytes. As expected,we found that differentiated THP-1 cells were no longer susceptible to Z-LLL-CHO-induced cytochrome c release, caspase activation,and Bcl-2 cleavage, and thus were refractory to apoptosis. These results established that proteasome inhibitors specifically target leukemia cells that are not differentiated. Similar resistance to apoptotic agents after differentiation of U937 and THP-1 cells was reported recently (37, 38).
The mechanisms by which proteasome inhibitors induce apoptosis in leukemic cells are not known. Accumulation of short-lived proteins that are critical for cell proliferation and cell cycle regulation appears to be linked to Z-LLL-CHO-mediated apoptosis in THP-1 cells. For example, a recent study by Kitagawa et al. (16)reported that proteasome inhibitor-induced apoptosis in human glioma cells was associated with the up-regulation of short-lived proteins,including p21Waf1, Mdm2, and p27Kip1. Furthermore, Manna and Aggarwal(39) reported that degradation of IkB, also a substrate of the proteasome degradation pathway, was accompanied by suppression of TNF-mediated apoptosis in human U937 cells. The relevance of these regulators in mediating apoptosis also is illustrated in our finding that THP-1 cells differentiated by bryo1 are highly resistant to Z-LLL-CHO-induced apoptosis, as described in this study. It has been shown that bryo1 treatment activates PKC and induced nuclear factor-κB activation in a number of leukemic and tumor cell lines (40, 41). In addition, tumor cells that constitutively express nuclear factor-κB were “resistant” to the apoptotic effects of TNF and a number of other apoptotic agents(42, 43). However, it should be pointed out that proteasome inhibitors, in addition to inducing apoptosis, have been reported to prevent apoptosis in sympathetic neurons upon deprivation of nerve growth factor (19) and in thymocytes treated with ionizing radiation, glucocorticoids, or phorbol ester(20), illustrating the complex nature of the proteasome systems in regulating apoptosis.
The biological significance of Bcl-2 cleavage is a matter of speculation. In addition to THP-1, we detected Bcl-2 cleavage and apoptosis in TF-1 and U937 human leukemia cells after treatment with Z-LLL-CHO. Bcl-2 cleavage also has been reported in HL-60 and other leukemic cells induced to undergo apoptosis, as reported previously (44). The observation that Bcl-2 was cleaved by activated caspase-3 was demonstrated using caspase-3 inhibitor. Cleavage of Bcl-2 may represent a means to effectively destroy and remove the antiapoptotic effect of Bcl-2. With Z-LLL-CHO treatment, the amount of Bcl-2 did not seem to decrease, although Bcl-2 was being cleaved, which indicated that Bcl-2 might be a proteasome substrate. However, the 22-kDa Bcl-2 fragment appears to be stable inside cells and becomes even more hydrophobic because of the loss of its hydrophilic NH2 terminus, suggesting that it may be functional. Relevant to this study, a recent work showed that the cleaved Bcl-2 fragment adapted a Bax-like activity (45). Furthermore, Bax has been implicated in the promotion of cytochrome c release by forming membrane pores on the mitochondria(46). The widespread occurrences of Bcl-2 cleavage seem to suggest that the cleaved fragment may have a feedback role in further promoting the release of cytochrome and apoptosis in THP-1 cells. In support of this view, we showed in this study that the acquisition of resistance to apoptosis in differentiated THP-1 cells was correlated with inhibition of Bcl-2 cleavage and cytochrome c release.
In a previous study, Lopes et al. (18)described a wild-type p53-dependent induction of apoptosis by proteasome inhibitors. In our hands, Z-LLL-CHO-induced apoptosis in THP-1 cells appeared to involve a p53-independent mechanism because THP-1 cells possess mutated inactive p53 (47). Furthermore, cell cycle analysis of Z-LLL-CHO-treated THP-1 cells did not show changes in the percentage of cells in various phases (data not shown), and THP-1 cells arrested in G1 and G2-M by serum starvation and colchicine were still responsive to Z-LLL-CHO-induced apoptosis (Fig. 9). Therefore, we believe that the action of proteasome inhibitors is not dependent on G1 and G2 cell cycle arrest, which is under the control of p53 (48). The relation of proteasome inhibitor-induced apoptosis to cell cycle progression status remains to be clarified. The notion that differentiation, not growth arrest, in bryo1-treated THP-1 and blood monocytes is the determinant of their refractoriness to apoptosis was also corroborated by a recent study by Drexler (15), who showed that differentiated HL-60 cells had reduced sensitivity toward proteasome inhibition-induced cell death.
Lactacystin, an irreversible proteasome inhibitor isolated from a Streptomyces metabolite, has been used to induce apoptosis in B-CLL-3 cells obtained from leukemia patients (49). Therefore, modulation of the function of proteasomes may be therapeutically advantageous in the treatment of cancers. The role of proteasomes in normal and tumor cells could provide a rational basis for the use of proteasome-targeting drugs. Our finding that proteasome inhibitors specifically target leukemic cells but not differentiated normal cells is significant because a major complication associated with chemotherapy is marrow cytotoxicity. The selective killing of nondifferentiating leukemic cells may provide a means for the purging of leukemic cells from the peripheral bloodstream in patients undergoing autologous bone marrow transplantation. Given that the differentiated cells are highly resistant to proteasome inhibitors, our data also suggest possible adverse effects of using these agents, and perhaps other inducers of apoptosis, in combination with differentiation therapy for leukemia. Clearly, the effects of proteasome inhibitors on the induction of apoptosis in leukemic cells deserve further study.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by Public Health Service Grant CA 73212 (to B. D. C.) and Outstanding Investigator Grant CA 44344-10 (to G. R. P.) awarded by the National Cancer Institute,Department of Health and Human Services, and the Wayne State University Graduate Research Assistantship (GRA) Award.
The abbreviations used are: TNF, tumor necrosis factor; bryo1, bryostatin 1; PKC, protein kinase C; PI,propidium iodide; AML, acute myelocytic leukemia; GM-CSF, granulocyte macrophage colony-stimulating factor.
C. Chen, unpublished observation.