[Frontiers in Bioscience 2, d232-241, June 1, 1997]

Table of Conents
 Previous Section   Next Section


Steven Grant.

Division of hermatology , Department of Pharmacology and Microbiology , Medical College of Virginia

Received 5/23/97; Accepted 5/28/97


5.1. Biochemical interactions

Initial studies examining the interaction between bryostatin 1 and ara-C were prompted by the discoveries that both bryostatin 1 (54) and ara-C (58) are highly inhibitory to leukemic cell self-renewal capacity, a biological determinant that correlates closely with clinical outcome in patients with acute leukemic (59). In addition, an accumulating body of evidence indicates that administration of DNA-damaging drugs in conjunction with differentiation-inducing agents often leads, in a sequence-dependent manner, to potentiation of cell death (60-63). Consequently, it is reasonable to propose that the combination of these agents might exhibit even greater antileukemic activity.

Initial empiric studies employing primary cultures of human AML cells did in fact demonstrate that pretreatment of cells with bryostatin 1 (10 nM; 24 hr), with or without the hematopoietic growth factor rGM-CSF, followed by ara-C resulted in a marked reduction in primary colony formation (L-CFU) (54). In addition, such combinations virtually abrogated the self-renewal ability of these cells (PE-2 capacity). Interestingly, the same drug exposures were considerably less inhibitory toward the in vitro growth of normal committed (CFU-GM) and early (HPP-CFU) hematopoietic progenitor cells (54). This suggests that the antileukemic selectivity of bryostatin 1, administered as a single agent, may also apply to its modulatory effects toward cytoxic drugs.

Subsequent studies were conducted to determine if bryostatin 1 might increase ara-C-related inhibitory effects by potentiating drug metabolism. In primary leukemic blast samples, bryostatin 1 exerted a heterogeneous effect on ara-CTP formation, although significant increases in ara-C activation were noted in a subset of samples (64). In the human leukemic cell line HL-60, pretreatment of cells with 10 nM bryostatin 1 for 24 hr increased the formation of ara-CTP after a 6-hr exposure to 10 µM ara-C, but only when cells were cultured under plateau phase conditions (cell density 106 cells/ml) (65). In contrast, under log phase conditions (cell density 5 x 105 cells/ml), no potentiation was observed. A possible explanation for these findings is that cell cycle-related changes in plateau phase cells (i.e., decrease in the S-phase fraction) or exhaustion of nutrients leads to reduction in activity of deoxycytidine kinase activity, the enzyme catalyzing the rate-limiting step in ara-C metabolism. Under these circumstances, adminstration of bryostatin 1 may provide a signal that prevents the accompanying inhibition of ara-C phosphorylation. It may also be relevant that the initial phosphorylation step in ara-C metabolism has been found to be dependent upon PKC, at least under in vitro conditions (18).

5.2 Signalling interactions

Further insights into the mechanism by which bryostatin 1 might modulate cellular susceptibility to ara-C arose from studies in logarithmically growing HL-60 cells. It was found that in such cells, pretreatment with 10 nM bryostatin 1 for 24 hr (and, to a lesser extent, phorbol dibutyrate) significantly increased the degree of DNA fragmentation and induction of apoptotic morphologic changes resulting from a 6-hr exposure to 10 µM ara-C (66,67). These effects correlated with the extent of PKC down-regulation, suggesting that one of the mechanisms by which such agents might potentiate drug-mediated lethality involves overcoming the protective effect of PKC. Support for this concept was provided by the observations that (1) acute exposure to bryostatin 1, which was associated with an increase in PKC activity, opposed ara-C-mediated apoptosis; and (2) increases in PKC activity induced by membrane-permeable synthetic diacylglycerides (e.g., diC8) or phospholipase C opposed ara-C-related apoptosis (67). Additional evidence that PKC opposes apoptotic events is provided by the observations that tumor-promoting phorbol esters (e.g., PMA) oppose drug-induced apoptosis in hematopoietic cells (68), and that PKC inhibitors induce leukemic cell apoptosis (69) and faciliate drug-induced cell death (70). Thus, the ability of ara-C to trigger leukemic cell apoptosis may be enhanced by two separate but related mechanisms: direct inhibition by PKC inhibitors, and down-regulation by chronic exposure to PKC activators such as bryostatin 1. Although the mechanism by which PKC activity promotes cell survival is unknown, it may be related to perturbations in the level of activity of opposing cell signalling pathways. For example, exposure of cells to ionizing radiation, TNF, and chemotherapeutic agents leads to activation of stress-activated protein kinases (SAPK/JNK; p46/p54)(71), whereas survival signals are associated with activation of extracellular receptor kinase-related pathways (EFK/MAPK/p42/p44) (72). PKC activity has recently been linked to activation of the latter pathway through the generation of sphingosine-1-phosphate (73). In view of evidence that cellular susceptibility to apoptosis may depend upon the net balance between stress- versus survival-related signalling pathways (74), it is conceivable that a reduction in PKC activity by bryostatin 1 favors a typical stress response after ara-C exposure.

5.3. Differentiation and apoptosis

In leukemia cells, a complex relationship exists between differentiation and apoptosis. Following exposure to inducers of maturation such as retinoic acid and PMA, leukemic cells undergo differentiation and ultimately die an apoptotic death, primarily as a late event. (75,76). On the other hand, induction of leukemic cell maturation (e.g., by PMA) can reduce their susceptibility to induction of apoptosis by agents such as VP-16 (77). In addition, leukemic cells exhibiting dysregulation of PKC respond to PMA by undergoing apoptosis rather than maturation (78). Based upon such evidence, it has been postulated that apoptosis represents an alternative pathway for cells unable to engage a normal differentiation program (79).

These considerations have implications for attempts to understand the interaction between bryostatin 1 and ara-C. For example, in HL-60 cells immune to bryostatin 1's differentiation-inducing actions, pretreatment with bryostatin 1 potentiates ara-C-induced apoptosis (65). It has recently been shown that co-administration of A23187 with bryostatin 1 partially restores the ability of HL-60 cells to undergo differentiation (80). Interestingly, when A23187 was combined with bryostatin 1, it abrogated bryostatin 1-related potentiation of ara-C-mediated apoptosis. Co-administration of the organotellurium compound AS101 also enhanced induction of HL-60 cell differentiation by bryostatin 1, although its effects were weaker than those of A23187 (81). The combination of bryostatin 1 and AS101 also failed to potentiate ara-C-induced apoptosis. These findings suggest that a reciprocal relationship exists between bryostatin 1-induced maturation and potentiation of apoptosis, in that interventions that restore bryostatin 1's capacity to induce differentiation prevent potentiation of cell death. It is important to note that while A23187 and AS101 reduced the ability of bryostatin 1 to augment ara-C-induced apoptosis, they resulted in a further decline in clonogenicity (80,81). This suggests that differentiation and apoptosis represent alternative pathways by which leukemic cell self-renewal capacity may be inhibited.

5.4.Sequence-dependent interactions

It has been shown by several investigators that exposure of leukemic cells to DNA damaging agents followed by a differentiation-inducing stimulus leads to potentiation of cell death (82-84). The mechanism underlying this phenomenon is unknown, but has been postulated to stem from differentiation-related inhibition of DNA repair (84). Studies in an HL-60 cell line essentially immune to bryostatin 1's differentiation-inducing actions demonstrated that exposure of cells to bryostatin 1 after ara-C failed to enhance apoptosis (65). In contrast, in the monocytic leukemic cell line U937, which is weakly differentiation-responsive to bryostatin 1 (52), pretreatment with bryostatin 1 failed to increase ara-C-related apoptosis, whereas exposure to ara-C followed by bryostatin 1 did lead to a significant increase in cell death (85). The sequence-dependent nature of the bryostatin 1-ara-C interactions raises the possibility that factors other than, or in addition to, biochemical perturbations (e.g., cell cycle-related events) may contribute to the observed enhancement of apoptosis.

5.5. Role of c-Myc

The proto-oncogene c-Myc, encodes a protein (c-MYC) intimately involved in cell cycle progression and differentiation (86). A recent study has identified the phosphatase cdc25 as a potentially important down-stream target of c-myc (87). Bryostatin 1 induces c-myc down-regulation in human leukemic cells susceptible to maturation induction (50,88), whereas it fails to reduce c-myc expression in cells unresponsive to its differentiating actions (50,88). Furthermore, inappropriate expression of c-myc has been linked to induction of apoptosis under certain adverse conditions (e.g., growth factor deprivation) (89). These observations raise the possibility that the inability of bryostatin 1 to induce c-myc down-regulation may contribute to its capacity to promote ara-C-related apoptosis. To test this possibility, antisense oligodeoxyribonucleotides (AS-ODN) directed against c-myc were employed to reduce c-MYC expression in bryostatin 1-treated cells, and to determine what effect this might have on ara-C-related apoptosis. It was found that a reduction in levels of c-MYC had no effect on ara-C-related apoptosis itself, but abrogated the ability of bryostatin 1 to potentiate ara-C-induced cell death (90). These findings support the notion that the failure of bryostatin 1 to reduce expression of c-MYC contributes to its modulatory effects on ara-C-related apoptosis.

5.6.Involvement of Bcl-2

Bcl-2 represents a member of a family of proteins that reciprocally regulate cell death (91). Anti-apoptotic proteins include Bcl-2, Bcl-xL, Mcl-1, and A1; pro-apoptotic proteins include Bax, Bad, Bak, and Bcl-xs, among others (92). Bcl-2 homodimerizes with itself, and also forms heterodimers with Bax, which has the net effect of antagonzing Bax-induced cell death (93). Increased expression of Bcl-2 (and Bcl-xL) protects leukemic cells from a wide variety of cytotoxic agents, including ara-C (30). One of the actions of these anti-apoptotic proteins appears to be to prevent activation of the protease cascade involved in the regulation of apoptosis (94). Chief among these is the Yama protease (CPP32), whose substrates include nuclear lamin and poly(ADP-ribosyl) polymerase(95). Thus, overexpression of Bcl-2 in human leukemia cells (HL-60) has been shown to prevent ara-C-mediated Yama activation and PARP degradation (96). It is noteworthy that the mitochondrial protein cytochrome c serves as a cofactor for Yama activation (97), and that Bcl-2 blocks translocation of cytochrome c from mitochondria to cytosol in response to various apoptotic stimuli (98). In view of similarities in structure between Bcl-2/Bcl-xL and certain bacterial pore-forming proteins (99), it is tempting to postulate that Bcl-2 and Bcl-xL act to block mitochondrial channels permitting cytosolic redistribution of cytochrome c, thereby preventing activation of the apoptotic protease cascade.

In hematopoietic cells, expression of Bcl-2 is linked to differentiation state. For example, as myeloid hematopoietic cells mature, expression of Bcl-2 declines (100). This phenomenon may account for apoptosis occurring in leukemic cells undergoing maturation in response to differentiation-inducing agents such as retinoic acid or dexamethasone (101,102). In fact, potentiation of ara-C-mediated apoptosis in human myeloid leukemia cells by differentiation-inducing stimuli has been attributed to Bcl-2 down-regulation (102). In addition, induction of maturation in lymphoid leukemia cells by bryostatin 1 has also been reported to result in decreased expression of Bcl-2 (103). In the human promyelocytic leukemia (HL-60) cell however, bryostatin 1 does not down-regulate Bcl-2 expression, at least over an initial 24-hr period, nor does it increase expression of Bax (104). Instead, bryostatin 1 up-regulates expression of the anti-apoptotic protein, Mcl-1. This suggests that the ability of bryostatin 1 to facilitate ara-C-related apoptosis in these cells involves factors other than a reduction in expression of the Bcl-2 protein.

Recently, phosphorylation of Bcl-2 and related proteins (e.g., BAD) has been proposed as a post-translational mechanism by which cell death may be regulated (105). For example, exposure of cells to taxol and other microtubule-active agents has been associated with Bcl-2 phosphorylation (106), and this event has been postulated to contribute to the induction of cell death (107). In the case of prostate cancer cells, taxol-associated phosphorylation of Bcl-2 is accompanied by reduced heterodimerization with Bax, providing a possible mechanism by which cell death may be facilitated (108). Interestingly, bryostatin 1 has been shown to induce phosphorylation of Bcl-2 in 32D hemtoapoieitic cells, although this phenomenon was associated with resistance to growth factor-induced apoptosis (109). In contrast, treatment with bryostatin 1 effectively restores the ability of ara-C to induce apoptosis in HL-60/Bcl-2 overexpressing cells to wild-type levels, an event associated with Bcl-2 phosphorylation (110). In addition, bryostatin 1 administration prevented Bcl-2 from blocking ara-C-mediated CPP32 activation and PARP cleavage. The differential effect of bryostatin 1 on apoptosis in these two systems may stem from (1) divergent effects in normal versus neoplastic cells; (2) different cell death-inducing stimuli (e.g., growth factor deprivation versus exposure to a cytotoxic agent); (3) phosphorylation site-specific events; or a combination of these factors. In any event, these findings suggest that post-translational modifications of Bcl-2 induced by bryostatin 1 may contribute to the ability of this agent to potentiate drug-induced apoptosis in leukemia.