Activation of protein kinase A (PKA) by 8-Cl-cAMP as a novel approach for antileukaemic therapy

Activation of PKA by cAMP agonists, such as 8-Cl-cAMP activation, selectively causes rapid apoptosis in v-abl transformed fibroblasts by inhibiting the Raf-1 kinase. Here we investigated whether 8-Cl-cAMP is useful for the treatment of chronic myelogenous leukaemia (CML), which is hallmarked by the expression of the p210bcr/abl oncogene. Autologous bone marrow transplantation is a feasible alternative for patients with no suitable donor, but hampered by the risk of relapse due to the persistence of leukaemia cells in the transplant. To study the effects of 8-Cl-cAMP on primary leukaemic cells, bone marrow cells (BMCs) from eight CML patients (one at diagnosis, three in chronic and four in accelerated phase) were treated. Ex vivo treatment of BMCs obtained in chronic phase of CML with 100 μM 8-Cl-cAMP for 24–48 h led to the selective purging of Philadelphia Chromosome (Ph1 chromosome) without toxic side effects on BMCs from healthy donors as measured by colony-forming unit (CFU) assays. BMCs from patients in accelerated phase showed selective, but incomplete elimination of Ph1 chromosome positive colony forming cells. The mechanism of 8-Cl-cAMP was investigated in FDCP-mix cells transformed by p210bcr/abl, a cell culture model for CML. The results showed that 8-Cl-cAMP reduced DNA synthesis and viability independent of Raf inhibition as Raf inhibitors had no effect. MEK inhibitors interfered with DNA synthesis, but not with viability. In summary, our results indicate that 8-Cl-cAMP could be useful to purge malignant cells from the bone marrow of patients with CML and certain other forms of leukaemias.

The activation of signalling pathways due to the constitutive expression of the Bcr-Abl oncogene plays a major role in the pathogenesis of some leukaemias, in particular chronic myelogenous leukaemia (CML). CML is the major adulthood leukaemia characterised by the Philadelphia chromosome (Ph1 chromosome, t:9/22), a chromosomal translocation where bcr sequences from chromosome 22 are juxtaposed to c-abl on chromosome 9 leading to the expression of an atypical fusion-protein p210 bcr-abl (Butturini et al, 1996). In its chronic phase CML is hallmarked by abnormally sustained cell survival rather than excessive proliferation and is relatively well controlled by cytoreductive chemotherapy. However, the chronic phase inevitably turns into an acute phase of blast crisis where leukaemic blast cells proliferate rapidly and aggressively with fatal consequences. p210 bcr-abl is a constitutively activated tyrosine kinase that activates numerous cellular signalling pathways including the Raf -MEK -ERK pathway, which is critical for malignant transforma-tion. A tyrosine kinase inhibitor for p210 bcr-abl , CGP-57148 now called STI571, has sparked great interest as it dramatically increased the number of patients achieving complete remission. However, almost half of the patients treated in the chronic phase remain Ph1 chromosome positive with the inherent risk of relapse. Indeed, many patients treated with STI571 in the acute phase relapse rapidly (La Rosee et al, 2002). This probably relates to the fact that STI571 is inhibiting proliferation rather than eliminating the leukaemic cells (Beran et al, 1998;La Rosee et al, 2002). Despite the addition of STI571 to the clinical arsenal bone marrow and peripheral blood stem cell transplantation (PBSCT) remains a mainstay of therapy. For autologous PBSCT stem cells are harvested from peripheral blood after stimulation with G-CSF at the time of clinical remission. A proportion of patients responds at least initially to autologous PBSCT (Reiffers et al, 1994). Regardless whether PBSCT or autologous bone marrow transplantation is used, it is crucial to eliminate leukaemic cells from the transplant in order to avoid the transfer of leukaemic cells back to the patient.
For CML the logical target that distinguishes normal cells from leukaemic cells is p210 bcr-abl . This fusion protein aberrantly activates a number of signalling pathways that act in concert to transform cells. Thus, these pathways emanating from p210 bcr-abl are all potential targets for therapeutic intervention. Prominent targets are the Raf-1 kinase and the c-Myc transcription factor. p210 bcr-abl induces c-Myc expression in haematopoietic cells (Sawyers et al, 1992;Sawyers, 1993;Weissinger et al, 1993). The deregulation of c-Myc expression has been shown to be required for transformation by oncogenic abl genes (Sawyers, 1993;Weissinger et al, 1993). p210 bcr-abl also activates Raf-1 and the MEK -ERK pathway. The canonical Raf-MEK -ERK pathway is often perceived as a linear signalling module that mediates cell proliferation, transformation and survival (Weissinger et al, 1997). We have shown previously that the activation of the cAMP dependent protein kinase A (PKA) with synthetic agonist drugs such as 8-Chloro-cyclic Adenosine Monophosphate (8-Cl-cAMP) results in the inhibition of Raf-1 kinase activity and rapid apoptosis induction in v-abl transformed fibroblasts. Apoptosis occurred despite a high constitutive activity of ERK suggesting that Raf-1 uses a different pathway to ensure cell viability (Weissinger et al, 1997).
These observations led us to explore the use of 8-Cl-cAMP for the treatment of CML bone marrow cells for ex vivo for the purging of leukaemic bone marrow or for the treatment of leukaemic patients. 8-Cl-cAMP is one of the most stable compounds that activate PKA (Schwede et al, 2000) and can be manufactured in quantities and quality sufficient for clinical use. In fact, the antitumour activity of 8-Cl-cAMP has been under study for a number of years, including clinical studies for the treatment of tumours (Cho-Chung et al, 1995;Tortora et al, 1995;Propper et al, 1999). To date, mainly solid tumours like breast carcinomas were studied. In this report we present data indicating that treatment with 8-Cl-cAMP can provide an effective method for purging bone marrow prior to autologous transplantation, targeting specifically the p210 bcr/abl transformed cells.

Patients
The studies were approved by the institutional ethics committees of Munich and Hannover. Bone marrow from healthy donors, eight patients with CML (five males and three females age range: 25 -49) in haematological chronic phase or at more advanced stages was obtained after informed consent.

Synthesis of 8-Cl-cAMP
Preparation of 8-Cl-cAMP was performed as described with minor modifications (Brentnall and Hutchinson, 1972;Schwede et al, 2000). Briefly, 30 g (85.47 mmol) cAMP, sodium salt, were suspended in 2000 ml DMF and reacted with 87 g (171 mmol) tetrabutylammonium iodotetrachloride for 18 h. The reaction mixture was poured into 3000 ml water and extracted with chloroform (3 Â 250 ml) to remove excess of reagent. The resulting solution was concentrated under reduced pressure. 8-Cl-cAMP was purified and isolated by means of column liquid chromatography using silica-base reversed phase material (Merck, Germany, LiChroprep s Rp-18) (Cummings et al, 1994). The product containing fractions were collected and evaporated to produce 13.95 g (36.18 mmol) 8-Cl-cAMP, sodium salt, with a purity of 499% (yield: 42.3%).

Long-term culture and colony-forming units (CFU)
The cells were resuspended in LTC-medium supplemented with 12.5% horse serum (Hyclone, Munich, Germany) and 10 À5 M hydrocortisone at a density of 2 Â 10 6 cells ml À1 and cultured for 4 -6 weeks on autologous feeder layers (LTC-medium, Iscoves Mod. Dulbecco's medium 340 mOsm, Gibco BRL, Bethesda, USA). Cells were cultured in duplicates and incubated with medium alone, 50 or 100 mM 8-Cl-cAMP for 24 h (summarised in Figure 1). Subsequently, cells were washed and resuspended in LTC medium. Nonadherent cells (NADC) were harvested in weekly intervals and resuspended in MethoCult GF H4434 (Stem Cell Technology, Munich, Germany). Colonies were counted after 7 days and evaluated according to standard clinical haematological technique.

Nonspecific toxicity testing of 8-Cl-cAMP
The nonspecific toxicity of 8-Cl-cAMP on bone marrow cells was tested using marrow mononuclear cells (MNC) from healthy donors as outlined in Figure 1. The MNCs were cultured at a density of 2 Â 10 6 cells ml À1 in LTC medium as described above in the presence or absence of 8-Cl-cAMP for 24 h. Subsequently, the cells were washed and grown in long-term culture medium. In total, 2 Â 10 5 cells ml À1 were used for CFU assays. At the end of the long-term culture (4 -6 weeks), the adherent cells (ADC) as well as the NADC were harvested and analysed in the same manner.

Cytogenetic analyses
After completion of the CFU assays, colonies were picked and analysed for Ph1 chromosome positive colonies. Cytogenetic analyses were essentially performed as described (Dube et al, 1981). Briefly, cell division was arrested in metaphase by the addition of 100 ml of a colchicine solution (1 mg ml À1 in a-minimal essential medium) to 1 ml of cell suspension for 12 h. The cells were subsequently transferred to poly-L-lysine-coated slides and incubated in 0.2 ml of 0.075 mol l À1 KCl at room temperature for 10 min. Cells were fixed by gently dropping 100 ml cold methanol on the slide. The excess of the fixative was removed with absorbent paper and slides were dried on a hot plate at 551C. After repeated cold fixation for 15 min, the cells were used for banding and cytogenetic analyses.
Growth curves 8-Cl-cAMP was added to the cultures as indicated on day 0 and was not replenished during the culture. FDCP-mix cells were plated in 24-well plates at a density of 2 Â 10 5 cells ml À1 . The MEK inhibitors (U0126, PD98059) and raf kinase inhibitors, raf kinase inhibitor 1 (Raf KI) and ZM 336372, were purchased from Calbiochem, UK. Survival and proliferation assays were performed as previously described (Pierce et al, 1998).

PKA-activation resulted in a transient growth inhibition of human bone marrow cells
The nonspecific toxicity of 8-Cl-cAMP was tested on human bone marrow cells (BMC) of healthy donors (Table 1). In total, 15 flasks of cells per donor were set up for long-term cultures after treatment with 8-Cl-cAMP, five for each condition (control, 50 mM and 100 mM 8-Cl-cAMP; Figure 1). BMCs taken at week 0 immediately after treatment with 8-Cl-cAMP showed a reduction of cell numbers and a concomitant reduction of colonies arising in the CFU assays. However, at all subsequent timepoints, ranging from 1 to 5 weeks, comparable numbers of CFU-initiating cells were obtained from treated and untreated BMC cultures. This is summarised in Table 1 showing CFUs obtained from week 5 cultures as example. Thus, 8-Cl-cAMP only caused a transient impairment of CFU capacity in BMCs from healthy donors that was readily reversed at longer time points. This is consistent with our earlier observations (Weissinger et al, 1997) that untransformed cells only respond to PKA activation with an initial, transient inhibition of proliferation.

Ph1 chromosome positive colony forming cells are selectively eliminated by 8-Cl-cAMP
Since no severe nonspecific toxicity was observed that would prohibit treatment with 8-Cl-cAMP, bone marrow MNCs of eight patients with CML were treated in the same manner (Table 2). Three patients were in the chronic phase of CML undergoing cytoreductive treatment, one patient was at diagnosis and four were in the accelerated phase or blast crisis. In all cases, treatment with 8-Cl-cAMP substantially reduced the number of Ph1 chromosome positive colonies. The incubation with 50 mM Data obtained after treatment of normal marrow MNC with 8-Cl-cAMP are summarised. The cells were resuspended at 2.5 Â 10 5 cells ml À1 as described in Figure 1. The effects of treatment with 8-Cl-cAMP (one single exposure of the cells for 24 h) on colony formation was observed at week 0 and after week 5.2 Â 10 5 cells ml À1 were plated in soft agar for colony formation in week 1 or after long-term culture. The nonadherent cell fraction did not yield significant amounts of colonies with or without 8-Cl-cAMP treatment after 5 weeks, thus only the adherent cells (ADC) are shown. Colonies were judged by their typical histological appearance according to haematological criteria: G ¼ granulocyte; Eo ¼ eosinophiles; M ¼ macrophages; E ¼ erythrocytes; GM ¼ mixed colonies, Granulocytes and macrophages.  8-Cl-cAMP resulted in a reduction of Ph1 chromosome positive colonies to 14 and 46%, respectively, but never led to a complete elimination of the Ph1 chromosome positive progenitor cells. Therefore, 100 mM 8-Cl-cAMP was used in further experiments.
In CFU assays prepared from cultures at week 0 all colonies were Ph1 chromosome positive, and no difference was observed between the treated and the untreated cells. However, treatment with 8-Cl-cAMP resulted in a significant reduction of Ph1 chromosome positive CFUs prepared after 5 weeks of culture depending on the stage of disease. In three patients undergoing cytoreductive therapy, the treatment with 100 mM 8-Cl-cAMP led to a complete loss of Ph1 chromosome positive colonies after 5 weeks of culture. In the absence of 8-Cl-cAMP only a small reduction of Ph1 chromosome positive colonies was observed after 5 weeks of culture, with more than 80% of the colonies remaining Ph1 chromosome positive. Cells from one untreated patient and four patients in advanced stages of CML exhibited a reduction of cells with a significant loss of Ph1 chromosome positive colonies ranging from 83 to 33%. Interestingly, after treatment with 8-Cl-cAMP Ph1 chromosome negative colonies appeared in CFU assays prepared from patients in the accelerated phase of CML, whereas in the untreated controls 100% of the colonies were Ph1 chromosome positive.

Studies to investigate the molecular mechanism of 8-Cl-cAMP
Given the encouraging results obtained with 8-Cl-cAMP in clinical samples and patients we investigated the molecular basis of the activity of 8-Cl-cAMP in particular with regard to its effects on cell proliferation and survival. Due to technical reasons such as freshness and instability of the material, these studies are extremely difficult in primary clinical samples. Therefore, we used bcr-abl transformed FDCP-mix p210 bcr/abl cells, a well characterised cell culture model system for CML (Pierce et al, 1998). These cells are conditionally transformed by expression of a temperature-sensitive p210 bcr/abl protein. They still remain IL-3 dependent although p210 bcr/abl sensitises them to the effects of IL-3, when cultured at the permissive temperature of 321C. The main effect of p210 bcr/abl is to enhance viability under conditions of low IL-3 levels (0.01 -0.1 ng ml À1 ) (Pierce et al, 1998).
Our previous studies with v-abl transformed fibroblasts (Weissinger et al, 1997) had indicated that PKA activation could downregulate the activity of Raf-1 and thereby cause apoptosis in these cells. This would provide a plausible explanation for the effects of 8-Cl-cAMP on CML cells. To test whether this hypothesis was also applicable to haematopoietic cells we compared 8-Cl-cAMP to selective pharmacological inhibitors of Raf-1 and MEK. As observed previously (Pierce et al, 1998) p210 bcr/abl did not significantly affect DNA synthesis when cells were compared to parental controls within 24 h after shifting them to the permissive temperature 321C (Figure 2). The two Raf kinase inhibitors (Raf KI and ZM336372) failed to interfere with IL-3 driven proliferation. Raf KI even accelerated proliferation in both the control cells and p210 bcr/abl cells exposed to 10 ng ml À1 IL-3. In contrast, both MEK inhibitors (U0126 and PD98059) interfered with DNA synthesis and this effect was slightly more pronounced in the p210 bcr/abl cells. In control cells 8-Cl-cAMP interfered with DNA synthesis only at high (10 ng ml À1 ) concentrations of IL-3, whereas it blocked proliferation in p210 bcr/abl cells at all concentrations.
A clear difference emerged when the effects on viability were assayed using trypan blue exclusion (Figure 3). FDCP-mix cells have been reported to die by apoptosis following cytokine removal (Williams et al, 1990). p210 bcr/abl protected cells from the cytotoxic effects of IL-3 withdrawal, maintaining the viability of almost 40% of cells 3 days after IL-3 withdrawal. Under these conditions the viability of the control cells was severely compromised. IL-3 was a very potent survival factor even at 0.1 ng ml À1 . Higher concentrations of IL-3 did not improve survival further. Interestingly, neither MEK inhibitors ( Figure 3A) nor Raf inhibitors ( Figure 3B) counteracted effects of p210 bcr/abl or IL-3 on cell viability.
In contrast, 8-Cl-cAMP significantly inhibited the cytoprotective effect of p210 bcr/abl , but not of IL-3 ( Figure 4A). Moreover, 8-Cl-cAMP preferentially induced cell death in p210 bcr/abl as compared to control cells. This effect was most pronounced 48 and 72 h after IL-3 withdrawal, suggesting that p210 bcr/abl sensitises cells to killing by 8-Cl-cAMP. IL-3 protected against 8-Cl-cAMP induced cytotoxicity suggesting that IL-3 can activate p210 bcr/abl independent survival pathways. As trypan blue exclusion ( Figure 4A) does not distinguish between necrotic and apoptotic cell death, we further tried to dissect the mode of 8-Cl-cAMP induced cell death. Apoptosis leads to cell surface phospholipid asymmetry resulting in the exposure of phosphatidylserine (PS) on the outer leaflet of the cytoplasmic membrane. Annexin V preferentially binds PS and has been used to detect apoptosis in the FDCP-mix cells (Francis et al, 2000). In contrast to necrosis, membrane integrity is maintained during apoptosis precluding staining of DNA by the membrane impermeable dye propidium iodide (PI). Thus, apoptosis is indicated by positive staining for annexin V and negative staining for PI. As shown in Figure 4B measuring viability as cells that escape apoptosis, that is, stain negative for annexin, largely parallels the data obtained with the trypan blue exclusion assay, exception that in this assay the parental FDCP-mix cells also show a significant decrease in viability in response to 8-Cl-cAMP 48 and 72 h after IL-3 removal. Measuring annexin positive and PI negative, that is, apoptotic cells ( Figure 4C) showed a higher rate of Figure 2 Effect of 8-Cl-cAMP, Raf kinase inhibitors and MEK inhibitors on the proliferation of FDCP-mix (A) and p210 bcr-abl transformed FDCP-mix cells (B). Cells were cultured at the permissive temperature in Fisher's medium with 20% horse serum and increasing concentrations of IL-3 (0, 0.01, 0.1, 1, 10 ng ml À1 ) as shown on the X-axis. 8-Cl-cAMP (100 mM), MEK inhibitors U0126 (10 mM) or PD98059 (50 mM), and Raf kinase inhibitors, Raf kinase inhibitor I (Raf KI, 10 mM) and ZM336372 (100 mM) were added and DNA synthesis was assessed by measuring [3H] thymidine incorporation after 16 h. Experiments were carried out in triplicates.
apoptosis in the control cells under almost all conditions. 8-Cl-cAMP did not significantly affect this rate. However, a significant increase in apoptosis was observed in 8-Cl-cAMP treated p210 bcr/ abl cells 48 h after IL-3 withdrawal. Necrotic cell death as measured by cells staining positive for annexin and PI ( Figure 4D) was enhanced by 8-Cl-cAMP under conditions of IL-3 withdrawal. Significant increases were observed in control cell 24 h, and in the p210 bcr/abl cells 48 and 72 h after IL-3 removal. These results confirm the data obtained by the trypan blue exclusion assay, and suggest that 8-Cl-cAMP mediated cytotoxicity includes both apoptosis and necrotic cell death.
In summary, the inhibitor experiments demonstrated that the pathways mediating proliferation can be dissociated from path-ways required for p210 bcr/abl driven cell survival. MEK -ERK signalling is required for DNA synthesis, but not for viability, whereas 8-Cl-cAMP can interfere with cell proliferation as well as survival. More importantly they show that 8-Cl-cAMP preferentially kills p210 bcr/abl cells.

DISCUSSION
In this report we have analysed the influence of PKA-activation on transformed cells from eight patients with CML. The expression of p210 bcr/abl is a hallmark of CML. Among other signalling pathways p210 bcr/abl also activates the Raf -MEK -ERK pathway. We have previously shown that the inhibition of Raf-1 by 8-Cl-cAMP led to apoptosis in v-abl transformed fibroblasts, while control cells or cells expressing the v-raf oncogene showed only a reversible growth inhibition (Weissinger et al, 1997). Here we demonstrate that the activation of PKA is a promising approach to selectively eliminate the Ph1 chromosome positive progenitor cells from marrow obtained from CML-patients. Despite an initial reduction of total cell numbers no long-term cytotoxicity was observed when marrow cells from normal donors were treated with 50 or 100 mM 8-CL-cAMP (Table 1). The initial decline in cell numbers is probably due to the inhibition of proliferation also observed in untransformed NIH3T3 fibroblasts or even in v-raf transformed fibroblasts (Weissinger et al, 1997). The reversible inhibition of normal cell proliferation could potentially be exploited to protect stem cells during chemotherapy. Thus, combining 8-Cl-cAMP with classical DNA damaging chemotherapeutic drugs may have the added benefits of assaulting the leukaemic cells by two routes while protecting the stem cells at the same time.
Seeking to understand the mechanism of growth inhibition and apoptosis induced in patient cells, we employed a well-characterised cell culture model of CML, that is, FDCP-mix cells expressing a temperature-sensitive p210 bcr/abl (Pierce et al, 1998). An obvious hypothesis emerging from our previous work with vabl transformed fibroblasts (Weissinger et al, 1997) was that the inhibition of Raf-1, but not MEK, would be crucial for the cytotoxic effects of 8-Cl-cAMP. Therefore, we compared the effects of 8-Cl-cAMP to Raf-1 and MEK inhibitors. The results clearly show that MEK activity is required for the proliferation of both normal and p210 bcr/abl cells. In contrast, MEK activity was not required for p210 bcr/abl or IL-3 mediated viability. Curiously, Raf-1 inhibitors did not inhibit proliferation or survival, and Raf KI even enhanced these parameters. These results suggest that Raf-1 does not play a significant role in mediating proliferation or survival in these cells. However, the unexpected effects of Raf kinase inhibitors may be explained by a paradoxical activation of Raf previously observed with ZM 336372 (Hall-Jackson et al, 1999). Alternatively, Raf kinase activity may be dispensable for maintenance of cell viability, as Raf-1 has been shown to prevent apoptosis independent of its kinase activity by binding to and inhibiting the activity of the proapoptotic kinase ASK-1 (Chen et al, 2001). In this scenario, the Raf inhibitors would not be expected to show any effects, since they are ATP analogues that block kinase activity but not binding to other proteins. Furthermore, myeloid cells can activate the ERK pathway independent of Raf (Buscher et al, 1995), which could explain why the Raf and MEK inhibitors have different effects. Thus, the inhibitory effects of 8-Cl-cAMP on the proliferation and viability of p210 bcr/abl cells cannot be explained by the inhibition of the catalytic activities of Raf-1 and MEK.
Importantly, 8-Cl-cAMP exhibited significant selective cytotoxicity for cells that express p210 bcr/abl . This was shown with the p210 bcr/abl transformed FDCP-mix cells as well as with primary bone marrow cells from leukaemic and normal donors. When marrow was obtained from patients in chronic phase of CML, a single incubation with 100 mM 8-Cl-cAMP for 24 h was sufficient to completely eliminate the Ph1 chromosome positive progenitor cells in vitro. In analogy to our previously published experiments (Weissinger et al, 1997), the effects of 8-Cl-cAMP treatment and PKA activation were irreversible in p210 bcr/abl transformed progenitor cells of CML. Interestingly, in week 0 no elimination of the Ph1 chromosome positive progenitor cells had occurred. One explanation is that the induction of cell death requires proliferating cells. We reason that dividing cells are eliminated, whereas differentiating progenitors may not be affected by the activation of PKA. The induction of cell death in the susceptible progenitor population is reflected by the loss of Ph1 chromosome positive colonies after treatment as summarised in Table 2.
In conclusion, our results demonstrate that 8-Cl-cAMP can be useful for the efficient elimination of Ph1 chromosome positive progenitor cells from bone marrow ex vivo without severe toxic effects on normal cells. This offers a new method to purge marrow/ stem cell populations from patients with Ph1 chromosome-positive leukaemias prior to autologous transplantation.