ALBERT

Description: The interleukin-3 (IL-3) dependent FL5.12 hematopoietic precursor cell line was isolated from murine fetal liver, an important site of early hematopoiesis, by panning with an anti-AA4 antibody and culturing the cells in medium containing IL-3. The AA4 cell antigen (human analogue, Clq Receptor) is a key cell surface marker expressed on early hematopoietic stem cells along with the IL-3R (CD123) and others. Upon Hoechst-33342 staining and flow cytometric analysis, FL5.12 cells normally have approximately 0.4% of the cells present in side population (SP) which are proposed to possess “stem cell like characteristics”. Some leukemic stem cells often display drug resistance properties. FL5.12 cells are normally sensitive to chemotherapeutic drugs such as doxorubicin, however, drug resistant cells (FL/Doxo) were isolated by subcloning the cells in doxorubicin. These drug resistant cells contain approximately 2-fold more SP cells than drug sensitive FL5.12 cells. Drug resistance was associated with 6-fold increased ERK activation and 11-fold decreased p53 induction after doxorubicin treatment. Furthermore the drug resistant cells displayed decreased apoptosis (1.8 fold) and caspase 3, 8 and 10 activation (4.2, 2.3, 2 fold respectively) upon culture with doxorubicin than the parental cells. A 3-fold higher level of proteasomal degradation of p53 was observed in the drug resistant cells which was due to 8-fold higher levels of MDM2. Drug resistant FL/Doxo cells were more sensitive to proteosome inhibitors (IC50 of 75 nM compared to 125 nM for parental cells). Treatment with proteosome inhibitors resulted in 12-fold and 2-fold more p53 detection in FL/Doxo and FL5.12 cells respectively. Synergistic responses to doxorubicin and proteosome inhibitor treatment were observed as the IC50 for doxorubicin was reduced 16- and 5-fold respectively with FL/Doxo and FL5.12 respectively. FL/Doxo cells were also highly sensitive to mTOR inhibitors and synergistic effects were observed upon combining doxorubicin and rapamycin as the IC50s decreased 40- and 6-fold for FL/Doxo and FL5.12 cells respectively, documenting the importance of the mTOR pathway in their drug resistance. Introduction of dominant negative (DN) p53 or activated MEK1 (MEK1 CA) further increased the resistance of the FL/Doxo cells to doxorubicin approximately 18- and 34-fold respectively. The DN p53 or MEK1 CA transduced cells contained 3.3- and 5.4-fold more SP positive cells than parental FL5.12 cells respectively. While both DN p53 and MEK1 CA increased drug resistance and the frequency of SP cells, this occurred by different mechanisms. DN p53 suppressed Bax induction after doxorubicin treatment and decreased the sensitivity of the cells to proteosome (1.6-fold), MEK (1.2-fold), MDM- 2 (5-fold) and mTOR (3-fold) inhibitors compared to the parental FL5.12 cells. MEK1 CA also decreased the sensitivity to proteosome inhibitors 3-fold, yet the cells remained sensitive to MEK, mTOR and MDM2 inhibitors. Thus while DN p53 or MEK1 CA both increased the frequency of SP hematopoietic “leukemia stem” like cells, they do this by different mechanisms which also alters the sensitivities of small molecule inhibitors. Mutation of p53 or activation of the Raf/MEK/ERK pathway in leukemia stem cells may render them resistant to chemotherapy as well as certain targeted therapeutic approaches.

Description: Abstract 594 Our group has recently reported on the growth inhibitory and pro-apoptotic effects of the MEK inhibitor PD0325901 in preclinical models of hematologic malignancies, particularly AML. However, the molecular mechanisms of AML sensitivity/resistance to MEK inhibitors remain elusive. Regardless of their sensitivity to PD0325901, none of the AML, ALL, and multiple myeloma cell lines examined harbored HRAS, BRAF, MEK, and PI3K mutations, while NB4, KMS18, and CEM had a mutated KRAS. In the absence of unequivocal genetic predictors of sensitivity/resistance to MEK inhibitors, one possibile alternative was to pursue rational, mechanism-based combinations with agents interfering with putative ‘escape' pathways. Therefore, we analyzed signaling along the ERK and AKT pathways in three different models of PD0325901 resistance: intrinsically resistant cell lines (U937), sensitive cell lines (OCI-AML3) that had been rendered resistant by prolonged exposure to another MEK inhibitor (CI-1040), and cytokine-exposed, non-responding FDC-P1 versus responding, v-fms-transformed FDC-P1 cells. Resistant U937 had low basal levels of phosphorylated ERK that were not completely abrogated by PD0325901 treatment even at high (1000 nM) concentrations; similarly, at least 100-fold higher PD0325901 concentrations were required to abrogate ERK phosphorylation in resistant IL-3-cultured FDC-P1 and OCI-AML3, as compared with their sensitive counterparts. Moreover, in the OCI-AML3 model, MEK inhibition-induced growth inhibition directly paralleled the ability of either PD0325901 or CI-1040 to abrogate ERK phosphorylation. Interestingly, PD0325901 induced AKT phosphorylation (S473) in all three models of resistance; phosphoproteomic analysis confirmed increased signaling through the PI3K/AKT/mTOR pathway upon MEK inhibition (10 nM PD0325901), with increased AKT1 (S473), mTOR (S2448), and S6Ka (T389) phosphorylation and increased PI3K expression. In OCI-AML3 resistance to MEK blockade-mediated growth inhibition clearly correlated with higher levels of AKT phosphorylation. Notably, comparative proteomic analysis of PD0325901-sensitive (OCI-AML3) and -resistant (U937) AML cell lines revealed upregulation of PI3K expression (+62%), increased AKT1 expression (+53%) and phosphorylation (T308, +86%; S473, +53%), inhibitory PTEN phosphorylation (S380+S382+S385, +46%), increased S6Ka and b expression (+48%), and increased S6 phosphorylation (S235, +54%) in U937 cells. From a functional standpoint, combined MEK inhibition (PD0325901) and mTOR blockade downstream of AKT (Temsirolimus) resulted in a striking growth-inhibitory synergism in OCI-AML3, with an average combination index at the ED50, ED75, and ED90 of 0.3±0.2. Similar results were obtained in the Flt3/ITD cell line MOLM-13. Overall, these results suggest that resistance to PD0325901-mediated growth inhibition may stem from a combination of decreased efficiency towards ERK inhibition and increased parallel signaling through AKT. The latter is an emerging theme in cancer cell signal transduction, as highlighted by recent reports of a functional cross-talk between the MEK/ERK and PI3K/AKT/mTOR in different models of cancer progression and response to individual pathway inhibitors. These findings bear potentially important consequences for clinical translation, as we show that combined blockade of both MEK and mTOR signaling, a therapeutic strategy that has shown promise in different solid tumor models, results in a strongly synergistic inhibition of leukemic cell growth. Disclosures: No relevant conflicts of interest to declare.

Description: In hematologic malignancies, constitutive activation of the Raf/MEK/ERK pathway is frequently observed, conveys a poor prognosis, and constitutes a promising target for therapeutic intervention. Indeed, we have recently demonstrated that selective MEK-I potently inhibit the growth of AML cell lines and ex vivo-cultured primary AML blasts (Blood2006, 108:254). However, these effects are mostly related to the inhibition of cell cycle progression, while apoptosis induction requires higher concentrations of the inhibitors and longer times of exposure. Thus, we investigated MEK-I-induced changes in phospho-protein expression and gene expression profiles, in order to identify relevant downstream targets and to design rational MEK-I-based combination strategies. Analysis of phosphorylation levels of 18 different target proteins performed in OCI-AML3 cells indicated that MEK blockade induces, among other effects, an over-activation of RAF and MEK, suggesting the interruption of a negative feedback loop. Moreover, gene expression profiling indicated that, in the same cellular model, MEK-I induced upregulation of the Flt-3 receptor. Based on these observations, as well as on recent evidence indicating that the Raf inhibitor sorafenib directly inhibits signaling through Flt-3 (JNCI2008, 100:184), experiments were performed in OCI-AML3 and MOLM-13 (which harbors a Flt3 ITD) cells to test the activity of MEK-I in combination with sorafenib. Simultaneous inhibition of Flt3/Raf and MEK resulted in the synergistic inhibition of cell growth, as measured by isobologram analysis (Chou–Talalay method) in both model systems, with combination indexes (CI) of 0.12 and 0.48 for OCI-AML3 and MOLM-13 cells, respectively. Neither sorafenib nor MEK-I induced apoptosis in either cell line when used alone; however, apoptosis was observed in up to 50% of the cells with the combined treatment. Based on our previous experience, as well as on the ability of MEK-I to modulate the expression, among others, of genes controlling mitochondrial homeostasis (e.g. PPIF, GRPEL1), we next investigated the impact of simultaneous inhibition of the MEK and Bcl-2 pathways in AML cells. Exposure of OCI-AML3 and MOLM-13 cells to a combination of MEK-I and the Bcl-2/Bcl-xL inhibitor, ABT-737 (kindly provided by Abbott Laboratories) synergistically inhibited cell growth, with CI ranging from 0.45 to 0.04 in OCI-AML3 and from 0.75 to 0.14 in MOLM-13, respectively. In both cellular models, ABT-737dose-dependently induced apoptosis, while MEK-I, at the concentrations used in combination experiments, did not appreciably increase apoptotic cell death; however, simultaneous Bcl- 2/Bcl-xL inhibition and MEK blockade resulted in the massive induction of apoptosis (up to 85% and 67% net apoptosis induction in OCI-AML3 and MOLM-13 cells, respectively). Such pro-apoptotic interaction was highly synergistic with CI of 0.18 and 0.16 in OCIAML3 and MOLM-13 cells, respectively. In contrast, combination with MEK-I did not appreciably sensitize the MEK-I-resistant cell line U937 to either sorafenib- or ABT-737- induced growth inhibitory and pro-apoptotic effects. Overall these results support the role of the Raf/MEK/ERK kinase module as a prime target for the molecular therapy of AML and suggest that both “vertical” and “lateral” combination strategies based on MEK inhibition may produce highly synergistic anti-leukemic effects.