The Novel Pan-PIM Kinase Inhibitor, PIM447, Displays Dual Antimyeloma and Bone-Protective Effects, and Potently Synergizes with Current Standards of Care

Multiple myeloma is characterized by a proliferation of malignant plasma cells in the bone marrow (BM) that secretes a monoclonal immunoglobulin (1). It is typically associated with osteolytic lesions, due to an increase in the number and bone-resorptive activity of osteoclasts (OC) together with osteoblast (OB) inhibition (2). Over the last decade, the introduction of new drugs, such as proteasome inhibitors and immunomodulatory agents, has improved the outcome of patients with multiple myeloma (3, 4); however, multiple myeloma is still considered incurable mainly due to the development of drug resistance. Therefore, the identification of new targets and the investigation of novel drugs against such targets are extremely important for the discovery of more effective treatments (5).

Recently, PIM kinases have been proposed as being new therapeutic targets for treating hematologic cancers (6). These are a family of serine/threonine kinases comprising three members (PIM1, PIM2, and PIM3) that regulate oncogenesis, survival pathways, drug resistance, and migration, among other functions (7). Several features make PIM inhibition an attractive therapeutic strategy, particularly against myeloma cells: (i) PIM2 is among the most highly overexpressed genes in multiple myeloma cells (8), (ii) this protein is required for maintaining multiple myeloma cell growth (9), and (iii) the expression of PIM2 in multiple myeloma cells is enhanced by mesenchymal stromal cells (MSC) and OCs as a survival and drug resistance mechanism (10). In addition, recent studies have also demonstrated the importance of PIM kinases in the biology of bone-remodeling cells. Accordingly, Kim and colleagues found that PIM1 positively regulates the ligand of the receptor activator of NF-κB (RANKL)–induced osteoclastogenesis (11). More recently, it has been shown that PIM2 acts as a negative regulator of osteoblastogenesis (12). Based on all these data, it is tempting to hypothesize that the inhibition of these proteins would result not only in an antitumor activity, but may also have a beneficial effect on multiple myeloma patients' bone disease.

PIM447 (N-(4-((1R,3S,5S)-3-amino-5-methylcyclohexyl)pyridin-3-yl)-6-(2,6-difluorophenyl)-5-fluoropicolinamide) is a potent and selective pan-PIM kinase inhibitor, derived from the tool compound LGB321 (6, 9), with pharmacokinetic and drug-like properties suitable for clinical development (13). In fact, preliminary results from a phase I clinical trial with PIM447 in patients with relapsed and/or refractory multiple myeloma have demonstrated single-agent activity and a tolerable safety profile (14). Although the preclinical efficacy of other PIM inhibitors has been studied in several hematologic malignancies including multiple myeloma (12, 15–18), only two of them, SGI-1776 and AZD1208, have reached clinical trials but none in multiple myeloma.

In the present work, we have investigated through the use of preclinical models, the activity and mechanism of action of PIM447 in myeloma cells, its potential synergism with standard-of-care treatments, and its effects on myeloma-associated bone disease. To the best of our knowledge, this is the first clinical pan-PIM-inhibitor that demonstrates dual antimyeloma and bone-protective effects, and this fact, together with the very strong synergy that PIM447 exhibits with standard-of-care treatments and the preliminary favorable clinical data, supports the future clinical development of this drug for the treatment of patients with multiple myeloma.

PIM447 was provided by Novartis Pharmaceuticals, Inc. Bortezomib was purchased from LC Laboratories, lenalidomide and pomalidomide from Selleckchem, and dexamethasone from Sigma-Aldrich.

The human myeloma cell lines, MM1S, MM1R, U266, and NCI-H929, were purchased from the ATCC, whereas RPMI-8226 and OPM-2 were obtained from DSMZ. The human myeloma cell line MM144 was a generous gift from Dr. S. Rudikoff (National Cancer Institute, National Institutes of Health, Bethesda, MD). The source of the human myeloma cell lines RPMI-LR5, U266-Dox4, and U266-LR7 has been previously described (19). The origin of MM1S-luc and RPMI-8226-luc cells (luciferase-expressing) was previously explained (20). All myeloma cell lines were cultured as previously described (19). The MM1S-luc cocultures with either the hMSC–TERT cell line (obtained from Dr. D. Campana, St. Jude Children's Research Hospital, Memphis, TN) or OCs were performed in the absence or presence of different concentrations of PIM447 as previously explained (20). Cell line identities have been tested and authenticated by short tandem repeat (STR) analysis with a PowerPlex 16 HS System kit (Promega) and online STR matching analysis (DSMZ institute; available from: http://www.dsmz.de/fp/register.php).

All primary BM and peripheral blood samples were obtained after the approval of the Complejo Asistencial Universitario de Salamanca Review Board and with the written informed consent of all participating subjects.

Cell viability of myeloma cells was evaluated by the MTT method (19). The IC₅₀ of PIM447 was calculated using SigmaPlot graphing software. The cell-cycle profile and apoptosis induction were evaluated as previously described (21).

BM samples from patients with multiple myeloma were lysed and cultured as previously described (21), in the absence or presence of PIM447. The percentage of Annexin-V–positive cells was analyzed by flow cytometry on myelomatous plasma cells (CD38^+bright, CD45^−/low, SSC^{low/intermediate}, CD56^−/+) and normal lymphocytes (CD45⁺⁺, SSC^low) populations.

MM1S or RPMI-8226 cells were treated for 48 to 72 hours with double and triple combinations of PIM447 and other antimyeloma agents (bortezomib, lenalidomide, pomalidomide, and dexamethasone). Cell viability was analyzed by MTT assays. The potency of each combination was quantified with Calcusyn software (Biosoft), which is based on the Chou–Talalay method, yielding a combination index (CI) with the following interpretation: CI > 1, antagonistic effect; CI = 1, additive effect; and CI < 1, synergistic effect.

Peripheral blood mononuclear cells (PBMCs) from healthy donors were differentiated in osteoclastogenic medium (containing 25 ng/mL M-CSF and 50 ng/mL RANKL, both from Peprotech) as previously described (22). Assays related to OCs formation and function included: F-actin ring formation (pre-OCs, 14 days differentiation), resorption capacity (17 days differentiation), and OC formation (21 days differentiation), and were performed as previously reported (23).

OBs were generated and assayed as previously described (20). Briefly, primary MSCs (passages 2–3) were cultured in osteogenic medium (containing 5 mmol/L β-glycerophosphate and 50 mg/mL ascorbic acid) and exposed to PIM447. After 11 days, alkaline phosphatase (ALP) activity was quantified by hydrolysis of p-nitrophenylphosphate into p-nitrophenol (Sigma-Aldrich), whereas mineralization was assessed by Alizarin Red staining of calcium deposits at day 21.

All gene expression microarray datasets were retrieved from Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/, accession number: GSE47552; ref. 24). CEL files were background-corrected and normalized using the Robust Multi-Array (RMA) algorithm implemented in the Affymetrix Expression Console, to estimate the log₂-normalized expression values for PIM1, PIM2, and PIM3.

Western blot was performed following standard procedures (25). Phospho-TSC2 levels were examined with a specific phospho-Akt substrate antibody (Cell Signaling Technology) following TSC2 immunoprecipitation, as previously described (9). The origin of primary antibodies was as follows: anti–Bcl-2, anti–Mcl-1, anti–p-Erk1/2, anti-NFATc1, and anti–cathepsin K from Santa Cruz Biotechnology; α-tubulin from Calbiochem; and all other antibodies from Cell Signaling Technology. The horseradish peroxidase–conjugated secondary antibodies were from GE Healthcare.

Total RNA was isolated using TRIzol reagent (Invitrogen). TaqMan Gene Expression Assays (Applied Biosystems) were performed according to the manufacturer's instructions: CA2 (Hs00163869_m1), ATP6V1A (Hs01097169_m1), and MMP9 (Hs00234579_m1). Normalized gene expression was calculated as 2^−ΔCt, being ΔCt = Ct (gene) − Ct (GAPDH).

Animal experiments were conducted according to institutional guidelines for the use of laboratory animals, and after granted permission from the University of Salamanca Animal Ethical Committee for animal experimentation. RPMI-8226-luc cells (8 × 10⁶) were injected intravenously into 6-week-old female NOD-SCID-IL-2Rγ^−/− (NSG) mice (Charles River Laboratories), and tumor development was monitored by noninvasive bioluminescence imaging (BLI), as previously described (20). After 4 weeks, animals were randomized into two groups (n = 12/group), one receiving vehicle (5 times/week by oral gavage), the other receiving PIM447 (100 mg/kg, 5 times/week by oral gavage). Serum levels of human Igλ (secreted by RPMI-8226-luc cells) were determined by ELISA (Bethyl Laboratories). For microcomputed tomography (microCT) analysis, one femur of each animal was fixed in 10% formalin and scanned using a microCT system (MicroCATII; Siemens) as described previously (22). The trabecular microarchitecture in the distal femur was analyzed using BoneJ (26). Carboxy-terminal telopeptide collagen cross-links (CTX) and N-terminal propeptide of type I procollagen (P1NP) were measured in mice sera by ELISA (Immunodiagnostic Systems).

Statistical analyses were performed using SPSS-v21.0 software (IBM Corp.).

We initially evaluated the basal levels of PIM kinases in multiple myeloma cell lines and primary myeloma cells. As shown in Fig. 1A, the three isoforms (PIM1, PIM2, and PIM3) were expressed in the 10 multiple myeloma cell lines analyzed. Notably, the level of expression of PIM2 was higher than the other two PIM kinases. Consistent with these results, we also observed in a series of 41 patients with newly diagnosed multiple myeloma that PIM2 levels were significantly higher than PIM1 and PIM3 in CD138⁺ myeloma cells (Fig. 1B). In addition, PIM1 levels were significantly higher than PIM3 levels (Fig. 1B).

Because PIM kinases are highly expressed in multiple myeloma cells, we evaluated the antimyeloma effect of the new pan-PIM kinase inhibitor, PIM447. Different multiple myeloma cell lines were treated with increasing concentrations of PIM447 (0.05–10 μmol/L) for 24, 48, and 72 hours. The dose–response curves obtained by MTT assay revealed two patterns of sensitivity: sensitive cell lines with IC₅₀ values at 48 hours ranging from 0.2 to 3.3 μmol/L (MM1S, MM1R, RPMI-8226, MM144, U266, and NCI-H929) and less sensitive cell lines with IC₅₀ values at 48 hours >7 μmol/L (OPM-2, RPMI-LR5, U266-Dox4, and U266-LR7; Fig. 1C). A lack of correlation was found between the levels of PIM kinases and the IC₅₀ values for PIM447 (P > 0.05; Fig. 1D). To explore the induction of apoptosis, three of the sensitive cell lines (MM1S, NCI-H929, and RPMI-8226) and two of the less sensitive cell lines (OPM-2 and RPMI-LR5) were treated with increasing doses of PIM447 for 24 and 48 hours (Fig. 2A), revealing a clear dose-response: while low doses of the drug (0.1–1 μmol/L) did not induce important levels of apoptosis in any of the cell lines tested, PIM447 at 5 μmol/L substantially increased Annexin-V levels (about 30%) in sensitive cell lines but not in OPM-2 and RPMI-LR5. In addition, the highest dose (10 μmol/L) induced apoptosis in all the cell lines but to a lesser extent in OPM-2 and RPMI-LR5. When MM1S cells were treated with 10 μmol/L PIM447 for different times, we observed a time-dependent increase in apoptotic cells (Fig. 2B). Moreover, treatment of MM1S cells with PIM447 promoted the cleavage of initiator caspases, such as caspases 8 and 9, and also the cleavage of the effector caspases 3 and 7, together with PARP cleavage (Fig. 2C). Similar results were found in RPMI-8226 and NCI-H929 cells (Supplementary Fig. S1A and S1B). To assess the potential effect of PIM447 on cell cycle, MM1S and OPM-2 cells were incubated with increasing concentrations (0.1–10 μmol/L) of this drug for 48 hours, and the cell cycle was analyzed by flow cytometry. PIM447 increased the percentage of cells in the G₀–G₁ phase and decreased the proliferative phases (S and G₂–M) of the cell cycle, in the two cell lines at all doses (Fig. 2D). Nevertheless, the effects at low concentrations (0.1–1 μmol/L) were more pronounced in MM1S cells than in OPM-2. Accordingly, the increase in the percentage of cells in G₀–G₁ after treatment with 0.1, 0.5, or 1 μmol/L PIM447 was, respectively, 10.2%, 16.66%, and 19.18% in MM1S; and 6.54%, 11.09%, and 9.51% in OPM-2. Similar results to those of MM1S were found in NCI-H929 (Supplementary Fig. S1C). Overall, these results indicate that, at least at low doses of the drug, OPM-2 cells are less susceptible to cell-cycle effects than sensitive cell lines.

The effect of PIM447 was also investigated ex vivo in cells isolated from BM samples from 10 patients with multiple myeloma, as explained in Materials and Methods. After 48 hours of exposure, PIM447 clearly induced apoptosis in myeloma cells (concentrations in the 5–10 μmol/L range) with low to moderate toxicity in lymphocytes (Fig. 2E).

Finally, we investigated whether PIM447 could overcome the protective effect conferred by the BM microenvironment on multiple myeloma cells. For this purpose, MM1S-luc cells were cocultured with MSCs or OCs in the presence of PIM447. Despite the proliferative advantage conferred by MSCs and OCs, PIM447 greatly reduced the viability of multiple myeloma cells under these circumstances (Fig. 2F).

Because current treatment of multiple myeloma is largely based on combinations of drugs with different mechanisms of action, we studied the effect of combining PIM447 with several standard-of-care treatments in multiple myeloma. The potency of each combination in MM1S cells was analyzed with Calcusyn software. With respect to double combinations, the synergy (CI < 1) of PIM447 with dexamethasone, lenalidomide, and pomalidomide is noticeable (Fig. 3A–C and Supplementary Tables S1–S3), being the combination with dexamethasone in the highly synergistic range (CI = 0.096) at PIM447 200 nmol/L + dexamethasone 10 nmol/L (Fig. 3A and Supplementary Table S1). On the other hand, triple combinations of PIM447 with bortezomib + dexamethasone (Fig. 3A and Supplementary Table S1), lenalidomide + dexamethasone (Fig. 3B and Supplementary Table S2), and pomalidomide + dexamethasone (Fig. 3C and Supplementary Table S3) also showed a very strong synergism. To confirm these results, the same combinations were tested on the RPMI-8226 cell line, rendering similar results (Supplementary Fig. S2).

PIM kinases are known to directly phosphorylate and regulate the proapoptotic Bcl-2 family member, Bad (27–29). Accordingly, treatment of MM1S, RPMI-8226, and NCI-H929 cells with 10 μmol/L PIM447 for different times reduced the phosphorylation of Bad on Ser112 without affecting the levels of total Bad (Fig. 4A and Supplementary Fig. S3A). In addition, treatment of MM1S cells with PIM447 reduced the levels of the Bad-regulated antiapoptotic protein Bcl-xl, but did not modify the levels of other antiapoptotic Bcl-2 family members such as Bcl-2 and Mcl-1 (Fig. 4A).

PIM2 modulates mTORC1 activity and promotes myeloma cell proliferation through phosphorylation of TSC2 (9). In this regard, just 1 hour of treatment with PIM447 strongly inhibited the phosphorylation of TSC2 in MM1S cells, whereas the phosphorylation of PRAS40, another mTORC1 modulator, at Thr246 required longer exposure (24 hours) to produce a decrease (Fig. 4B). In addition, treatment of MM1S cells with PIM447 reduced the phosphorylation of downstream mTORC1 targets such as 4EBP1 at Thr 37/46, P70S6 at Thr 389, and S6RP at Ser 235/236 (Fig. 4B). Similar results were found for RPMI-8226 and NCI-H929 cells (Supplementary Fig. S3B).

To gain further insights into the mechanism of PIM447, we analyzed c-Myc, a PIM-regulated transcription factor (30). Treatment of MM1S cells with PIM447 reduced the levels of total c-Myc and phospho-c-Myc at Ser62, a key residue involved in c-Myc stabilization (ref. 30; Fig. 4C). The levels of phospho-c-Myc (Ser62) also decreased in RPMI-8226 and NCI-H929 cells after treatment with PIM447 (Supplementary Fig. S3C). In addition, the levels of Mad-1, a protein that antagonizes Myc-mediated transcriptional activity, increased in MM1S after PIM447 exposure (Fig. 4C). Treatment of MM1S cells with PIM447 also reduced the levels of cyclins D2 and E1, two well-known regulators of G₁-to-S phase transition, previously described as being transcriptionally induced by c-Myc (refs. 31–33; Fig. 4C).

As previously commented, the sensitivity to PIM447 differs among different multiple myeloma cell lines. In order to elucidate the molecular basis for these differences, we explored the molecular effects of PIM447 in the low sensitive cell line OPM-2. Treatment of OPM-2 with 10 μmol/L PIM447 reduced the levels of phospho-S6RP (Ser 235/236) and phospho-4EBP1 (Thr 37/46) after just 1 hour of treatment, indicating the inhibition of the mTORC1 pathway (Supplementary Fig. S3D). Interestingly, PIM447 did not reduce the levels of phospho-Bad (Ser112) in OPM-2, whereas c-Myc and phospho-c-Myc (Ser62) only decreased after 24 hours of exposure to the drug (Fig. 4D). Because the effects of PIM447 on these proteins were detected earlier in sensitive cell lines, we then analyzed the potential correlation between their basal levels in multiple myeloma cell lines and the sensitivity to PIM447. The analysis showed a strong positive correlation between basal levels of Bad and the IC₅₀ for PIM447 (r = 0.8880; P = 0.0006; Fig. 4E). A positive correlation was also found between basal levels of phospho-Bad and IC₅₀ values, although in this case, the correlation was not so strong (r = 0.6044; P = 0.0642; Fig. 4F). Finally, sensitivity to PIM447 does not seem to be associated with basal levels of c-Myc (r = 0.4222; P = 0.2242) or phospho-c-Myc (r = 0.4270; P = 0.2184; Supplementary Fig. S4A and S4B).

Because PIM1 is known to be involved in murine RANKL-induced osteoclastogenesis (11), we investigated the effect of PIM447 on the formation and function of OCs from human origin. First of all, we observed an upregulation of PIM1 and PIM3 during OC differentiation (Fig. 5A). To evaluate the effect of PIM447 on OC formation, human PBMCs were exposed to several concentrations of the drug during OC differentiation. PIM447 reduced the number of TRAP⁺ multinucleated cells (IC₅₀ = 2 μmol/L) derived from PBMCs of healthy donors (Fig. 5B). Moreover, representative micrographs in Fig. 5B illustrate that cell densities were not significantly reduced in cultures exposed to up to 5 μmol/L PIM447, suggesting a selective effect of PIM447 on inhibition of OC formation rather than on cell viability.

To evaluate changes in OC functionality, we examined the effect of PIM447 on osteoclastogenic cultures established on calcium substrate–coated slides. A dose-dependent reduction of the area of resorptive pits was observed with PIM447 treatment (Fig. 5C). Of note, PIM447 inhibited OC-mediated resorption at doses clearly lower than those required to inhibit OC formation, suggesting that this drug not only inhibits osteoclastogenesis but also directly affects OC functionality.

To gain insight into the mechanisms mediating the aforementioned effects, we tested the effect of PIM447 on different transcription factors/molecules involved in OC differentiation and activity. Although levels of PU.1 and p-ERK1/2 were not significantly modified, PIM447 treatment reduced the levels of NFATc1, a master transcription factor in the differentiation of OCs, and cathepsin K, a protease responsible for degradation of organic bone matrix (Fig. 5D). We also observed that pre-OCs differentiated under PIM447 exposure were associated with partial disruption of the F-actin ring (Fig. 5E). Expression of molecules involved in matrix resorption, such as the matrix metalloproteinase 9 (MMP9) and the vacuolar-H⁺-ATPase catalytic subunit A1 (ATP6V1A), was significantly diminished after PIM447 treatment (Fig. 5F), therefore stating the role of PIM447 in preventing bone resorption. Expression of carbonic anhydrase II (CA2) also decreased although not reaching statistical significance.

Because PIM2 has recently been identified as a negative regulator for osteoblastogenesis in multiple myeloma and as an important target for treatment in myeloma bone disease (12), we also evaluated whether PIM447 was capable of promoting in vitro OB differentiation and activity. Primary MSCs from myeloma patients (n = 5) were maintained in osteogenic medium in the presence of different PIM447 concentrations, and ALP activity was measured at day 11 as a surrogate marker of early OB function. As seen in Fig. 5G, a significant increment in ALP activity was observed after treatment with the pan-PIM kinase inhibitor. We also observed a modest increase in matrix mineralization at the end of the osteogenic differentiation period (day 21) as assessed by Alizarin Red staining of calcium deposits (Fig. 5H).

We examined whether the in vitro effects of PIM447 were also present in vivo in a disseminated murine model of human myeloma. Compared with the vehicle group, PIM447 clearly controlled tumor progression as measured by bioluminescence (Fig. 6A) and serum levels of hIgλ secreted by RPMI-8226-luc cells (Fig. 6B). Importantly, PIM447 was well tolerated, as the body weight of mice did not decrease by more than 10% (Supplementary Fig. S5).

Representative microCT images at the metaphysis of distal femurs showed tumor-associated bone loss in vehicle mice compared with normal bone (mice not injected with myeloma cells); in contrast, PIM447-treated animals showed a trabecular microarchitecture similar to that of normal bone (Fig. 6C). Moreover, the analysis of bone morphometric parameters indicated that PIM447 increased bone volume density and trabecular number and reduced trabecular separation relative to vehicle control (Fig. 6D). In accordance with these findings, the serum levels of the bone resorption marker CTX were significantly diminished in PIM447-treated mice (Fig. 6E). Although there was a trend for PIM447 augmenting serum levels of the bone formation marker P1NP with respect to the vehicle-treated group, this was in the limit of statistical significance (P = 0.06; Fig. 6F).

The search for new targets is of utmost importance in multiple myeloma because the disease is still incurable and the therapeutic options currently available are limited. PIM kinases are a family of serine/threonine kinases composed of three members (PIM1, PIM2, and PIM3) that have recently been proposed as new therapeutic targets in multiple myeloma (6). Specifically, PIM2 is one of the most highly overexpressed genes in multiple myeloma (8), and its role as an antiapoptotic and cell growth mediator in myeloma cells has been described (9, 10). In line with these data, our results demonstrate that the level of expression of PIM2 in myeloma cells is higher than the expression of PIM1 and PIM3, suggesting that PIM2 has a major role in the biology of multiple myeloma. Nevertheless, the three PIM kinases have shown some functional redundancy in transgenic Eμ-myc mice (34, 35). Because the functional overlap of PIM isoforms is also likely to occur in myeloma, it is reasonable to hypothesize that the use of pan-PIM inhibitors may be more effective than targeting individual isoforms. In this work, we have evaluated in preclinical models the efficacy of the novel pan-PIM kinase inhibitor, PIM447, in myeloma cells and bone disease.

PIM447 shows antimyeloma activity in primary myeloma cells and multiple myeloma cell lines, with different degrees of sensitivity. It is of note that these differences do not depend exclusively on basal levels of PIM kinases, because we found a lack of correlation between the expression of these proteins and the IC₅₀ values for PIM447. On the other hand, interactions between myeloma cells and their microenvironment are a crucial factor in myeloma growth and drug resistance (36). Here, we have shown in vitro that PIM447 reduces myeloma cell viability even in the presence of MSCs and OCs, suggesting that this drug is able to overcome microenvironment-mediated drug resistance. Moreover, the clear antitumoral effect of PIM447 observed in the in vivo murine model of disseminated multiple myeloma demonstrates the efficacy of this agent in the context of the BM microenvironment.

PIM kinases phosphorylate and regulate multiple targets involved in different functions such as cell cycle, apoptosis, and metastasis (7). Our results with multiple myeloma cell lines and patient samples indicate that the cytotoxic effects of PIM447 are mediated through cell-cycle disruption and induction of apoptosis. The effects at low doses of the drug are mainly due to cell-cycle arrest rather than apoptosis induction, what is concordant with the clinical data, as many of the patients treated with the drug in monotherapy achieved stabilization of the disease, being some of them quite durable (14). Treatment of myeloma cells with PIM447 provokes an increase of cell-cycle G₀–G₁ phase and a decrease of S phase, suggesting a cell-cycle blockade at the G₁-to-S transition. Moreover, PIM447 also downregulates the expression of cyclins D2 and E1, two fundamental regulators of G1-to-S phase progression (37, 38). In addition, PIM447 reduced the levels of the transcription factor c-Myc, a direct target of PIM kinases (7), and increased the expression Mad-1, a protein that antagonizes Myc-mediated transcriptional activity (39). Because cyclins D2 and E1 have been described in fibroblast models to be transcriptionally induced by c-Myc (31–33), it is reasonable to speculate that the reduced expression of these cyclins after treatment with PIM447 may be a consequence of the reduction of c-Myc levels.

In addition to cell-cycle blockade, the mechanism of PIM447 in myeloma involves the induction of apoptosis, as indicated by the increase of Annexin-V–positive cells after treatment, and the cleavage of caspases and their substrate, PARP. It should be noted, however, that the levels of apoptosis are quite moderate in less sensitive versus sensitive cell lines, even at high doses and long times of exposure to the drug. It is well established that PIM kinases phosphorylate Bad at Ser112 as an antiapoptotic mechanism (27, 28). Accordingly, treatment of sensitive multiple myeloma cell lines with PIM447 reduced phospho-Bad (Ser112) levels, suggesting that apoptosis induction is mediated, at least in part, by dephosphorylated Bad which is known to bind and thereby inactivate the antiapoptotic proteins Bcl-2 and Bcl-xl (40). In addition, treatment with PIM447 also reduced Bcl-xl levels in MM1S, but not Mcl-1 which has been shown to be downregulated by PIM inhibition in CLL and MCL (15, 41). On the contrary, the levels of phospho-Bad (Ser112) did not decrease after treatment with PIM447 in less sensitive cells, suggesting a potential mechanism for apoptosis resistance. Moreover, our results indicate that basal levels of Bad and phospho-Bad could predict the sensitivity to PIM447, because a positive correlation was found between basal levels of these proteins and the IC₅₀ values. Although there is no clinical information validating this correlation, the results presented here strongly support the study of changes induced in the phosphorylation of Bad as a potential biomarker of response to this family of agents.

Together with the described mechanisms, our results also indicate that PIM447 represses mTORC1 signaling in myeloma cells. In this regard, treatment with PIM447 reduces the phosphorylation of TSC2 and PRAS40, two mTORC1 inhibitors in their unphosphorylated state (42, 43). The effect on phospho-TSC2 was detected earlier than the effect on phospho-PRAS40, suggesting that PIM447 primarily modulates mTORC1 activity in myeloma cells through TSC2 (9). As a result of mTORC1 inhibition, there was a decrease in phospho-4EBP1 and phospho-P70S6, two of its downstream targets implicated in the translation of proteins involved in survival, cell-cycle progression, and the translation machinery itself (42, 43). These results indicate that the inhibition of protein translation would contribute to PIM447-induced cell death, as previously observed with other PIM inhibitors (15–17, 44).

A second major finding of our work is the beneficial effect of PIM447 on myeloma bone disease. It is known that the interaction between myeloma cells and their microenvironment contributes to bone disease as a consequence of increased osteoclastogenesis and suppressed osteoblastogenesis (45). Moreover, osteolytic lesions are the most common complications of myeloma, developing in more than 80% of patients, and associated with decreased overall survival (46). This highlights the importance of finding treatments not only targeting malignant plasma cells but also having a beneficial effect on bone disease. On the one hand, it has been reported that PIM1 positively regulates RANKL-induced murine osteoclastogenesis via NFATc1 induction (11). In this work, we have shown that both PIM1 and PIM3 expressions increase during OC differentiation from human PBMCs, and, consequently, our in vitro experiments show that PIM447 inhibits OC formation and resorptive activity. Mechanistically, these effects seem to be mediated, at least in part, by disruption of the F-actin ring and downregulation of several molecules involved in OC differentiation and function, including NFATc1, the major transcription factor integrating RANKL signaling in terminal OC differentiation. On the other hand, PIM2 expression has been found to be upregulated in BM MSCs and pre-OBs from patients with myeloma in the presence of inhibitory factors of osteoblastogenesis in myeloma (i.e., IL3, IL7, TNFα, TGFβ, activin A) or after coculture with multiple myeloma cells, and PIM2 inhibition was able to resume in vitro osteoblastogenesis (12). In accordance with the later results, PIM447 treatment of MSCs from myeloma patients also significantly increased ALP activity and augmented mineralization in in vitro assays. These in vitro effects on bone had their correlate in our mouse model of disseminated multiple myeloma. As observed in the microCT images, the bone trabecular architecture at the metaphyses of distal femurs in PIM447-treated animals was similar to that of normal bone. In fact, these results match with significantly augmented bone parameters of bone volume density and trabecular number and reduced trabecular separation. Moreover, significantly lower serum levels of CTX (bone resorption marker) were observed in PIM447-treated animals than in the vehicle control, being indicative of an in vivo effect of this drug in reducing OC resorptive activity. PIM447 also augmented levels of the bone formation marker P1NP with respect to those mice treated with vehicle, although differences did not reach statistical significance. Taken together, our data seem to be indicative of PIM447 having antimyeloma activity and preventing bone loss, by both antiresorptive and bone-anabolic effects. This is in line with previous reports showing that PIM2 kinase is an important target of treatment for tumor progression and bone loss in myeloma (12).

One final important point is the very strong synergistic effect observed when PIM447 is combined with standard treatments such as dexamethasone, bortezomib + dexamethasone, lenalidomide + dexamethasone, and pomalidomide + dexamethasone. This makes this drug attractive to try to improve the efficacy of these standards of care, particularly with oral agents, as it would result in effective all-oral combinations.

Our preclinical results demonstrate the relevance of the new pan-PIM-kinase inhibitor, PIM447, in multiple myeloma as shown by the dual antitumoral and bone-protective effects displayed by this drug. Moreover, in addition to these beneficial effects, PIM447 strongly synergizes in vitro with standard-of-care treatments. In summary, the present work supports the use of PIM447 in patients with multiple myeloma, particularly in combination with the current standards of care.

M.V. Maetos is a consultant/advisory board member for Novartis. E.M. Ocio reports receiving a commercial research grant from and is a consultant/advisory board member for Novartis Pharmaceuticals. No potential conflicts of interest were disclosed by the other authors.

Conception and design: T. Paíno, A. Garcia-Gomez, M. Garayoa, E.M. Ocio

Development of methodology: T. Paíno, A. Garcia-Gomez, L. González-Méndez, L. San-Segundo, S. Hernández-García, E.M. Algarín, M. Martín-Sánchez

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L. González-Méndez, L. San-Segundo, D. Corbacho-González, C. Ortiz-de-Solorzano, M.-V. Maetos

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T. Paíno, A. Garcia-Gomez, L. González-Méndez, L. San-Segundo, S. Hernández-García, A.-A. López-Iglesias, E.M. Algarín, M. Martín-Sánchez, D. Corbacho, C. Ortiz-de-Solorzano, L.A. Corchete, N.C. Gutiérrez, M. Garayoa, E.M. Ocio

Writing, review, and/or revision of the manuscript: T. Paíno, A. Garcia-Gomez, C. Ortiz-de-Solorzano, N.C. Gutiérrez, M. Garayoa, E.M. Ocio

Study supervision: T. Paíno, M. Garayoa, E.M. Ocio

The authors thank Phil Mason for his help in reviewing English grammar.

This work was supported by funding from the RTICC-Hematology Group (RD12/0036/0058), Spanish FIS (PI11/01465 and PI15/02156) and FEDER Funds, Ministerio de Economía y Competitividad grants DPI2012-38090-C03-02 and DPI2015-64221-C2-2, AECC (GCB120981SAN), Junta de Castilla y León, Consejerías de Sanidad (GRS 862/A/13, GRS 1175/A/15 and BIO/SA05/14) y Educación (FIC335U14) and the Centro en Red de Medicina Regenerativa y Terapia Celular de Castilla y León.

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.