Pim2 is required for maintaining multiple myeloma cell growth through modulating TSC2 phosphorylation

Pim kinases are a family of serine/threonine kinases that includes three highly homologous members (Pim1, Pim2, and Pim3). Pim kinases were initially identified as frequent Proviral Integration sites of Moloney (PIM) murine leukemia virus in virus-induced lymphomas from infected mice.¹  Further transgenic mouse models and overexpression studies in cell lines confirmed that Pim kinases promote tumorigenesis, particularly in hematologic tissues, either alone or synergistically with other oncogenes, such as Myc and Bcl-2.^2,3  More recent findings also implicate Pims as promoting the growth of solid tumors, such as prostate cancer and gastric and liver carcinomas.^3-5  Importantly, the upregulation of Pims correlates with a poor prognosis in multiple cancer types, which suggests the therapeutic potential of Pim inhibitors in cancer.⁶ 

In contrast to many other kinases whose activities are tuned by phosphorylation status, the Pim kinases are constitutively active and lack regulatory domains.⁷  Instead, the Pim kinases are tightly regulated at both transcriptional and translational levels.⁷  The signals that induce Pim gene expression are diverse, including various cytokines, growth factors, and mitogenic stimuli in different cell types.⁷  Janus kinase/signal transducer and activator of transcription and nuclear factor κB pathway activation are among the most extensively studied Pim upstream regulators.^8,9  Once expressed, Pim kinases localize to both the cytosol and the nucleus to phosphorylate many important signaling molecules that promote cancer cell survival and proliferation. For example, Pim can directly phosphorylate Bcl-2-associated death promoter (BAD), which then interacts with and neutralizes antiapoptotic Bcl-2.¹⁰ Pim kinases are also capable of regulating cell cycle progression in various cellular contexts. It was found that Pim1 phosphorylates p21Cip1/Waf1 to regulate its stability and localization.¹¹  In addition, Pim kinases can either directly phosphorylate p27Kip1 to modulate its nuclear export and proteasome-dependent elimination or indirectly inhibit p27Kip1 transcription by phosphorylation and inactivation of FoxO1a and FoxO3a.¹²  Thus, the oncogenic functions of Pim kinases may be attributed to their involvement in regulating these relevant signaling cascades in cancer.

Pim kinases have been implicated in hematologic malignancies.^6,13  However, the expression and importance of each individual Pim kinase in different hematologic cancers have not been investigated systematically. In this study, we demonstrated that Pim2 is most highly expressed in multiple myeloma (MM) cells and that Pim2 expression is required for maintaining MM cell proliferation. In addition, we identify a novel Pim2-TSC1/2-mTOR-C1 signaling cascade that is essential for MM cell proliferation.

Microarray analysis of Pim1, Pim2, and Pim3 messenger RNA (mRNA) expression in Cancer Cell Line Encyclopedia (CCLE) cell lines was performed by using Spotfire software with publically available data (http://www.broadinstitute.org/ccle/home; Accession number: GSE36139; http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE36139).

Details are available in supplemental Materials and methods, available on the Blood Web site.

p-P70-T389 (9234), T-P70 (9202), p-S6RP-235/236 (4857), T-S6RP (2217), p-BAD-112 (5284), T-BAD (9239), poly adenosine 5′-diphosphate ribose polymerase (PARP) (9542), Pim2 (4730), TSC2 (3635), p-Akt substrate (9614), p-RPAS40-246 (2640), T-PRAS40 (2610), p-AMPK-172 (2531), T-AMPK (2532), mammalian target of rapamycin (mTOR) (2972), and Raptor (2280) antibodies were purchased from Cell Signaling Technology (Danvers, MA). Deptor (09-463) and Raptor (05-1470; used for immunoprecipitation [IP]) antibodies were purchased from Millipore (Billerica, MA). Actin (A5441) and Tubulin (T5168) antibodies and AICAR (A9978) were purchased from Sigma-Aldrich (St. Louis, MO).

Human wild-type (WT) and kinase dead Pim2 (short isoform) was in pT-REx-DEST30 vector and TSC2 mutants were in pDONR221 vector and pCMV vector. All mutants were generated by using a site-directed mutagenesis kit (Stratagene, Santa Clara, CA). p-LKO short hairpin RNA (shRNA) vector and packaging vectors were purchased from Sigma-Aldrich and used for all knockdown experiments. Target sequences for Pim2 were CCAGTCATTAAAGTCCAGTAT (match position-1226) and GCTTGACTGGTTTGAGACACA (match position-491). Target sequences for TSC2 were CGACGAGTCAAACAAGCCAAT (match position-4551), GCTCATCAACAGGCAGTTCTA (match position-1437), CGCTATAAAGTGCTCATCTTT (match position-2170), CCAACGAAGACCTTCACGAAA (match position-413), and GAGGGTAAACAGACGGAGTTT (match position-112). Target sequences for Deptor were GCCATGACAATCGGAAATCTA (match position-877), GCAAGGAAGACATTCACGATT (match position-1101), and CCTACATGATAGAACTGCCTT (match position-1578).

TSC2 knockout was generated by zinc finger nuclease (ZFN) system (CKOZFND1074) from Sigma-Aldrich following supplier’s instructions. Details are available in the supplemental Materials and methods.

p-LKO lentiviral vector containing different shRNA sequences was transfected into 293 cells with packaging vector pCMV-VSV-G and pCMVΔR.89 by lipofectamine (Invitrogen). After 48 hours, virus-containing medium was collected and used for infection of target cells in the presence of polybrene (Sigma-Aldrich).

For kinase assay, immunoprecipitates were washed 3 times with 1% NP-40 buffer and 2 times with buffer containing 50 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES; pH 7.5) and 5 mM MgCl₂. Kinase assay was performed with purified Pim2 (preincubated with dimethylsulfoxide or LGB321 for 60 minutes at room temperature) in kinase buffer (50 mM HEPES [pH 7.5], 5 mM MgCl₂, 1 mM dithiothreitol, 0.05% bovine serum albumin, and 20 μM adenosine triphosphate) at room temperature. Then reactions were stopped by adding sample buffer, boiled for 5 minutes, and analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis and immunoblotting.

For TSC2 IP studies, cell extracts were prepared in 1% NP-40 lysis buffer and immunoprecipitated with anti-TSC2 antibody. Precipitated protein was washed 3 times before running on 4% to 12% Bis-Tris gels (Invitrogen). For Raptor IP, cell extracts were prepared in 0.3% CHAPS buffer.

All studies were done in an Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC)–accredited animal facility and in compliance with the Institute for Laboratory Animal Research (ILAR) Guide for the Care and Use of Laboratory Animals. Details are available in supplementary Materials and methods.

To better understand the role of individual Pim family member in various hematologic malignancies, we analyzed Novartis internal and publically available databases for the mRNA expression of Pim1, Pim2, and Pim3. In particular, we examined relative levels of expression in different hematologic cancers, solid tumors, and their normal tissue counterparts. We found that the mRNA of all three Pims was more prominently expressed in hematologic cancers compared with solid tumors and that among hematologic cancers, Pim2 is most highly expressed in MM (data not shown). A similar analysis of the CCLE database¹⁴  confirmed high Pim2 mRNA expression in hematologic cancer cell lines (Figure 1A) and MM cells in particular (Figure 1B). Importantly, Pim2 protein was also expressed at much higher levels in MM cell lines (n = 12) compared with acute monoblastic leukemia (AML) cell lines (n = 7) (Figure 1C). A similar mRNA expression analysis for Pim1 reveals a higher level of expression in AML and chronic myeloblastic leukemia (CML) cells relative to other hematologic cancers (supplemental Figure 1A-C). In contrast, Pim3 showed no distinct expression pattern across cell lines of multiple solid tumors and hematologic cancer subtypes (supplemental Figure 1B,D). These data suggested that Pim2 might play an important role in MM, while Pim1 may be more relevant in AML and CML. In this study, we tested the hypothesis that MM depends on Pim2 activity.

We first used specific shRNAs to knockdown Pim2 expression in MM cells. These shRNAs significantly reduced Pim2 protein with concomitant inhibition of p-BAD (Ser-112), a direct downstream substrate of Pim2, in MM cell lines (Figure 2A). Pim2 knockdown led to a significant inhibition of MM cell proliferation (Figure 2B). To evaluate possible off-target inhibitory effects on MM growth, we expressed shRNA1226-resistant WT-Pim2 or a kinase dead (KD-Pim2) short isoform Pim2 in KMS-11.luc cells and found that expression of exogenous WT-Pim2, but not KD-Pim2, rescued KMS-11.luc from the effects of Pim2 knockdown (Figure 2C). These data demonstrate that Pim2 kinase activity is critical for MM cell proliferation.

The dependency on Pim2 activity for MM cell proliferation, together with many reports of Pim kinases involvement in tumorigenesis and pro-survival functions, strongly suggest its potential as a therapeutic target in MM.^2,5  Therefore, we developed a potent and selective small-molecule pan-Pim kinases inhibitor, 3-(S)-amino-piperidine pyridyl carboxamide (LGB321), and tested its efficacy in MM cell lines. LGB321 is a potent ATP competitive inhibitor of Pim1, Pim2, and Pim3, with inhibition constant of 1.0, 2.1, and 0.8 picomolar, respectively (supplemental Figure 2). We assayed the effect of LGB321 on the proliferation of MM cell lines with diverse genetic background, including KMS-26, KMS-11.luc, H929, KMS-34, and others. LGB321 significantly inhibited the proliferation of these cells in a dose-dependent manner (Figure 3A; supplemental Table 1). A recent study has identified Deptor, an mTOR complex component, as being highly expressed in 28% of MM samples, particularly in samples with c-MAF/MAFB translocations.¹⁵  The high Deptor levels may potentially provide prosurvival benefit through maintaining PI3K/Akt activation.¹⁵  Interestingly, we observed no correlation of Deptor expression levels with sensitivity to Pim compound treatment in MM cell lines, because we found that KMS-11.luc, KMS-26, MM1.S, and RPMI-8266 cells (all with a high or intermediate level of Deptor expression and c-MAF/MAFB translocation) are similarly sensitive to Pim inhibition as were other MM cells (supplemental Figure 3A; supplemental Table 1). The reduced growth upon Pim inhibition was mainly due to decreased cell proliferation, not apoptosis, since we did not observe substantial increase of cleaved PARP by Pim2 inhibition (Figure 3B; supplemental Figure 3B). Given that we observed a dramatic inhibition of MM cell proliferation by Pim inhibition, we examined the effect of Pim compounds on mTOR-C1 pathway activity, which is known to be critical for proliferation. In addition, it has been reported that Pim kinases modulate mTOR-C1 in certain malignancies, such as pre–T-cell lymphoblastic lymphoma and acute T-cell lymphoblastic leukemia.¹⁶  Indeed, p-p70S6K and its downstream substrate p-S6RP, major downstream targets of mTOR-C1, were strongly repressed in response to Pim inhibition (Figure 3C; supplemental Figure 3C). In addition, Pim2 knockdown by shRNAs consistently led to a significant suppression of mTOR-C1 activity (Figure 3D). We also observed inhibition of phosphorylation of 4EBP1 and eIF4G (data not shown), two other direct mTOR-C1 substrates.^17,18  Thus, Pim inhibition by either LGB321 or Pim2-specific shRNAs leads to a significant repression of mTOR-C1 activity, a well-known regulator of cell proliferation. We also tested the effect of LGB321 on the phosphorylation of cell cycle regulators p21Cip1/Waf1 and p27Kip1, both known substrates of Pim kinases,¹²  but did not observe significant inhibition in these MM cells (data not shown).

Disrupted binding of Raptor or enhanced binding of Deptor with mTOR-C1 complex are potential mechanisms by which Pim2 might regulate mTOR-C1 pathway activity in MM cells.^15,19  However, we observed no significant alterations of the interaction of Raptor and Deptor with mTOR-C1 complex after Pim inhibition (supplemental Figure 4A-B). Similarly, Pim inhibition showed no significant effect on the phosphorylation of AMPK and PRAS40, both regulators of mTOR-C1 and previously reported to be downstream of Pim(s)^20,21  (supplemental Figure 4C-D). We then asked whether Pim activates mTOR-C1 by negatively regulating the TSC2 tumor suppressor. TSC2 is a guanosine triphosphate (GTP)ase-activating protein for the small G-protein Rheb which, when bound to GTP, activates mTOR-C1.^22,23  Thus, TSC2 is a negative regulator of mTOR-C1 activity.^22,23  TSC2 is extensively regulated by phosphorylation on multiple sites, including consensus recognition sites (RXRXXS/T) for the protein kinase A, G and C family to which Pim and Akt belong.^24,25  To test this hypothesis, we treated KMS-11 and KMS-26 cells with LGB321 and examined phospho-TSC2 levels with a specific phospho-Akt substrate antibody following TSC2 IP. We observed a significant reduction of p-TSC2 upon Pim inhibition (Figure 4A). To further explore the role of Pim2 in the phosphorylation of TSC2 and regulation of the mTOR-C1 pathway, we expressed Pim2 in 293A cells, which lack endogenous expression of Pim2 but have a functional TSC1/2-mTOR-C1 signaling cascade. Pim2 expression in 293A cells led to the activation of mTOR-C1 pathway, as indicated by increased p-S6RP (Figure 4B), and this increase was inhibited by Pim inhibition with LGB321 (Figure 4B). More importantly, we also observed a higher level of p-TSC2 in Pim2-expressing 293A cells, and this increase was also blocked by LGB321 (Figure 4B). To investigate whether TSC2 plays a critical role in mediating Pim’s regulation of mTOR-C1, we used shRNAs to knock down TSC2 in KMS-11, KMS-26, KMS-34, and H929 cells. TSC2 knockdown resulted in inhibition of growth in all MM cells tested (supplemental Figure 5). This result suggests that mTOR-C1 hyperactivation after loss of TSC2 is detrimental to MM cell growth, which might be related to the feedback inhibition of Akt induced by hyperactivation of mTOR-C1.²⁶  To circumvent this limitation, we generated TSC2^−/− and TSC2^+/− 293A cells using ZFN technology.²⁷  As expected, knockout of TSC2 led to a significant increase of mTOR-C1 activity as shown by the increased levels of p-p70S6K (Figure 4C). When Pim2 was expressed in these TSC2^−/− cells, we observed no further increase in mTOR-C1 activity (Figure 4D). In addition, LGB321 could not inhibit the high levels of p-p70S6K in TSC2^−/− cells (Figure 4D). TSC2^+/+ and TSC2^+/− cells were fully responsive to both Pim2 expression and LGB321 inhibition (Figure 4D). These results demonstrate that Pim2 modulation of mTOR-C1 is TSC2 dependent.

Next, we wanted to identify the specific Pim2 site(s) on TSC2 that modulate its activity. Phospho-motif analysis identified 8 sites on TSC2 as potential Pim2 sites (Figure 5A). We generated two sets of TSC2 mutants: the first set contained single serine/threonine-to-alanine substitution for each site and the second set contained a single WT serine/threonine with all other 7 sites mutated to alanine (Figure 5A). Both sets of TSC2 mutants were transfected into TSC2^−/− 293A cells with or without Pim2 coexpression. Pim2 coexpression with WT-TSC2 led to an increase of p-TSC2 compared to expression of WT-TSC2 alone (Figure 5B-C). Expression of TSC2-S1798A resulted in loss of TSC2 phosphorylation by Pim2, while all other single-site mutants were phosphorylated as the WT-TSC2 (Figure 5B). Conversely, Pim2 can phosphorylate TSC2 only when Ser-1798 is preserved (Figure 5C). Together, the results indicate that Pim2 phosphorylates TSC2 primarily on Ser-1798. We then raised a phospho-specific antibody that recognizes p-Ser-1798 on TSC2 (supplemental Figure 6) and used it in a Pim2 in vitro kinase assay to demonstrate that WT-TSC2 and TSC2-S1798WT, but not the TSC2-S1798A, can be phosphorylated by Pim2 (Figure 5D). More importantly, if Pim2 protein is preincubated with LGB321, it is no longer able to phosphorylate WT-TSC2 and TSC2-S1798WT (Figure 5D). Taken together, these data strongly indicated that Pim2 phosphorylates TSC2 primarily on Ser-1798. We then hypothesized that the TSC2-S1798A mutant, like WT-TSC2, would be functional in repressing mTOR-C1 activity, but Pim2 would not be able to relieve this suppression. Indeed, both WT-TSC2 and TSC2-S1798A suppressed p-S6RP, and Pim2 relieved the repression by WT-TSC2, but not by TSC2-S1798A (Figure 5E). Thus, these results suggest that Pim2 phosphorylates TSC2 on Ser-1798 to relieve the suppressive effect of TSC2 on mTOR-C1 activity. Most importantly, Pim2 inhibition by LGB321 in KMS-11 and KMS-26 cells led to reduced phosphorylation on Ser-1798 (Figure 5F). Thus, our data demonstrate that Pim2 regulates mTOR-C1 activity by phosphorylating TSC2 on Ser-1798 to relieve its suppression on mTOR-C1.

It was recently shown that Pim1 can regulate mTOR-C1 through modulating the phosphorylation of PRAS40, a suppressive component of mTOR-C1 complex, in FDCP1 cells.²⁰  Interestingly, Pim inhibition showed no significant effect on p-PRAS40 (supplemental Figure 4D) and did not disturb the interaction of PRAS40 with mTOR-C1 complex in MM cells (data not shown). We hypothesized that in MM cells, PRAS40 phosphorylation is maintained by other kinases, such as Akt. In an attempt to elucidate the role of Pim2 and Akt in the regulation of mTOR-C1 in MM cells, we took advantage of our specific Pim inhibitor (LGB321) and PI3K inhibitor (BKM120).²⁸  The specificity of these inhibitors in MM cells is indicated by dephosphorylation of their respective downstream targets, BAD and Akt (Figure 6A). We found that inhibition of Pim or Akt alone is sufficient to significantly reduce mTOR-C1 signaling, while inhibition of both leads to a more prominent suppression of mTOR-C1 activity (Figure 6A). Interestingly, Pim inhibition showed a pronounced effect on repressing p-TSC2, but a minor effect on p-PRAS40 (Thr-246) (Figure 6A). Conversely, PI3K/Akt inhibition elicited a significant suppression on p-PRAS40 (Thr-246), while showing a minimal effect on p-TSC2 (Figure 6A). These results suggest that in MM cells, Pim2 primarily modulates mTOR-C1 activity by phosphorylating TSC2, while Akt primarily does it by phosphorylating PRAS40 (Figure 6B).

The in vivo activity of LGB321 was then evaluated in KMS-11.luc cells in a subcutaneous xenograft model. First, LGB321 pharmacokinetics was assessed from the concentration in plasma following a single oral dose of LGB321 at 10, 20, 50 or 100 mg/kg (Figure 7A). We then examined the modulation of p-S6RP and p-BAD in tumor lysate, normalized to total S6RP and total BAD. Unlike p-BAD whose inhibition is persistent through 24 hours following a single dose of 50 mg/kg or 100 mg/kg, p-S6RP inhibition was sustained for 24 hours only with a dose of 100 mg/kg (Figure 7B). We next tested the efficacy of LGB321, choosing a dosing regimen of 10, 20, or 50 mg/kg twice per day or 100 mg/kg once per day. A dose-dependent inhibition of tumor growth was observed, and complete stasis was achieved with both 50 mg/kg twice per day and 100 mg/kg once per day dosing (Figure 7C). These results suggest that the maximal in vivo effect correlates with significant and sustained inhibition of p-S6RP, not p-BAD. Tumor regression was not observed in the KMS-11.luc model, consistent with our previous results showing that inhibition of Pim kinases alone is not sufficient to induce apoptosis in this cell line. All regimens tested were well tolerated, as judged by minimal body weight loss (within a margin of 5% of initial weight) observed for all dosing groups (Figure 7D).

In this study, we identified that Pim2 expression is elevated in MM compared with other hematologic cancers. It is believed that Pim2 is a key regulator of signaling pathways involved in normal B-cell homeostasis.²⁹  Pim2 has been found to be upregulated and correlated with the progression of malignancies of B-cell origin, including chronic lymphatic leukemia, diffuse large B-cell lymphoma, and mantle cell lymphoma.^30,31  Our study here provides the first comprehensive analysis of the expression of each individual Pim family member across different hematologic malignancy cell lines. We found that Pim2 level is most prominent in MM cell lines, while Pim1 may be more relevant in AML and CML cell lines. We further demonstrate that Pim2 is required for MM cell proliferation because functional interference of Pim2 by either shRNA knockdown or pharmacologic inhibition leads to severe repressed proliferation. Our data suggest that the effect of Pim2 on proliferation is largely due to its regulation of mTOR-C1 pathway. Pims have previously been implicated in regulating proliferation through modulating cell cycle progression by phosphorylating p21Cip1/Waf1 and p27Kip1.^11,12,32  Interestingly, we found no changes in the phosphorylation status of either p21Cip1/Waf1 or p27Kip1, although we did observe a delay of G1→S phase transition in MM cells upon Pim inhibition (data not shown). Instead, our data suggest that the effect on cell cycle by Pim inhibition is more likely attributed to the repression of mTOR-C1, which is well-known to regulate cell cycle by controlling the translation of proteins essential for driving cell cycle progression.^32-34  Indeed, mTOR-C1 inhibition by rapamycin and other rapalogs leads to an accumulation of cells in the G1 phase of the cell cycle.³⁵ 

Mechanistically, we discovered that TSC2, a tumor suppressor and key negative regulator of mTOR-C1, is a novel Pim2 substrate, and we demonstrate that Pim2 directly phosphorylates TSC2 on Ser-1798 and relieves the suppression of TSC2 on mTOR-C1. The mTOR-C1 pathway integrates inputs from diverse signals, including energy, nutrients, stress, growth factor, and oxygen status, many of which converge on the TSC1/2 complex.³⁶  PI3K/Akt is by far the best characterized signaling pathway regulating the TSC1/2 complex. The activation of PI3K/Akt leads to phosphorylation of TSC2 on Ser939 and Thr1462 to relieve its suppression on mTORC-C1.^37,38  Pim and Akt kinase families have been implicated as playing complementary roles in promoting survival of hematologic cells through common substrates such as BAD.⁶  Our findings provide another example of their interrelationship because they can both regulate mTOR-C1 through modulating the phosphorylation of TSC2, although with different site preferences. In addition, both Pim and Akt are known to directly phosphorylate PRAS40, resulting in the dissociation of PRAS40, an inhibitory component, from the mTOR-C1 complex.^20,39  However, in the context of MM, we found that p-PRAS40 is more responsive to PI3K/Akt pathway modulation and less sensitive to Pim kinase regulation. Our study supports the finding that in MM cells, Pim2 functions in parallel with Akt in maintaining mTOR-C1 activity, with TSC2 as the major effector of Pim2 and PRAS40 as the main effector of Akt. Previous studies have established feedback activation of p-Akt and p-ERK upon mTOR-C1 suppression by rapalog.^40,41  Interestingly, we did not observe significant effects on p-ERK in MM cells upon Pim inhibition (supplemental Figure 7). More intriguingly, we consistently observed less of an increase or no increase of p-Akt (S473) by Pim inhibitor as compared with RAD001 treatment, even though both led to similarly significant suppression of mTOR-C1 (supplemental Figure 7). It is possible that Pim inhibition may attenuate mTOR-C2 activity through its modulation of TSC2. It has been previously been demonstrated that TSC1/2 complex promotes the activity of mTOR-C2, an upstream activator of Akt, independent of its inhibitory effects on mTOR-C1.^26,42  Further comprehensive studies will be needed to elucidate the mechanistic understanding of this observation.

The importance of Pim kinases in oncogenesis, particularly in hematologic malignancies, has led to an intensive campaign in both academia and industrial settings of developing potent Pim(s) inhibitors.^43-45  Some of the compounds have demonstrated activities in both cell lines and xenograft models, including both hematologic and solid tumor models.^46-48  Here, our findings strongly support the potential of targeting Pim2 in MM patients and suggest that a potent small-molecule Pim2 (or pan-Pim) inhibitor holds promise as an effective targeted therapy in MM. Indeed, our potent and selective Pim kinase inhibitor LGB321 showed significant suppression of MM cell growth both in vitro and in vivo.

Interestingly, we found that the effect of blocking Pim2 activity is mainly due to the inhibition of proliferation and not increased apoptosis, even though we observed a significant reduction of p-BAD. This suggested that the antiapoptotic functions of Pim kinases are more likely to be observed in the presence of death-promoting stimuli.^5,10  Thus, Pim inhibitors may augment the effect of apoptosis-promoting agents such as thalidomide and bortezomib in MM.

The antiproliferative effect of Pim inhibition is in part due to a significant suppression on mTOR-C1 pathway. Consistent with the mTOR-C1 pathway playing an integral role in cancer progression, multiple clinical trials are currently evaluating the efficacy of rapalogs in various tumor types, including hematologic malignancies. Interestingly, a previous study has found that Pim2 kinase can confer resistance of primary hematopoietic cells to rapamycin treatment.⁴⁹  Therefore, Pim inhibitors, compared with rapalogs, may generate better efficacy, while minimizing side effects due to the distinctive high expression of Pim2 in MM. Nevertheless, activation of PI3K/Akt or MAPK/ERK pathway either by mTOR-C1–mediated feedback regulation or by other signaling alterations may provide resistance mechanisms to Pim(s) targeted therapy. Therefore, rationally designed combination strategies of Pim inhibitor with PI3K/Akt or MAPK/ERK pathway inhibitors in clinical settings may achieve better clinical efficacy and potentially prevent or suppress resistance.

The authors sincerely thank Dr Leon Murphy, Dr Beat Nyfeler, and Dr Abdallah Fanidi for their input in scientific discussion, Wei Gao, Ken Crawford, and Laura Tandeske for helping with plasmids and protein preparation, and Charles Voliva, Natasha Aziz, Nancy Pryer, and Emma Lees for their support.

This work was supported by the Presidential Postdoctoral Program and the Department of Oncology in Novartis Institutes for Biomedical Research.

Contribution: J.L., J.H., and P.D.G. developed the concept, designed the experiments, and prepared the manuscript; J.L. performed most of the experiments and analyzed data; T.Z. and Y.D. contributed to shRNA knockdown studies; X.-H.N., J.H., and J.C. contributed to proliferation assays; J.S. contributed to constructs generation; Y.W. and J.L.L. performed in vivo experiments; J.Y. contributed to data analysis; and K.S. provided conceptual advice and edited the manuscript.

Conflict-of-interest disclosure: J.L., T.Z., Y.D., X.-H.N., J.C., J.S., J.Y., Y.W., J.L.L., J.H., and P.D.G. are current employees at Novartis Institute for Biomedical Research. The remaining authors declare no competing financial interests.

Correspondence: Pablo D. Garcia, 4560 Horton St, Emeryville, CA, 94608; e-mail: pablo.garcia@novartis.com.