MK-2206, an AKT Inhibitor, Promotes Caspase-Independent Cell Death and Inhibits Leiomyoma Growth

Abstract

Uterine leiomyomas (ULs), benign tumors of the myometrium, are the number one indication for hysterectomies in the United States due to a lack of an effective alternative therapy. ULs show activation of the pro-survival AKT pathway compared with normal myometrium; however, substantial data directly linking AKT to UL cell survival are lacking. We hypothesized that AKT promotes UL cell survival and that it is a viable target for inhibiting UL growth. We used the investigational AKT inhibitor MK-2206, currently in phase II trials, on cultured primary human UL and myometrial cells, immortalized leiomyoma cells, and in leiomyoma grafts grown under the kidney capsule in mice. MK-2206 inhibited AKT and PRAS40 phosphorylation but did not regulate serum- and glucocorticoid-induced kinase and ERK1/2, demonstrating its specificity for AKT. MK-2206 reduced UL cell viability and decreased UL tumor volumes. UL cells exhibited disruption of mitochondrial structures and underwent cell death that was independent of caspases. Additionally, mammalian target of rapamycin and p70S6K phosphorylation were reduced, indicating that mammalian target of rapamycin complex 1 signaling was compromised by AKT inhibition in UL cells. MK-2206 also induced autophagy in UL cells. Pretreatment of primary UL cells with 3-methyladenine enhanced MK-2206-mediated UL cell death, whereas knockdown of ATG5 and/or ATG7 did not significantly influence UL cell viability in the presence of MK-2206. Our data provide molecular evidence for the involvement of AKT in UL cell survival and suggest that AKT inhibition by MK-2206 may be a viable option to consider for the treatment of ULs.

Uterine leiomyoma (ULs), also known as uterine fibroids, are nonmalignant myometrial tumors. ULs are common, occurring in up to 80% of reproductive age women, and symptomatically affecting 20%–30% of women. ULs disrupt daily life with abdominal pain, increased abdominal girth, heavy menstrual bleeding, frequent urination, painful intercourse, and pregnancy complications (1). Treatments for UL are lacking due to a dearth of biological information about tumor signaling. Consequently, UL is the top indication for hysterectomy in the United States, and total health care costs are estimated at $34 billion annually (2). Nonsurgical options for women with UL are needed.

AKT signaling has been broadly implicated as important for UL biology. AKT is a serine/threonine kinase and oncogene that is a signaling hub to control proliferation, cell cycle, and apoptosis (3). UL tumors express higher phosphorylated AKT than matched myometrium (4–8). Previously, we reported that AKT inhibition using API-59 effectively reduced cell viability and induced UL cell apoptosis (9). However, detailed signaling linking AKT to UL cell survival and the in vivo effects of AKT inhibition on UL tumor progression remain unstudied.

Cellular death occurs through multiple pathways. Apoptosis is the most understood cell death pathway and is the most commonly assessed when testing potential therapies. However, necrosis and other modes of caspase-independent cell death can occur pathologically and in response to chemotherapies (10, 11). Furthermore, both apoptosis and necrosis can be attenuated by AKT signaling, highlighting the importance of distinguishing between apoptotic and nonapoptotic cell death in response to AKT inhibition for disease treatment (12).

MK-2206 is a potent allosteric inhibitor of all 3 AKT isoforms with a dissociation constant K_d in the nanomolar range (13). MK-2206 is effective in reducing xenograft growth in models of breast, prostate, nonsmall cell lung, and ovarian cancers (14, 15) and has shown antitumor properties in a small phase I trial, while causing only minor side effects (16). Currently, phase I and II trials are being conducted with MK-2206 in solid tumors and blood cancers (www.clinicaltrials.gov).

In the current studies, we sought to determine whether MK-2206 could effectively reduce UL cell viability for use as a potential therapy for uterine fibroids. We show that MK-2206 inhibits AKT activity and promotes cell death in the absence of caspase activation. We demonstrate, for the first time, that AKT can be targeted in vivo for UL therapy using MK-2206 in a xenograft model of tumor growth. These data strongly suggest that targeting AKT represents a potential therapeutic avenue for treating ULs.

Materials and Methods

Reagents

The MK-2206 was generously provided by Merck Sharp & Dohme Corp and the National Cancer Institute, National Institutes of Health. Z-VAD-FMK, 3-methyladenine (3-MA), etoposide, and hygromycin B were purchased from Sigma.

Tissue collection and cell culture

Leiomyoma and myometrial tissues were collected from premenopausal women undergoing hysterectomy or myomectomy (leiomyoma only) at Northwestern University Prentice Women's Hospital (Chicago, Illinois), according to an IRB-approved protocol. Women included in the study were not taking hormonal contraceptives or GnRH antagonist or agonist at least 3 months prior to tissue collection. Consent was granted by all women included in the study. Primary leiomyoma and myometrial cells were isolated and cultured as previously described (9) and were passage 3 or less. Immortalized leiomyomas (DD-HLM) were obtained from Dr Ayman Al-Hendy. Cell lines were maintained in DMEM/F12 supplemented with 10% fetal bovine serum, containing estrogen and progesterone, and 1% pen-strep at 37°C and 5% CO₂.

Cell culture treatments

Prior to apoptosis, viability, or proliferation assays cells were serum starved 24 hours without phenol red, and treated in the presence of 10% fetal bovine serum. For the caspase inhibition experiment, primary leiomyoma cells were pretreated for 3 hours with 40 μM Z-VAD-FMK prior to adding 10 μM MK-2206 or 300 μg/ml hygromycin B for 48 hours. For the autophagy inhibition experiment using 3-MA, primary leiomyoma cells were pretreated with 5 mM 3-MA for 3 hours prior to adding 10 μM MK-2206 for 48 hours. Vehicle indicates dimethylsulfoxide.

Protein isolation, Western blotting, and antibodies (Table 1)

Following treatment, cells were washed (PBS, 2×) before adding M-PER mammalian protein extraction buffer (Invitrogen) supplemented with protease and phosphatase inhibitors (Sigma Aldrich). Primary antibodies: phosphorylated AKT (p-AKT) S473 (1:2000), T308 (1:1000); pan-AKT (1:2000); p-PRAS40 T246 (1:2000); PRAS40 (1:2000); serum- and glucocorticoid-induced kinase (SGK) (1:1000); p-ERK1/2 T202, Y204 (1:2000); ERK1/2 (1:2000); cleaved caspase 3 (Asp175) (1:1000); LC3A/B (1:2000); ATG5 (1:2000); ATG7 (1:2000); p-mTOR S2448 (1:1000); mammalian target of rapamycin (mTOR) (1:1000); p-p70S6K T389 (1:2000); p70S6K (1:2000); cleaved PARP (Asp 214) (1:1000) were from Cell Signaling Technology. p-SGK1 T256 (1:1000) was from Santa Cruz Biotechnology. Actin (1:10 000) was from Sigma. α-Tubulin (1:10 000) was from Millipore Corp. Membranes were developed using chemiluminescence (Pierce Chemical Co.) detection agents and Fuji-Film Intelligent Dark Box. Densitometry was performed using ImageJ software. Results are expressed as the fold change compared with vehicle and normalized by the loading control or total protein levels for phosphorylated proteins.

Transmission electron microscopy

Immortalized leiomyoma cells were grown on Thermanox plastic coverslips. The Northwestern University Cell Imaging Facility processed the samples and assisted with imaging.

Immunofluorescence

Primary leiomyoma cells were grown on glass coverslips. Following treatment, cells were washed (PBS, 2×), fixed (10% formalin), and permeabilized (0.1% Triton-X). Cells incubated with primary antibody (LC3 1:200; Cell Signaling) and antirabbit fluorescent secondary antibody (1:250; Invitrogen). Slides were visualized with a Zeiss Inverted Fluorescent microscope.

Immunohistochemistry

Paraffin-embedded sections were deparaffinized and stained using the Envision DAB HRP kit (Dako) or hematoxylin and eosin. Antibodies were cleaved caspase 3 (Asp 175) 1:50 from Cell Signaling and LC3A/B 1:2500 from MBL International Corp. Slides were visualized with a Zeiss Inverted microscope.

Small interfering RNA (siRNA) transfection

Primary leiomyoma cells were transfected with siRNA (100 nM ATG5 and ATG7) (Dharmacon) using Lipofectamine RNAiMax (Invitrogen). Serum starvation and MK-2206 treatments began 48 hours after the initial transfection. Protein was isolated and cell viability was measured after 48 hours of MK-2206 treatment.

Cell viability assay

Primary leiomyoma were plated at 3000 cells per well in a 96-well plate. WST1 reagent (1:10; Biovision) was added to wells the final 2 hours of treatment. Data were analyzed according to the manufacturer's instructions.

Adenovirus infection

Immortalized leiomyoma cells were infected with 150 multiplicity of infection cytomegalovirus (CMV) and Myr-AKT for 24–48 hours prior to 24-hour serum starvation followed by MK-2206 treatment. CMV and Myr-AKT was provided by Dr. Romana Nowak (University of Illinois-Urbana-Champaign).

Leiomyoma xenografting and in vivo MK-2206 treatment

Primary leiomyoma tissues were collected and cultured as above. Primary cell preparation and xenografting were performed as described previously (17). One million primary leiomyoma cells were used per tumor. MK-2206 was dissolved in 30% captisol and administered (360 mg/kg) once weekly by oral gavage for 3 weeks. Vehicle was 30% captisol solution without MK-2206. Following the last MK-2206 dose, mice were humanely destroyed; kidneys with tumor grafts and lungs were harvested and fixed or frozen. All animal experiments were approved by the Northwestern University Animal Care and Use Committees. All experiments were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Animals.

Tumor volume measurements

Images of leiomyoma grafts were taken while still embedded under the kidney capsule. Graft volume was calculated as previously described (17).

Statistics

Results are presented as mean ± SEM unless otherwise stated. Data were analyzed using the paired Student's t test or one-way ANOVA as appropriate. A value of P < .05 was considered statistically significant.

Results

MK-2206 selectively inhibits AKT in ULs

UL tissues have been reported to express higher p-AKT than myometrial tissues (4–6). Increased phosphorylation of AKT was confirmed in primary leiomyoma cell in culture and suggests that AKT is differentially regulated between UL and myometrial cells (Figure 1A).

Figure 1.

Inhibition of AKT with MK-2206 in UL Cells. A, Protein from primary human UL and myometrial cells was analyzed by Western blot. Bar graphs represent relative density of the protein bands. *, P < .05; n = 3 matched patients. B, Primary UL and myometrial cells were treated with increasing concentrations of MK-2206 for 24 hours prior to Western blot. Bar graphs represent relative density of the protein bands. **** (leiomyoma), #### (myometrial) indicates P < .0001 and **, P < .01 using one-way repeated measures ANOVA. C and D, Immortalized UL cells were treated with MK-2206 for 24 hours prior to Western blot. Bar graphs represent relative density of the protein bands. ***, P < .001 using a one-way repeated measures ANOVA. All results represent the mean ± SEM. n = 3. Leio, leiomyoma; Myo, myometrial.

Given that p-AKT was increased in UL cells, the therapeutic potential of targeting AKT in ULs was explored. MK-2206, an AKT inhibitor, previously demonstrated to reduce AKT activity, reduce sc tumor volume in mice, and to have a mild side effect profile in phase I clinical trials, was used in this study (14–16). First, MK-2206's optimal concentration and specificity for AKT inhibition in UL cells was assessed. Increasing concentrations of MK-2206 effectively inhibited p(Ser473)-AKT and p(T246)-PRAS40 (proline-rich AKT substrate of 40 kDa), an AKT substrate, in primary UL and myometrial cells at 100 nM with full p-AKT ablation at 1 μM (Figure 1B). Furthermore, 1 μM MK-2206 reduced p(Thr308)-AKT in primary leiomyoma cells (Supplemental Figure 1A published on The Endocrine Society's Journals Online web site at http://endo.endojournals.org).

To characterize the specificity of MK-2206 for AKT, primary and immortalized leiomyoma cells were used (18). Immortalized leiomyoma cells showed sufficient inhibition at 100 nM MK-2206 and complete inhibition at 1 μM MK-2206 of p-AKT on both S473 and T308 and p(T246)-PRAS40 (Figure 1C). Although AKT signaling was inhibited, glycogen synthase kinase 3 β (GSK3b) phosphorylation, a well characterized AKT substrate, was unaffected (Figure 1C). GSK3b is regulated by ribosomal protein S6 kinase and serum and glucocorticoid induced kinase (SGK1) as well as AKT, and the activities of these kinases may compensate for GSK3b phosphorylation in the presence of MK-2206. Furthermore, MK-2206 did not inhibit non-AKT targets SGK1 or ERK1/2 phosphorylation at any concentration tested in both primary and immortalized leiomyoma cells (Figure 1D and Supplemental. Figure 1 A). These results demonstrate the specificity of MK-2206 for AKT.

MK-2206 reduces UL cell viability independently of caspases

AKT regulates survival through apoptosis (19) and autophagy (20). Therefore, the effects of MK-2206 on apoptosis and autophagy in ULs were measured. MK-2206 treatment decreased primary UL and myometrial cell viability (Figure 2A) in a dose-dependent manner with a significant decrease at 10 μM MK-2206. However, 10 μM and 25 μM MK-2206 did not induce robust caspase 3 or poly ADP ribose polymerase (PARP) cleavage at 48 hours in primary or immortalized UL cells (Figure 2, B and C). Importantly, hygromycin b, an antibiotic, did induce both caspase 3 and PARP cleavage at 24 hours, indicating that certain conditions do activate caspases in primary and immortalized UL cells (Figure 2, B and C). To confirm that the reduction in UL viability in response to MK-2206 did not involve caspases, primary UL cells were treated with the pan-caspase inhibitor, Z-VAD-FMK. Treatment of primary UL cells with hygromycin b decreased cell viability and inhibition of caspases with Z-VAD-FMK maintained UL cell viability (Figure 2D). In contrast, MK-2206 treatment significantly reduced primary UL cell viability, and caspase inhibition was not able to prevent this decrease (Figure 2D). In addition, caspase inhibition potentiated cytotoxicity in the presence of MK-2206, which is similar to previous reports of enhanced nonapoptotic cell death in response to caspase inhibition (21). In addition to the decrease in cell viability, electron microscopy of immortalized UL cells treated with MK-2206 revealed disrupted mitochondria at both 1 μM and 10 μM MK-2206 (Figure 2E), demonstrating that MK-2206 is damaging UL cells, which eventually leads to cell death.

Figure 2.

Effect of MK-2206 on UL Cell Viability. A, Primary UL and myometrial cells were incubated with MK-2206 for 48 hours prior to cell viability assay using WST1. **** (leiomyoma), #### (myometrial), P < .0001. B, Primary UL cells were treated with MK-2206 for 48 hours prior to Western blot. Red arrowhead indicates cleaved caspase 3 bands. H-hygromycin b (150 μg/mL) was used as a positive control for caspase induction; n = 2. C, Immortalized UL cells were treated as in B; n = 3. D, Primary UL cells were treated with MK-2206 for 48 hours after an initial 3-hour Z-VAD-FMK pretreatment and cell viability was assessed by WST-1 incubation. Vehicle, dimethylsulfoxide; n = 4; n.s., P > .05; *, P < .05. E, Immortalized UL cells were grown on plastic Thermanox coverslips and treated with MK-2206. Cells were fixed and processed for electron microscopy. Red arrowhead indicates normal mitochondria. Black arrowhead indicates disrupted mitochondria. (−) indicates the absence of MK-2206. (+) indicates the presence of MK-2206; n = 2. All results represent the mean ± SEM.

MK-2206 induces autophagy

Previous work has shown that MK-2206 induces autophagy in glioblastoma and leukemia cells by inhibiting mTOR complex 1 (mTORC1) activity downstream of AKT (22, 23). Thus, we explored whether MK-2206 could inhibit mTORC1 activity (p(S2448)-mTOR) in UL cells. MK-2206 reduced p(S2448)-mTOR and p(T389)-p70S6K, an mTORC1 substrate (Figure 3A) in the immortalized UL cells. The reduction of p(T389)-p70S6K was confirmed in primary UL cells from 6 patients (Figure 3B), demonstrating MK-2206 inhibits the primary regulator of autophagy, mTORC1, in UL cells.

Figure 3.

Effect of MK-2206 on Autophagy in UL Cells. A and B, Immortalized (panel A, n = 3) or primary (panel B, n = 6) UL cells were treated with 10 μM MK-2206 for 24 hours prior to Western blot. Bar graphs represent relative density of the indicated proteins. C, Primary UL cells were treated with MK-2206 prior to Western blot. Bar graphs represent relative density of indicated protein; n = 4. D, Primary UL cells were grown on glass coverslips, treated for 24 hours, and stained for LC3. Vehicle, dimethylsulfoxide; n = 3. E, Immortalized UL cells were infected with adenovirus-CMV or adenovirus-myristolated AKT (Myr-AKT) prior to treatment for 24 hours and Western blotting. Bar graphs represent relative density of indicated protein; n = 4. *, P < .05; n.s., P > .5. (−) indicates the absence of MK-2206. (+) indicates the presence of MK-2206. All results represent the mean ± SEM.

An essential step in autophagy is the cleavage of microtubule-associated protein 1A/1B-light chain 3 (LC3) I to produce LC3II. LC3II then localizes to autophagosome membranes and is essential for autophagosome formation (24). MK-2206 treatment in primary UL cells increased the LC3II cleavage product as observed by Western blotting (Figure 3C) and visualized as punctate staining by immunocytochemistry (Figure 3D). To demonstrate that autophagy induction was specific to inhibition of the AKT pathway in response to MK-2206, immortalized UL cells were infected with a myristolated-AKT (Myr-AKT) adenovirus, which is resistant to the effects of MK-2206, and assessed autophagy. Myr-AKT constitutively targets AKT to the cell membrane and replaces the need for phosphatidylinositol (3,4,5)-trisphosphate to bind the pleckstrin homology (PH) domain of AKT for activation (25). Because MK-2206 targets the PH domain of AKT, Myr-AKT circumvents MK-2206's mode of action. In control CMV-infected cells, 1 μM MK-2206 decreased p(Ser473)-AKT levels as expected whereas MK-2206 failed to decrease p(Ser473)-AKT levels in cells infected with Myr-AKT (Figure 3E). In addition, MK-2206 induced autophagy in UL cells infected with control CMV, whereas Myr-AKT expression in the presence of MK-2206 rescued cells from autophagy as assessed by the lack of LC3II cleavage product (Figure 3E). These data indicate that MK-2206 induces autophagy in UL cells through AKT inhibition.

Autophagy can be prosurvival or prodeath under normal physiologic conditions (24) and in many types of cancer (26). To determine the role of autophagy in the induction of UL cell death by MK-2206, autophagy was inhibited by knocking down the essential autophagy proteins ATG5 and ATG7 in addition to using a chemical inhibitor, 3-MA, prior to assessing cell viability. Primary UL cells transfected with control siRNA and treated with 10 μM MK-2206 underwent autophagy as assessed by LC3II and significantly reduced cell viability (Figure 4, A and B). Conversely, knockdown of both ATG5 and ATG7 concurrently in the presence of MK-2206 attenuated autophagy induction but failed to rescue primary UL cells from death (Figure 4, A and B). Similarly, immortalized UL cells transfected singularly with ATG5 or ATG7 siRNA reduced LC3II induction in response to MK-2206 treatment, whereas immortalized UL cells transfected with control siRNA and treated with MK-2206 showed robust LC3II (Supplemental Figure 2A). MK-2206 treatment reduced immortalized UL cell viability in control transfected cells, whereas knock-down of ATG5 or ATG7 did not significantly rescue immortalized UL cell viability in the presence of MK-2206 (Supplemental Figure 2B). Additionally, autophagy initiation was inhibited with 3-MA. Cells were pretreated with 3-MA and then treated with MK-2206, which enhanced the ability of MK-2206 to reduce UL cell viability (Figure 4C), suggesting a potential role of autophagy in cell survival. However, the further reduction in UL cell viability was modest and there was no significant effect when ATG5/7 were knocked down, indicating that autophagy does not significantly rescue or contribute to MK-2206 driven UL cell death under the conditions used in this study. Further experimentation is needed to elucidate the role of autophagy upon inhibition of AKT in UL cells.

Figure 4.

MK-2206-Induced UL Cytotoxicity Is Independent of Autophagy. A and B, Primary UL cells were transfected with nontargeting (control) siRNA or siRNA targeting both ATG5 and ATG7 for 48 hours prior to treating with 10 μM MK-2206 for 48 hours; n = 3. A, Western blot. B, Cell viability assay using WST1. C, Primary UL cells were pretreated with 5 mM 3-MA prior to adding 10 μM MK-2206 for 48 hours. Cell viability was assessed by WST-1; n = 3. *, P < .05; n.s., P > .5; Vehicle, dimethylsulfoxide. All results represent the mean ± SEM. sicont, small interfering control.

MK-2206 decreases growth of leiomyoma grafts

The effect of MK-2206 on human leiomyoma tumors was investigated using the previously described xenograft model (17). Primary UL cells were suspended in collagen and surgically implanted under the kidney capsule of immunocompromised female mice (17). Myometrial cells suspended in collagen and implanted under the kidney capsule do not form adequate tissue growths; therefore, myometrial cells were not used (17). After 2 weeks, mice were treated for 3 weeks with MK-2206 (Figure 5A). Mice did not exhibit any reductions in body weight over the course of the experiment, suggesting minimal MK-2206 toxicity (Supplemental Figure 1A). After 3 weeks of MK-2206 treatment, leiomyoma tumors from mice treated with MK-2206 showed a significantly lower leiomyoma tumor volume (Figure 5, B and C) compared with vehicle-treated mice. To verify the activity and efficacy of MK-2206 in vivo, mouse lung tissue was collected and assessed for p(Ser473)-AKT and p(T246)-PRAS40. Compared with control mice, MK-2206-treated mice showed reduced levels of p(Ser473)-AKT, indicating that the dose and treatment schedule effectively inhibited AKT activity (Supplemental Figure 1B). Immunohistochemical staining of the leiomyoma tumors showed areas of pynknosis, esonophilic cytoplasm, and detachment from the surrounding cells as assessed by pathologic examination of the hematoxylin-eosin staining, all indicative of cell death in the MK-2206-treated grafts (Figure 5D, red arrows). Furthermore, staining for cleaved caspase 3 was absent in the MK-2206-treated grafts (Figure 5D). Although LC3 immunohistochemistry cannot distinguish between LC3I and LC3II, increased intensity of LC3 using immunohistochemistry can be used to assess autophagy in tissues (27–29). Autophagy marker LC3 was present in both xenografts from vehicle- and MK-2206-treated mice, but the intensity of LC3 staining was increased in MK-2206 treatment group, indicating autophagy induction. These observations support the in vitro data demonstrating MK-2206 induces autophagy but does not induce caspase-dependent apoptosis in UL xenografts (Figure 2), while also demonstrating that in vivo AKT inhibition with MK-2206 is sufficient to reduce primary leiomyoma tumor volume.

Figure 5.

Effect of MK-2206 on Leiomyoma Tumors in Mice. A, Timeline of leiomyoma xenograft experiment. B, Tumor volume after 3 weeks of MK-2206; patients cells used in 17 mice per treatment group (n = 4). C, Representative image of kidneys with grafts from one patient at the time of harvest. D, Immunohistochemical staining of xenografts. Arrows indicate dying cells as identified by a clinical pathologist (J. J. Wei) using the criteria of pyknosis, eosinophillic, cytoplasm, and detachment from surrounding area. Cleaved caspase 3-positive control − endometrial cancer xenograft treated with MK-2206+progesterone (50). Vehicle, 30% captisol dissolved in sterile water. *, P < .05. All results represent the mean ± SEM. H and E, hematoxylin and eosin.

Discussion

ULs account for 200 000 hysterectomies each year in the United States (30). Unfortunately, women have few options other than hysterectomy to manage their ULs, and existing therapies cannot be used long term or are ineffective (31). Therefore, therapies that can be used in place of surgery are attractive alternatives. In this manuscript, AKT inhibition using the experimental AKT inhibitor MK-2206 promotes leiomyoma cell death independently of caspases, but with mitochondrial disruption. Despite the induction of autophagy and reduction of mTORC1 signaling, autophagy did not have a robust effect on UL cell death in response to MK-2206. Importantly, in vivo MK-2206 attenuated UL tumor growth without caspase activation, supporting the potential of targeting AKT with MK-2206 as an alternative treatment for UL.

Previous data support that the phosphatidylinositol 3-kinase (PI3K) and AKT pathways are important for leiomyoma cell survival. Although the myometrium expresses basal p-AKT, leiomyoma tumors and cells exhibit significantly increased AKT phosphorylation compared with normal myometrium (4–6, 9). Furthermore, higher p-AKT was correlated with larger tumors, indicating that AKT may be involved in leiomyoma growth (8). The cause of increased AKT phosphorylation in leiomyoma cells is unknown. Nonetheless, receptor tyrosine kinases, steroid hormone receptors, and their ligands are overexpressed in leiomyoma, possibly activating PI3K and AKT (32, 33). Recently, In addition to expression data, the biological roles of PI3K and AKT have not been thoroughly explored. One study showed that inhibition of PI3K and mTOR reduced immortalized leiomyoma cell proliferation, but survival and cell death were not assessed (34). Previously, we showed that the AKT inhibitor API-59 induced apoptosis in primary leiomyoma cells, but the in vivo effects of AKT inhibition are unknown (9). Finally the lack of therapeutic options for uterine leiomyoma requires that alternatives be identified using leiomyoma biology. Therefore, the functions of AKT in primary and immortalized leiomyoma cells as well as in an in vivo model were further explored.

MK-2206, an experimental AKT inhibitor was used to target AKT in ULs. An allosteric AKT inhibitor, MK-2206, is currently in phase II clinical trials for solid and blood cancers in adults as well as a phase I clinical trial for pediatric cancer patients (www.clinicaltrials.gov). Completed phase I clinical trials showed that MK-2206 has mild side effects, including skin rash and gastrointestinal (GI) upset (16, 35). Furthermore, preclinical animal data suggested that MK-2206 negatively affected xenograft volume of breast, ovarian, prostate, and nonsmall cell lung cancers (14, 15, 36). Given that MK-2206's side effects were mild in human studies and the compound could prevent xenograft growth in vivo, MK-2206 was tested in ULs.

We found in this work and in a previous publication that inhibiting AKT in leiomyoma cells using 2 AKT inhibitors reduced leiomyoma cell survival via distinct pathways. MK-2206 induced caspase-independent cell death, whereas our previous work showed that AKT inhibitor API-59 induced caspase-dependent apoptosis (9). This difference could be due to the compounds' modes of actions and differential substrate regulation. API-59 inhibits AKT activity by binding to its catalytic site but does not affect AKT phosphorylation. In addition, API-59 compound is an ellipticine that intercalates DNA, which involves additional mechanisms that could trigger apoptosis (37, 38). On the other hand, MK-2206 is an allosteric inhibitor that binds to the PH domain of AKT and reduces AKT phosphorylation (13). We previously reported that API-59 promotes FOXO1 activity (9), a well-characterized AKT substrate involved in apoptosis, whereas MK-2206 failed to regulate FOXO1 (data not shown). This differential response between inhibitors of a similar pathway highlights the importance of individually analyzing inhibitors with the potential to treat human tumors.

The mechanisms associated with MK-2206-driven caspase-independent cell death remain unknown. Studies have shown that MK-2206 induces minimal apoptosis as a single agent in multiple cancers (14, 15, 39), and MK-2206 decreases neuroblastoma cell viability without caspase activation (40). Determining the signaling downstream of AKT inhibition with MK-2206 will contribute greatly to UL biology and may predict new drug targets.

Signaling connecting AKT to caspase-independent cell death is poorly understood. There are multiple caspase-independent cell death mechanisms, which have AKT regulated components and are potential pathways for UL cell death in response to MK-2206, such as necrosis and cell death mediated by noncaspase proteases. AKT is well documented to protect cells from mitochondrial outer membrane permeabilization, which is central to both caspase-dependent and -independent cell death pathways (41, 42). For example, mitochondria membrane transition pore opening, key to necrosis, requires bax and p53, both of which are negatively regulated by AKT (43). Cyclophilin D, critical for the mitochondria membrane transition pore during necrosis, is required for prosurvival AKT signaling during cardiac ischemia/reperfusion preconditioning (12, 44). Importantly, ULs express increased bax and p53 proteins, which could poise UL cells for necrosis in the absence of prosurvival AKT (4, 45). Given that AKT inhibition caused mitochondrial disruption, investigation of the mitochondrial membrane potential and translocation of mitochondrial proteins, such as apoptosis-inducing factor, a well characterized mediator of caspase-independent cell death, are underway (46). A detailed signaling pathway connecting AKT to caspase-independent cell death and that promotes survival of UL cells will be the focus of future studies.

In UL cells MK-2206 induces autophagy in UL cells. Single knockdown of ATG5 or ATG7 in immortalized UL cells as well as double knockdown of ATG5 and ATG7 in primary UL cells, failed to significantly promote UL death or enhance cell viability. However, inhibition of the onset of autophagy by pretreating with 3-methyladenine (3-MA) followed by MK-2206 treatment did promote a further decrease in UL cell viability than MK-2206 alone. These discordant results suggest that there are off-target effects of using 3-MA or that a more complex mechanism involving autophagy exists in UL cells. For example, 3-MA prevents the initiation of autophagy by inhibiting PI3K class III. However, 3-MA can also inhibit PI3K class I and exhibits additional effects independent of autophagy inhibition, which could influence UL cell viability (47, 48). Therefore, assessing the chemical inhibition and siRNA results together, we conclude that the role of autophagy in UL cell survival in response to MK-2206 remains unclear and that further experimentation is required. It has been demonstrated in glioblastoma and leukemia cells that autophagy induced by MK-2206 was prosurvival (22, 23). However, recent studies show that autophagy that has little effect on cell survival can be associated with caspase-independent cell death, highlighting the intricate functions of autophagy in response to stimuli and in multiple cell types (49, 50). In our studies, mTORC1 signaling was reduced and may play a role for UL cell survival. The distinct function of autophagy induced by AKT inhibition and mTORC1 in UL cells is under investigation.

The data in this study and previous data support that AKT is a promising target for leiomyoma therapy. Increased AKT phosphorylation in cultured leiomyoma cells, as well as a reduction of leiomyoma cell viability with AKT inhibition, was confirmed in this study. Importantly, MK-2206 treatment in vivo inhibited primary human leiomyoma xenograft growth without substantial toxicity. Leiomyoma xenografts showed areas of cell death in the absence of caspase 3 cleavage similar to the in vitro data using primary and immortalized UL cells. Importantly, AKT inhibition in vivo using MK-2206 resulted in a lower leiomyoma tumor volume despite the presence of combined estrogen (E₂) and progesterone (P₄). E₂ and P₄ are essential for leiomyoma tumor growth and activate multiple signaling pathways (9, 49). The efficacy of MK-2206 to cause smaller leiomyoma tumors, even in the presence of E₂ and P₄, demonstrate the central role that AKT plays in leiomyoma growth and survival. Although myometrial cells express phosphorylated AKT and MK-2206 reduced myometrial cell viability in vitro, it was previously demonstrated that neither myometrial cells nor tissue engrafted under the kidney capsule of immunocompromised female mice establish and grow (17). In contrast, leiomyoma cells and tissues grow well using this model, highlighting the in vivo differences between myometrium and leiomyoma (17). These data strongly support the importance of AKT in uterine fibroid cell survival and demonstrate, for the first time, that AKT is a promising candidate for targeted therapy for uterine fibroids.