PTEN deletion or mutation occurs in 30% to 60% of patients with glioblastoma (GBM) and is associated with poor prognosis. Efficacious therapy for this subgroup of patients is currently lacking. To identify potential target(s) to selectively suppress PTEN-deficient GBM growth, we performed a three-step synthetic lethal screen on LN18 PTEN wild-type (WT) and knockout (KO) isogeneic GBM cell lines using a library containing 606 target-selective inhibitors. A MCL1 inhibitor UMI-77 identified in the screen exhibited excellent suppression on the proliferation, colony formation, 3D spheroid, and neurosphere formation of PTEN-deficient GBM cells. Mechanistically, loss of PTEN in GBM cells led to upregulation of MCL1 in posttranslational level via inhibition of GSK3β, and consequently confer cells resistance to apoptosis. Pharmacologic inhibition or knockdown of MCL1 blocked this PI3K–GSK3β–MCL1 axis and caused reduction of several antiapoptotic proteins, finally induced massive caspase-3 cleavage and apoptosis. In both subcutaneous and orthotopic GBM models, knockdown of MCL1 significantly impaired the in vivo growth of PTEN-deficient xenografts. Moreover, the combination of UMI-77 and temozolomide synergistically killed PTEN-deficient GBM cells. Collectively, our work identified MCL1 as a promising target for PTEN-deficient GBM. For future clinical investigations, priority should be given to the development of a selective MCL1 inhibitor with efficient brain delivery and minimal in vivo toxicity.

Glioblastoma multiforme (GBM) is the most lethal and common type of malignant brain tumor (accounting for 47.7% of all malignant brain and other CNS tumors; ref. 1). Chemotherapy for GBM currently relies heavily on temozolomide, an alkylating agent used as a first-line treatment for GBM. The lack of other effective therapeutics to treat GBM when the tumor becomes irresponsive to temozolomide is a major hurdle that needs to be overcome.

The heterogeneous nature of GBM makes it refractory to treatment (2) and constitutes a challenge for drug development. After decades of futile efforts to find effective targeted therapeutic agents for GBM treatment, nowadays clinical and basic researchers are more predisposed to finding treatments that are tailored for a subgroup of patients with GBM with the same genomic feature, allowing these patients to receive the maximum therapeutic benefit.

PTEN is a tumor suppressor gene located at chromosome 10q23. Mutations or deletions of PTEN are frequent events in GBM (30%–60%, 36% according to TCGA; ref. 3) and correlate with poor prognosis (4–6). Although loss of PTEN occurs as a late event, it plays a critical role in GBM progression as it gains multiple tumor-related survival advantages via activation of the PI3K–AKT–mTOR signaling pathway. Therefore, blocking the PI3K pathway seems to be a reasonable strategy to treat PTEN-deficient GBM. However, several studies found that patients with PTEN loss did not benefit from PI3K inhibitors (7–9). Our previous works also suggest that even for PI3K/mTOR inhibitors penetrating blood–brain barrier relatively well, their in vivo antitumor activities against PTEN-deficient GBM were limited (10, 11). Although the underlying mechanism for that could be complex, but the versatile roles of PTEN act both in the cytoplasm and nucleus in addition to its most known function as a membrane lipid phosphatase may provide an explanation.

An alternative solution to treat PTEN-deficient GBM is to identify target(s) that have synthetic lethal interactions with PTEN. Synthetic lethality occurs when the simultaneous perturbation of two genes results in cellular or organismal death, and it also occurs between genes and small molecules. In fact, a few candidate targets have been identified to cause synthetic lethality when PTEN and the interacting gene are disrupted or blocked in prostate (12), breast (13–15), and lung cancers (16), but studies based on GBM models are so far lacking.

In this study, we aimed to find effective targeted agent(s) for PTEN-deficient GBM by screening a library composed of target-selective small molecules. We found that PTEN-deficient GBM cells were vulnerable to UMI-77, an MCL1 (myeloid cell leukemia-1) inhibitor. Our results also revealed the underlying mechanism and a synergistic effect between UMI-77 and temozolomide.

Cell culture and compounds

Human GBM cell lines (LN229, LN18, T98G, SNB19, A172, and U118MG) and HEK293T cells were obtained from the Cell Bank of the Chinese Academy of Sciences. U87 and U251 cells were a gift from Dr. X. Hou (Xuzhou Medical University, Jiangsu, China). All the above cell lines were authenticated by Biowing Biotech and maintained in DMEM (Gibco) supplemented with 10% FBS and penicillin (100 U/mL)–streptomycin (100 μg/mL) with less than 15 passages. The GSC cell line NY-33 was a gift from Dr. X. Wu (Tianjin Medical University, Tianjin, China) and was maintained in neurobasal medium (Life Technologies) supplemented with B27, l-glutamine, sodium pyruvate, basic FGF (20 ng/mL), and EGF (20 ng/mL; R&D Systems). To avoid Mycoplasma contamination, Mycoplasma test was performed before formal experiments. Except that PEI (Polyfectine) and Polybrene were purchased from Sigma Aldrich. TRIzol Reagent was purchased from TaKaRa. Temozolomide, MG-132, AZD5991, and UMI-77 were purchased from Selleck. MIM1, ABT-737, A-121047, S63845, Z-DEVD-FMK, and cycloheximide were purchased from MedChemExpress.

Generation of PTEN KO GBM cells

sgRNAs were designed using an online design tool (http://crispr.mit.edu/) to generate the optimal PTEN-targeted sequences. Two PTEN-targeted sgRNA sequences (Supplementary Materials and Methods) were synthesized and cloned into a LentiCRISPR v2 vector (#52961, Addgene) using a previously described protocol (17). HEK293T cells were used for lentivirus production by transfection with the above plasmids together with the packaging plasmids pVSVg (#8454, Addgene) and psPAX2 (#12260, Addgene) following 2 days of culture. Then, the culture medium of LN18, LN229, and T98G cells was replaced with the above lentivirus-containing supernatants with the addition of 1:1,000 of polybrene and was centrifuged for 2 hours at 37°C for viral infection. After selection by puromycin for 3 to 7 days, the infected KO cells were serially diluted into a single clone, and validation of PTEN deletion was performed until enough cells were obtained.

Synthetic lethality drug screening

A target-selective inhibitors library (#L3500), including 606 target-selective inhibitors lyophilized and dissolved in DMSO, was purchased from Selleck Chemicals. A total of 2,000 PTEN WT/KO LN18 cells per well were seeded in 96-well plates 24 hours prior to screening. The drug screening was composed of three steps. First, compounds with excellent potency against PTEN KO cells (inhibition of proliferation ≥ 50%) were selected as candidates for the next step. In the second step, namely the synthetic lethality screening, the proliferation inhibition of these candidates on both PTEN WT and KO cells was assessed, and only those compounds that caused an inhibition of proliferation ≥70% in PTEN KO but ≤50% in PTEN WT cells passed this screening. Finally, for compounds that passed the first two screenings, their antiproliferative activities were tested at multiple concentrations in PTEN WT and KO cells.

Immunoblotting

Cells were lysed in RIPA buffer supplemented with protease inhibitors (Supplementary Materials and Methods). Protein concentrations were determined by a BCA assay (Thermo Fisher Scientific). Equal amounts of proteins (20–30 μg) were resolved by SDS-PAGE gels, transferred to a PVDF membrane, and incubated with appropriate primary and secondary antibodies. Antibodies against PTEN (#9559), MCL1 (#39224), p-AKTser473 (#4060), p-AKTthr308 (#13038), p-GSK3βser9 (#5558), survivin (#2808), and caspase-3 (#9664) were purchased from Cell Signaling Technology. Signals were visualized using ECL (Yeasen Biotech).

Quantitative real-time PCR

Cellular RNA was isolated using TRIzol Reagent and reverse transcribed to complementary DNA (cDNA) using a cDNA Synthesis Kit (Vazyme). Quantitative real-time PCR was performed on an ABI 7500 RT-PCR System using specific primers (Supplementary Table S1). The data were normalized to GAPDH.

Antiproliferation, colony, 3D spheroids, and neurosphere formation assay

For the antiproliferation assay, 1,000 cells/well were seeded in 96-well plates 24 hours prior to treatment. After the addition of drug for 72 hours, cell viability was determined using an Alamar Blue assay (Yeasen Biotech). For the colony formation assay, cells were seeded in 12-well plates at a density of 400 to 1,000 cells/well. After the addition of drug for 10 to 14 days, all cells were fixed and stained using a paraformaldehyde (4% v/v) and Giemsa staining solution (KeyGEN Biotech). For the 3D spheroid formation assay, 400 cells/well were seeded in an ultra-low binding 96-well plate with a round bottom. After the addition of the drug for 9 days, images of spheroids were captured under a microscope. For the neurosphere formation assay, 5,000 and 1,000 NY-33 cells/well received UMI-77 and MCL1-specific shRNAs, respectively, and were seeded in 24-well plates and cultured for 8 to 12 days before image capture.

Inhibition of protein synthesis and proteasome-dependent degradation in GBM cells

Protein synthesis inhibitor cycloheximide was first dissolved in DMSO and then diluted in culture medium to form a concentration of 100 μg/mL for treatment. Proteasome inhibitor MG-132 was first dissolved in DMSO and then diluted in culture medium to form concentrations of 5, 10, and 20 μmol/L for treatment. The cycloheximide treatment durations were 0, 4, 8, and 16 hours and the combination of cycloheximide and MG-132 treatment was 16 hours. After treatment, cell lysates were collected and analyzed by Western blotting.

Cell apoptosis assay by flow cytometry and by immunofluorescent microscopy

For cell apoptosis analysis, 1 × 106 cells were washed with PBS buffer twice and were stained with Annexin V-APC and propidium iodide (PI) using an Apoptosis Detection Kit (Yeasen Biotech) at room temperature for 15 minutes. The resuspended cells were subjected to FACS analysis by a BD FACSVerse flow cytometer, and data were analyzed by FlowJo 7.6. For immunofluorescent microscopy, a same kit and procedure was applied on cells attached to the culture plate, after staining for 15 minutes, cells were observed with a fluorescent microscope (ZEISS Axio Observer A1).

Generation of MCL1 knockdown cells

Two MCL1-targeted shRNA sequences (Supplementary Materials and Methods) were designed using the Sigma online website, synthesized, and cloned into a pLKO.1 shRNA vector (#10878, Addgene). HEK293T cells transfection, virus production, shRNA infection, and generation of stable MCL1 knockdown LN18/LN229 cells were performed following the method described in the section of Generation of PTEN KO GBM cells.

IHC staining and apoptosis antibody array

The experimental procedures are described in the Supplementary Materials and Methods.

GBM xenograft study

For the subcutaneous xenograft model, 5 × 106PTEN WT, PTEN WT shMCL1, PTEN KO, and PTEN KO shMCL1 LN229-Luc cells were injected into the right flanks of athymic BALB/c nude mice (n = 4). The tumor volume was measured daily using the formula ½ × length × width2. For intracranial xenografts, mice were anesthetized and placed in a stereotact; 3 × 104 cells resuspended in 4 μL complete medium were injected 2 mm lateral and 1 mm anterior to bregma, 2 mm below the skull (n = 5). Bioluminescence from the tumor was monitored once a week using the IVIS Spectrum imaging system (PerkinElmer). The mice were sacrificed when they showed weight loss (>20%) or neurologic symptoms.

Analysis of GDSC drug sensitivity data and analysis of combination interactions between UMI77 and temozolomide

Detailed procedures are described in the Supplementary Materials and Methods.

Statistical analysis

Data analysis was performed using a two-tailed unpaired Student t test. Except for the xenograft study, the data in figures are represented as the mean ± SD. Tumor volumes or the bioluminescence intensity in the xenograft studies are represented as the mean ± SEM. All in vitro experiments were performed at least in triplicate with more than two independent experiments. P values are denoted in figures by *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

Identification of small-molecule inhibitors against PTEN-deficient GBM cells using synthetic lethal screen

We first established isogenic PTEN WT and KO GBM cell lines for targeted drug screening. Among eight commonly used GBM cell lines, only LN18 and LN229 contained no mutations in the PTEN gene (Fig. 1A), and intact PTEN proteins were lost in five cell lines (Fig. 1B). Although T98G contains a missense point mutation, a functional unknown PTEN protein was still expressed. We used the CRISPR/Cas9 technique to delete a fragment in the conserved C2 domain of PTEN (Fig. 1C) in LN18 and LN229 cells and selected stable PTEN KO clones for further investigation. Western blotting and qPCR results confirmed PTEN deletions in the two GBM cell lines (Fig. 1D).

Figure 1.

Generation of isogenic PTEN KO and WT GBM cell lines and the procedure of drug screening. A and B, Genetic mutation (A) and protein expression of PTEN (B) in eight GBM and one GSC cell line. C, sgRNA design strategy and targeting sites in the PTEN gene. D,PTEN KO was confirmed by Western blotting and RT-PCR. E, A flowchart of drug screening. F, Dot plot of 202 candidates identified in the potency screen. Each dot represents a compound. G, Seven candidates, including UMI-77, were identified in the synthetic lethality screen. H, List of the top candidates after the synthetic lethality screen.

Figure 1.

Generation of isogenic PTEN KO and WT GBM cell lines and the procedure of drug screening. A and B, Genetic mutation (A) and protein expression of PTEN (B) in eight GBM and one GSC cell line. C, sgRNA design strategy and targeting sites in the PTEN gene. D,PTEN KO was confirmed by Western blotting and RT-PCR. E, A flowchart of drug screening. F, Dot plot of 202 candidates identified in the potency screen. Each dot represents a compound. G, Seven candidates, including UMI-77, were identified in the synthetic lethality screen. H, List of the top candidates after the synthetic lethality screen.

Close modal

The purpose of the screening was to identify the potential targeted small molecules possessing synthetic lethal interactions upon PTEN deletion. For that, we chose a target-selective inhibitor library because it is easy to reveal the underlying mechanism. We divided the experiment into three steps: potency screening, synthetic lethality screening, and multiple dose screening (Fig. 1E). In the first step, the potency of each compound against the proliferation of PTEN KO LN18 cells was evaluated. A total of 202 compounds exhibited excellent inhibitory activity against PTEN KO cells (Fig. 1F). In the second step, the compounds that caused more than 70% but less than 50% decreases in proliferation in PTEN KO and WT cells, respectively, were selected as candidates for further investigation (Fig. 1G). In total, seven compounds were singled out after the second screening, and among them, the MCL1 inhibitor UMI-77 outperformed the others, exhibiting strong inhibitory activity against PTEN-deficient cells (Fig. 1H).

PTEN loss sensitized GBM cells to UMI-77 or shRNA-mediated MCL1 knockdown

Next, we confirmed the synthetic lethal effects of the seven candidates at different concentrations in PTEN WT/KO LN18 cells. Four of them showed significant differences in their antiproliferative activities between PTEN WT and KO cells, but three failed in this step. The four compounds were the inhibitor of apoptosis (IAP) inhibitor GDC-0152 (18), the IAP inhibitor birinapant (19), the PPAR inhibitor T0070907 (20), and the MCL1 inhibitor UMI-77 (Fig. 2A). Finally, we chose the MCL1 inhibitor UMI-77 for further investigation for its excellent ability to suppress the proliferation of PTEN KO LN18 cells but not that of WT cells. We found that UMI-77 was also able to selectively inhibit the proliferation of PTEN KO LN229 cells and even T98G cells (which express a PTEN point mutation) relative to KO cells (Fig. 2B).

Figure 2.

Assessment of the PTEN-specific synthetic lethal effects of UMI-77 and MCL1-specific shRNAs. A, Antiproliferative effects of GDC-0152, birinapant, T0070907, and UMI-77 on PTEN WT/KO LN18 cells. B, Antiproliferative effects of UMI-77 on PTEN WT/KO LN229 and T98G cells. Right, the structure of UMI-77. C and D, Colony formation (C) and 3D spheroid formation (D) of PTEN WT/KO LN18 and LN229 cells exposed to the indicated concentrations of UMI-77 for 14 and 9 days, respectively. Scale bar, 50 μm. E and F, Neurosphere formation of NY-33 cells treated with UMI-77 (E) or MCL1-specific shRNAs (F). Scale bar, 0.1 mm. Evaluation of the impact of ABT-737, a BCL2 inhibitor, on colony formation (G) and cell proliferation (H) on PTEN WT/KO GBM cells. I and J, shRNA-mediated MCL1 knockdown on proliferation (I) and colony formation (J) of PTEN WT/KO LN18 cells. K, Western blots confirmed the knockdown of MCL1 in LN18/LN229 cells. EV, shRNA empty vector. L, Growth inhibition curves and IC50 values of eight GBM cell lines that received UMI-77 treatment.

Figure 2.

Assessment of the PTEN-specific synthetic lethal effects of UMI-77 and MCL1-specific shRNAs. A, Antiproliferative effects of GDC-0152, birinapant, T0070907, and UMI-77 on PTEN WT/KO LN18 cells. B, Antiproliferative effects of UMI-77 on PTEN WT/KO LN229 and T98G cells. Right, the structure of UMI-77. C and D, Colony formation (C) and 3D spheroid formation (D) of PTEN WT/KO LN18 and LN229 cells exposed to the indicated concentrations of UMI-77 for 14 and 9 days, respectively. Scale bar, 50 μm. E and F, Neurosphere formation of NY-33 cells treated with UMI-77 (E) or MCL1-specific shRNAs (F). Scale bar, 0.1 mm. Evaluation of the impact of ABT-737, a BCL2 inhibitor, on colony formation (G) and cell proliferation (H) on PTEN WT/KO GBM cells. I and J, shRNA-mediated MCL1 knockdown on proliferation (I) and colony formation (J) of PTEN WT/KO LN18 cells. K, Western blots confirmed the knockdown of MCL1 in LN18/LN229 cells. EV, shRNA empty vector. L, Growth inhibition curves and IC50 values of eight GBM cell lines that received UMI-77 treatment.

Close modal

We further confirmed this finding by colony formation, 3D spheroids, and neurosphere formation assays. Although PTEN KO LN18 and LN229 cells formed similar or even more colonies with no treatment or low concentrations of UMI-77, the PTEN KO cells appeared to be more sensitive to UMI-77 at 10 μmol/L than PTEN WT cells (Fig. 2C). This effect became even more significant in 3D culture conditions because the rapid expansion of spheroids formed by PTEN KO cells was fully blocked by 5 μmol/L UMI-77 (Fig. 2D). Moreover, the self-renewal of GBM cells, as measured by neurosphere formation, was impaired by treatment with UMI-77 (Fig. 2E) or MCL1 knockdown (Fig. 2F) in NY-33 cells, a PTEN-deficient GSC line. Because MCL1 is a member of the B-cell lymphoma 2 (Bcl-2) family of apoptosis-regulating proteins, we wondered whether such synthetic lethal interactions between PTEN and MCL1 can also extend to other BCL-2 family members. Therefore, we tested the responses of GBM cells to ABT-737, a multitarget inhibitor of BCL-xL, BCL-2, and BCL-w but not of MCL1 (21, 22). In both colony formation (Fig. 2G) and antiproliferation (Fig. 2H) experiments, PTEN KO and WT cells exhibited no difference upon ABT-737 treatment, suggesting that the synthetic lethal interaction between PTEN and MCL1 is selective.

We doubted whether such potent activity of UMI-77 against PTEN-deficient GBM cells could be attributed to the off-target effect of the drug instead of selective MCL1 inhibition. To exclude this possibility, we designed two shRNAs targeting MCL1 and infected LN18 or LN229 cells with either an empty pLKO.1 (EV) vector or MCL1-specific shRNAs (MCL1 sh1 and sh2). We observed similar differentiated responses between PTEN WT and KO cells with UMI-77 treatment. MCL1 knockdown resulted in a 1.3- to 1.8-fold reduction in viable PTEN WT LN18 cells, but the fold reduction was further increased to a 3.2- to 4.4-fold reduction in PTEN KO cells by MCL1 shRNA1 and shRNA2, respectively (Fig. 2I). In addition, the colony formation of PTEN KO GBM cells was more severely impaired than PTEN WT cells upon MCL1 knockdown (Fig. 2J). Western blotting confirmed MCL1 knockdown in the cell lines used above (Fig. 2K).

Next, we treated 8 GBM cell lines with UMI-77 to assess whether established GBM cells with intact PTEN or mutated PTEN responded differently to MCL1 inhibition. As shown in Fig. 2L, the mean IC50 values of PTEN-deficient GBM and PTEN intact cell lines were 9.11 and 13.91 μmol/L, respectively, so there is a clear trend that PTEN-deficient GBM cell lines are more sensitive to MCL1 inhibition.

PTEN-deficient GBM cells were sensitive to other MCL1 inhibitors

We took advantage of GDSC, an online drug sensitivity database (https://www.cancerrxgene.org), for the identification of drugs with increased sensitivities in cell lines carrying PTEN mutations. In a collection of pan-cancer cell lines, 8 of 162 candidate drugs were identified because their sensitivities were significantly increased in cells carrying PTEN mutations. MIM1, a selective MCL1 inhibitor, is the only agent that does not target PI3K or AKT among the top eight candidates (Fig. 3A). In a subgroup in which only GBM cells lines were selected, even though no drug showed significantly increased sensitivity against PTEN-mutated cells, MIM1 appeared to be the most sensitive drug to kill cells carrying PTEN mutations (Fig. 3B). Therefore, we reckoned that PTEN-deficient GBM cells were also vulnerable to MIM1, so we evaluated the antiproliferative activity of MIM1 on our isogenic PTEN WT and KO GBM cells. Indeed, LN18 PTEN KO cells were more sensitive to MIM1 than PTEN WT cells, and LN229 PTEN KO cells were sensitive to MIM1 only at high concentrations (Fig. 3C). Moreover, we also checked three other MCL1 selective inhibitors, A-1210477, S63845, and AZD5991, for their ability to suppress PTEN-deficient GBM cell proliferation. Again, both PTEN KO LN18 and LN229 cells were more vulnerable to A-1210477, S63845, and AZD5991 than corresponding WT cells, although the effects of these MCL1 inhibitor were less prominent than these of UMI-77 (Fig. 3D and E).

Figure 3.

PTEN-mutated/deficient GBM cells were sensitive to other MCL1 inhibitors. Analysis of the interactions between mutations of PTEN and drug sensitivities in Pan-cancer cell line collection (A) and GBM cell lines collection (B) from GDSC drug sensitivity database. The volcano plot showed that MCL1 inhibitor MIM1 was among the top candidates possessing high sensitivity and significance to PTEN-mutated cells in both Pan-cancer (162 drugs, A) and GBM cell lines (157 drugs, B). C–F, Validation of the antiproliferation activity of MCL1 inhibitor MIM1 (C), A-120477 (D), S63845 (E), and AZD5991 (F) in PTEN WT/KO LN18 and LN229 cells. The structures of these compounds were shown in the right side of (C–F).

Figure 3.

PTEN-mutated/deficient GBM cells were sensitive to other MCL1 inhibitors. Analysis of the interactions between mutations of PTEN and drug sensitivities in Pan-cancer cell line collection (A) and GBM cell lines collection (B) from GDSC drug sensitivity database. The volcano plot showed that MCL1 inhibitor MIM1 was among the top candidates possessing high sensitivity and significance to PTEN-mutated cells in both Pan-cancer (162 drugs, A) and GBM cell lines (157 drugs, B). C–F, Validation of the antiproliferation activity of MCL1 inhibitor MIM1 (C), A-120477 (D), S63845 (E), and AZD5991 (F) in PTEN WT/KO LN18 and LN229 cells. The structures of these compounds were shown in the right side of (C–F).

Close modal

Loss of PTEN led to MCL1 upregulation via inhibition of GSK3β

We noticed a clear difference in MCL1 expression between PTEN WT and KO GBM cell lysates in the above experiments (Fig. 2K). Therefore, we evaluated the expression of MCL1 in clinical samples from a cohort of 139 patients with glioma. IHC analysis revealed that the patients with glioma with high expression of MCL1 correlated with short overall survival, suggesting a crucial role of MCL1 in GBM (Supplementary Fig. S1). Whereas a statistically significant correlation could not be detected (P = 0.32, r = −0.09) in this cohort, we noticed a tendency that higher MCL1 levels were detected in PTEN-negative than PTEN-positive glioma tissues (Fig. 4A and B). It is interesting because there was no solid evidence revealing a link between PTEN and MCL1. We searched the literature and found that GSK3β could be the link connecting PTEN and MCL1 because it is a known substrate of AKT (23) and an upstream regulator of MCL1 (24, 25). Indeed, loss of PTEN led to GSK3β suppression, thus increasing the level of p-GSK3β and consequently elevating the expression of MCL1 (Fig. 4C), indicating that GSK3β is a mediator between PTEN and MCL1. Another interesting finding is that the p-AKT signaling was severely suppressed upon knockdown of MCL1 in PTEN KO cells, but such declines in PTEN WT cells were only marginal.

Figure 4.

Upregulation of MCL1 in PTEN-deficient GBM and relevant signaling pathway analysis. A and B, Representative MCL1 and PTEN staining (A) and a summary of their expression in 139 patient-derived glioma tissue arrays (B). C–H, Cell lysates collected from the following experiments were analyzed by Western blotting and probed with antibodies against PTEN, MCL1, p-AKT, p-GSK3β, and total GSK3β or other indicated antibodies: PTEN WT/KO LN18 and LN229 cells were stably transduced with empty control or MCL1 shRNA lentivirus (C); PTEN WT/KO LN18 and LN229 cells were treated with 10 and 20 μmol/L UMI-77, A1210477, and S63845 (D); PTEN WT/KO LN18 and LN229 cells were treated with BKM120 at concentrations of 0, 2, 4, and 10 μmol/L (E); PTEN WT/KO LN18 and LN229 cells were stably transduced with empty control or GSK3β shRNA lentivirus (F). G, qPCR analysis of the PTEN and MCL1 mRNA levels in PTEN WT/KO LN18 and LN229 cells. H, LN18 and LN229 cells were treated with protein synthesis inhibitor 100 μg/mL cycloheximide for 0, 4, 8, and 16 hours or together with MG132 (at indicated concentrations) for 16 hours prior to Western blotting analysis. I,PTEN WT/KO LN18 and LN229 cells were stably transduced with empty control or GSK3β shRNA lentivirus, and then were treated with 100 μg/mL cycloheximide for 0, 4, 8, and 16 hours. The cell lysates were analyzed by Western blotting using the indicated antibodies.

Figure 4.

Upregulation of MCL1 in PTEN-deficient GBM and relevant signaling pathway analysis. A and B, Representative MCL1 and PTEN staining (A) and a summary of their expression in 139 patient-derived glioma tissue arrays (B). C–H, Cell lysates collected from the following experiments were analyzed by Western blotting and probed with antibodies against PTEN, MCL1, p-AKT, p-GSK3β, and total GSK3β or other indicated antibodies: PTEN WT/KO LN18 and LN229 cells were stably transduced with empty control or MCL1 shRNA lentivirus (C); PTEN WT/KO LN18 and LN229 cells were treated with 10 and 20 μmol/L UMI-77, A1210477, and S63845 (D); PTEN WT/KO LN18 and LN229 cells were treated with BKM120 at concentrations of 0, 2, 4, and 10 μmol/L (E); PTEN WT/KO LN18 and LN229 cells were stably transduced with empty control or GSK3β shRNA lentivirus (F). G, qPCR analysis of the PTEN and MCL1 mRNA levels in PTEN WT/KO LN18 and LN229 cells. H, LN18 and LN229 cells were treated with protein synthesis inhibitor 100 μg/mL cycloheximide for 0, 4, 8, and 16 hours or together with MG132 (at indicated concentrations) for 16 hours prior to Western blotting analysis. I,PTEN WT/KO LN18 and LN229 cells were stably transduced with empty control or GSK3β shRNA lentivirus, and then were treated with 100 μg/mL cycloheximide for 0, 4, 8, and 16 hours. The cell lysates were analyzed by Western blotting using the indicated antibodies.

Close modal

The above observation might provide a clue explaining why PTEN KO GBM cells are more sensitive to UMI-77. Next, we also tested the p-AKT signaling of GBM cells upon treatment with UMI-77 and two other MCL1 inhibitors. In line with the effect of MCL1 silencing, all three MCL1 inhibitors caused significant declines in both p-AKTser473 and p-AKTthr308 in PTEN KO GBM cells but only weak changes in corresponding PTEN WT cells (Fig. 4D), suggesting that MCL1 silencing/blockage specifically attenuates AKT signaling in PTEN-deficient GBM cells. Next, we investigated whether a PI3K blockade also affects MCL1. Treatment with BKM120, a selective pan-PI3K inhibitor, efficiently decreased the expression of both p-GSK3β and MCL1 in a dose-dependent manner (Fig. 4E), confirming that the PI3K/AKT signaling determines the expression of MCL1 via GSK3β. Moreover, we investigated whether there is a direct link between GSK3β and MCL1 by knockdown GSK3β in GBM cells. Indeed, in the PTEN KO cells with a highly activated PI3K signaling, knockdown of GSK3β by specific shRNA released its negative control upon MCL1 and consequently caused an elevation of MCL1, confirming GSK3β as a direct negative mediator of MCL1 (Fig. 4F).

To further interrogate the mechanism that how MCL1 was regulated in the PI3K−GSK3β−MCL1 axis, we quantified the mRNA level of MCL1 in PTEN WT and KO cells and found no significant differences between WT and KO GBM cells (Fig. 4G). Therefore, the regulation of MCL1 by PI3K/GSK3β is probably posttranslational. Indeed, PTEN KO GBM cells received cycloheximide, a protein synthesis inhibitor, exhibited marked reduction of MCL1 (Fig. 4H) and addition of MG-132 (26), a proteasome inhibitor, prevented the reduction of MCL1, so the ubiquitination mediated proteasomal degradation of MCL1 could be a major controller and putatively a determinant of MCL1 protein level. Next, we asked whether PI3K–GSK3β signaling is critical for this process. As shown in Fig. 4I, the half-life of MCL1 was markedly extended in GSK3β knockdown relative to control cells expressing normal GSK3β. On the basis of the these evidences, we conclude that GSK3β may negatively regulate MCL1 by promoting its ubiquitination-mediated proteasomal degradation.

MCL1 blockage reduced the expression of several antiapoptotic proteins, elevated the expression of cleaved caspase-3, and increased apoptosis of PTEN-deficient GBM cells

MCL1 is an antiapoptotic protein, so we reasoned that the synthetic lethal effect could be caused by activation of the apoptotic cascade. We used an Apoptosis Antibody Array Kit (Supplementary Fig. S2A) to analyze the differentially expressed proteins in PTEN WT/KO LN18 cells upon UMI-77 treatment. Interestingly, most of the antiapoptotic proteins analyzed were upregulated, while seven apoptotic proteins were downregulated in PTEN-deficient GBM cells, suggesting that PTEN-deficient cells tended to escape from apoptosis. However, this trend was completely reversed after treatment with UMI-77: the levels of survivin, claspin, and 6 other antiapoptotic proteins markedly decreased (Fig. 5A), and the levels of apoptotic proteins, including cleaved caspase-3, phosphor-P53, and Bad, increased (Fig. 5B). Several proteins demonstrating strong alterations across four samples were selected and quantified (Supplementary Fig. S2B). Among them, we noticed that two proteins were subjected to remarkable changes in PTEN KO but not WT GBM cells upon UMI-77 treatment and confirmed their expression by Western blotting. Survivin, a member of the IAP family that inhibits caspase activation, was not altered in UMI-77–treated PTEN WT cells but was significantly decreased in PTEN KO cells treated with UMI-77. In contrast, cleaved caspase-3 was markedly elevated in PTEN KO cells but not WT cells that received UMI-77 (Fig. 5C and D; Supplementary Fig. S2C).

Figure 5.

Analysis of the apoptotic pathway and apoptosis of GBM cells that received UMI-77 treatment. A and B, Analysis of alterations in the antiapoptotic (A) and apoptotic proteins (B) across cell lysates from PTEN WT/KO±UMI-77 LN18 cells using an apoptosis antibody array. C and D, Immunoblots of cleaved caspase-3 and survivin in PTEN WT/KO LN18 and LN229 cells that received no treatment or UMI-77 (10 μmol/L). E and F, Flow cytometry analysis of apoptotic PTEN WT/KO LN18 and LN229 cells transduced with an empty vector or MCL1-specific shRNAs. G, Quantification of the apoptotic faction of cells in each sample. H,PTEN WT and KO LN229 cells received no treatment or UMI-77 (10 μmol/L) were stained with FITC-conjugated Annexin V (green) and PI (red). Cells were observed under a fluorescent microscope and representative photos were given. Scale bars, 0.05 mm. I, Prior to UMI-77 treatment (10 μmol/L), PTEN KO LN18 and LN229 cells were pretreated with 30 μmol/L Z-DEVD-FMK, a caspase-3 inhibitor, for 1 hour and then treated with UMI-77 for 24 hours. The cell lysates were collected and analyzed by Western blotting using the indicated antibodies.

Figure 5.

Analysis of the apoptotic pathway and apoptosis of GBM cells that received UMI-77 treatment. A and B, Analysis of alterations in the antiapoptotic (A) and apoptotic proteins (B) across cell lysates from PTEN WT/KO±UMI-77 LN18 cells using an apoptosis antibody array. C and D, Immunoblots of cleaved caspase-3 and survivin in PTEN WT/KO LN18 and LN229 cells that received no treatment or UMI-77 (10 μmol/L). E and F, Flow cytometry analysis of apoptotic PTEN WT/KO LN18 and LN229 cells transduced with an empty vector or MCL1-specific shRNAs. G, Quantification of the apoptotic faction of cells in each sample. H,PTEN WT and KO LN229 cells received no treatment or UMI-77 (10 μmol/L) were stained with FITC-conjugated Annexin V (green) and PI (red). Cells were observed under a fluorescent microscope and representative photos were given. Scale bars, 0.05 mm. I, Prior to UMI-77 treatment (10 μmol/L), PTEN KO LN18 and LN229 cells were pretreated with 30 μmol/L Z-DEVD-FMK, a caspase-3 inhibitor, for 1 hour and then treated with UMI-77 for 24 hours. The cell lysates were collected and analyzed by Western blotting using the indicated antibodies.

Close modal

Next, we evaluated the induction of apoptosis by either UMI-77 treatment or MCL1 knockdown. The apoptotic cell fractions were similar between treated and untreated PTEN WT cells but were significantly elevated in UMI-77-treated PTEN KO cells relative to in untreated cells (Supplementary Fig. S3F–S3H), and In line with this result, knockdown of MCL1 also induced cell apoptosis, but the apoptotic cell fractions of PTEN KO cells were much greater than those of PTEN WT cells (Fig. 5EG). As a direct evidence, we found more Annexin V–positive early apoptotic cells in PTEN KO than WT cells upon treatment of UMI-77 (Fig. 5H).

Finally, to exclude the possibility that the attenuation of PI3K signaling was caused by apoptosis, we pretreated PTEN KO LN229 cells with Z-DEVD-FMK, a caspase-3 inhibitor prior to UMI-77 treatment. As shown in Fig. 5I, Z-DEVD-FMK successfully inhibited the cleavage of caspase-3, but the levels of p-AKT, p-GSK3β, and survivin were still markedly dropped upon treatment of UMI-77 relative to those untreated cells, suggesting that apoptosis is not the cause of MCL1 inhibitor–induced PI3K signaling attenuation.

Knockdown of MCL1 significantly impaired the in vivo growth of PTEN-deficient GBM xenografts

To investigate whether there is also a synthetic interaction between PTEN and MCL1 in vivo (where the tumor microenvironment is more complex than in formulated culture medium), we evaluated the subcutaneous and intracranial tumor growth of MCL1-silenced PTEN WT/KO LN229 xenografts and their empty vector controls. Unlike in vitro culture, subcutaneous PTEN KO LN229 xenografts grew much rapidly and formed larger tumors than PTEN WT xenografts after 22 days of implantation (Fig. 6A). Surprisingly, knockdown of MCL1 completely restrained the uncontrollable growth of PTEN KO xenografts to a rate comparable with that of tumors formed by PTEN WT xenografts. In comparison, PTEN WT xenograft growth was hardly affected by MCL1 knockdown (Fig. 6B). Western blotting analysis of the tumor tissue confirmed MCL1 knockdown (Fig. 6C). Such an effect was also observed in the intracranial xenograft models (Fig. 6D and E) and led to significantly extended survival. In contrast, the PTEN WT mice failed to gain a meaningful survival benefit from MCL1 knockdown (Fig. 6F).

Figure 6.

Evaluation of the in vivo effect of MCL1 knockdown on xenograft tumor growth. A, Tumor bulks isolated 22 days after subcutaneous implantation of PTEN WT/KO ± MCL1-specific shRNA (sh1) LN18 cells (n = 4). B, Tumor growth curves from four different xenografts were plotted, and PTEN and MCL1 expression (C) in tumor tissues was confirmed by Western blotting. D, Bioluminescence imaging of mouse brains 29 days after intracranial implantation of PTEN WT/KO ± MCL1-specific shRNA (sh1) LN229-luciferase cells (n = 5). The scale of the luminance bar (right side of images) was kept equal among all images. E and F, Tumor growth curves (E) and survival curves (F) of each group were plotted.

Figure 6.

Evaluation of the in vivo effect of MCL1 knockdown on xenograft tumor growth. A, Tumor bulks isolated 22 days after subcutaneous implantation of PTEN WT/KO ± MCL1-specific shRNA (sh1) LN18 cells (n = 4). B, Tumor growth curves from four different xenografts were plotted, and PTEN and MCL1 expression (C) in tumor tissues was confirmed by Western blotting. D, Bioluminescence imaging of mouse brains 29 days after intracranial implantation of PTEN WT/KO ± MCL1-specific shRNA (sh1) LN229-luciferase cells (n = 5). The scale of the luminance bar (right side of images) was kept equal among all images. E and F, Tumor growth curves (E) and survival curves (F) of each group were plotted.

Close modal

The combination of UMI-77 and temozolomide synergistically killed PTEN-deficient GBM cells

Previous work suggested that the expression of MCL1 may determine the responsiveness of GBM cells to temozolomide (27, 28). We wondered whether UMI-77 works synergistically with temozolomide, the first-line therapy for GBM. The addition of UMI-77 significantly enhanced the cytotoxicity of temozolomide (Supplementary Fig. S3A), and PTEN KO cells were more sensitive than WT cells to the combination therapy (Supplementary Fig. S3B). Isobologram and combination index analysis showed that UMI-77 and temozolomide caused synergistic effects in both PTEN WT and KO GBM cells (Supplementary Fig. S3C). We further quantified the combination effects in excess of Loewe additivity using Horizon's proprietary Chalice software. Representative growth inhibition and Loewe excess dose matrices for the combined effect of UMI-77 and temozolomide are shown in Supplementary Figs. S3D (LN18 cells) and S3E (LN229 cells), and the synergy scores were calculated by the software. In both LN18 and LN229 cells, the synergy scores of the combination treatment of UMI-77 and temozolomide were higher in PTEN KO cells than in PTEN WT cells. Treatment with UMI-77 and temozolomide in combination, but not with temozolomide alone, significantly increased the apoptotic cell fraction in PTEN KO cells relative to PTEN WT cells (Supplementary Fig. S3F–S3H).

To confirm the combined effect of temozolomide and MCL1 inhibitor, we also tested another MCL1 inhibitor AZD5991. In line with the results of UMI-77, combination of temozolomide and AZD caused marked suppression (Supplementary Fig. S4A) and impressive synergistic effect (Supplementary Fig. S4B and S4C) in PTEN KO GBM cells.

Finally, we applied the combination therapy of temozolomide and UMI-77 to eight established GBM cell lines. Despite the differentiated responses it caused, the synergy scores of the cell lines carrying PTEN mutations were generally higher than those carrying WT PTEN (Supplementary Fig. S5A–S5C). Western blotting analysis revealed that the levels of p-AKT, p-GSK3β, MCL1, Bax, and survivin were dropped but the cleaved caspase-3 was elevated upon combination treatment relative to single treatment of temozolomide or UMI-77 in PTEN KO GBM cells (Supplementary Fig. S5D). Together, these results suggest that the UMI-77 and temozolomide combination treatment caused stronger synergy and apoptosis in PTEN KO cells than in WT GBM cells.

In the current study, we only focused on PTEN-deficient or mutated GBM and aimed to find an effective therapy for it. Through target-selective inhibitor library-based screening, we successfully identified four candidates selectively inhibiting the proliferation of PTEN-deficient GBM cells from 606 targeted small molecules. Among them, MCL1 inhibitor UMI-77 exhibited promising activities against various PTEN-deficient GBM/GSC models. Further investigation found that PTEN-deficient GBM cells were vulnerable to both MCL1-specific shRNA and other MCL1 inhibitors, but not BCL2 inhibitor. Mechanistically, UMI-77 blocked MCL1 and suppressed p-AKT, surviving, and several antiapoptotic proteins, causing an overall downregulation of the antiapoptotic level, which was critical for the survival of PTEN-deficient cells. Knockdown of MCL1 led to significant tumor growth inhibition and survival prolongation in PTEN-deficient GBM xenograft models. These evidences together support the usefulness of MCL1 inhibitor(s) in treatment of PTEN-deficient GBM.

Loss of PTEN is a critical event in GBM evolution because PTEN is the only known lipid phosphatase counteracting the PI3K oncogenic pathway. The consequent activation downstream of the PI3K–AKT pathway benefits tumor cells in multiple aspects, including proliferation, survival, metabolism, and energy supply. PTEN also has versatile roles in the cytoplasm and nucleus by acting as a general phosphatase or by directly interacting with other proteins (29) and regulating their activities. Loss of PTEN also causes a series of PI3K-independent alterations in impaired DNA repair, genomic instability (30, 31), abnormal cell cycle, and mitosis (32, 33). These alterations reshape the global biological network not only promoting proliferation and survival but also bringing potential vulnerability to tumor cells. In our study, we discovered that PTEN-deficient GBM cells upregulated several antiapoptotic proteins, including MCL1, to protect against vulnerability and to avoid the apoptotic fate that would normally occur for clearance of cells with severe abnormalities. Nevertheless, if the antiapoptotic shield is broken down, PTEN-deficient GBM cells would become extra vulnerable and sensitive to treatment. Luckily, we found that UMI-77 could not only inhibit the activity of MCL1 but could also suppress several antiapoptotic proteins in PTEN-deficient GBM cells. This consequently led to massive caspase-3 cleavage and reprogrammed GBM cells to enter the apoptotic fate. Further evidence supporting this hypothesis includes the following: (i) the top three candidate compounds identified in the drug screen, UMI-77, GDC-0152, and birinapant, are all antiapoptotic protein inhibitors, suggesting that PTEN-deficient GBM cells are highly dependent on antiapoptotic maintenance; (ii) the MCL1 inhibitor MIM1 was among the top candidates causing markedly increased sensitivity in PTEN-mutated cell lines (Fig. 3A–C), suggesting that it could be a general mechanism for all cancer cells lacking intact PTEN.

The antiapoptotic protein MCL1, a member of the Bcl-2 family, plays an important role in tumor cell survival and drug resistance (34). A recent study by Wu and colleagues reported that MCL1 was positively expressed in more than half of glioma tissues and its expression was significantly higher in gliomas than in tumor-adjacent tissues. In addition, they found that silencing MCL1 reduced PI3K/AKT signaling (35). Our results showed that not only MCL1 silencing caused a remarkable drop of p-AKT level, MCL1 inhibition by various MCL1 inhibitors also decreased cellular p-AKT and p-GSK3β level. Thus, the antitumor activity of MCL1 inhibitor was aggravated. Because of the crucial role of PI3K–AKT pathway in cell survival, it is not surprising that the suppression of PI3K–AKT signaling would result in dramatic alterations of the apoptotic–antiapoptotic balance of cells. Interestingly, survivin was the most significantly downregulated antiapoptotic protein upon treatment of UMI-77. This could be a consequence of MCL1 inhibition triggering the release of SMAC/DIABLO, which in turn bound to and dampened the activity of cellular IAPs including survivin (36), as we also noticed the marked drop of CIAP1/2 and XIAP, members of the IAP family (Fig. 5A).

Another interesting finding is the synergistic interaction between temozolomide and MCL1 inhibitor as combination treatment for GBM. The potent combined effects of the two drugs were repetitively observed in different GBM cells carrying PTEN mutation and with different MCL1 inhibitors (Supplementary Figs. S3–S5). Because previous studies reported temozolomide treatment downregulated the mRNA (37) and protein levels (27) in GBM cells, we wondered whether the synergy between temozolomide and UMI77 were mediated by MCL1. Indeed, an extra reduction of MCL1 in GBM cells was caused by the combined treatment compared with temozolomide or UMI-77 single treatment. Interestingly, combined treatment also caused extra reductions of p-AKT, GSK3β, and survivin even though they were remaining unaltered in cells treated with temozolomide alone (Supplementary Fig. S5D).

Finally, several ongoing MCL1 inhibitor clinical trials are listed on clinicaltrials.gov, but none of them have been implicated to treat GBM or any type of glioma. There are still some obstacles to be conquered before a clinical evaluation of the toxicity and efficacy of MCL1 inhibitors in patients with GBM can be launched. For example, the chemical structure of UMI-77 contains a carboxyl group, which may limit its delivery across the blood–brain barrier. Given the promising antitumor effect of UMI-77 in PTEN-deficient GBM, further studies should be focused on optimizing its chemical structure, improving its brain delivery and evaluating its efficacy in suitable GBM models.

No potential conflicts of interest were disclosed.

C. Chen: Data curation, formal analysis, investigation, methodology. S. Zhu: Data curation, formal analysis, investigation, writing-original draft. X. Zhang: Data curation, formal analysis, investigation. T. Zhou: Data curation, formal analysis, supervision, investigation. J. Gu: Data curation, formal analysis, investigation. Y. Xu: Data curation, formal analysis, investigation. Q. Wan: Data curation, formal analysis, investigation. X. Qi: Data curation, formal analysis, investigation. Y. Chai: Data curation, formal analysis, investigation. X. Liu: Conceptualization, resources, supervision. L. Chen: Conceptualization, resources, funding acquisition, project administration. J. Yan: Conceptualization, resources, supervision, project administration, writing-review and editing. Y. Hua: Conceptualization, resources, validation, writing-review and editing. F. Lin: Conceptualization, formal analysis, supervision, investigation, methodology, writing-original draft, project administration, writing-review and editing.

This study was supported by the National Natural Science Foundation of China (81672962), the Jiangsu Provincial Innovation Team Program Foundation, and the Joint Key Project Foundation of Southeast University and Nanjing Medical University. We thank Dr. Yujie Sun for helpful suggestions on the manuscript.

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.

1.
Ostrom
QT
,
Gittleman
H
,
Truitt
G
,
Boscia
A
,
Kruchko
C
,
Barnholtz-Sloan
JS
. 
CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2011–2015
.
Neuro Oncol
2018
;
20
:
iv1
86
.
2.
Sottoriva
A
,
Spiteri
I
,
Piccirillo
SG
,
Touloumis
A
,
Collins
VP
,
Marioni
JC
, et al
Intratumor heterogeneity in human glioblastoma reflects cancer evolutionary dynamics
.
Proc Natl Acad Sci U S A
2013
;
110
:
4009
14
.
3.
Cancer Genome Atlas Research N
. 
Comprehensive genomic characterization defines human glioblastoma genes and core pathways
.
Nature
2008
;
455
:
1061
8
.
4.
Yang
Y
,
Shao
N
,
Luo
G
,
Li
L
,
Zheng
L
,
Nilsson-Ehle
P
, et al
Mutations of PTEN gene in gliomas correlate to tumor differentiation and short-term survival rate
.
Anticancer Res
2010
;
30
:
981
5
.
5.
Srividya
MR
,
Thota
B
,
Shailaja
BC
,
Arivazhagan
A
,
Thennarasu
K
,
Chandramouli
BA
, et al
Homozygous 10q23/PTEN deletion and its impact on outcome in glioblastoma: a prospective translational study on a uniformly treated cohort of adult patients
.
Neuropathology
2011
;
31
:
376
83
.
6.
Han
F
,
Hu
R
,
Yang
H
,
Liu
J
,
Sui
J
,
Xiang
X
, et al
PTEN gene mutations correlate to poor prognosis in glioma patients: a meta-analysis
.
OncoTargets Ther
2016
;
9
:
3485
92
.
7.
Weigelt
B
,
Warne
PH
,
Downward
J
. 
PIK3CA mutation, but not PTEN loss of function, determines the sensitivity of breast cancer cells to mTOR inhibitory drugs
.
Oncogene
2011
;
30
:
3222
33
.
8.
Juric
D
,
Castel
P
,
Griffith
M
,
Griffith
OL
,
Won
HH
,
Ellis
H
, et al
Convergent loss of PTEN leads to clinical resistance to a PI(3)Kalpha inhibitor
.
Nature
2015
;
518
:
240
4
.
9.
Koul
D
,
Fu
J
,
Shen
R
,
LaFortune
TA
,
Wang
S
,
Tiao
N
, et al
Antitumor activity of NVP-BKM120–a selective pan class I PI3 kinase inhibitor showed differential forms of cell death based on p53 status of glioma cells
.
Clin Cancer Res
2012
;
18
:
184
95
.
10.
Lin
F
,
Buil
L
,
Sherris
D
,
Beijnen
JH
,
van Tellingen
O
. 
Dual mTORC1 and mTORC2 inhibitor Palomid 529 penetrates the blood-brain barrier without restriction by ABCB1 and ABCG2
.
Int J Cancer
2013
;
133
:
1222
33
.
11.
Lin
F
,
de Gooijer
MC
,
Hanekamp
D
,
Chandrasekaran
G
,
Buil
LC
,
Thota
N
, et al
PI3K-mTOR pathway inhibition exhibits efficacy against high-grade glioma in clinically relevant mouse models
.
Clin Cancer Res
2017
;
23
:
1286
98
.
12.
Ding
Y
,
Li
N
,
Dong
B
,
Guo
W
,
Wei
H
,
Chen
Q
, et al
Chromatin remodeling ATPase BRG1 and PTEN are synthetic lethal in prostate cancer
.
J Clin Invest
2019
;
129
:
759
73
.
13.
Zhao
D
,
Lu
X
,
Wang
G
,
Lan
Z
,
Liao
W
,
Li
J
, et al
Synthetic essentiality of chromatin remodelling factor CHD1 in PTEN-deficient cancer
.
Nature
2017
;
542
:
484
8
.
14.
Wang
W
,
Oguz
G
,
Lee
PL
,
Bao
Y
,
Wang
P
,
Terp
MG
, et al
KDM4B-regulated unfolded protein response as a therapeutic vulnerability in PTEN-deficient breast cancer
.
J Exp Med
2018
;
215
:
2833
49
.
15.
Li
K
,
Yan
H
,
Guo
W
,
Tang
M
,
Zhao
X
,
Tong
A
, et al
ATM inhibition induces synthetic lethality and enhances sensitivity of PTEN-deficient breast cancer cells to cisplatin
.
Exp Cell Res
2018
;
366
:
24
33
.
16.
Chatterjee
N
,
Pazarentzos
E
,
Mayekar
MK
,
Gui
P
,
Allegakoen
DV
,
Hrustanovic
G
, et al
Synthetic essentiality of metabolic regulator PDHK1 in PTEN-deficient cells and cancers
.
Cell Rep
2019
;
28
:
2317
30
.
17.
Bauer
DE
,
Canver
MC
,
Orkin
SH
. 
Generation of genomic deletions in mammalian cell lines via CRISPR/Cas9
.
J Vis Exp
2015
:
e52118
.
18.
Yang
L
,
Shu
T
,
Liang
Y
,
Gu
W
,
Wang
C
,
Song
X
, et al
GDC-0152 attenuates the malignant progression of osteosarcoma promoted by ANGPTL2 via PI3K/AKT but not p38MAPK signaling pathway
.
Int J Oncol
2015
;
46
:
1651
8
.
19.
Benetatos
CA
,
Mitsuuchi
Y
,
Burns
JM
,
Neiman
EM
,
Condon
SM
,
Yu
G
, et al
Birinapant (TL32711), a bivalent SMAC mimetic, targets TRAF2-associated cIAPs, abrogates TNF-induced NF-kappaB activation, and is active in patient-derived xenograft models
.
Mol Cancer Ther
2014
;
13
:
867
79
.
20.
Handeli
S
,
Simon
JA
. 
A small-molecule inhibitor of Tcf/beta-catenin signaling down-regulates PPARgamma and PPARdelta activities
.
Mol Cancer Ther
2008
;
7
:
521
9
.
21.
Chauhan
D
,
Velankar
M
,
Brahmandam
M
,
Hideshima
T
,
Podar
K
,
Richardson
P
, et al
A novel Bcl-2/Bcl-X(L)/Bcl-w inhibitor ABT-737 as therapy in multiple myeloma
.
Oncogene
2007
;
26
:
2374
80
.
22.
Garner
TP
,
Lopez
A
,
Reyna
DE
,
Spitz
AZ
,
Gavathiotis
E
. 
Progress in targeting the BCL-2 family of proteins
.
Curr Opin Chem Biol
2017
;
39
:
133
42
.
23.
Hermida
MA
,
Dinesh Kumar
J
,
Leslie
NR
. 
GSK3 and its interactions with the PI3K/AKT/mTOR signalling network
.
Adv Biol Regul
2017
;
65
:
5
15
.
24.
Wang
R
,
Xia
L
,
Gabrilove
J
,
Waxman
S
,
Jing
Y
. 
Downregulation of Mcl-1 through GSK-3beta activation contributes to arsenic trioxide-induced apoptosis in acute myeloid leukemia cells
.
Leukemia
2013
;
27
:
315
24
.
25.
Kang
XH
,
Zhang
JH
,
Zhang
QQ
,
Cui
YH
,
Wang
Y
,
Kou
WZ
, et al
Degradation of Mcl-1 through GSK-3beta activation regulates apoptosis induced by bufalin in non-small cell lung cancer H1975 cells
.
Cell Physiol Biochem
2017
;
41
:
2067
76
.
26.
Zhu
W
,
Zhan
D
,
Wang
L
,
Ma
D
,
Cheng
M
,
Wang
H
, et al
Proteasome inhibitor MG132 potentiates TRAIL-induced apoptosis in gallbladder carcinoma GBC-SD cells via DR5-dependent pathway
.
Oncol Rep
2016
;
36
:
845
52
.
27.
Gratas
C
,
Sery
Q
,
Rabe
M
,
Oliver
L
,
Vallette
FM
. 
Bak and Mcl-1 are essential for Temozolomide induced cell death in human glioma
.
Oncotarget
2014
;
5
:
2428
35
.
28.
Tian
T
,
Li
A
,
Lu
H
,
Luo
R
,
Zhang
M
,
Li
Z
. 
TAZ promotes temozolomide resistance by upregulating MCL-1 in human glioma cells
.
Biochem Biophys Res Commun
2015
;
463
:
638
43
.
29.
Milella
M
,
Falcone
I
,
Conciatori
F
,
Cesta Incani
U
,
Del Curatolo
A
,
Inzerilli
N
, et al
PTEN: multiple functions in human malignant tumors
.
Front Oncol
2015
;
5
:
24
.
30.
Puc
J
,
Keniry
M
,
Li
HS
,
Pandita
TK
,
Choudhury
AD
,
Memeo
L
, et al
Lack of PTEN sequesters CHK1 and initiates genetic instability
.
Cancer Cell
2005
;
7
:
193
204
.
31.
Ming
M
,
He
YY
. 
PTEN in DNA damage repair
.
Cancer Lett
2012
;
319
:
125
9
.
32.
Brandmaier
A
,
Hou
SQ
,
Shen
WH
. 
Cell Cycle Control by PTEN
.
J Mol Biol
2017
;
429
:
2265
77
.
33.
Salazar-Roa
M
,
Malumbres
M
. 
Fueling the Cell Division Cycle
.
Trends Cell Biol
2017
;
27
:
69
81
.
34.
Thomas
LW
,
Lam
C
,
Edwards
SW
. 
Mcl-1; the molecular regulation of protein function
.
FEBS Lett
2010
;
584
:
2981
9
.
35.
Wu
DM
,
Hong
XW
,
Wen
X
,
Han
XR
,
Wang
S
,
Wang
YJ
, et al
MCL1 gene silencing promotes senescence and apoptosis of glioma cells via inhibition of the PI3K/Akt signaling pathway
.
IUBMB Life
2019
;
71
:
81
92
.
36.
Stolfi
C
,
Pallone
F
,
Monteleone
G
. 
Molecular targets of TRAIL-sensitizing agents in colorectal cancer
.
Int J Mol Sci
2012
;
13
:
7886
901
.
37.
Chu
CW
,
Yang
MC
,
Chou
CH
,
Huang
WS
,
Hsiao
BX
,
Wang
YT
, et al
GSK3betamediated Ser156 phosphorylation modulates a BH3like domain in BCL2L12 during TMZinduced apoptosis and autophagy in glioma cells
.
Int J Mol Med
2018
;
42
:
905
18
.