An LXR agonist promotes glioblastoma cell death through inhibition of an EGFR/AKT/SREBP-1/LDLR-dependent pathway

Abstract

Glioblastoma (GBM) is the most common malignant primary brain tumor of adults and one of the most lethal of all cancers. Epidermal growth factor receptor (EGFR) mutations (EGFRvIII) and phosphoinositide 3-kinase (PI3K) hyperactivation are common in GBM, promoting tumor growth and survival, including through sterol regulatory element-binding protein 1 (SREBP-1)-dependent lipogenesis. The role of cholesterol metabolism in GBM pathogenesis, its association with EGFR/PI3K signaling, and its potential therapeutic targetability are unknown. In our investigation, studies of GBM cell lines, xenograft models, and GBM clinical samples, including those from patients treated with the EGFR tyrosine kinase inhibitor lapatinib, uncovered an EGFRvIII-activated, PI3K/SREBP-1-dependent tumor survival pathway through the low-density lipoprotein receptor (LDLR). Targeting LDLR with the liver X receptor (LXR) agonist GW3965 caused inducible degrader of LDLR (IDOL)-mediated LDLR degradation and increased expression of the ABCA1 cholesterol efflux transporter, potently promoting tumor cell death in an in vivo GBM model. These results show that EGFRvIII can promote tumor survival through PI3K/SREBP-1-dependent upregulation of LDLR and suggest a role for LXR agonists in the treatment of GBM patients.

Figure 1. EGFRvIII EGFRvIII/EGFR signaling promotes LDLR expression through a PI3K/Akt-mediated, SREBP-1 dependent pathway

A) 5 × 10⁵ U87MG and U87MG/EGFRvIII cells were implanted into the flank of immunodeficient SCID/Beige mice. After 21 days, tumors were harvested and immunoblot analysis for p-EGFR, EGFR, p-Akt and LDLR was performed. M is mature form of LDLR, which is glycosylated. P is precursor of LDLR, non-glycosylated. B) U87/EGFRvIII tumors were treated for 7 days with erlotinib (150 mg/kg by oral gavage) and harvested; immunoblot analysis was performed using the indicated antibodies. C) GBM39 cells, a human serially-passaged model of endogenous EGFRvIII expression, were implanted in the flanks of SCID mice and treated for 7 days with erlotinib (150 mg/kg by oral gavage). Immunohistochemical analysis (IHC) of p-EGFR and LDLR was performed. Scale bar = 20 um. D) Western blot was performed to analyze GBM39 tumor lysates using indicated antibodies. E) U87/EGFR cells were placed in serum free medium for 24 hours, stimulated with EGF (20 ng/ml) for the indicated times, and immunoblot analysis was performed using the indicated antibodies. P is precursor of SREBP-1; N is N-terminus of SREBP-1, which is active form. F) Effect of the EGFR inhibitor erlotinib (10 uM), the PI3K inhibitor LY294002 (20 uM), the Akt inhibitor Akti-1/2 (5 uM), or the mTORC1 inhibitor rapamycin (1 nM) for 12 hours on SREBP-1 cleavage and LDLR protein levels in U87-EGFR cells. Cells were pretreated with inhibitors for 30 minutes before EGF 20 ng/ml stimulation; immunoblot analysis was performed using the indicated antibodies. G) U87/EGFRvIII cells were infected using SREBP-1 shRNA lentivirus for 48 hrs at the indicated doses. Immunoblot analysis was performed using the indicated antibodies. H) U87/EGFR cells were transfected using SREBP-1 siRNA (10 nM) for 24 hours, then serum-free overnight, cells were stimulated with EGF (20 ng/ml) for 16 hours. Immunoblot analysis was performed using the indicated antibodies. I) shRNA lentiviral knockdown of SREBP-2 had no effect on LDLR levels. Immunoblot analysis was performed using the indicated antibodies. P is precursor of SREBP-1; C is C-terminus of SREBP-2.

Figure 2. EGFR/PI3K/Akt signaling and nuclear SREBP-1 staining correlate with elevated LDLR expression in GBM patient samples

A) Analysis of p-EGFR, p-Akt, nuclear SREBP-1, and LDLR abundance (reddish brown) in two tissue microarrays (TMA) of 252 tumor cores and 91 matched normal tissues from 140 primary (de novo) GBM patients using immunohistochemical analysis (IHC). Top row is normal brain, the middle row is a negative tumor (GBM A) and the bottom row shows positive staining in a tumor (GBM B). Inset shows nuclear SREBP-1 staining indicated by green arrow. Images are magnified 20×; tissue is counterstained with hematoxylin. Scale bar = 20 um. B) Quantitative analysis of LDLR expression in GBM vs. adjacent normal tissues. C) Correlation analysis between LDLR expression and p-EGFR, p-Akt or SREBP-1 in 140 GBM patients. D) Correlation analysis between LDLR expression and p-EGFR/p-Akt/SREBP-1 signaling pathway in 140 GBM patients. E) Immuohistochemical staining of p-EGFR, p-Akt, SREBP-1 and LDLR before and after 7–10 days of lapatinib treatment in two representative GBM patients. Scale bar = 20 um.

Figure 3. GBM cells depend on extracellular cholesterol levels for growth

A, B) U87 and U87/EGFRvIII cells were cultured in 1% FBS or 1% lipoprotein depleted serum (LPDS) medium for 3 days; live and dead cells were counted using trypan blue assay. Relative cell growth was calculated by comparing cell number with cells in 1% FBS condition (A). Dead cell percentage was calculated by comparing dead cells number with total cell amount (B). C) U87 and U87/EGFRvIII cells were cultured in 1% and 5% of FBS or LPDS medium; low density lipoprotein was added to the cultured media (5 ug/ml, 1:1,000 dilution) for 3 days. Cell number was counted using a hemocytometer. D) Micrographs showing morphological differences in U87 and U87/EGFRvIII cells cultured in 5% FBS and 5% LPDS medium; addition of LDL (5 ug/ml) recovered cell phenotype and cell growth, LDL stock 5 mg/ml. Scale bar = 20 um.

Figure 4. The LXR agonist GW3965 promotes GBM cell death in vitro with enhanced efficacy in EGFRvIII-expressing tumor cells

A) U87 and U87/EGFRvIII cells were treated with LXR agonist GW3965 for 4 days in 1% LDPS medium; cell viability was measured using WST-1 assay daily and normalized with day 0. Indicated cell survival percentage was calculated by normalizing with control at day 4. B, C) U87 and U87/EGFRvIII cells were treated with GW3965 for 2 days in 1% LPDS medium, cells were fixed using paraformaldehyde for 15 minutes at room temperature. Cell apoptosis was assayed by TUNEL staining (Roche) and counterstained with 4’, 6-diamidino-2-phenylindole (DAPI). DMSO, dimethy sulfoxide. Scale bar, 20 um. D) Apoptotic cells in B and C were quantified. E) U87/EGFRvIII cells were treated using GW3965 for 4 days in 1% LPDS medium with or without LDL (50 ug/ml). The images represent cell morphology change after GW3965 treatment. Scale bar = 20 um. F, G) Dead cells and live cells were counted in panel E using trypan blue. LDL was added into media at a range of does as indicated. LDL stock 5 mg/ml.

Figure 5. The LXR agonist GW3965 up-regulates expression of the cholesterol transporter gene ABCA1 and the E3 ubiquitin ligase IDOL and reduces LDLR levels

A, B) U87/EGFRvIII cells were treated with GW3965 5 uM for the indicated times. Total cellular RNA was extracted, and ABCA1 and IDOL gene expression was quantified using real-time PCR. C) U87/EGFRvIII cells were treated with GW3965 for 24 hrs at a range of doses as indicated. Immunoblot analyses were performed using the indicated antibodies. D) Cells were treated with GW3965 at 5 uM for the indicated times. Immunoblot analyses were performed using the indicated antibodies. E) Multiple cancer cell lines were treated with GW3965 at 5 uM for the indicated times. Immunoblot analysis was performed using the indicated antibodies.

Figure 6. Reduction of LDLR protein levels is required for apoptotic effect of GW3965

A) U87/EGFRvIII cells were infected with LDLR shRNA lentivirus or scrambled control and selected by puromycin with 2 ug/ml for 10 days. Cells were treated with GW3965 at a range of doses as indicated for 48 hrs. Immunoblot analysis was performed using the indicated antibodies. B, C) U87/EGFRvIII control shRNA and LDLR knockdown cells were treated with GW3965 at a range of doses as indicated for 3 days in 1% FBS medium. The representative images show cell morphology after GW3965 treatment (B), then cells were counted using trypan blue assay (C). Scale bar = 20 um. D) U87/EGFRvIII cells were infected by adenovirus Ad-LacZ or Ad-Idol for 48 hrs. Immunoblot analysis was performed using LDLR antibody. E, F) U87/EGFRvIII cells were infected with Ad-LacZ or Ad-Idol for 8 hrs, then treated with GW3965 in a range of dose as indicated for 3 days. Representative images after GW3965 treatment (E); tumor cells were counted using trypan blue assay (F). Scale bar = 20 um.

Figure 7. GW3965 treatment inhibited tumor growth in vivo

A) 5 × 105 U87/EGFRvIII cells were implanted into the flank of immunodeficient SCID/Beige mice (n=8/group); after tumor size reached 80 mm3, GW3965 was administrated at 40 mg/kg by oral gavage daily for 12 days. Tumors were harvested and immunoblot analyses was performed using the indicated antibodies. B) Tumor size was measured daily. Tumor growth was demonstrated by fold change compared with the tumor size on first day of treatment, P<0.05. The experiments were repeated twice. C) Representatives of U87/EGFRvIII xenograft tumors after GW3965 treatment for 12 days. D) Tumor tissues were stained by TUNEL assays (Roche) to determine cell apoptosis after GW3965 treatment. Scale bar = 20 um. E) Quantification of TUNEL staining. F) Model demonstrating the pathways by which EGFR/RTK signaling mediates cholesterol metabolism and the therapeutic drug targets. The scheme shows that EGFR/PI3K/Akt signaling regulates LDLR mediated by SREBP-1; GBM cells maintain cholesterol homeostasis through uptake, biosynthesis, and efflux. It also indicates that LXR agonist GW3965 disturbs cancer cell cholesterol homeostasis by up-regulating efflux through transporter gene ABCA1, and by reducing LDL uptake through degrading LDLR mediated by IDOL.