Concurrent targeting of BMI1 and CDK4/6 abrogates tumor growth in vitro and in vivo

Despite recent advances in cancer management and therapy, resistance to cytotoxic medications remains a major clinical challenge; hence, combination-based anti-cancer treatment regimens are currently gaining momentum. PTC-209 reduced BMI1 protein expression, while palbociclib inhibited CDK4, Rb, and pRbSer795 protein expression in MDA-MB-231 cells. PTC-209 and palbociclib exhibited dose-dependent cytotoxic effects against MDA-MB-231 (breast), HCT116 (colon), and PC-3 (prostate) models, which was more profound in the combination group. Transcriptome and pathway analyses revealed inhibition of insulin signaling, focal adhesion, DNA damage response, and Wnt/pluripotency signaling pathways as well as cell proliferation, and cellular movement functional categories by PTC-209. Transcriptome and pathway analyses revealed palbociclib to mainly affect cell cycle progression and survival. Upstream analysis identified several networks affected by PTC-209 (EZH2, IFNB1, TRIB3, EGFR, SREBF1, IL1A, ERG, TGFB1, MAX, MNT) and palbociclib (RABL6, MITF, RARA, TAL1, AREG, E2F3, FOXM1, ESR1, ERBB2, and E2F). PTC-209 and palbociclib reduced colony and sphere formation, cell migration, and cell viability, which was further enhanced in the combination group. Concordantly, combination of PTC-209 and palbociclib exhibited more profound effects on MDA-MB-231 tumor formation in vivo. Our data suggest concurrent targeting of BMI1 and CDK4/CDK6 might provide novel therapeutic opportunity for breast, colon, and prostate cancer.


Maintenance of cancer cell lines. Human TNBC (MDA-MB-231), prostate (PC-3) and colon (HCT116)
cancer cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cancer cells were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Pen-Strep), all were purchased from Thermo Scientific (Thermo Scientific, Rockford, IL, USA). Cells were cultured as an adherent monolayer at 37 °C under 5% CO 2 in a humidified incubator.
Measurement of cell viability. Cell viability was measured using alamarBlue (Thermo Scientific, Rockford, IL, USA) assay according to the manufacturer's recommendations as described before 25 . In brief, 4 × 10 3 cancer cells were seeded in a flat-bottom 96-well plates in the presence of PTC-209 or palbociclib (0.07-10 μM) as single agents or in combination (0.07-10 μM). On day 4, 20 µl of alamarBlue was added to each well at a final concentration of 10% and then plates were incubated in the dark at 37 °C for 2 hrs. Quantification of fluorescent signal intensity was performed using a BioTek Synergy II (BioTek Inc., Winooski, VT, USA) fluorescent plate reader at an excitation and emission wavelengths of 540 nM and 590 nM, respectively. Cell viability of cancer cell lines is presented as percent viability of treated cells compared to DMSO vehicle-treated control cells. PTC-209, palbociclib and combinational (PTC-209 plus palbociclib) were used at 5 μM concentration in subsequent experiments, unless indicated otherwise. Each experiment had a minimum of three replicates and was conducted at least twice.
Colony formation assay. The colony forming ability of cancer cells treated with the indicated small molecule inhibitor was determined using clonogenic assay as described before 26 . Briefly, 4 × 10 4 cells were seeded in 2 ml culture medium in 12-well flat-bottom tissue culture plate, followed by subsequent serial dilution (1:1 to 1:32) in the presence of PTC-209 (5.0 μM), palbociclib (5.0 μM), or combination of PTC-209 and palbociclib (5.0 μM each) compared to DMSO vehicle control. The media was changed with fresh media supplemented with the appropriate inhibitor every 3-4 days. On day 10, the plates were washed and stained with Diff-Quik Staining Kit (Siemens Healthcare Diagnostics, USA), and were subsequently scanned and the number of colonies were observed under inverted microscope.
Sphere formation assay. MDA-MB-231 cells were trypsinized, counted, re-suspended in culture media, and were plated in 60 mm low cell binding dishes (Nunc, Thermo Scientific, Rockford, IL, USA) at 3 × 10 4 cells/ well as we previously described 27 in the presence of PTC-209 (5.0 μM), palbociclib (5.0 μM), or combination of both inhibitors (5.0 μM each). Cells treated with DMSO were used as experimental control. On day 13, the size and number of multicellular tumor spheroids in each well were counted using an inverted microscope (Axio Observer-A1, Carl Zeiss, Germany) at 10X magnification.
Scratch assay. MDA-MB-231 cell migration was assessed using the scratch assay as described before 28 . Cells were seeded in 12-well plates and once they reached confluence, cells were treated with the indicated inhibitors for overnight. Vertical wounds were subsequently created using a sterile 200-µl pipette tip. To eliminate cell debris, adherent cells were washed twice with fresh media before supplementation with fresh medium containing the appropriate inhibitor at 5.0 μM compared to cells cultured in the presence of DMSO vehicle control. Cell migration was subsequently monitored and scored at 24 hrs and 48 hrs using phase contrast microscope (Axio Observer-A1, Carl Zeiss, Germany) at 5X magnification.
Transwell migration assay. A transwell migration system utilizing 8.0 µm pore polyethylene terephthalate (PET, BD Falcon, MA, USA) inserts was used for the migration experiments as we previously described 29 . Briefly, inserts were placed in 12-well notch plate, and 5 × 10 4 cells were suspended in DMEM supplemented with 1.0% serum and were seeded in the upper chamber, while DMEM supplemented with 10% serum was used as attractant in the lower chamber. Twenty four hours later, non-migrating cells were gently removed using cotton swab, Detection of apoptosis/necrosis using fluorescence microscopy. The acridine orange and ethidium bromide (AO/EB) fluorescence staining method was used to assess apoptosis/necrosis in cells after exposure to 5.0 μM of PTC-209, palbociclib, or combination of both. After 72 hrs of treatment, control and treated cancer cells were washed twice with PBS and subsequently stained with dual fluorescent staining solution containing 100 μg/ml AO and 100 μg/ml EB (AO/EB, Sigma Aldrich, St. Louis, MO, USA) for two minutes; subsequently, the cells were observed and imaged under Nikon Eclipse Ti2 fluorescence microscope (Nikon, Tokyo, Japan). The differential uptake of AO/EB allows the identification of viable and non-viable cells. Principally, Acridine Orange was used to visualize the number of cells undergone apoptosis, while EB positive cells indicated necrotic cells. www.nature.com/scientificreports www.nature.com/scientificreports/ Gene expression profiling using microarray and pathway analysis. One hundred nanograms of total RNA was labeled using the low input Quick Amp Labeling Kit (Agilent Technologies, CA, USA) and then hybridized to the Agilent Human SurePrint G3 Human GE 8 × 60 k microarray chip as we previously described 32 . The extracted data were normalized and analyzed using GeneSpring 13.0 software (Agilent Technologies). Pathway analyses were performed using the single experiment pathway analysis feature in GeneSpring 13.0 as described earlier 20 . Two-fold cutoff and p < 0.05 (Benjamini-Hochberg multiple testing corrected) were used to determine significantly changed transcripts. Ingenuity pathway and functional annotation analyses were conducted using Ingenuity pathway (Ingenuity Systems, http://www.ingenuity.com) 33,34 . Differentially expressed genes exhibiting ≥ 2 ≤ and corrected P value < 0.05 were subjected to core analysis using the human database. Enriched networks categories were algorithmically generated based on their connectivity and ranked according to the Z score. A Z score ≤ or ≥2.0 was considered significant.
In vivo xenograft assay. In vivo experiments were approved by the Institutional Animal Care Committee, King Saud University and were conducted as we previously described 26,35 . All methods were performed in accordance with the relevant guidelines and regulations. Briefly, four to six weeks old, nude mice (bred at the Experimental Surgery and Animal Laboratory, King Khalid University Hospital) were utilized for the xenograft experiments and were housed under conditions of a 12 hour light/dark cycle, 20-24 °C, and 60-70% humidity. MDA-MB-231 cells were exposed to PTC-209, palbociclib, or combination of the two inhibitors at 5.0 μM for 72 hrs. Cells were then trypsinized, washed with PBS, and 2 × 10 6 cells were subcutaneously injected into the right left flank of female nude mice in a 100 μl mixture (1:1 v/v of PBS/matrigel). The animals were monitored twice weekly and tumor volume was measured using caliper. At the end of these experiments, the mice were sacrificed, and the tumors were excised, fixed in 10% buffered formalin, embedded in paraffin, and were subsequently sectioned. Sections were stained with haematoxylin & eosin. Human-specific vimentin staining was used to detect human MDA-MB-231 cells in the xenograft.
Statistical analysis. Data were presented as mean ± standard error of the mean (S.E.M). Statistical analyses and graphing were performed using GraphPad Prism 8.0 software (GraphPad, San Diego, CA, USA). P-values were calculated using the two-tailed t test. Data are derived from at least two independent experiments ran in triplicate, unless stated otherwise.  Fig. 1b lower panel), which was associated with significant reduction in CDK4, Rb, and pRb Serine795 protein expression (Fig. 1b, upper panel). The combination of the two drugs exerted more inhibitory effects on the growth of MDA-MB-231 cells compared to either drug alone (0.07-10 μM, Fig. 1c). We subsequently assessed the ramifications of PTC-209, palbociclib or the combination of both inhibitors on the clonogenic and sphere formation potential of MDA-MB-231 cells. Data presented in Fig. 1d,e revealed significant inhibition of MDA-MB-231 colony and sphere formation, respectively, which was more profound in the PTC-209 plus palbociclib combination group. Concordantly, combination of PTC-209 and palbociclib was most effective in inhibiting MDA-MB-231 cell migration using the classical boyden chamber (Fig. 1f) and scratch assays (Fig. 1g). In order to gain more insight on the mechanism by which PTC-209 and palbociclib inhibited MDA-MB-231 cell growth, we conducted cell cycle analysis and AO/EB staining under different treatment conditions. Data presented in Fig. 1h,i revealed significant increase in apoptosis (sub-G0) and arrest in the G0/G1 phase of the cells cycle in PTC-209, palbociclib, and the combination treatment group. We subsequently extended our investigations to the HCT116 (colon) and PC-3 (prostate) cancer models, since those two models are known to harbor BMI1 + CSC populations 37,38 . Additionally, CDK4 and CDK6 are known to be expressed by both cancer models 39,40 . Combination of PTC-209 and palbociclib exhibited dose-dependent growth inhibitory effects on HCT116 and PC-3 cancer cells (Fig. 2a). Concordantly, combination of PTC-209 and palbociclib was most efficacious in inhibiting the colony formation, as well as the migration of the HCT116 and the PC-3 cancer cells, respectively ( Fig. 2b-d). Concordant with data obtained using the MDA-MB-231 model, induction of apoptosis and G1 cell cycle arrest was similarly observed in the HCT116 and the PC-3 models (Fig. 2e) in response to PTC-209 and palbociclib, or combination treatment.

Multiple dysregulated cellular pathways in PTC-209 treated MDA-MB-231 cells.
To gain more insight into the molecular mechanisms by which PTC-209 affects cancer cellular and functional processes, we performed global mRNA expression profiling on the MDA-MB-231 breast cancer model treated with PTC-209 compared to vehicle-treated control cells. As depicted in Fig. 3a, hierarchical clustering based on differentially expressed mRNA transcripts revealed clear separation between the two treatment groups. A total of 1439 transcripts were upregulated, while 1520 transcripts were downregulated in response to BMI1 inhibition using PTC-209 (2.0 FC, P(corr) < 0.05, Supplementary Table 1). The distribution of the top 10 enriched pathway designations for the differentially expressed genes is shown in Fig. 3b. The expression of selected number of genes from the microarray data was subsequently validated using quantitative reverse transcription PCR (qRT-PCR, Fig. 3c). We subsequently subjected the differentially expressed transcripts in response to PTC-209 treatment into ingenuity www.nature.com/scientificreports www.nature.com/scientificreports/ pathway analysis software. As shown in Fig. 3d, disease and function analysis revealed remarkable inhibition of cellular movement, growth, and proliferation in PTC-209 treated cells. Top significantly affected disease and function categories (≤2.0 ≤ activation Z score) are shown in Fig. 3e. The cellular movement and cellular growth and proliferation functional categories are depicted in Fig. 3f,g, respectively.
Palbociclib predominantly target pathways regulating cell cycle and cell proliferation. The signaling pathways affected by palbociclib treatment were subsequently analyzed using global mRNA expression profiling and pathway analysis. As shown in Fig. 4a and Supplementary Table 2, palbociclib treatment led to substantial changes in gene expression of MDA-MB-231 cells. Palbociclib treatment predominantly affected genes involved in cell cycle regulation and DNA replication (Fig. 4b). Validation of selected genes from the microarray data using qRT-PCR is shown in Fig. 4c. The list of differentially expressed genes in palbociclib-treated MDA-MB-231 cells was subsequently subjected to IPA, which revealed significant suppression of genes regulating cell cycle progression, cell survival, while genes involved in promoting cell death were enriched (Fig. 4d). The list of most affected functional categories is shown in Fig. 4e. Cell cycle and cell death and survival functional categories are illustrated in Fig. 4f,g, respectively.
Upstream regulator analysis on the differentially expressed genes revealed dramatic effects of palbociclib on several vital networks in cancer cells. Notably, the most inhibited networks by palbociclib were those driven by RABL6, MITF, RARA, TAL1, AREG, E2F3, FOXM1, ESR1, ERBB2, and E2F upstream regulators (Fig. 5a,  Supplementary Table 3). Although the expression of several of those upstream regulators was not affected by palbociclib at the transcriptional level, substantial inhibition of downstream targets was observed in our microarray data suggesting inhibition of those networks by palbociclib. Notably, FOXM1 levels were severely downregulated in response to palbociclib treatment (−10.8 FC). The FOXM1 network is illustrated in Fig. 5b. On the other hand, among the top activated networks was NUPR1 and TP53 (Fig. 5a). NUPR1 is a stress-induced chromatin-associated protein, which is necessary for the expression of stress-response genes, including those involved in DNA repair, cell cycle regulation, apoptosis, and autophagy 41 . NUPR1 was upregulated in response to palbociclib treatment (8.8 FC), which is likely as a response to DNA damage and induction of cell death inflicted by palbociclib. Several downstream targets of TP53 were also affected by palbociclib treatment (Supplementary  Table 4), which collectively indicated activation of TP53 in palbociclib-treated cells. Illustration of the TP53 network is shown in Fig. 5c. On the other hand, PTC-209 inhibited several networks, including EZH2, IFNB1, TRIB3, EGFR, SREBF1, IL1A, ERG, TGFB1, MAX, and MNT (Fig. 5d, Supplementary Table 5). EZH2 is the core component of the PRC2, which implies disruption of BMI1 by PTC-209 not only affects the PRC1 complex, but it also affects the PRC2 complex. Among the induced networks, we observed activation of NUPR1 network, which is similar to the palbociclib data. Activation of NUPR1 might imply induction of DNA damage and induction of  Fig. 6b. Histological examination of xenograft tumors revealed a greater degree of cell death and lesser mitotic figures in the combination drug group, while a number of pleomorphic cells were observed in control group (Fig. 6c).

Discussion
Despite recent advances in the development of molecular and cellular-based anti-cancer therapies, chemotherapy remains a main treatment choice for large number of human cancers 42 . However, the evolution of chemoresistence during the treatment process represent a major obstacle for successful chemotherapy 43 . Tumor-initiating cells, also termed cancer stem cells (CSCs), represent a minor subpopulation of cancer cells with intrinsic self-renewal, differentiation and tumor initiation properties, which has also been implicated in drug resistance 44 . In current study, we present the first evidence on the efficacy and elucidated the molecular mechanism by which combination of PTC-209 and palbociclib inhibited the growth of a panel of cancer models (breast: MDA-MB-231; colon: HCT116; prostate: PC-3). Combination of PTC-209 and palbociclib inhibited cell proliferation, sphere and colony formation, cell migration in vitro, and tumor formation in vivo. Tumor sphere generated in vitro is widely acceptable as indicative of self-renewal potential, one of the characteristics of TICs 45 . In current study, we observed remarkable reduction in spheres and colony formation in the combined treatment group; in addition, sphere formed in the combination group were also less compact compared to the control group, which corroborates earlier findings where PTC-209 was shown to reduce sphere formation of biliary tract and colorectal cancer cells 21,46 .
PTC-209 treatment inhibited a number of signaling pathways including those involved in insulin signaling, DNA damage response and wnt/pluoripotency in the MDA-MB-231 breast cancer model. Additionally, PTC-209 inhibited cellular movement and proliferation cellular processes. Metabolic stress has been implicated in the acquisition of stem cell-like characteristics and pharmacological modulation of Wnt signaling pathway disrupting glycolysis in cancer cells 47 . Earlier studies also correlated overexpression of IGF2BP2 in breast cancer and esophageal adenocarcinoma and short patient survival 48 , and as biomarker in several cancer types 35,49 . IGF2BP2 has been implicated in the maintenance of cancer stem cells (CSCs) 50 . Concordantly, blockade of the insulin signaling www.nature.com/scientificreports www.nature.com/scientificreports/ pathway inhibited the growth and metastasis in several cancer types including breast cancer both in vitro and in vivo 51,52 . Therefore, our data established targeting the insulin signaling pathway as one potential mechanism by which PTC-209 inhibited tumorigenicity of MDA-MB-231 model. GADD45, member of the DNA damage response pathway, has been implicated in several signaling pathways and cellular functions, including MAPK signaling, cell cycle regulation, DNA repair, genomic stability, and apoptosis 53,54 . Our data revealed marked increase in the expression of genes involved in DNA damage response such as GADD45A, GADD45B and RAD51 in treated groups. Overexpression of GADD45 family proteins was shown to induce apoptosis in leukemia, lung cancer and HeLa cells 53,55 . Our finding from current study is concordant with our previous data implicating BMI1 in protecting cancer cells from radiation therapy-induced apoptosis 20 . Taken together, our data revealed inhibition of BMI1 using PTC-209 to induce increased DNA damage and to induce apoptosis in cancer cells.
On the other hand, palbociclib treatment of MDA-MB-231 cells lead to substantial changes in gene expression affecting mainly cell cycle progression. In several cancer models, E2F was shown to activate transcriptional programs and to promote cell cycle progression and DNA damage response. Retinoblastoma, a tumor suppressor protein, could prevent abnormal cell growth by suppressing E2F activity and inhibiting the expression of CDK4