Downregulation of Bmi1 in breast cancer stem cells suppresses tumor growth and proliferation

Targeting cancer stem cells during initial treatment is important to reduce incidence of recurrent disease. Bmi1 has been associated with cancer stem cell self-renewal and aggressive disease. The purpose of this study was to determine the effects of downregulation of Bmi1 in breast cancer stem cells in order to target and eliminate the stem cell population in the tumor mass. Bmi1 was downregulated using two approaches in the mouse breast cancer stem cell line FMMC 419II—a small molecule inhibitor (PTC 209) and stable transfection with a Bmi1 shRNA plasmid. The functional effect of Bmi1 downregulation was tested in vitro and in vivo. Each approach led to decreased Bmi1 expression that correlated with an inhibition of cancer stem cell properties in vitro including cell cycle arrest and reduced mammosphere forming potential, and a decrease in tumor mass in vivo after either intra-tumoral or systemic nanoparticle-targeted delivery of anti-Bmi1. These results show that inhibiting Bmi1 expression in breast cancer stem cells could be important for the complete elimination of tumor and potentially preventing disease relapse.

another 24 hours to synch up the growth cycle. Bmi1 shRNA transfected cells were trypsinized and washed in PBS. Cells were then centrifuged, fixed, and permeablized by slow addition of ice-cold methanol to the pellet under constant vortexing and incubation at 37°C for 20 minutes. Cells were then washed in PBS+0.1% BSA and centrifuged at 200 x g. Cell pellets were stained with PI/RNAse staining buffer (BD Biosciences) and run on the FACSAria flow cytometer. Doublets were excluded and the DNA content of the single cells was analyzed to obtain the percentage of cells in each phase of the cell cycle (Supplementary Table 1). Analysis of the data was done using FlowJo VX.0.7 data analysis software (FlowJo, LLC, Ashland, OR).

Cell proliferation and viability assay
Proliferation and viability were measured using the Vybrant MTT (3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide) Cell Proliferation kit from Thermo Fisher Scientific. Cells treated with 2 μM and 5 μM PTC 209 or Bmi1 shRNA transfected cells were trypsinized and washed with media. Cells were then counted, plated at appropriate concentrations in 96-well plates, and incubated at 37°C for 48 hours. After incubation, MTT assay was performed using the manufacturer's protocol. Briefly, the MTT reagent was added to each well at 0.5 mg/ml and incubated for 3 hours. The formazan crystals were then solubilized with DMSO and the absorbance read at 540 nm as an indicator of cell growth and proliferation (Supplementary Figure 1). Each sample was plated in triplicate and results are displayed as mean ± S.E.M.

Mammosphere formation assay
The mammosphere formation assay was performed according to Lo et al. [2]. Mammosphere culture medium was prepared by addition of 20 ng/ml EGF, 10 ng/ml basic FGF, 5 μg/ml insulin, and 0.4% bovine serum albumin to DMEM. Cells were resuspended in this medium at a concentration of 1 cell/μl and 200 μl was seeded into 10 wells of an ultra-low attachment 96-well plate. The top and bottom rows were filled with 200 μl of 1X PBS and the plates were sealed with laboratory tape to prevent evaporation. Plates were incubated at 37°C for 2 weeks. After incubation, tumorspheres were counted under a phase-contrast microscope at 20X magnification (Supplementary Table 2) and the diameters of the tumorspheres were also measured.

Nanoparticles (NPs) synthesis
Briefly, 100 μl of PLGA-PEG and maleimide-PLGA-PEG in a ratio of 9:1 (80 mg/ml) were mixed with 20 μl of PTC 209 (20 mg/ml in DMSO). This solution was added drop by drop to 10 ml of 1% PVA solution under constant magnetic stirring. It was then sonicated in a probe sonicator for 30 seconds and stirred for another 2 hours. Finally, the solution was dialyzed to remove free PTC 209 and residual DMSO. Anti-CD49f was conjugated to the surface of the PTC 209-encapsulated NPs by first thiolating the anti-CD49f using Traut's reagent. The thiolated anti-CD49f readily reacts with the maleimide group on the NPs (Supplementary Figure 2). This nanoformulation was lyophilized in the presence of cryoprotectant 1-2% sucrose or mannitol to get it in a powdered form and then was re-dispersed in DI water for further studies.

Dye-conjugated nanoparticles (NPs)
To facilitate in vivo imaging to determine cellular uptake, the NPs were labeled with a Cy7-NHS dye as per our previously described method [3] to attach an amine arm to the terminal COOH of the NPs.

Dynamic laser light scattering (DLS)
The size distribution of PLGA-PEG NPs in aqueous dispersions was determined using a Malvern zeta sizer (Malvern Co, Westborough, MA, USA). We observed an average size of 225 nm in diameter (Supplementary Figure 3).

Entrapment efficiency (EE)
EE of the PLGA-PEG NPs encapsulating docetaxel was determined by filtering a known amount of PLGA-PEG NPs through a 100 kDa filter membrane to separate the free PTC 209 from the encapsulated PTC 209. We

In vivo imaging with Cy7-labeled anti-CD49f nanoparticles
Anti-CD49f NPs or unconjugated NPs were labeled with Cy7 and injected subcutaneously (data not shown, essentially the same results as IP) or IP (Supplementary Figure 4) into normal FVB mice or FVB mice with tumors produced by injection of 1 x 10 5 FMMC 491II cells in 50% Matrigel in the lower mammary fat pads. IVIS imaging was done 1 week after injection of the cells when tumors were very small (less than 3 X 3 X 2 mm). IVIS images were taken before injection, immediately after injection and 1, 2, 3, 4 and 24 hours after injection. Tumors were then harvested to study the fluorescence intensity. With either NP injection site, there is immediate localization of both the NP Cy7 and anti-CD49f Cy7 NPs to the tumor sites. However, there is much less and transient (less than 4 hours) labeling of the tumor site with NP Cy7 compared with the heavier and longer lasting (over 24 hours) labeling using anti-CD49f-NP Cy7. There is no prolonged labeling of any other tissue site. These data support the hypothesis that the anti-CD49f coated NPs are selectively localized to tumor tissues and not to any normal tissues.

Histology
Tumor tissues collected at termination were fixed in 4% paraformaldehyde for 4 hours and then transferred to distilled water. Tissues were then processed by paraffin embedding and were cut into 6 μm sections that were placed on microscopy slides. For H and E staining, slides were hydrated and then stained in hemtoxylin for 10 minutes. The stain was then developed by passing the slides through 3 acidified ethanol washes and 5 minute incubation in ammonia. Slides were then stained in eosin for 40 seconds and dehydrated by passing through 3 washes of 100% ethanol and 3 washes of xylene before coverslips were affixed. Tumor sections were then visualized using bright field microscopy at 10X and 20X magnification.