The potent and selective cyclin-dependent kinases 4 and 6 inhibitor ribociclib (LEE011) is a versatile combination partner in preclinical cancer models

Inhibition of cyclin-dependent kinases 4 and 6 (CDK4/6) is associated with robust antitumor activity. Ribociclib (LEE011) is an orally bioavailable CDK4/6 inhibitor that is approved for the treatment of hormone receptor–positive, human epidermal growth factor receptor 2–negative advanced breast cancer, in combination with an aromatase inhibitor, and is currently being evaluated in several additional trials. Here, we report the preclinical profile of ribociclib. When tested across a large panel of kinase active site binding assays, ribociclib and palbociclib were highly selective for CDK4, while abemaciclib showed affinity to several other kinases. Both ribociclib and abemaciclib showed slightly higher potency in CDK4-dependent cells than in CDK6-dependent cells, while palbociclib did not show such a difference. Profiling CDK4/6 inhibitors in large-scale cancer cell line screens in vitro confirmed that RB1 loss of function is a negative predictor of sensitivity. We also found that routinely used cellular viability assays measuring adenosine triphosphate levels as a proxy for cell numbers underestimated the effects of CDK4/6 inhibition, which contrasts with assays that assess cell number more directly. Robust antitumor efficacy and combination benefit was detected when ribociclib was added to encorafenib, nazartinib, or endocrine therapies in patient-derived xenografts.


Enzyme-linked immunosorbent assay to quantitate Rb phosphorylation
4H1 Rb antibody #9309 (CST; catalog number: 9309) was added to clear MaxiSorp ® plates (ThermoScientific) at a 25 ng/well concentration in 50 μL of Dulbecco's Phosphate Buffered Saline (Gibco; catalog number: 14190-144) and incubated overnight with rocking at 4°C. After washing with Tris buffered saline solution-Tween 20 (TBST; Teknova; catalog number: T9501), 250 μL of SuperBlock ® Blocking Buffer (Pierce; catalog number: 37535) was added, and the mixture was incubated for 1 hour. After blocking, 10 μL of cell lysates containing 10 μg of total protein were added in triplicate wells; 40 μL of phosphate buffered saline (PBS; Gibco; catalog number: 10010-023) containing 10% SuperBlock Blocking Buffer was added to the wells to make a final volume of 50 μL. Plates were sealed, incubated for 2 hours at room temperature, and then washed 3 times with TBST. Then, 50 μL of a 1:1000 dilution of Phospho-Rb (Ser780; catalog number: 9307) antibody in PBS/10% SuperBlock Blocking Buffer was added, and the mixture was incubated overnight at 4°C. Plates were washed 3 times in TBST, then 50 μL of 1:2500 dilution of donkeyanti-rabbit HRP (Amersham; catalog number: NA934V) in PBS/10% SuperBlock Blocking Buffer was added, and the plates were shaken for 30 minutes at room temperature. After washing, 100 μL of 1-Step™ Ultra-TMB enzymelinked immunosorbent assay (Pierce; catalog number: 34028) was added, and the mixture was incubated for 30 minutes in the dark. Sulfuric acid (100 μL of 2 M) was added to the plate to stop the reaction, and absorbance was read on a SpectraMax ® reader (Molecular Devices) at 450 nM within 2 hours.

Fluorescence-activated cell sorting
To identify the distribution of cell-cycle phases in treated cells, we measured their DNA content by fluorescence-activated cell sorting or flow cytometry. Analysis of cellular DNA content by propidium iodide (PI; MP Biomedicals) staining can discriminate between cells that are in the G 0 /G 1 , S, or G 2 /M phase. Cells were transferred to a V-bottomed, 96-well polypropylene plate (using trypsin to remove adherent cells, if necessary) and centrifuged to remove the supernatant media, then 100 μL of a hypotonic lysis buffer (0.1% sodium citrate [Sigma; catalog number: S-4641], 0.1% Triton X-100, 25 μg/mL of PI, and 10 μg/mL of RNAse [Roche; catalog number: 1 579 681]) was added to the cells and incubated at room temperature in the dark for 1 hour and stored at 4°C overnight, if necessary. Last, DNA content was analyzed using the BD™ LSR II System and FACSDiva™, version 5.0.1 (BD biosciences) and ModFit LT™ 3.1 (Verity Software House, Inc).

In vitro viability assays for Ribociclib by CellTiter-Glo ® or microscopy
To determine the effects of ribociclib on cell proliferation in vitro, 750 to 1500 cells per well were seeded in 80 μL of medium in 384-well plates, briefly centrifuged to promote an even distribution of cells across the entire well and incubated at room temperature for 30 minutes. All plates were incubated at 37°C in 5% CO 2 for 24 hours before addition of the compound. Then, a 400× compound stock was added using a Peak Analysis & Automation robot equipped with a 200-nL pin tool to achieve a 1× compound concentration. Ribociclib was tested at 8 dose points and 1:3 dilution steps from 4.5 nM and 10 μM. After 72 hours of treatment for a minimum of 3 replicate wells, single-agent effects were assessed by both quantification of cellular ATP levels by CellTiter-Glo (Promega; catalog number: G7573), as described in the manufacturer's protocol, or by microscopy. For quantification by microscopy, cells were fixed and permeabilized for 45 minutes in 4% paraformaldehyde (Electron Microscopy Sciences; catalog number: 15714) and 0.12% TX-100 (Electron Microscopy Sciences; catalog number: 22140) in PBS. After washing the cells 3 times with PBS, DNA and tubulin were stained for 30 minutes with Hoechst 33342 (ThermoFisher; catalog number: H3570) at a final concentration of 4 μg/mL and anti-α-tubulin fluorescein isothiocyanate (FITC) antibody (Sigma; catalog number: F2168) at a final dilution of 1:2500, respectively. Cells were washed again 3 times with PBS, and then the plates were heat-sealed and stored at 4°C until imaging. To quantify cell numbers (equivalent to the number of nuclei revealed by the Hoechst signal), all cells per well were captured in a single image by fluorescence microscopy using an IN Cell Analyzer 2000 (GE Healthcare) equipped with a 4× objective and DAPI excitation/emission filters and a CCD camera. To assess cell size, 4 images per well were captured with a 10× objective using DAPI (for Hoechst 33342) and FITC (for anti-α-tubulin) excitation/emission filters.

Image analysis for Ribociclib in vitro effects
Images were acquired using the IN Cell Analyzer 2000 software and analyzed using adapted methods [1] and using the R/Bioconductor package EBImage [2]. Images captured with the 4× objective lens in the DAPI channel (for Hoechst/DNA) were segmented by adaptive thresholding and counted. By analyzing 17 additional object/nuclei features in the DNA channel (ie, shape and intensity features), debris/ fragmented nuclei were identified. The distributions of the additional features between staurosporine-positive controls and DMSO-negative controls were compared manually. Features that could differentiate between the conditions (eg, a shift in the distribution of a feature measurement comparing DMSO with staurosporine) were used to define the "debris" population and subtract it from raw nuclei counts to obtain measurements of cell proliferation (ie, cell count). Images captured with the 10× objective lens in the DAPI channel (for Hoechst/DNA) were segmented by adaptive thresholding, and the cell size was estimated from fluorescence levels in the α-tubulin channel by extension from the nuclei in the DAPI channel using a propagation algorithm [3].

In vivo studies
Animal studies comply with Novartis Global Standards and Principles for the Care and Use of Animals, in accordance with National Research Council 2011 standards and are conducted under protocols approved by Institutional Animal Care and Use Committee (protocol #12-ONC-016). Animals were housed in a temperatureand humidity-controlled vivarium with a 12-hour light/ dark cycle and were provided with food and water ad libitum. In all studies, tumor volume was calculated using the following formula: (length × width 2 )/2. Percent treatment/control (T/C) values were calculated using the following formula: %T/C = 100 × ΔT/ΔC, if ΔT >0. Percent regression was calculated using the following formula: %Reg = 100 × ΔT/T initial , if ΔT <0. In all equations, T = mean tumor volume of the drug-treated group on the final day of the study; ΔT = mean tumor volume of the drug-treated group on the final day of the study − mean tumor volume of the drug-treated group on initial day of dosing; T initial = mean tumor volume of the drug-treated group on initial day of dosing; C = mean tumor volume of the control group on the final day of the study; and ΔC = mean tumor volume of the control group on the final day of the study − mean tumor volume of the control group on the initial day of dosing. The percent change in body weight (BW) was calculated as (BW current − BW initial )/BW initial × 100. Data are presented as the percent change in BW from the day of treatment initiation and expressed as mean ± standard error of the mean.

Pharmacokinetic, pharmacodynamic, and efficacy studies in the JeKo-1 rat model
For pharmacokinetic (PK), pharmacodynamic (PD), and efficacy studies in the mantle cell lymphoma model, male nude rats (NTac:NIH-Foxn1 rnu [Taconic]) were sublethally irradiated, and then JeKo-1 cells with Matrigel™ (BD Biosciences) supplementation were implanted subcutaneously. Rats bearing tumors ≥800 mm 3 were randomized into treatment groups for daily oral treatment with ribociclib. For the PK/PD study, rats received 5 days of once-daily ribociclib, and then tumors were harvested at designated time points for measurement of Rb phosphorylation by enzyme-linked immunosorbent assay as described previously. Blood was also collected to measure drug concentrations in plasma. For the efficacy study, rats received once-daily ribociclib for 28 days, and tumor volumes were measured regularly to determine the effects of the treatment on tumor growth.

Combination of Ribociclib with Encorafenib
The HMEX1906 xenograft model was confirmed to carry the BRAF V600E mutation and deletion of CDKN2A (p16; data not shown). Female athymic nude mice (Crl:NU[Ncr]-Foxn1 nu [Charles River Laboratories]) were implanted subcutaneously with 100 μL of minced HMEX1906 tumor (passage 6) suspended in 50% Matrigel in the right dorsal axillary region. Treatment for the tumor growth efficacy experiment was initiated when the tumor volume reached an average size of 420 mm 3 (range, 252-642 mm 3 , 35 days post tumor implantation). The mice were sorted into the treatment groups shown in Figure 6A (n=8 per group). Ribociclib was administered at doses of 250 mg/kg and 75 mg/kg once-daily, using ribociclib-succinate (salt, powder form) formulated in 0.5% methylcellulose. Encorafenib formulated in 0.5% carboxymethylcellulose and 0.5% Tween 80 was administered at 5 mg/kg twice daily. Body weight and tumor volume were recorded twice weekly. The mice received continuous daily drug treatments until tumors relapsed (tumor volume >1000 mm 3 ).

Combination of Ribociclib with Letrozole
Tumor fragments of the HBCx-34 patient-derived xenograft model were implanted subcutaneously in 5-weekold female athymic nude mice Crl:NU(Ncr)-Foxn1 nu (Charles River Laboratories) weighing 18 to 25 g. To stimulate initial rate of intake and tumor growth, HBCx-34 tumor-bearing mice received supplementary estrogen diluted in drinking water (β-estradiol, 8.5 mg/L) from the date of tumor implant to the date of inclusion (ie, for 37 days). Supplementary estrogen was then removed during the drug treatment period. Drug treatment for the tumor growth efficacy experiment was initiated 37 days after implantation of the tumor. Ribociclib was administered at 75 mg/kg in combination with 2.5 mg/kg letrozole for 56 days. Tumor volumes and BWs were evaluated biweekly during the treatment period.

Breast cancer subtype determination in patientderived xenograft models
Patient-derived xenograft samples were evaluated for HER2 (ERBB2), estrogen receptor (ER; ESR1), and progesterone receptor (PR; PGR) expression. HER2 status was determined by quantifying protein expression with immunohistochemistry (IHC; antibody: Dako polyclonal rabbit antibody [A0485]) and by quantifying gene copy number using Affymetrix™ Genome-Wide Human Single Nucleotide Polymorphism 6.0 Arrays. In addition, expression of ER (antibody: Spring Bioscience, monoclonal rabbit antibody [M3014]) and PR (antibody: Spring Bioscience, monoclonal rabbit antibody [M3024]) were quantified by IHC. Xenograft tumors were determined to be HER2 positive if either of the following was observed: (1) uniform intense membranous staining observed by IHC in at least 30% of cells (3+), or (2) complete membranous staining, either uniform or weak, in at least 10% of cells (2+) and gene copy number greater than 10. Xenograft tumors were determined to be ER positive or PR positive if as few as 1% of tumor cells showed weak immunostaining by IHC. Patient-derived xenograft models were then assigned to 1 of 4 categories: (1) HER2 positive (xenograft tumors scored as HER2 positive regardless of ER and PR status), (2) ER positive (xenograft tumors scored as HER2 equivocal or negative that were ER positive regardless of PR status), (3) PR positive (xenograft tumors scored as HER2 equivocal or negative that were ER negative and PR positive), or (4) triple negative (xenograft tumors scored as HER2 equivocal or negative that were ER and PR negative).