PROTAC-mediated CDK degradation differentially impacts cancer cell cycles due to heterogeneity in kinase dependencies

Abstract

Background: Cyclin-dependent kinase 4 and 6 (CDK4/6) inhibition yields differential cellular responses in multiple tumor models due to redundancy in cell cycle. We investigate whether the differential requirements of CDKs in multiple cell lines function as determinant of response to pharmacological agents that target these kinases.

Methods: We utilized proteolysis-targeted chimeras (PROTACs) that are conjugated with palbociclib (Palbo-PROTAC) to degrade both CDK4 and CDK6. FN-POM was synthesized by chemically conjugating pomalidomide moiety with a multi-kinase inhibitor, FN-1501. Patient derived PDAC organoids and PDX model were utilized to investigate the effect of FN-POM in combination with palbociclib.

Results: Palbo-PROTAC mediates differential impact on cell cycle in different tumor models, indicating that the dependencies to CDK4 and 6 kinases are heterogenous. Cyclin E overexpression uncouples cell cycle from CDK4/6 and drives resistance to palbo-PROTAC. Elevated expression of P16INK4A antagonizes PROTAC-mediated degradation of CDK4 and 6. FN-POM degrades cyclin E and CDK2 and inhibits cell cycle progression in P16INK4A-high tumor models. Combination of palbociclib and FN-POM cooperatively inhibit tumor cell proliferation via RB activation.

Conclusion: Resistance to CDK4/6 inhibition could be overcome by pharmacologically limiting Cyclin E/CDK2 complex and proves to be a potential therapeutic approach.

Fig. 1. Cellular response to different palbociclib-PROTACS.

a Chemical structures of three different Palbociclib-PROTACS, where the pomalidomide moiety is attached to palbociclib via structurally distinct linkers. b Effect of three different palbo-PROTACS on the degradation of CDK4 and CDK6 in 1222 cells following 18 H treatment. c Western blot analysis to indicate the impact of palbo-PROTACS on CDK4 and CDK6 following 48 H exposure. d Live cell imaging to monitor the proliferation of cells following the treatment with BSJ-02-162 (250 nM) and concurrent depletion of CDK4 and CDK6 kinases. Graphs represent mean and SD calculated from triplicates. Experiment was done at two independent times. e Comparative analysis of the effect of palbociclib and BSJ-02-162 on the indicated proteins in 1222 cell lines at two different concentrations (0.1 and 0.2 µM). f Immunoprecipitation of cyclin D1 was performed in 1222 cells following 48 H treatment with BSJ-02-162 (0.2 µM). Co-immunoprecipitated proteins were determined by western blotting. g Effect of CDK4 and CCND1 knockdowns on the cellular response of BSJ-02-162 (0.2 µM) in 1222 cells. The effect on the indicated proteins were determined by immunoblotting. h Effect of CDK4 knockdown following the depletion of CDK4 in the absence and presence of BSJ-02-162 (0.2 µM) on the proliferation of 1222 cells. Graphs represent mean and SD calculated from triplicates. Experiment was done at two independent times. (***p < 0.001 as determined two-way ANOVA).

Fig. 2. Differential effects of BSJ-02-162 in a panel of cell lines.

a IC50 values of BSJ-02-162 on the indicated cell lines based on its impact on cell proliferation following 120 H treatment. b Oncoprint indicates the different alterations on the indicated genes in different cell lines. c Western blotting analysis to determine the basal expression of different cell cycle proteins on the indicated cell lines. The total protein was validated based on the loading control, βActin. d Live cell imaging to indicate the differential effect of BSJ-02-162 (0.25 µM) in three different ER+ breast cancer cell lines, MCF7, T47D and CAMA-1 following the 120 H treatment. e Western blotting to determine the effect of BSJ-02-162 at two different concentrations (0.1 and 0.2 µM) on CDK4 and CDK6 in MCF7 and CAMA-1 after 48 H exposure. f Effect of CDK4, CDK6 and concurrent knockdowns (CDK4/6) on the proliferation of MCF7 and CAMA1 cells based on live cell imaging. Mean and SD were calculated from triplicates and the experiments were done at two independent times. g Differential effect of CDK4 knockdown on MCF7 and CAMA-1 cells on the cell cycle proteins based on immune blotting. h Live cell imaging in T47D cells to determine the effect of CDK4, CDK6 and concurrent knockdowns. Mean and SD were calculated from triplicates from two independent experiments. (***p < 0.001 as determined by two-way ANOVA). i Effect of CDK6 KD on the cellular response to BSJ-02-162 was determined based on live cell imaging in 1222 and MCF7 cells. The experiment was done at two independent times and the mean and SD were calculated form triplicates. (***p < 0.001 as determined by two-way ANOVA). j Biochemical analysis to validate the CDK6 KD and its impact on cell cycle proteins following the treatment with BSJ-02-162 (200 nM) in MCF7 and 1222 cells.

Fig. 3. Resistance to BSJ-02-162 via CCNE1 and CDKN2A overexpression.

a Western blot analysis on the indicated proteins from HCC1806 cells treated with different concentrations of BSJ-02-162 for 48 H. b Live cell imaging to monitor the cell proliferation in HCC1806 following the depletion of CDK4 and CDK6. Mean and SD were calculated based on triplicates from two independent experiments. c Effect of CCNE1 KD in combination with BSJ-02-162 (250 nM) in HCC1806 cells. Growth curve represents two independent experiments. Mean and SD were calculated from triplicates. d Western blotting in HCC1806 cells to determine the effect of BSJ-02-162 on its cellular targets and cell cycle proteins following the deletion of Cyclin E. e Impact of Cyclin E overexpression on the cellular response to BSJ-02-162 in MCF7 cells based on the expression of cell cycle proteins. f Effect of CDK4 and CDK6 knockdowns on the proliferation of MB157 cells. Error bars represent mean and SD from two independent experiments. g Immunoblotting on the indicated proteins in MB157 cells following the treatment with BSJ-02-162. h Immunoprecipitation of P16INK4A from MB157 cells and the co-immunoprecipitated proteins were determined by western blotting. i Impact of P16INK4A knockdown on the CDK4 degradation by BSJ-02-162 in MB157 was determined based on western blotting. j Immunoprecipitation of P16INK4A from HCC1806-WT and p16 overexpressing cells (HCC1806-p16). Co-immunoprecipitated proteins are determined by western blotting. k Effect of BSJ-02-162 on the cell cycle proteins in MCF7 and HCC1806 cells following the overexpression of p16. l Structural model of BSJ-02-162 bound to a p16-CDK6 complex.

Fig. 4. Cellular response to FN-POM in a panel of cell lines.

a Effect of CDK2 KD on the proliferation of MB157 and MB436 cell lines based on live cell imaging. Error bars represent SD calculated from triplicates. Experiment was done at two independent times. (***p < 0.0001 as determined by two-Way ANOVA) (b) Chemical structure of FN-POM, which is a CDK2 PROTAC based on a CDK inhibitor, FN-1501. c Scheme indicating the synthesis of FN-POM. d Dose dependent effect of FN-POM on GSPT1, cyclin E and CDK2 at the indicated time intervals from MB157 and T47D cells. The protein expression was determined based on the band intensities from western blotting. Error bars represent mean and SD from three independent experiments. (*p < 0.05, **p < 0.01, ***p < 0.0001 as determined by 1-Way ANOVA). e Biochemical analysis from MB157 cells to determine the effect of FN-POM, Pomalidomide and CC-885 on the degradation of indicated proteins at the indicated concentrations following 16 h exposure. f Effect of pevonedistat (100 nM) and PF06873600 (1 µM) in modulating the FN-POM-mediated degradation of Cyclin E and CDK2 in MB157 cells as determined by western blotting. MB157 cells were treated with FN-POM (50 nM) in combination with pevonedistat (100 nM) for 48 h. MB157 cells were pretreated for 24 H with PF06873600 (1 µM) and then exposed to 50 nM of FN-POM up to 24 H. g Differential effect of FN-POM on the indicated proteins between MB157-WT and MB157-RB-del cells following 48 H treatment.

Fig. 5. Cellular response to FN-POM in combination with palbociclib.

a Effect of FN-POM (500 nM) on the proliferation of 1222 cells in combination with palbociclib (250 nM). Growth curve represents mean and SD from triplicates. (***p < 0.001 as determined by two-way ANOVA). b Western blot analysis in 1222 and 3226 on the indicated proteins to determine the combination effect of palbociclib and FN-POM. 1222 cells were treated with FN-POM (250 nM) in combination with palbociclib (250 nM). 3226 cells were treated with FN-POM (1 µM) in combination with palbociclib (250 nM) for 48 h. c Live cell imaging in MCF7-WT cells following the treatment with FN-POM (500 nM) in combination with palbociclib (200 nM). d MCF7 cells that harbor ectopic overexpression of Cyclin E were treated with FN-POM (500 nM) in combination with palbociclib (200 nM) for the indicated period of time. Error bars represent mean and SD from triplicates. (***p < 0.01 as determined by two-way ANOVA). e Biochemical analysis on the indicated proteins in MCF7 and T47D wild-type cells and their isogenic counterparts that harbor ectopic overexpression of Cyclin E. MCF7 and T47D WT/E1 cells were treated with palbociclib (250 nM) in combination with FN-POM (250 nM) for 48 h. f Colony formation in 1222, MCF7-WT and MCF7-E1 cells that were treated with palbociclib (500 nM) in combination with FN-POM (500 nM) for the indicated number of days. Representative images from two independent experiments were shown. g Organoids derived from 1222 cell line were treated with palbociclib (500 nM) in combination with FN-POM (500 nM). The growth of organoids was determined based on average organoid area. Error bars represent mean and SD from triplicates. Experiments were at two independent times. (***p < 0.001 as determined by two-way ANOVA). h H&E staining on tumor tissues excised from 3226 PDX treated with vehicle and Palbo/FN-POM for 5 days. Representative images were shown (scale bar = 100 micron). Tumor tissues were also stained for KI67, pRB(S807/811) and RB. i Western blot analysis from tumor tissues excised from 3226 PDX, treated with Vehicle, Palbo (100 mg/kg), FN-POM (30 mg/kg) and combination of both the drugs.