PROTACs are bifunctional molecules comprised of an E3 Ubiquitin ligase-binding moiety, chemically linked to a parent molecule that binds to a cellular target ¹. This study illustrates the utilization of two different PROTACs each targeting distinct cell cycle regulatory kinases, CDK4/6 and CDK2 for proteasomal degradation. Based on our investigation, CDK4/6-targeting PROTAC not only facilitates kinase degradation but also serves as a pharmacological tool to decipher the differential requirements of these kinases within a range of tumor models. This observation is critical to understand that distinct cell cycle states play a pivotal role in conferring resistance to pharmaceutical agents that target CDK4 and 6 kinases. Furthermore, the synthesis and development of a CDK2-PROTAC unveils a new therapeutic approach especially pertinent to models that are resistant to CDK4/6 inhibitors. Finally, this study provides preclinical evidence to support an emerging treatment strategy that aims at pharmacologically co-targeting CDK4/6 and CDK2 kinases to expand the therapeutic window of CDK4/6 inhibitors in various tumor types ².

We began this study by investigating the cellular effects of a previously developed PROTAC, BSJ-02-162 where the ubiquitin ligase-binding moiety, pomalidomide is linked to a CDK4/6 inhibitor, palbociclib ³. We used a PDAC derived cell line as a model system. The rationale behind choosing this model is its dependency on both CDK4 and 6 kinases for cell cycle progression, despite demonstrating a moderate response to the pharmaceutical inhibitor, palbociclib ⁴. Therefore, we hypothesized that the CDK4/6-degrader, BSJ-02-162 would yield a more effective cell cycle arrest as compared to palbociclib. Unexpectedly, the BSJ-02-162-mediated growth arrest was not as potent as the concurrent depletion of both CDK4 and 6 as determined by live cell imaging. The biochemical analysis uncovered that although BSJ-02-162 effectively degraded CDK6, the impact on CDK4 was limited and the residual kinase was in complex with cyclin D1, thereby retaining its catalytic activity. Earlier structural investigation revealed that palbociclib predominantly binds the monomeric CDK4, wherein the trimeric complex, consisting cyclin D1 and P27KIP1/p21CIP1 hinders the binding⁵. Notable in this study, depletion of both p27KIP1 and p21CIP1 augmented the PROTAC-mediated degradation of CDK4. Consequently, our findings suggest that the degradation of CDK4 by PROTAC could potentially serve as an indicator of palbociclib binding to its target.

In the process of evaluating the efficacy of BSJ-02-162 across a panel of different cancer cell lines, we observed a correlation between the cellular response and the specific reliance of tumor models on CDK4 for cell cycle progression. Notably, an ER+ breast cancer cell line, CAMA-1 is extremely susceptible to depletion of CDK4, resulting in a durable growth arrest and emerged as the most sensitive model to BSJ-02-162. Conversely, models of triple negative breast cancer (TNBC) displayed robust resistance to BSJ-02-162, a phenomenon attributed to their CDK4 and 6 kinase-independent cell cycle machinery due to RB loss and/or CCNE1 amplification. Importantly, our observations also unveil that resistance to BSJ-02-162 can also be an outcome of ineffective degradation of CDK4 and 6 kinases, indicating an impaired binding of palbociclib to its target. This mechanism is driven by the overexpression of p16INK4A, an endogenous CDK4/6 inhibitor. Based on biochemical and structural analysis our data provided a clear evidence that the overexpression of P16INK4A sequesters the binding pockets in CDK4 and 6 kinases and impedes the binding of palbociclib. Hence the CDK4/6-targeting PROTAC provides insights into the underlying mechanisms that drive resistance to pharmacological inhibitors.   

Our group has previously reported that high levels of endogenous CDK4/6 inhibitor, p16INK4A renders tumor models to exclusively rely on CDK2 for G1/S transition ⁶. Consistent to this RNAi-mediated deletion of CDK2 gene yielded a robust growth arrest in p16INK4A-high TNBC models. To pharmacologically mimic the CDK2 deletion, we synthesized a novel PROTAC, FN-POM by linking the pomalidomide moiety to a kinase inhibitor, FN-1501, which is known to target cyclin D1/CDK4, cyclin E/CDK2 and cyclin A/CDK2 kinase complexes ⁷. However, based on our analysis FN-POM preferentially degrades cyclin E and CDK2 over cyclin D1 and CDK4. Additionally, the impact of FN-POM on cell cycle is most pronounced in cells that are susceptible to loss of cyclin E and CDK2.

The primary objective of the preclinical investigation of FN-POM was to elucidate its structure-activity relationship. We performed multiple biochemical assays to validate its proposed pharmacological attributes. Our investigation revealed that the parent molecule FN-1501 possessed no effect on cyclin E1/CDK2 degradation in all the cell lines tested, confirming the essential role of pomalidomide moiety in instigating proteasomal degradation. While pomalidomide is known to effectively degrade various neo-substrates of cereblon such as GSPT1 and ZNF276, FN-POM’s impact on these proteins was limited ^8,9. Moreover, the potency of pomalidomide in degrading cereblon’s neo-substrates did not extend to cyclin E/CDK2. Overall these observations revealed an important conclusion that neither of the chemical moieties in PROTAC exhibited potent activity independently, while the chemically linked molecule, FN-POM possessed the enhanced degrading efficacy. We also demonstrated that the FN-POM-mediated degradation of Cyclin E/CDK2 is mediated via its binding with CDK2 based on in vitro kinase assay and cell-based competition assay.

Since, the efficacy of FN-POM was more pronounced in p16INK4A-high cell lines, we speculated that pharmacologically inhibiting CDK4/6 will mimic this phenomenon in p16INK4A deficient tumor models. To test this hypothesis, we combined palbociclib and FN-POM, which resulted in a durable cell cycle arrest observed both in PDAC organoids and in vivo PDX models. Our study concludes that limiting the CDK2 activity could be a potential approach in enhancing the therapeutic benefits of CDK4/6 inhibitors.  

References:

1            Lai, A. C. & Crews, C. M. Induced protein degradation: an emerging drug discovery paradigm. Nat Rev Drug Discov 16, 101-114, doi:10.1038/nrd.2016.211 (2017).

2            Arora, M. et al. Rapid adaptation to CDK2 inhibition exposes intrinsic cell-cycle plasticity. Cell 186, 2628-2643 e2621, doi:10.1016/j.cell.2023.05.013 (2023).

3            Jiang, B. et al. Development of Dual and Selective Degraders of Cyclin-Dependent Kinases 4 and 6. Angew Chem Int Ed Engl 58, 6321-6326, doi:10.1002/anie.201901336 (2019).

4            Kumarasamy, V., Vail, P., Nambiar, R., Witkiewicz, A. K. & Knudsen, E. S. Functional Determinants of Cell Cycle Plasticity and Sensitivity to CDK4/6 Inhibition. Cancer Res 81, 1347-1360, doi:10.1158/0008-5472.CAN-20-2275 (2021).

5            Guiley, K. Z. et al. p27 allosterically activates cyclin-dependent kinase 4 and antagonizes palbociclib inhibition. Science 366, doi:10.1126/science.aaw2106 (2019).

6            Knudsen, E. S. et al. CDK/cyclin dependencies define extreme cancer cell-cycle heterogeneity and collateral vulnerabilities. Cell Rep 38, 110448, doi:10.1016/j.celrep.2022.110448 (2022).

7            Wang, Y. et al. Discovery of 4-((7H-Pyrrolo[2,3-d]pyrimidin-4-yl)amino)-N-(4-((4-methylpiperazin-1-yl)methyl)p henyl)-1H-pyrazole-3-carboxamide (FN-1501), an FLT3- and CDK-Kinase Inhibitor with Potentially High Efficiency against Acute Myelocytic Leukemia. J Med Chem 61, 1499-1518, doi:10.1021/acs.jmedchem.7b01261 (2018).

8            Matyskiela, M. E. et al. A novel cereblon modulator recruits GSPT1 to the CRL4(CRBN) ubiquitin ligase. Nature 535, 252-257, doi:10.1038/nature18611 (2016).

9            Gao, S., Wang, S. & Song, Y. Novel immunomodulatory drugs and neo-substrates. Biomark Res 8, 2, doi:10.1186/s40364-020-0182-y (2020).