Noncovalent CDK12/13 dual inhibitors-based PROTACs degrade CDK12-Cyclin K complex and induce synthetic lethality with PARP inhibitor

Highlights

• A novel type of CDK12 degraders based on noncovalent CDK12/13 dual inhibitors.

• PP-C8 induced potent and selective CDK12-CycK degradation without affecting CDK13.

• PP-C8 suppressed DDR-associated genes and induced synthetic lethality with Olaparib.

Abstract

Cyclin-dependent kinase 12 (CDK12) plays a crucial role in DNA-damage response gene transcription and has recently been validated as a promising target in cancer therapy. However, existing CDK12 inhibitors potently inhibit its closest isoform CDK13, which could cause potential toxicity. Therefore, the development of CDK12 inhibitors with isoform-selectivity against CDK13 continues to be a challenge. By taking advantage of the emerging PROteolysis-TArgeting Chimeras (PROTACs) approach, we have synthesized a potent PROTAC degrader PP-C8 based on the noncovalent dual inhibitors of CDK12/13 and demonstrated its specificity for CDK12 over CDK13. Notably, PP-C8 induces profound degradation of cyclin K simultaneously and downregulates the mRNA level of DNA-damage response genes. Global proteomics profiling revealed PP-C8 is highly selective toward CDK12-cyclin K complex. Importantly, PP-C8 demonstrates profound synergistic antiproliferative effects with PARP inhibitor in triple-negative breast cancer (TNBC). The potent and selective CDK12 PROTAC degrader developed in this study could potentially be used to treat CDK12-dependent cancers as combination therapy.

Introduction

Cyclin-dependent kinases (CDKs) are central enzymes that control cell cycle progression and regulate gene transcription in eukaryotic cells [1,2]. Cyclin-dependent kinase 12 (CDK12) is one of the transcription-associated CDKs, and its activity is tightly regulated by interactions with the partner protein cyclin K (CycK) [3,4]. By forming complex with CycK, CDK12 phosphorylates Ser2 in a C-terminal domain (CTD) of RNA polymerase II (RNAPII) and primarily regulates transcription of DNA-damage response (DDR) genes, thereby maintains cell genomic stability [5]. CDK12 and its closest homolog CDK13 share an substantially indistinguishable central kinase domain and participate in overlapping cellular processes, such as pre-mRNA splicing and processing [6]. Despite this, CDK12 and CDK13 regulate the expression of distinct downstream genes and control several different biological processes. CDK12 predominantly regulates the expression of genes involved in DNA damage response (DDR) pathways [7,8]. On the other hand, CDK13 makes less contribution to this set of genes but is crucial in regulating genes related to protein translation and also serves important biological functions in normal cells [9]. Recently, CDK12 inhibition with small molecules was reported to induce synthetic lethality in cancer cells, and extensive studies have verified CDK12 as a compelling cancer therapeutic target [[9], [10], [11], [12], [13], [14], [15], [16], [17]]. To regulate CDK12 specific biological functions and minimize the potential off-target toxicity, small molecules that selectively target CDK12 without interfering with CDK13 are highly desired. To date, only a handful of small-molecule inhibitors have been discovered to target CDK12 in cells, including the irreversible covalent inhibitor THZ531 as well as the noncovalent inhibitor SR-4835. However, all of them potently inhibit CDK13 due to the highly homologous kinase domain between CDK12 and CDK13 [[18], [19], [20], [21], [22]]. Moreover, the CDK12/13 dual inhibitors failed in clinical trials because of unexpected toxicity [9]. Therefore, developing isoform-selective CDK12 inhibitors for the treatment of cancer is attractive but challenging.

An alternative approach to block CDK12 activity is directly reducing its cellular levels via targeted protein degradation (TPD) [23]. PROteolysis-TArgeting Chimera (PROTAC) is an emerging revolutionary TPD modality. By linking target protein to a Cullin-RING ligase, in theory the PROTAC degrader could induce any protein of interest for proteasomal degradation (Fig. 1A). The last two decades witnessed the PROTAC mediated degradation of a number of human disease-associated proteins, including those undruggable transcription factors and traditionally hard-to-target protein complexes [24,25]. Due to the cooperative interactions between the target protein and E3 ligase in a transient PROTAC-induced ternary complex, the PROTAC-mediated protein degradation may display enhanced target selectivity compared with protein inhibition [[26], [27], [28], [29], [30]]. Moreover, unlike the conventional small molecule inhibitors primarily targeting kinase catalytic domain, PROTAC degraders destruct the entire protein and abolish both kinase-dependent and kinase-independent functions [31]. Therefore, we envisioned that PROTAC degraders based on CDK12/13 dual inhibitors such as 1 and SR-4835 might acquire exceptional selectivity against either isoform than its parental inhibitors (Fig. 1B). Meanwhile, PROTAC-mediated degradation could remove the entire protein other than inhibit its kinase functions alone. In this perspective, a PROTAC degrader targeting CDK12 could be employed in studying the potentially non-canonical cellular functions of CDK12 and its activating partner CycK as a complex [31]. Recently, Jiang et al. disclosed the first isoform selective CDK12 PROTAC degrader BSJ-4-116 based on a covalent CDK12 inhibitor THZ531 (Fig. 1B) [30]. The rationally designed PROTAC BSJ-4-116 demonstrates potent and durable CDK12 degradation without affecting the CDK13 and CycK protein levels. Furthermore, BSJ-4-116 overcomes the inhibitor-induced drug resistance of CDK12^C1039F mutation. However, long-term high-dose exposure with BSJ-4-116 in tumor cells leads to PROTAC resistance through CDK12 point-mutations. BSJ-4-116 highlights the potential advantages of degradation over inhibition in selectively degrading CDK12, yet a new class of CDK12 degraders with diverse target protein ligands is still valuable in light of the PROTAC-induced resistance and for treating a broader scope of cancers. Furthermore, CDK12 participates in multiple cellular processes after forming a complex with CycK, and the CDK12-CycK complex has been recognized as a druggable target. Hence, we hypothesized that degrading CDK12-CycK complex other than CDK12 alone would presumably provide more profound biological effects and find broader use in treating cancers [32]. A group of promiscuous CDK inhibitors were found to induce degradation of CycK in the “molecular glue” mode-of-action, in which CDK12 displaced the DCAF in E3 ligase complex and worked as the “substrate receptor” [[32], [33], [34], [35]]. For instance, a recent study identified a “molecular glue” molecule NCT02 that triggers significant degradation of cellular CycK as well as its binding partners, CDK12 and CDK13. Genetic depletion identified degrading CDK12 rather than CDK13 was the druggable vulnerability of a subgroup of colorectal cancers [32]. Considering the molecular glues induced CycK degradation accompanied by dual depletion of CDK12/13, we aim to develop PROTAC degraders that selectively target the CDK12-CycK complex while reducing the potential toxicity caused by CDK13 suppression.

Here, we report the development of a new class of PROTAC degraders targeting the CDK12-CycK complex, based on the noncovalent CDK12/13 dual inhibitors 1 and SR-4835. Biological evaluations identified PP-C8 induced rapid and highly selective CDK12-CycK degradation in two cancer cell lines without detectable CDK13 degradation. Of note, in contrast to the proceeding BSJ-4-116 which only leads to CDK12 degradation, PP-C8 degrades CycK simultaneously [30]. Compared with the molecular glues targeting CycK, PP-C8 maintains comparable potency in suppressing CDK12 specific downstream signaling and DDR-related genes expression, yet with improved degradation selectivity against CDK13. Importantly, PP-C8 demonstrates profound synthetic lethal effects with PARP inhibition in triple-negative breast cancer cell (TNBC) growth. We envision that the PROTAC degraders developed in this study would be valuable tools in uncovering the potential kinase-independent functions of CDK12-CycK complex. With further optimization, PP-C8 could also be a promising therapeutic candidate for treating CDK12 dependent cancers as a synthetic lethal partner.