Profiling the Activity of the Potent and Highly Selective CDK2 Inhibitor BLU-222 Reveals Determinants of Response in CCNE1-Aberrant Ovarian and Endometrial Tumors

Abstract

BLU-222 is an investigational, potent, highly selective, orally bioavailable cyclin-dependent kinase 2 (CDK2) inhibitor in clinical development. BLU-222 demonstrated robust antitumor activity in select CCNE1-high ovarian and endometrial cancer models. We used a combination of CRISPR whole-genome screens coupled with targeted genetic and pharmacologic approaches in ovarian and endometrial cell lines to identify biological determinants to predict BLU-222 monotherapy activity. Rb and p16 expression were biomarkers that enriched for CDK2-dependency/BLU-222 sensitivity in CCNE1-overexpressed, nonamplified cells. Furthermore, intact Rb and low p16 expression predicted a BLU-222 and CDK4/6 inhibitor combination response. BLU-222 demonstrated robust activity in combination with carboplatin or paclitaxel in CCNE1-aberrant models, rendering chemotherapy-resistant tumors strongly sensitive to the combination. These findings demonstrate that response to CDK2 inhibition by BLU-222 can be further predicted using a combinatorial biomarker signature that could refine patient selection criteria in CCNE1-high patients and support clinical development. Significance: The identification of biomarkers of response to the CDK2-selective inhibitor BLU-222 and effective combinations with CDK4/6 inhibitors or chemotherapy could enable precision medicine strategies for CDK2 inhibition in ovarian and endometrial cancer. See related article by Dommer and colleagues, p. 1310.

Graphical abstract

Figure 1.

BLU-222 is a potent and selective CDK2i. A, Chemical structure of BLU-222. B, BLU-222 kinome tree. C, Enzymatic, NanoBRET, and cellular potency and selectivity of BLU-222. Kinome selectivity tested at 3 µmol/L by KINOMEscan; S(10) is the number of kinases inhibited at <10 percent of control divided by the total number of human wild-type kinases. Biochemical assays were conducted at 1 mmol/L ATP in the presence of appropriate cyclin (indicated). Fold selectivity is noted in brackets. D, CDK2 cellular potency measured by inhibition of pRb Thr821/826 by AlphaLISA after 18-hour dosing in OVCAR-3. pLamin S22 inhibition was used as a cellular readout of CDK1 inhibition by AlphaLISA after 2-hour dosing. E, CDK4/6 cellular potency in T47D, a CDK4/6-dependent breast cell line. In T47D, pRb Ser807/811 is controlled by CDK4/6 and therefore was used as a readout for CDK4/6 activity after 18-hour dosing.

Figure 2.

CCNE1-amplified ovarian and endometrial cancer cells are sensitive to BLU-222. A and B, Cell confluency measured by IncuCyte in BLU-222–treated CCNE1-amplified (CN ≥6, mRNA high; A) or CCNE1-normal (CN <3, mRNA normal; B) cell lines. C and D, BLU-222 antiproliferative effect measured by CyQUANT after 5-day treatment. Each dot represents the BLU-222 GI₅₀ of a single cell line. C and D, Cells were categorized by CCNE1 CN (C) or CCNE1 (D) mRNA expression derived from CCLE. The BLU-222 strong responder threshold was GI₅₀ ≤200 nmol/L. OE, overexpressed. E and F, In vivo growth kinetics of xenograft models in response to BLU-222. Mean tumor volume ± SEM is plotted for OVCAR-3 T2A (CCNE1-amplified, Rb-intact, p16-high expressor; E) in NOD/SCID female mice and ES-2 (CCNE1 normal; F) xenograft in BALB/c nude female mice. Statistical deviation from vehicle assessed by two-way ANOVA. *, P < 0.05; ***, P < 0.001. G and H, Western blots for indicated proteins in BLU-222–treated cells (24 hours): CCNE1-amplified (G) and CCNE1-normal (H) cell lines. I and J, Cell-cycle phase analysis after 24-hour BLU-222 treatment in CCNE1-amplified (I) and CCNE1-normal (J) cell lines.

Figure 3.

Rb and p16 are determinants of response to BLU-222 in CCNE1-high cells. A, A CRISPR whole-genome library deletion screen was used to identify markers of resistance to BLU-222 in OVCAR-3 cells by monitoring outgrowth in the presence of BLU-222 (200 nmol/L). Resistance screen results represented by a waterfall plot of gene dependency Z-scores. Markers with strong magnitude of effect that reached statistical significance (t test P < 0.05) are marked in blue. B and C, BLU-222 antiproliferative response was assessed by 5-day CyQUANT in OVCAR-3 sgRB1 lines (B) and sgCDKN2A lines (C). D–F, BLU-222 GI₅₀s from cell lines tested in C were categorized by CCNE1 mRNA and Rb/p16 protein expression status as measured by Western blotting. G, Western blot (WB) quantifications of cyclin E1, Rb, and p16 in endometrial PDX study vehicle tumors. Representative Western blot image in Supplementary Fig. S10. Blue font, predicted responders to single-agent BLU-222; gray font, predicted nonresponders. H–L, In vivo growth kinetics in response to BLU-222 60 mpk twice daily in CCNE1-high uterine PDX models. Rb and p16 protein expression status as indicated. Mean tumor volume ± SEM is plotted. M, TGI in endometrial PDX models categorized by cyclin E1, Rb, and p16 status based on protein expression quantification in G.

Figure 4.

Inhibition of CDK4/6 activity sensitizes CCNE1-high cells to BLU-222. A, Western blots for indicated proteins in OVCAR-3 sgCDKN2A deletion cell lines. Total Rb and pRb Ser807/811 were probed sequentially from one membrane; pRb Thr821/826 and pRb Ser 780 were probed sequentially from a second membrane. B, CDK4 coimmunoprecipitation demonstrates increased cyclin D1 association upon CDKN2A deletion in OVCAR-3. IP, immunoprecipitate; WCL, whole cell lysate. C, Western blots for indicated proteins in vector control (+NTC) or CDKN2A-overexpressing (+CDKN2A) cell lines. AN3CA and OV-90 express endogenously low levels of p16. KLE expresses endogenously high levels of p16 and is CDK2i sensitive. Total Rb and pRb Ser807/811 were probed sequentially from a single membrane. D–F, The antiproliferative effect of BLU-222 and ribociclib measured by CyQUANT in +NTC or +CDKN2A cell lines. IC₅₀s are indicated. G, Cell-cycle analysis in CDKN2A isogenic OVCAR-3 cells treated with BLU-222 (250 nmol/L), ribociclib (3 µmol/L), or both for 24 hours. H–J, Drug synergy analysis of BLU-222 and ribociclib in ovarian and uterine cell lines. CCNE1, Rb, and p16 designations determined as in Fig. 3.

Figure 5.

BLU-222 in combination with ribociclib induces strong antitumor activity in p16-low models. The BLU-222 + ribociclib combination effect was measured in vitro by 5-day CyQUANT and subsequent zero interaction potency (ZIP) synergy analysis and using in vivo CDX model tumor growth kinetics in OVK18 (A–C), a CCNE1-high, Rb-intact, and p16-low ovarian cell line; MFE-296 (D–F), a CCNE1-high, Rb-intact, and p16-low endometrial cell line; and OVCAR-3 (G–I), a CCNE1-high (CCNE1-amplified), Rb-intact, and p16-high ovarian cell line. Mean tumor volume is plotted ± SEM. Statistical deviation between groups as determined by two-way ANOVA at study endpoint. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 6.

BLU-222 in combination with chemotherapeutic agents demonstrates therapeutic benefit in ovarian and endometrial tumor models. A and B, In vivo growth kinetics in CCNE1-amplified OVCAR-3 T2A xenograft in NOD/SCID female mice dosed for 46 days (A) and ES-2 CCNE1-normal xenograft in BALB/c nude female mice dosed for 21 days (B). C–G, In vivo tumor growth kinetics in endometrial PDX models. Mean tumor volume ± SEM is plotted. Statistical deviation as determined by two-way ANOVA on latest day all the groups remained intact (ST3052, day 58; ST2526, day 52; ST1386, day 59; ST259 day 28; and ST189, day 13). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 7.

Model for BLU-222 response in a CDK2-dependent cell. A, In a CDK2-insensitive cell line, CDK4/6 is the major regulator of the G₁–S transition, with a nonessential role for CDK2. If CDK2 is inactivated in this setting, there is no effect on cell-cycle progression because CDK4/6 can compensate for loss of CDK2 activity and inactivate Rb. B, High levels of p16 render CDK4/6 inactive, and the G₁–S-phase transition becomes dependent on CDK2-mediated phosphorylation/inactivation of Rb. In this setting, cells are sensitive to pharmacologic inhibition of CDK2 by BLU-222 as a monotherapy to inhibit downstream activation of S-phase–related genes and arrest the cell cycle but are insensitive to pharmacologic inhibition of CDK4/6. Created in BioRender.com. House, N. (2025) https://BioRender.com/u22y649.