Discovery of an Aldo-Keto reductase 1C3 (AKR1C3) degrader

Abstract

Aldo-keto reductase 1C3 (AKR1C3) is a protein upregulated in prostate cancer, hematological malignancies, and other cancers where it contributes to proliferation and chemotherapeutic resistance. Androgen receptor splice variant 7 (ARv7) is the most common mutation of the AR receptor that confers resistance to clinical androgen receptor signalling inhibitors in castration-resistant prostate cancer. AKR1C3 interacts with ARv7 promoting stabilization. Herein we report the discovery of the first-in-class AKR1C3 Proteolysis-Targeting Chimera (PROTAC) degrader. This first-generation degrader potently reduced AKR1C3 expression in 22Rv1 prostate cancer cells with a half-maximal degradation concentration (DC₅₀) of 52 nM. Gratifyingly, concomitant degradation of ARv7 was observed with a DC₅₀ = 70 nM, along with degradation of the AKR1C3 isoforms AKR1C1 and AKR1C2 to a lesser extent. This compound represents a highly useful chemical tool and a promising strategy for prostate cancer intervention.

Fig. 1. Structures of selected AKR1C3 inhibitors, PROTAC and Lenalidomide.

Chemical structures of representative small molecule AKR1C3 inhibitors (1–3), PROTAC warhead 4, the assembled PROTAC 5 with warhead (Blue), triazole-PEG2 linker, and E3-ligase lenalidomide (green, 6).

Fig. 2. Degradation of AKR1C3 in prostate cancer cells by small molecule inhibitors.

Representative Western blots of AKR1C3, AKR1C1/C2 and ARv7 protein expression in 22Rv1 prostate cancer cells treated with (a) AKR1C3 inhibitor 3 at 43 nM; (b) AKR1C3 inhibitor 3 at 1 µM; (c) AKR1C3 inhibitor warhead 4 at 1 µM and (d) DMSO; for 0, 24, 48 and 72 h. Images representative of at least two technical replicates performed in duplicate. Quantification in Supplementary Fig. S1.

Fig. 3. Docking predictions of PROTAC 5 binding to AKR1C3.

The PROTAC warhead (gold) is predicted to bind into the same SP1 pocket of AKR1C3 (PDB ID: 3UG8) as the known inhibitor indomethacin (green, overlay). A PEG2 linker with a triazole anchor point predicts sufficient length to engender solvent exposure of the E3 ligase ligand.

Fig. 4. Synthesis of AKR1C3 degrader warhead 14.

Reagents and conditions: (a) tert-butylchlorodiphenylsilane (TBDPSiCl), imidazole, tetrahydrofuran (THF), rt; (b) 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC·HCl), 1-hydroxybenzotriazaole hydrate (HOBt hydrate), N,N-diisopropylethylamine (DIPEA), dichloromethane (DCM), 0 °C – rt; (c) methyl acrylate, Pd(OAc)₂, P(Ph)₃, triethylamine (NEt₃), toluene, 110 °C; (d) phenyl boronic acid, Pd(dppf)Cl₂⋅CH₂Cl₂, Cs₂CO₃, toluene, 110 °C; (e) tetrabutylammonium fluoride (TBAF), THF, 0 °C, 40 min; (f) propargyl bromide, Cs₂CO₃, dimethylformamide (DMF), 60 °C.

Fig. 5. Synthesis of E3 ligase ligand lenalidomide and linker 18.

Reagents and conditions: (a) tert-butyl bromoacetate, sodium hydride (NaH), THF, 0 °C (30 min) to rt; (b) trifluoroacetic acid (TFA), DCM, rt; (c) SOCl₂, DCM, rt; (d) lenalidomide, N-methyl-2-pyrrolidone (NMP), rt.

Fig. 6. Synthesis of PROTAC degrader 5.

Reagents and conditions: (a) sodium ascorbate, copper sulfate pentahydrate (CuSO₄⋅5H₂O), CH₂Cl₂:MeOH:H₂O, rt; (b) 1 N NaOH, MeOH:THF, reflux.

Fig. 7. Degradation of AKR1C3, AKR1C1/C2 and ARv7 by PROTAC 5.

a Time study of degradation of AKR1C3, AKR1C1/C2, and ARv7 upon treatment of 5 (10 nM) at different time points (0, 2, 4, 6, 12, 16, 24, 48, and 72 h). Blots are representative of two separate experiments; (b) combined quantification of a and replicates for AKR1C3 and AKR1C1/C2 expression. Data obtained from two technical replicates performed in duplicate is represented as the mean ± standard deviation. *p < 0.05 by two-tailed, unpaired Mann–Whitney test; (c) Quantification of a for AKR1C3 expression; (d) quantification of a for AKR1C1/C2 expression; (e) quantification of a for ARv7 expression. Relative expression is normalized to actin.

Fig. 8. Mechanistic studies of PROTAC 5 degradation effect in 22Rv1 prostate cancer cells.

a Quantification (n = 3) and (b) Selected Western blot image of AKR1C3 expression after 72 h treatment of degrader 5 versus small molecule AKR1C3 inhibitors 3 and 4 at 10 nM concentration; (c) Quantification (n = 12 DMSO, n = 4 MG132, n = 12 PROTAC, n = 3 MG132 & PROTAC) and (d) selected Western blot image of effect on protein degradation with 2 h pretreatment of DMSO or the proteasome inhibitor MG132 (3 µM). Cells were then treated with PROTAC 5 at 10 nM for 4; (e) Quantification (n = 14 DMSO, n = 7 Lenalidomide, n = 14 PROTAC, n = 7 Lenalidomide & PROTAC) and (f) selected Western blot image of effect on protein degradation with 2 h pretreatment of DMSO or lenalidomide (3 µM). Cells were then treated with PROTAC 5 at 10 nM for 72 h. Data obtained from at least three replicates is represented as the mean ± SEM. ns, not significant; **p < 0.01; ***p < 0.001; ****p < 0.0001 by unpaired two-tailed t test.

Fig. 9. Activity of PROTAC 5 and Warhead 4 on cell viability of prostate cancer cells displaying differential expression of AKR1C3.

a Effect of warhead compound 4 at indicated concentrations on cell viability of 22Rv1 cells (n = 9); (b) Effect of PROTAC 5 at indicated concentrations on cell viability of 22Rv1 cells alone and in combination with enzalutamide (ENZ), (n = 20 except ENZ 25 µM, n = 47); (c) Effect of PROTAC 5 at indicated concentrations on cell viability of 22Rv1 cells grown in CSS media (high AKR1C3 expression) alone and in combination with ENZ (n = 9); (d) Effect of PROTAC 5 and ENZ at indicated concentrations on cell viability of LNCaP cells (AKR1C3 null), (n = 18 ENZ 50 µM, 25 µM and control, n = 9 each concentration); (e) Effect of PROTAC 5 and ENZ at indicated concentrations on cell viability of LNCaP1C3 cells (stably transfected with AKR1C3), (n = 18 ENZ 50 µM, 25 µM and control), n = 9 each concentration. Data is represented as the mean ± SEM of at least three experiments ran in triplicate. *p < 0.05, **p < 0.01; ***p > 0.001; ****p < 0.0001 by unpaired two-tailed t test.