18F-Labeled PET Probe Targeting Enhancer of Zeste Homologue 2 (EZH2) for Cancer Imaging

Abstract

The enzyme enhancer of zeste homologue 2 (EZH2) plays a catalytic role in histone methylation (H3K27me3), one of the epigenetic modifications that is dysregulated in cancer. The development of a positron emission tomography (PET) imaging agent targeting EZH2 has the potential to provide a method of stratifying patients for epigenetic therapies. In this study, we designed and synthesized a series of fluoroethyl analogs based upon the structure of EZH2 inhibitors UNC1999 and EPZ6438. Among the candidate compounds, 20b exhibited a high binding affinity to EZH2 (IC₅₀ = 6 nM) with selectivity versus EZH1 (IC₅₀ = 200 nM) by SAM competition assay, and furthermore, EZH2 inhibition was demonstrated in the pancreatic cancer cell line PANC-1 (IC₅₀ = 9.8 nM). [¹⁸F]20b was synthesized successfully and showed 5-fold higher uptake in PANC-1 cells than in MCF-7 cells. MicroPET imaging in a PANC-1 cell xenograft mouse model indicates that [¹⁸F]20b has specific binding to EZH2, which was identified by ex vivo Western blot analysis of the tumor tissue.

Figure 1

Structures of previous published EZH2 inhibitors.

Figure 2

Structures of EZH2 inhibitor analogs designed and synthesized in this study: (A) UNC1999 series and (B) EPZ6438 series.

Figure 3

Effect of UNC1999 and EPZ6438 compounds on EZH2 activity in PANC-1 cells. Representative Western blot analysis for H3K27me3 or H3K4me3 on histone extracts from PANC-1 cells treated with UNC1999, EPZ6438, or modified compounds for 3 days. All blots were stripped and reprobed to assess the loading control, histone 3 (H3). Cell culture experiments were performed twice (n = 2) for H3K27me3 and once (n = 1) for H3K4me3. Bars represent mean ± standard deviation. No statistical analysis was performed on this data.

Figure 4

Increasing concentrations of EPZ6438, 20b, 21b, 22a, and 22b on H3K27me3 in PANC-1 cells. (A) Representative Western blot analysis for H3K27me3 on histone extracts from PANC-1 cells treated with increasing amounts of 21b, 22a, and 22b for 3 days. All blots were stripped and reprobed to assess the loading control, histone 3 (H3). Inhibition curves for (B) parental EPZ6438, and compounds (C) 21b, (D) 22b, and (E) control 22a. The % H3K27me3 represents the amount of H3K27me3:H3 relative to each blot’s DMSO control. IC₅₀ values, if applicable, are listed on their corresponding curve. There was no significant difference (p = 0.5795) between EPZ6438 and 21a’s abilities to inhibit EZH2 activity. Similar Western blot analysis (F) and inhibition curves (G) for lead compound 20b.

Figure 5

(A) Reagents and conditions: a. K_2.2.2, K₂CO₃, ACN, 115 °C, 10 min; b. K₂CO₃, ACN, 130 °C (MW), 10 min. (B) Cell uptake [¹⁸F]20b (1 MBq) in 100 μL of PBS buffer, PANC-1, or MCF-7 cells (600,000 each vial) in 1 mL of incubation medium, n = 5, and EPZ6438 (100 μg) in 10 μL of DMSO. (C,D) Correlating EZH2 activity to PET signal intensity for mice with PANC-1 tumor. microPET image at 100–120 min in NOD/SCID mice (n = 2) with PANC-1 cells tumor xenografts were administered a 10 MBq IV dose of [¹⁸F]20b and blocked by EPZ6438. (E,F) EZH2 expression in tumor tissue samples from PANC-1 tumors were analyzed by Western blots.