Applications and Limitations of Oxime-Linked "Split PROTACs"

Abstract

Proteolysis targeting chimeras are of keen interest as probe molecules and drug leads. Their activity is highly sensitive to the length and nature of the linker connecting the E3 Ubiquitin Ligase (E3 Ubl) and target protein (TP) ligands, which therefore requires tedious optimization. The creation of "split PROTACs" from E3 Ubl and TP ligands modified with residues suitable for them to couple when simply mixed together would allow various combinations to be assessed in a combinatorial fashion, thus greatly easing the workload relative to a one-by-one synthesis of many different PROTACs (proteolysis targeting chimeras). We explore oxime chemistry here for this purpose. We show that PROTAC assembly occurs efficiently when the components are mixed at a high concentration, then added to cells. However, in situ coupling of the TP and E3 Ubl ligands is inefficient when these units are added to cells at lower concentrations.

Keywords: BRD4; VHL; bioconjugation; oxime; proteolysis targeting chimeras.

Figure 1.

Schematic of the split PROTAC concept.

Figure 2.

(A) General scheme for generation and screening of VHL degraders using oxime ligation. (B) Structure of an active VHL degrader and it’s assembly from component parts. (C) Chromatogram showing the coupling of the aldehyde and alkoxyamine components of the split PROTAC. WJ099 (retention time 7.080), WJ634 (retention time 7.450), unpurified WJ638 (retention time 8.528) and purified WJ638 (retention time 8.531). Unpurified WJ638 is defined as the product of incubation of 10 mM WJ634 + 10 mM WJ099 in DMSO for 12 h at room temperature. This solution is then used without purification. The detail of LC method is described in SI. (D) SDS-PAGE and Western blot analysis of the level of pVHL30 in HeLa cells treated with the indicated reagents for 12 hours. (E) Determination of the DC₅₀ of unpurified WJ638 towards pVHL30. (F) SDS-PAGE and Western blot analysis of VHL in cells treated with different concentrations of purified WJ638 for 12 h. The original blot was showed in the SI. (G) DC₅₀ of purified WJ638. The amount of pVHL30 (the strongest band) present normalized to the vehicle control (DMSO) is indicated as “%”. All of the experiments were performed in triplicate. The Western blots shown are typical of all of the results obtained.

Figure 3.

Assessment of the stability of the oxime linkage in WJ638. (A) Format of the competition assay employed to measure oxime stability in vitro. WJ638 (50 μM) was incubated with 5 mM methoxyamine in DPBS buffer at 37°C for 0–48 h. The retention times of the products are indicated. (B) Chromatographic analysis of the breakdown of WJ638 over time in vitro using the assay shown in A. Peaks with retention time 7.448, 8.105 and 8.564 corresponds to Frag1, Frag2 and WJ638 respectively. (C) SDS-PAGE and Western blot analysis of the level of pVHL30 in the presence of purified WJ638 at the times indicated (hours). DMSO and CM11 were used as vehicle and positive controls, respectively. The amount of pVHL30 (the strongest band) present normalized to the vehicle control (DMSO) is indicated as “%”. All of the experiments were performed in triplicate. The Western blot shown is typical of all of the results obtained.

Figure 4.

(A) Conditions employed for the in situ assembly of WJ638 (B) Chromatographic analysis of the formation of WJ638 in vitro. WJ634 (retention time 7.455), WJ099 (7.464 retention time). (C) SDS-PAGE and Western blot analysis of the level of pVHL30 (strongest band) in cells treated with the indicated reagents. In all cases, the WJ634 and WJ099 were added to cells at the concentrations indicated, not pre-incubated at a higher concentration. (D) SDS-PAGE and Western blot analysis of the effect of VHL032 (a monomeric VHL ligand that cannot couple to form a split PROTAC) on the WJ638-mediated degradation of pVHL30. The cells were incubated with the indicated reagents for 12 hours prior to analysis. The amount of pVHL30 (the strongest band) present normalized to the vehicle control (DMSO) is indicated as “%”. All of the experiments were performed in triplicate. The Western blots shown are typical of all of the results obtained.

Figure 5.

(A) General scheme to synthesize BRD4-targeted split PROTAC candidates. (B) Structures of an active BRD4-targeted split PROTAC (WJ051) and its component pieces. (C) Chromatographic analysis of the efficiency of formation of WJ051 in vitro. Blue: WJ100. Black: WJ015, Red: Crude product of mixing WJ100 and WJ015 (10 mM each) in DMSO for 12 hr. at room temperature. Purple: purified WJ051. (D) SDS-PAGE and Western blot analysis of BRD4 levels in cells treated with HPLC-purified WJ051 at the concentration indicated for 12 hours. (E) SDS-PAGE and Western blot analysis of BRD4 levels in cells treated with unpurified WJ051 at the concentration indicated for 12 hours. Vinculin was used as loading control. The BRD4 level (a total of the two isoforms band in the western blot) was normalized to DMSO control is indicated as “%”. All of the experiments were performed in triplicate. The Western blots shown are typical of all of the results obtained. The original blots were showed in the SI.

Figure 6.

SDS-PAGE and Western blot analysis of BRD4 levels in cells treated with the indicated reagents for 12 hours. (A) Effect of the individual components of WJ051 on the degradation of BRD4 by purified WJ051. (B) Effect of the component pieces WJ015 and WJ100 on BRD4 levels in the absence of WJ051, and the effect of adding both pieces together to cells incubated with purified WJ051. Vinculin was used loading control. The BRD4 level (a total of the two isoforms band in the western blot) was normalized to DMSO control is indicated as “%”. The experiments were performed in duplicate. THE Western blot shown is representative of both results.