A modular PROTAC design for target destruction using a degradation signal based on a single amino acid

Abstract

Proteolysis targeting chimeras (PROTACs) are bivalent molecules that bring a cellular protein to a ubiquitin ligase E3 for ubiquitination and subsequent degradation. Although PROTAC has emerged as a promising therapeutic means for cancers as it rewires the ubiquitin pathway to destroy key cancer regulators, the degradation signals/pathways for PROTACs remain underdeveloped. Here we append single amino acids, the simplest degradation signal, to a ligand specific for estrogen-related receptor α (ERRα) and demonstrate their utility in ERRα knockdown via the N-end rule pathway and also their efficiency in the growth inhibition of breast cancer cells. The modular design described offers unique advantages including smaller molecular size with shortest degradation sequences and degradation speed modulation with different amino acids. Our study expands the repertoire of limited ubiquitin pathways currently available for PROTACs and could be easily adapted for broad use in targeted protein degradation.

Introduction

Proteolysis targeting chimeras (PROTAC)² has emerged as an effective means for cancer therapy by shuttling cancer regulators to the proteasome for degradation (1–4). The principle of PROTAC-mediated protein knockdown strategy is based on ubiquitin-mediated proteolysis; a ubiquitin ligase (E3) recognizes its target and covalently attaches ubiquitin, a conserved 76-residue protein, onto the substrate, leading to subsequent degradation by the proteasome (1, 5, 6). PROTACs contain two moieties, which separately bind a relevant drug target and a ubiquitin ligase, with a flexible linker in the middle (1, 4). Thereby PROTACs rewire the ubiquitin-proteasome pathway to rapidly eliminate, rather than inhibit, specific target proteins on contact, which could minimize the risk that acquired defects develop over time and also phenotypes are reversed, common shortcomings of targeted therapies. Furthermore, as PROTACs require only transient drug-target binding without the need of direct inhibition of substrate activity, this approach could be easily used to manipulate previously undruggable cancer targets via selective degradation.

Of the two main components of PROTACs, whereas the ligands for various targets have been developed and refined over the years, the degradation motifs adapted for PROTACs remain underdeveloped (4, 7). Thus far, the degradation moieties specific for a few ubiquitin ligases (e.g. VHL, cereblon, MDM2, and XIAP) have been validated for PROTAC-mediated destruction of a variety of key cancer proteins (7, 8). However, one major caveat with these degradation signals employed is the lack of degradation speed control, which can be a problem as too fast or slow substrate turnover may be harmful in some cases. It is also important to reduce the sizes of PROTACs that currently are much larger than typical drugs (often less than 500 Da), which could improve cellular permeability and potency. Moreover, the ubiquitin ligases Cereblon and VHL, which are most frequently employed in PROTACs currently, may not be easily adopted for cancers in kidney, lung, and brain, due to their low expression/activity in these cells (9–11). Therefore, how to expand the repertoire of ubiquitin ligases and make this strategy more generally applicable remains a critical, pressing issue.

Results and discussion

We reason that a short binding motif to a potent ubiquitin ligase may greatly improve the utility of the PROTAC approach. The shortest degradation motif is a single amino acid, a destabilizing residue (e.g. Arg, His, Lys, Leu, and Ile) at the N terminus of a protein that targets the substrate to a ubiquitin ligase termed Ubr1 for rapid degradation via the N-end rule pathway, the first ubiquitin-dependent degradation pathway identified (12–14). We synthesized adaptor molecules starting with Arg or His and fused with a flexible linker (15) to a ligand for estrogen-related receptor α (ERRα) (3) (Fig. 1A), a nuclear receptor that is a major regulator of several critical metabolic pathways and also is a prognostic marker of breast cancer (16). ERRα inhibition has been shown to reduce the proliferation of breast tumor cells in vitro and in vivo. We found that both compounds Arg-T^ERRa and His-T^ERRa reduced the endogenous ERRα level in the MCF-7 metastatic breast cancer cell line (Fig. 1B). To assess whether the ERRα decrease was due to protein turnover, we treated the cells with cycloheximide to block protein synthesis and collected the samples at various time points for Western blot analysis. Both Arg-T^ERRa and His-T^ERRa triggered ERRα degradation (Fig. 1C and Fig. S1).

Figure 1.

Single amino acid-based PROTACs induce ERRα turnover. A, structure of PROTACs Arg-T^ERRa and His-T^ERRa. Arg and His were linked to the ligand previously demonstrated specifically for ERRα (3). B, Arg-T^ERRa and His-T^ERRa trigger ERRα reduction in a dose-dependent manner. Varying amounts of Arg-T^ERRa or His-T^ERRa were added to MCF-7 cells for 48 h. ERRα levels were determined by Western blot analysis. Equal amounts of protein extracts were used and ascertained by blotting with GAPDH antibody in the experiments. All experiments were done at least 3 times. C, ERRα stability was examined in the absence or presence of PROTACs by a protein expression shut-off assay. MCF-7 cells with or without PROTACs (10 μm Arg-T^ERRa and 5 μm His-T^ERRa) were treated with 200 μg/ml of cycloheximide to shut off protein synthesis. Samples were then collected at the indicated time points. Extracts were processed for immunoblotting with ERRα antibody. GAPDH serves as a loading control (lower panels). The experiments were done more than three times and quantified as shown in Fig. S1. A representative image is shown here.

We then evaluated whether PROTAC-induced ERRα degradation is mediated by the proteasome. We found that Arg-T^ERRa– or His-T^ERRa–triggered ERRα turnover was compromised upon the treatment of the proteasome inhibitor MG132 (Fig. 2, A and B, and Fig. S2), suggesting the involvement of the proteasome in ERRα turnover. To ascertain that ERRα is degraded by the N-end rule pathway, we took advantage that dipeptides starting with a destabilizing residue (e.g. Arg, His) can block the degradation of N-end rule substrates through their competition for the ubiquitin ligase Ubr1 (13, 14). We found that the PROTAC-exposed ERRα level was increased upon the addition of the dipeptide Arg-Ala or His-Ala, but not the control peptide Ala-Ala bearing a stabilizing residue (Fig. 2, C and D, and Fig. S3).

Figure 2.

ERRα turnover is blocked by the inhibitors of the proteasome and the N-end rule pathway. A and B, ERRα degradation involves the proteasome. MCF-7 cells were incubated with 10 μm Arg-T^ERRa or 5 μm His-T^ERRa for 32 h and then mixed with DMSO control or the proteasome inhibitor MG132 (25 μm). ERRα stability was determined after the addition of cycloheximide as described in the legend to Fig. 1C. C and D, Arg-T^ERRa- or His-T^ERRa-induced ERRα reduction was reversed by dipeptides Arg-Ala and His-Ala, respectively. MCF-7 cells incubated with 10 μm Arg-T^ERRa (C) or 5 μm His-T^ERRa (D) were treated with 100 μm dipeptides Arg-Ala (RA), His-Ala (HA), or control Ala-Ala (AA). Cell extracts were subjected to Western blot analysis to determine the levels of ERRα and control GAPDH.

We next examined the biological effects of the PROTAC molecules in MCF-7 cells. Both compounds led to decreased proliferation of MCF-7 cells (Fig. 3A). As cell migration is key to tumorigenesis, we then evaluated whether the compounds affect cell motility by the scratch wound healing method and the ThinCert cell migration assay. We found both Arg-T^ERRa and His-T^ERRa reduced MCF-7 cancer cell migration and wound repair (Fig. 3, B and C, and Fig. S4). Furthermore, we monitored the expression of several markers of epithelial-mesenchymal transition (EMT) that is crucial for cancer cell migration and invasion. PROTACs led to increased expression of E-cadherin and reduced expression of N-cadherin, SNAIL, and fibronectin (Fig. 3D), indicating EMT is repressed upon ERRα depletion. Combined, these results suggest that Arg-T^ERRa and His-T^ERRa could effectively modulate ERRα and be explored for anticancer purpose.

Figure 3.

Biological effects of Arg-T^ERRa and His-T^ERRa in MCF-7 cells. A, PROTACs Arg-T^ERRa and His-T^ERRa inhibit breast cancer cell proliferation. MCF-7 cells were treated with DMSO, 10 μm Arg-T^ERRa, or 5 μm His-T^ERRa. Cell proliferation was determined by the MTT assay after drug exposure on the indicated days. B, Arg-T^ERRa and His-T^ERRa impede wound healing. Confluent monolayers of MCF-7 cells were scraped by a pipette tip to generate wounds and then treated with DMSO, 10 μm Arg-T^ERRa, or 5 μm His-T^ERRa. Wound repair was photographed at the indicated time points. Scale bar, 100 μm. C, Arg-T^ERRa and His-T^ERRa reduce breast cancer cell migration. MCF-7 cells were allowed to invade Transwell chambers for 48 h in the presence or absence of PROTACs. Invaded cells were then fixed, stained, and photographed. Scale bar, 20 μm. D, effects of Arg-T^ERRa and His-T^ERRa on EMT markers. After the exposure to 10 μm Arg-T^ERRa or 5 μm His-T^ERRa, cell extracts were analyzed by Western blotting to examine the expression of various EMT markers including fibronectin, N-cadherin, SNAIL, and E-cadherin.

Our study demonstrates that the N-end rule-based PROTACs efficiently trigger target destruction and inhibit the proliferation of breast cancer cells. The advantage of our approach lies in the use of the simplest degradation signal-single amino acid, which can be any one of the 13 destabilizing residues in the N-end rule pathway (13, 14) that in turn may lead to different degradation speeds. The PROTAC method presented here overcomes limitations of existing approaches with significantly smaller molecular size of E3 targeting moiety, likely leading to higher permeability and better efficacy, and a modulatable potent degradation activity, allowing better control and broader application. Furthermore, the N-end rule degradation pathway is universally present and constitutively active, making it ideally suited for the PROTAC approach. This modular design expands the repertoire of limited ubiquitin pathways currently available for PROTACs and could be easily adapted for broad use in targeted protein degradation.

Experimental procedures

Cell cultures

The MCF-7 cell line (ATCC, Manassas, VA) was maintained at 37 °C with 5% CO₂ in Dulbecco's modified Eagle's medium with 2 mm l-glutamine that was modified to contain 1 mm sodium pyruvate, 0.1 mm nonessential amino acids, and 1.5 g/liter of sodium bicarbonate. It was also supplemented with 10% fetal bovine serum, 10 μg/ml of bovine insulin, and 100 units/μl of penicillin/streptomycin. All drug treatment studies were supplemented with 2% FBS. Functional assays, including MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide), migration assays were done in 10% FBS in Dulbecco's modified Eagle's medium.

Peptides and compounds

Details of Arg-T^ERRa and His-T^ERRa syntheses are described under supporting information. Dipeptides Arg-Ala and His-Ala were purchased from Bachem Americas (Torrance, CA). MG132 is a proteasome inhibitor obtained from Calbiochem (Gibbstown, NJ). Cycloheximide and dipeptide Ala-Ala were purchased from Sigma.

Cell proliferation assay

The MTT method was used to measure living cells through mitochondrial dehydrogenase activity (Sigma). Cells were plated in a 96-well plate, 5000 cells/100 μl/well. After 24 h, the cells were treated with PROTACs in fresh medium. At the indicated time point, the medium was removed and DMSO was added as MTT solubilization solution, followed by 100 μl of stop solution. Absorbance was measured at 550 nm.

Western blot analysis

Cells were lysed in RIPA buffer (Sigma) with the addition of protease inhibitors tablet and phosphatase inhibitors mixture (Sigma). Lysates were resolved by SDS-polyacrylamide gels, transferred onto nitrocellulose membranes, and then probed with antibodies as indicated. Antibodies against ERRα, E-Cadherin, N-Cadherin, Snail, and GAPDH were obtained from Cell Signaling (Danvers, MA). Fibronectin antibody was obtained from Sigma.

Wound-healing assay

Cells were allowed to grow to near confluence in 6-well dishes. A uniform scratch was then made down the center of the plate using a 200-μl micropipette tip, followed by washing twice with PBS. The same marked field of the scratch wound was photographed using an Olympus light microscope (×4 objective) at the indicated time points. The width of the scratch wound was measured at three different areas with ImageJ software.

Migration assay

MCF-7 cells (1 × 10⁴) were seeded on an 8 μm pore size Thin-cert for 24-well plates (Greiner Bio-One) in serum-free media. PROTAC compounds were added to the bottom chamber as chemoattractant. After 48 h, cells on the top of the membrane were removed with a cotton swab. The migrated cells at the bottom side were washed with PBS, fixed with 70% ethanol, and stained using 0.1% crystal violet to visualize the migrated cells. Migrated cells attached to the lower side of the membrane were enumerated using a light microscope at ×10 magnification.

Author contributions

K. S., P. S., H. C., B. C., S. F. M., and T. L. data curation; K. S. and H. R. formal analysis; K. S. validation; K. S., P. S., H. C., B. C., S. F. M., and T. L. investigation; K. S., P. S., H. C., B. C., S. F. M., T. L., and H. R. methodology; K. S., P. S., H. C., and T. L. writing-original draft; P. S., H. C., B. C., S. F. M., T. L., and H. R. writing-review and editing; S. F. M., T. L., and H. R. supervision; S. F. M. and H. R. project administration; T. L. and H. R. funding acquisition; H. R. conceptualization.