Characterization of TR-107, a novel chemical activator of the human mitochondrial protease ClpP

Abstract

We recently described the identification of a new class of small-molecule activators of the mitochondrial protease ClpP. These compounds synthesized by Madera Therapeutics showed increased potency of cancer growth inhibition over the related compound ONC201. In this study, we describe chemical optimization and characterization of the next generation of highly potent and selective small-molecule ClpP activators (TR compounds) and demonstrate their efficacy against breast cancer models in vitro and in vivo. We selected one compound (TR-107) with excellent potency, specificity, and drug-like properties for further evaluation. TR-107 showed ClpP-dependent growth inhibition in the low nanomolar range that was equipotent to paclitaxel in triple-negative breast cancer (TNBC) cell models. TR-107 also reduced specific mitochondrial proteins, including OXPHOS and TCA cycle components, in a time-, dose-, and ClpP-dependent manner. Seahorse XF analysis and glucose deprivation experiments confirmed the inactivation of OXPHOS and increased dependence on glycolysis following TR-107 exposure. The pharmacokinetic properties of TR-107 were compared with other known ClpP activators including ONC201 and ONC212. TR-107 displayed excellent exposure and serum t_1/2 after oral administration. Using human TNBC MDA-MB-231 xenografts, the antitumor response to TR-107 was investigated. Oral administration of TR-107 resulted in a reduction in tumor volume and extension of survival in the treated compared with vehicle control mice. ClpP activation in vivo was validated by immunoblotting for TFAM and other mitochondrial proteins. In summary, we describe the identification of highly potent new ClpP agonists with improved efficacy against TNBC, through targeted inactivation of OXPHOS and disruption of mitochondrial metabolism.

Keywords: agonist; cell proliferation; mitochondria; oxidative phosphorylation; protease; small molecule; triple-negative breast cancer.

FIGURE 1

New TR compound analogs potently inhibit breast cancer cell growth in a ClpP‐dependent manner. (A) Comparison of novel TR compound chemical structures to ONC201 and analogs. Cell viability assays of (B) Wildtype (WT) and (C) ClpP knockout (ClpP‐KO) MDA‐MB‐231 and SUM159 cells. Cells were treated with indicated drug concentrations for 72 h and imaged following addition of Hoechst stain as described in Section 2. Values represent mean ± SEM normalized to DMSO control, representative of N = 3. (D) IC50 values in WT cells for compounds represented in (B). Values represent mean IC50 value (nM), N = 3. *ONC201 IC50 in MDA‐MB‐231 obtained from data shown in Figure 4D, SUM159 cells previously published. (E) Cell viability assay of C3Tag cells following TR‐107 and paclitaxel treatment. Cells were treated with drug concentrations as indicated for 72 h and imaged following Hoechst stain addition. Representative of N = 2. Average IC50 values indicated to the left and right of the graph for paclitaxel and TR‐107, respectively.

FIGURE 2

New TR compounds are potent activators of mitochondrial ClpP. (A) In vitro ClpP activity assays was performed as described in Section 2 using fluorescent casein‐FITC. Increase in fluorescent intensity relative to DMSO controls is shown; values represent mean ± SEM for technical replicates. EC50 values are presented in the table to the right. (B) Left panel—surface plasmon resonance (SPR) sensorgrams of TR‐107 binding to recombinant human ClpP coupled on the chip. Right panel—binding curves are shown as response units (RU) versus ligand concentration at steady state and fitted to a one‐site Langmuir binding model. The apparent K _d value obtained from the fit is shown.

FIGURE 3

TR‐107 induces loss of mitochondrial proteins in TNBC cells in a ClpP‐ dependent manner. Triple‐negative breast cancer cells MDA‐MB‐231 (A, C) and SUM159 (B, D) were treated with 10 mM ONC201 or 100 nM TR‐107 for indicated time points (A, B) or indicated doses (C, D) of ONC201 or TR‐107 for 24 h and immunoblot was performed for various mitochondrial metabolic proteins. Representative of N = 3.

FIGURE 4

TR‐107 reduces mitochondrial metabolic functions in MDA‐MB‐231 cells. Oxygen consumption rate (OCR) of MDA‐MB‐231 cells treated with DMSO or TR‐107 at indicated concentrations for 24 h was measured by Seahorse XF Analyzer as described in Section 2. Trace of oxygen consumption data following TR‐107 treatment in (A) WT and (B) ClpP‐KO MDA‐MB‐231 cells, representative of N = 3 experiments. Indicated compounds were added at concentrations described in Section 2. (C) Bar charts of data shown in (A) and (B), values represent mean ± SEM, N = 3. p‐value <.05 (*), .001 (**). (D) Cell viability assays (Hoechst stain, 72 h) of MDA‐MB‐231 cells treated with ONC201 (left) or TR‐107 (right) at indicated concentrations in media containing either glucose (black) or galactose (red). Values represent mean ± SEM, representative of N = 3. Average IC50 values are presented in table (right).

FIGURE 5

TR‐107 prevents tumor growth in MDA‐MB‐231 xenograft model. (A) Box‐ and‐whisker plot of tumor volume (mm³) of MDA‐MB‐231 xenografts following TR‐107 treatment at indicated concentrations at Day 26 (Table S2). N = 8 (Group 1) and 10 (Groups 2, 3). (B) Average tumor volume following TR‐107 administration as described in Methods. The tumor volume threshold (1500 mm³) is indicated by dashed line. N = 10 per group. Values represent mean ± SEM. (C) Kaplan‐Meier graph of mouse survival following TR‐107 administration as described in Methods and Table S2. Long‐rank (Mantel‐Cox) p‐value of 0.0163. Individual mice were recorded as leaving the study following death, tumor volume exceeding threshold (1500 mm³), or end of study period. (D) Box‐and‐whisker plot of relative mitochondrial DNA (mtDNA) copy number present in mouse xenograft lysate following conclusion of mouse study. N = 9 (Group 1) and N = 10 (Group 3). (E) Immunoblot (top) from tumor lysates of mouse xenografts following administration of vehicle control or 8 mg/kg TR‐107. Each animal ID represented an individual mouse. Two images are shown for HMGCS1 representing a short exposure (“short exp”) and long exposure (“long exp”). Quantification of these blots is shown in the box‐and‐whisker plot (bottom) with indicated p‐values (<.05 (*) and .01 (**)).