Targeting the Mitochondrial Protease ClpP for Anticancer Therapy

Abstract

Cancer cells depend on mitochondrial reprogramming for growth, but this raises reactive oxygen species (ROS), increasing reliance on protein quality control (PQC) repair mechanisms. The mitochondrial proteome is maintained through a robust PQC composed of chaperones and proteases, including the mitochondrial matrix protease caseinolytic protease P (ClpP). ClpP has recently emerged as a potential therapeutic target against cancer. Notably, imipridones act as ClpP agonists and have shown potent anticancer activity by inhibiting mitochondrial Electron Transport Chain (ETC) function. In this study, we developed a new generation ClpP agonist, compound 9 (MS6076), which exhibits enhanced ClpP binding, more potent disruption of mitochondrial ETC and lethality in breast cancer models compared to the imipridone ONC212. Furthermore, we show that compound 9 induced cell death in cancer cells resistant to ONC212. The discovery and characterization of compound 9 therefore add to the expanding arsenal of imipridones to target ClpP in cancer.

Graphical Abstract

INTRODUCTION

Cancer cells are characterized by reprogramming of their mitochondria for their anabolic growth through production of onco-metabolites for the generation of lipids, amino acids and nucleotides^1–8 and redirect ATP production to glycolysis.^6,9 However, such reprogramming comes at the cost of increased reactive oxygen species (ROS), which can lead to DNA damage, lipid peroxidation, and protein oxidation ultimately leading to mitochondrial dysfunction and cell death if kept unchecked.^10–13 Accumulation of misfolded proteins in the mitochondria activates a retrograde signaling cascade known as the mitochondrial unfolded protein response (UPRmt), which promotes the transcription of genes involved in increased mitochondrial antioxidant machinery, biogenesis, mitophagy and protein quality control (PQC).^14–19

While in the cytosol, protein degradation is facilitated through the ubiquitin-proteosome pathway, this system is absent from the mitochondrial matrix.^20–22 Instead, the mitochondrial proteome is maintained through the activity of a network of chaperones and proteases that either refold or degrade misfolded proteins.^23–25 One such essential protease is the serine caseinolytic protease P (ClpP), which exists as a tetradecameric ring.²⁶ Upon the binding of a ClpX hexamer, two tetradecameric ClpP rings associate to form a mature barrel-shaped protease ClpXP (thereafter referred to as ClpP).^27,28 ClpP has been shown to be frequently overexpressed in various cancers including breast, ovary, lung, uterus, testis, and prostate and primary acute myeloid leukemia (AML) human samples^29–31 and promote tumor growth through the regulation of OXPHOS and mitochondrial metabolism.^31,32 Both inhibition and overactivation of ClpP have been shown to inhibit cancer growth. Inhibition of ClpP has been previously used as a viable strategy for reducing tumor growth by disrupting OXPHOS, induction of mitochondrial ROS (mtROS), and initiation of apoptosis.³³

However, overactivation of ClpP has gained considerable attention as an anticancer therapeutic strategy and a number of ClpP agonists have been developed. These agonists can be classified into 2 types: imipridones (ONC201,^34–36 ONC212³⁷) including its derivatives (e.g., CLPP-1071, TR-107, NCA029) (Figure 1),^38–44 and nonimipridone compounds.^45–49 Imipridone ONC201 was initially developed as a TRAIL receptor inducer³⁵ but was later identified as a ClpP agonist, which can target AML cells while sparing non-malignant hematopoietic cells.³⁴ In this seminal study, the authors reported that ONC212 promotes the degradation of subunits of the electron transport chain, notably DAP13.³⁴ ONC201 has been advanced into multiple clinical trials to treat solid tumors and hematological malignancies. Notably, ONC201 is being tested in a phase 3 clinical trial for the treatment of H3K27 M glioma.⁵⁰

Figure 1.

Chemical structures of representative imipridone ClpP agonists.

In the current study, building on the structures of ONC201 and ONC212, we discovered compound 9 (MS6076), which displays higher binding affinity to ClpP. Furthermore, we assessed the efficacy of compound 9 using the degradation of DAP13 and mitochondrial respiration as readouts, and viability assays in various breast cancer cell lines and demonstrated that compound 9 is a more potent CIpP agonist compared to ONC212. We also show that compound 9 affects the viability of a cancer cell line that is resistant to ONC212. Overall, we discovered and characterized an improved ClpP agonist, expanding the arsenal of imipridones to target ClpP in cancer.

RESULTS AND DISCUSSION

Design and Evaluation of New ClpP Agonists.

Human mitochondrial ClpP exists as a heptamer in the absence of ClpX.⁵¹ Binding of the protein with ONC201 shifts the equilibrium to a heptamer-heptamer double-ringed tetradecameric structure.³⁴ In the heptamer, the pockets are formed by two adjacent monomers, with each pocket occupied by one ONC201 molecule (Figure 2A, red circle).³⁴ Pharmacological agonism of ClpP leads to the widening of the hollow chamber and increased activity of ClpP, independent of ClpX.²⁷ While the discovery of ONC201 is groundbreaking, one limitation is the observation that concentrations in the micromolar range are required to significantly affect cancer cell survival.³⁴ Analysis of the published cocrystal structure of ClpP and ONC201 (PDB ID: 6DL7) revealed that the methyl group at the ortho position of the phenyl ring, a hydrophobic group, is surrounded with nearby residues with a known propensity to form hydrogen bonds, such as Arg78, Glu82, Gln107, and Ser108 (Figure 2B).³⁴ Therefore, we hypothesized that introducing an amine group to ONC201 with a flexible linker from this methyl group may enable accessibility to these residues and create additional interactions with ClpP.

Figure 2. The crystal structure of human mitochondrial ClpP in complex with ONC201.

(A) Cocrystal structure of ClpP in complex with ONC201 (PDB ID: 6DL7)³⁴ showing the seven ClpP monomers forming a heptamer ring complex. ONC201 molecules bind to the pockets formed by the adjacent monomers. One of the ONC201 molecules in its binding pocket is highlighted in a red circle. (B) Close-up view of the interaction between ONC201 (shown as yellow sticks) and ClpP protein (gray surface). The polar interactions are shown as orange dotted lines. The tertiary amine group buried deep in the pocket forms a hydrogen bond with Tyr-118, and the carbonyl group forms hydrogen bonds with two water molecules. The nearby Arg78, Glu82, Gln107, and Ser108 residues with a propensity to form hydrogen bonds are labeled and shown as sticks and colored surfaces.

Since alteration of substituents of ONC201 (X and Y groups in Table 1) have been previously described,^{38,39,43,44,52–54} we used this information to guide the selection of the best X and Y group combinations, while altering the Z group, which replaces the methyl group at the ortho position of the phenyl ring, to achieve improvement in the affinity for ClpP (Table 1). A trifluoromethyl group or a halogen group, such as fluoro, chloro, or bromo has been shown to be effective as group X. Therefore, we chose either trifluoromethyl (CF₃) or fluoro (F) group as the X group. F was chosen over chloro (Cl) or bromo (Br) group, due to its smaller size and molecular weight as well as to prevent the possible complications that could arise from Cl or Br substitutions during the chemical synthesis. Furthermore, cyano (CN) group was shown to be effective as a Y group as exemplified by compounds CLP-1071, TR-107, and NCA029 (Figure 1).^{39,43,44,53,54} Having identified X (CF₃ or F) and Y (H or CN) as potential optimal substituents, we designed and synthesized compounds altering the Z group as the first set of compounds (Table 1). As an initial starting point, we chose a five-atom linker to introduce the amine group, to possibly achieve enough flexibility to reach surrounding ClpP residues (Table 1).

Table 1.

Structures of ONC201, ONC212 and Newly Designed Imipridone Agonists 1–8

To evaluate the binding affinity of our compounds, we used the thermal shift assay (TSA) as the primary screening assay. TSA is based on the observation that at a certain temperature, a protein not in complex with a small-molecule ligand undergoes denaturation while the protein in complex with a small-molecule ligand is more resistant to denaturation due to its additional stability.⁵⁵ TSA using whole-cell lysates allows for a comparative assessment of the binding affinity of multiple compounds as previously reported.⁵⁶ To determine an optimal cell line to conduct the screening of compounds, we first assessed the relative level of ClpP in eight different breast cancer cell lines (Figure S1A). Although the basal ClpP level was heterogeneous among cell lines, the triple negative breast cancer cell line MDA-MB-468 showed a relatively high level (Figure S1B). Additionally, since targeted therapy is lacking for triple-negative breast cancers, we selected MDA-MB-468 cells for compound screening and characterization.^57,58

Compound 1, which contains the same X and Y substitutions as ONC212 but the pentylamino group in the Z position (Table 1) showed marginally improved binding to ClpP compared to ONC212, suggesting that the substitution in the Z position can positively enhance ClpP binding (Figure 3A, 3B). Compounds 2–4 contain the same Z group as compound 1 but vary in the substitutions at the X and Y positions. An approximately 2-fold improvement in ClpP binding of compound 2 featuring the CF₃ group in the X and the CN group in the Y position was observed compared to ONC212 and compound 1 (Figure 3A, 3B). Compound 3 retains the same substitution in the Y position as compound 1, but replaced CF₃ with F in position X. The TSA revealed that this change in the X position resulted in a weaker ClpP binding for compound 3 compared to compound 1 (Figure 3A, 3B). A similar effect was observed for compound 4, which is structurally identical to compound 2 except replacing the CF₃ group with the F group in position X. Thus, we concluded that the pentylamino group at the Z position could slightly improve ClpP binding, while the CF₃ group at the X position and the CN group at the Y position further improve ClpP binding. Compounds 5 and 6 were synthesized with an acetyl amide group at the end of the amino tail (Table 1) and led to much reduced binding to ClpP than the primary amine (compounds 1–4), which further indicates that the amino group plays a key role (Figure 3A, 3B). Meanwhile, compounds 7 and 8 were synthesized featuring different tertiary amine groups in the Z position (Table 1). The ClpP binding of both compounds is recovered to the same level as that of compound 2 (Figure 3A, 3B), suggesting that the amino group, including primary amino and tertiary amino but not the amide group, may potentially form hydrogen bonds with surrounding polar residues of ClpP and play a role in stabilizing the binding of the compounds to ClpP. While promising, the improvement in ClpP binding of these compounds is moderate.

Figure 3. Thermal shift assays reveal increased ClpP binding for several of the newly designed ClpP agonists.

(A) TSA analysis of ClpP stabilization following treatment with increasing concentrations of compounds 1–8 and ONC212 on MDA-MB-468 lysates. WB results shown are representative of two independent experiments and include a no-compound control (0.0) and unheated control (UHC). (B) Quantification of ClpP stabilization by each compound show in panel A. The calculated values represent the mean ± SD from two independent experiments.

Therefore, to induce more favorable interactions with ClpP, we shorten the Z group from five to three atoms to position the terminal amino group closer to the target residues. Furthermore, the carbon atom attached to the phenyl ring was replaced with a more hydrophilic oxygen atom to ease the chemical synthesis of the derivatives and potentially improve the solubility of the compounds. These modifications resulted in the generation of compounds 9–16 (Figure 4A). Compound 9 (MS6076) showed superior stabilization of ClpP by TSA compared to all other compounds including ONC212 (Figure 4B, 4C) as a significant amount of ClpP was stabilized even at a concentration of 0.313 μM compared to ONC212, which failed to stabilize ClpP at a concentration below 2.5 μM. The acetyl-capped derivative, compound 10, showed decreased ClpP binding compared to 9 but similar to ONC212 (Figure 4B, 4C). This is consistent with our earlier finding that a terminal amine may be superior to an amide. We next explored the different modifications on the amino group and synthesized compounds 11–16 (Figure 4A). While compounds 11–16 stabilized ClpP at a concentration of less than 2.5 μM (Figure 4B), given the superior ClpP stabilization of compound 9 compared to other compounds, we selected compound 9 as the lead compound for further biological characterization.

Figure 4. Compound 9 shows superior ClpP stabilization.

(A) Structures of the second generation ClpP agonists 9–16. (B) TSA results of ClpP in MDA-MB-468 lysates incubated with indicated concentrations of compounds 9–16 and ONC212 including a no-compound control (0.0) and unheated control (UHC). WB results shown are representative of two independent experiments. (C) Quantification of ClpP stabilization by TSA shown in panel B. The calculated values represent the mean ± SD from two independent experiments. (D) Quantification of purified hClpP stabilization shown by DSF assay (corresponding melting curves are shown in Figure S2). The calculated values represent the mean ± SD from four independent experiments. *P < 0.05, ns: not significant.

TSA is a biologically relevant assay, as compound binding is measured in the context of cellular lysates containing numerous other proteins. Moreover, the TSA performed in live cells requires the compound to cross both the plasma and mitochondrial membranes, providing an additional layer of physiological relevance to the observed ClpP stabilization. As additional support for binding of compound 9 to ClpP, we also performed differential scanning Fluorimetry (DSF) with purified human ClpP (hClpP). This biochemical assay, which cannot be affected by other factors such as other proteins in cell lysate or indirect stabilization in cells, clearly showed the binding of ONC212 and compound 9 with a very significant thermal shift (Figures 4D and S2).

Compound 9 Increases Degradation of DAP13 and Accelerates Mitochondrial Dysfunction.

While TSA on cell extracts demonstrated superior ClpP stabilization of compound 9 compared to ONC212 and the other derivatives tested, this assay does not assess permeability. Additionally, since ClpP localizes to the mitochondrial matrix, to be biologically relevant, compound 9 needs to penetrate three membranes: the cell membrane, the outer mitochondrial membrane, and the inner mitochondrial membrane. As such, we evaluated the permeability as well as ClpP binding of compound 9 using the TSA in live cells. We found that while ONC212 stabilized ClpP at a concentration as low as 0.63 μM (Figure 5A, 5B), compound 9 stabilized ClpP at 0.08 μM (Figure 5A, 5B), suggesting enhanced permeability and ClpP binding.

Figure 5. Compound 9 exhibits enhanced proteolytic activity of ClpP and mitochondrial dysfunction and induces apoptosis.

(A) TSA results of ONC212 and compound 9 in MDA-MB-468 cells treated with the indicated compound for 30 min. WB results shown are representative of two independent experiments and include a no-compound control (0.0) and unheated control (UHC). (B) Normalized quantification of the ClpP stabilization by each compound as determined by TSA in panel A. The calculated values represent the mean ± SD from two independent experiments. (C) MDA-MB-468 cells were treated with either 9 or ONC212 at concentrations of 25, 50, 100, 150, and 200 nM for 24 h. The DAP13 protein level was analyzed using WB. WB results shown are representative of two independent experiments. (D) Quantification of DAP13 degradation shown in panel C. The calculated values represent the mean ± SD from two independent experiments. (E) Basal, ATP-linked, and maximal respiration analyzed by Seahorse assay of MDA-MB-468 cells treated with 50 nM of ONC212 or compound 9 for 12, 24, and 48 h. Comparisons between the control sample and ONC212 are shown in black. Comparisons between control and compound 9 are shown in red. Comparisons between compounds ONC212 and 9 are shown in blue. The calculated values represent the mean ± SD from three independent experiments. Two-tailed unpaired t test was used. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, ns: not significant. (F) Western blot analysis of apoptosis markers in MDA-MB-468 cells treated with vehicle, 1 μM of ONC212, 1 μM of compound 9, or 0.5 of μM Doxorubicin for 72 h. WB results shown are representative of two independent experiments.

Since ONC212 and ONC201 have been reported to promote the degradation of noncanonical targets such as the mitochondrial matrix protein DAP13,³⁴ in order to assess the effect of compound 9 on the proteolytic activity of ClpP, we performed a DAP13 degradation assay. We found that while ONC212 and compound 9 both degraded DAP13 at high concentrations such as 100, 150, and 200 nM (Figure 5C, 5D), at 25 nM, compound 9 still degraded more than 50% of DAP13 while ONC212 did not significantly degrade DAP13 (Figure 5D). This result suggests that the greater ClpP binding of compound 9 translates into an increased activation of its proteolytic activity.

The degradation of proteins involved in the electron transport chain, such as DAP13, by ONC201 and ONC212 have been shown to disrupt mitochondrial function.^34,59 Therefore, we performed a time course for the effect of compound 9 on mitochondrial function using a Seahorse assay (Figure 5E). We found that compound 9 reduced basal respiration, ATP generation and maximal respiration at all time points tested, while the ONC212 significantly affected these mitochondrial functions only at 48 h. Collectively, these results indicate that compound 9 enhances the proteolytic activity of ClpP and disrupts mitochondrial function more effectively than ONC212. Disruption of the cellular energy network is likely a key component of cancer cell death induced by compound 9 (see below).

Since previous reports showed that treatment with ONC201 induced apoptosis in a caspase 3-dependent manner,^60,61 we tested the effect of compound 9 on caspase 3, cleaved caspase 3, as well as PARP and cleaved PARP using doxorubicin as a positive control.⁶¹ We found a reduction in protein levels of full-length caspase 3 and PARP in cells treated with compound 9 or ONC212 for 72 h, while compound 9 increased the cleaved caspase 3 and cleaved PARP levels, similar to the positive control doxorubicin (Figure 5F), indicating that similar to ONC212, compound 9 effectively induces apoptosis. Furthermore, we performed flow cytometry experiments by using eFluor 780,⁶² which is a viability dye to irreversibly label dead cells upon treatment with compounds (Figure S3). We used eFluor 780 to identify late-stage apoptosis and necrosis instead of early apoptosis determined by Annexin V.⁶³ As shown in Figure S3, both compound 9 and ONC212 effectively induced late-stage apoptosis and compound 9 was more effective than ONC212 in this assay.

Compound 9 Exhibits Potent Cancer Cell Killing in Breast Cancer Models and ONC212-Resistant Cells.

To determine if this increased activation of ClpP by compound 9 results in increased cancer cell death, we performed a viability assay in the MDA-MB-468 triple-negative breast cancer cell line. We found that at 72 h, compound ONC212 had an IC₅₀ of 35.2 nM while compound 9 had an IC₅₀ which was approximately 3.5-fold lower at 10.6 nM (Figure 6A), suggesting superior efficacy against cancer cells. In order to determine whether this effect is also observed in other cancer cell lines, we expanded the viability assay to four additional breast cancer cell lines (Figure 6B). This analysis revealed that in the triple-negative cell line MDA-MB-231, while the IC₅₀ of ONC212 is 50.5 nM, the IC₅₀ of compound 9 is 13.9 nM, therefore also showing an approximately 3.5-fold improvement. In SK-BR-3 cells, we observed the IC₅₀ values of 114.7 nM and 11.5 nM for ONC212 and compound 9, respectively (approximately 10-fold improvement). In MCF-7 cells, the IC₅₀ of ONC212 was 63.7 nM, and the IC₅₀ of compound 9 was 7.0 nM (approximately 9-fold improvement). In BT-474, the IC₅₀ of ONC212 was 32.7 nM, and the IC₅₀ of compound 9 was 8.0 nM (approximately 4-fold improvement). These results indicate that across tested breast cancer cell lines, compound 9 shows an improved antiproliferative effect.

Figure 6. Compound 9 exhibits potent cancer cell killing in breast cancer models and ONC212-resistant cells.

(A) Viability curves of MDA-MB-468 cells treated with ONC212 or compound 9 at the indicated concentrations for 72 h. (B) Viability curves of MDA-MB-231, SK-BR-3, MCF-7, and BT-474 cells treated with ONC212 or compound 9 at indicated concentrations for 72 h. (C) Schematic representation of the strategy used to select ONC212 resistant MDA-MB-468 cells. (D) Viability curves of wild-type or resistant MDA-MB-468 cells treated with ONC212 at the indicated concentrations for 72 h. (E) Viability curves of MDA-MB-468 cells resistant to ONC212 treated with either ONC212 or compound 9 at indicated concentrations for 72 h. The calculated values in panels A, B, D and E represent the mean ± SD from three independent experiments.

To further evaluate compound 9, we aimed at testing its efficacy in cells that are resistant to ONC212. To do so, we first selected MDA-MB-468 cells that remain viable in the presence of 50 nM ONC212 (Figure 6C). We found that while ONC212 inhibited the growth of wild-type MDA-MB-468 cells with an IC₅₀ of 35.2 nM, it suppressed the growth of the resistant cells with a much higher IC₅₀ of 350.8 nM (Figure 6D). We then assessed the effect of compound 9 on the viability of these resistant cells and found that an IC₅₀ of 27.7 nM in response to compound 9 compared to an IC₅₀ of 10.6 nM of wild-type cells (Figure 6E). This result indicates that although the resistant cells also developed partial resistance to compound 9, the increase in the IC₅₀ value for compound 9 was only 2.6-fold compared to the approximately 10-fold increase for ONC212, suggesting that compound 9 may potentially be utilized to overcome acquired resistance to first-line treatment with ClpP agonists.

ONC212-resistant cells retained partial sensitivity to both ONC212 and compound 9, as high concentrations of either drug continued to induce apoptosis. The observed right-shift in the viability curve, rather than a complete loss of response, indicates that resistance is gradual and adaptive rather than absolute. This adaptation likely reflects proteome remodeling that enables cells to tolerate the proteostatic stress caused by ClpP hyperactivation.

Mechanistically, resistance can be conceptualized as either reducing the magnitude of the cellular insult or enhancing cellular safeguards. The effective insult may be reduced if cells downregulate ClpP expression or acquire mutations that diminish its activity, thereby limiting the pool of functional enzyme available for hyperactivation. Conversely, resistant cells may counterbalance the insult by augmenting protective pathways. This could involve stronger activation of antioxidant defenses or other proteostatic programs that mitigate mitochondrial stress.

Together, these observations support a model in which resistance to ONC212 arises from adaptive mechanisms that either blunt the ClpP insult or reinforce cellular defenses against it. The enhanced potency of compound 9 enables it to surpass these adaptations, highlighting its promise as a second-line therapy for tumors that develop resistance to ONC212. Furthermore, we assessed in vivo mouse pharmacokinetic (PK) properties of compound 9. We determined the plasma concentrations of compound 9 in Swiss albino mice following a single intraperitoneal (IP) administration at 50 mg/kg dose and showed that compound 9 is bioavailable in mice via IP injection (Figure S4) and was well tolerated by the treated mice. Hence, compound 9 is suitable for in vivo studies.

Chemical Synthesis of Compounds 1–16.

The synthetic routes for preparation of compounds 1–8 are outlined in Scheme 1. Compounds 1–4 were generated by reacting 20a/b with 21c/d following the reported procedures.⁶⁴ The intermediates 20a/b were obtained from 17a/b, via Sonogashira coupling reaction to give 18a/b followed by hydrogenation to produce amines 19a/b. 19a/b were further reacted with methyl 2-(methylthio)-4,5-dihydro-1H-imidazole-1-carboxylate via S_NAr reaction and hydrolysis of the methyl carboxylate in one-pot to give 20a/b (Scheme 1A). Compounds 5 and 6 were generated through acetylation of compounds 1 and 2 (Scheme 1A), while compound 7 was generated from compound 2 via reductive amination reaction with formaldehyde (Scheme 1B, top). Compound 8 was generated from compound 2 via reductive amination with ethyl 2-oxoacetate, followed by Boc-protection, hydrolysis of ester, and amide coupling reaction with methyl amine to give 22. Boc protecting group of 22 was then removed followed by reductive amination with formaldehyde, to give desired compound 8 (Scheme 1B, bottom).

Scheme 1. Synthesis of Compounds 1–8^a.

^aReagents and conditions: (a) tert-butyl pent-4-yn-1-ylcarbamate, Pd(PPh₃)₂Cl₂, CuI DMF/TEA = 10/1, 80 °C, 2 h, 88% and 90%, respectively. (b) Raney nickel, NH₃, MeOH, rt, overnight, quantitative yield. (c) Methyl 2-(methylthio)-4,5-dihydro-1H-imidazole-1-carboxylate, AcOH, MeOH, 70 °C, overnight. (d) (1) NaOMe, MeOH, 60 °C, overnight; (2) TFA, DCM, rt, 30 min, yields 40%–60% over 4 steps from 17a/b. (e) Ac₂O, TEA, DCM, rt, 2h, 86% and 92%, respectively. (f) (1) Ethyl 2-oxoacetate, NaBH₃CN, AcOH, MeOH, rt, overnight; (2) NaHCO₃ (aqueous), Boc₂O, MeOH, rt, 2 h. (g) (1) 1 M NaOH (aqueous), MeOH, rt, overnight; (2) EDCI, HOAT, methylamine, NMM, DMF, rt, overnight, 40% over 2 steps. (h) TFA, DCM, rt, 30 min. (i) HCHO (37% wt in H₂O), NaBH₃CN, AcOH, MeOH, rt, overnight, 65% over 2 steps.

The synthetic routes for the preparation of compounds 9–16 are outlined in Scheme 2. Intermediates 25a and 25b were prepared in a similar way starting from 2-hydroxy-4-(trifluoromethyl)benzonitrile, the phenol group of which was reacted with alkyl bromide (R-Br) via S_N2 reaction to give 23a and 23b. The CN group of 23a and 23b was then reduced via hydrogenation and further reacted with methyl 2-(methylthio)-4,5-dihydro-1H-imidazole-1-carboxylate to yield desired 25a and 25b (Scheme 2A). Compound 9 was synthesized from intermediates 25a and 21d according to the same procedures described for the synthesis of compounds 1–4 (Scheme 2A). Compound 10 was generated through acetylation of compound 9 (Scheme 2A). Intermediate 26 was generated by reacting 25b with 21d, followed by hydrolysis of dimethyl acetyl and then reductive amination reaction with different amines, to afford compounds 12–15 (Scheme 2B, top). Compound 11 was generated from compound 9 via reductive amination with formaldehyde (Scheme 2B, middle). Compound 16 was generated through acetylation of compound 15 (Scheme 2B, bottom).

Scheme 2. Synthesis of Compounds 9–16^a.

^aReagents and conditions: (a) 23a: tert-butyl (2-bromoethyl)carbamate, Cs₂CO₃, CH₃CN, 80 °C, overnight, 34%; 23b: 2-bromo-1,1-dimethoxyethane, K₂CO₃, NaI, DMF, 100 °C, overnight, 46%. (b) Raney nickel, NH₃, MeOH, H₂, rt, overnight. (c) Methyl 2-(methylthio)-4,5-dihydro-1H-imidazole-1-carboxylate, AcOH, MeOH, 70 °C, overnight. (d) NaOMe, MeOH, 60 °C, overnight, 53% for 26. (e) TFA, DCM, rt, 30 min 54% over 2 steps for 9. (f) Ac₂O, TEA, DCM, rt, 2 h, 92%. (g) (1) TsOH·H₂O, acetone, reflux, 4 h; (2) amine, NaBH₃CN, AcOH, MeOH, rt, overnight; (3) only for 15, TFA, DCM, rt, 30 min. (h) HCHO (37% wt in H₂O), NaBH₃CN, AcOH, MeOH, rt, overnight, 92%.

CONCLUSIONS

Malignant cancer cells continue to utilize their mitochondria to produce necessary macromolecules and oncometabolites, which contribute to growth and proliferation.^6,9 In order to maintain the health and integrity of their mitochondria, cancer cells overexpress mitochondrial matrix protease ClpP, which promotes the degradation of misfolded and damaged proteins.²⁶ However, this overexpression of ClpP can be therapeutically exploited as an anticancer treatment by modulating ClpP’s activity.^34,38 Previous investigations found that the pharmacological agonism of ClpP by small molecules increased ClpP’s proteolytic activity.³⁴ Improved ClpP’s activity also increased the range of its substrates, now targeting nonmisfolded mitochondrial matrix proteins such as DAP13.³⁴ Induced hyperactivation of ClpP ultimately led to increased levels of mitochondrial reactive oxygen species, disruption of the electron transport chain complexes, and selective cancer cell death in a concentration-dependent manner.³⁴ Thus, the activation of ClpP through small molecule agonists represents a promising therapeutic strategy for inducing cancer cell death.

In this study, we conducted structure–activity relationship (SAR) studies and characterized various novel ClpP agonists (Table 1, Figures 3 and 4). Among the tested derivatives, we discovered compound 9, which showed superior binding for ClpP compared to ONC212 (Figure 4). We further showed that the difference in ClpP binding between the two compounds persists in a live cell context (Figure 5A, 5B). We also demonstrated that DAP13, a previously characterized target of agonized ClpP, was degraded more readily by compound 9 (Figure 5C, 5D). We also analyzed the integrity of the cellular respiration chain in cells treated with compound 9. We found that at equal treatment concentration, compound 9 led to greater disruption of the mitochondrial complexes on a faster time scale compared to ONC212 (Figure 5E). We also showed that compound 9 effectively induced apoptosis in cancer cells (Figures 5F and S3).^60,61 Additionally, we showed that compound 9 led to greater cancer cell killing than ONC212 in various breast cancer cell models (Figure 6A, 6B). Lastly, we derived resistant cancer cells to ONC212 that had a 10-fold greater IC₅₀ compared to wild-type MDA-MB-468 breast cancer cells (Figure 6C). Importantly, these ONC212-resistant cells were only modestly resistant to compound 9, suggesting that compound 9 could be used as a second-line treatment for tumors that become desensitized to ONC212 (Figure 6D, 6E).

Overall, we designed, synthesized and evaluated a set of new ClpP agonists. Our SAR studies identified compound 9 as the most effective ClpP agonist to date. Compound 9 displayed superior ClpP binding/stabilization, anticancer cell viability, and greater metabolic disruption. Given that ONC201 is currently undergoing clinical trials for the treatment of H3K27 M mutant glioma, more potent ClpP agonists such as compound 9 have potential as next-generation therapies, either as standalone treatments or as second-line options for tumors that develop resistance to first-line ClpP agonists.⁵⁰ Targeting cancer by disrupting mitochondrial functions through pharmacological modulation of ClpP activity with small molecules represents an innovative and promising therapeutic approach and compound 9 adds to the expanding arsenal of ClpP-targeting small-molecule therapeutics.

EXPERIMENTAL SECTION

General Chemistry Methods.

Reagents and solvents were purchased from commercial sources without further purification, unless otherwise indicated. The progress of reactions was monitored by thin-layer chromatography (TLC) and/or LC-MS. All final compounds for biological evaluation were purified by preparative high-performance liquid chromatography (HPLC) on an Agilent Prep 1200 series with H₂O containing 0.1% TFA (A) and 10% of methanol or acetonitrile (B) as mobile phases. NMR spectra were obtained from a Bruker (Billerica, MA) DRX nuclear magnetic resonance (NMR) spectrometer (400 and 600 MHz for ¹H NMR, 151 MHz for ¹³C NMR, 376 MHz for ¹⁹F NMR). Chemical shifts for all compounds are reported in units of parts per million (ppm, δ) relative to the residual CDCl₃ (7.26 ppm 1H), MeOD (3.31 ppm 1H) or DMSO-d₆ (2.50 ppm 1H). ¹H NMR data are reported in the following format: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet), coupling constant, and integration. High-performance liquid chromatography (HPLC) spectra for all compounds were acquired using an Agilent 1260 infinity II series system with a diode array detector. Chromatography was performed on a 2.1 × 50 mm Exteud-C18 1.8 μm column with water containing 0.1% formic acid as solvent A and acetonitrile containing 0.1% formic acid as solvent B at a flow rate of 0.8 mL/min. The gradient program was as follows: 5–99% B (1–1.5 min), 99% B (1.5–2.5 min) and 5% B (2.5–4 min). High resolution mass spectra were recorded on an Agilent G6230LC/TOF with an electrospray ionization (ESI) source. All final compounds had >95% purity using the HPLC methods described above.

tert-Butyl (5-(2-Cyano-5-(trifluoromethyl)phenyl)pent-4-yn-1-yl)carbamate (18a).

To a solution of 2-bromo-4-(trifluoromethyl)benzonitrile (1.0 g, 4.0 mmol, 1.0 equiv), tert-butyl pent-4-yn-1-ylcarbamate (0.81 g, 4.4 mmol, 1.1 equiv) in 11 mL DMF/TEA (ratio = 10/1) were added Pd(PPh₃)₂Cl₂ (140 mg, 0.20 mmol, 0.05 equiv) and CuI (38 mg, 0.20 mmol, 0.05 equiv). Bubbled with nitrogen gas for 5 min and the mixture was stirred at 80 °C for 2 h. Cool to room temperature, diluted with 50 mL EA, and washed with brine for 3 times. Concentrated and the residue was purified by chromatography column (Hexane–hexane/EA = 3/1). m = 1.24 g, yield: 88%. ¹H NMR (400 MHz, CDCl₃) δ 7.75 (d, J = 7.4 Hz, 2H), 7.61 (d, J = 8.0 Hz, 1H), 4.72 (s, 1H), 3.32 (s, 2H), 2.57 (t, J = 7.2 Hz, 2H), 1.86 (t, J = 7.1 Hz, 2H), 1.44 (s, 9H). LC-MS m/z: [M + Na]⁺ calcd for C₁₈H₁₉F₃N₂O₂Na⁺, 375.13; found, 375.29.

tert-Butyl (5-(2-Cyano-5-fluorophenyl)pent-4-yn-1-yl)-carbamate (18b).

This compound was synthesized in a similar way to 18a, except for using 2-bromo-4-fluorobenzonitrile instead of 2-bromo-4-(trifluoromethyl)benzonitrile. m = 1.36 g, yield: 90%. ¹H NMR (400 MHz, Chloroform-d) δ 7.61 (dd, J = 8.7, 5.4 Hz, 1H), 7.19 (dd, J = 8.9, 2.6 Hz, 1H), 7.07 (ddd, J = 8.6, 7.9, 2.6 Hz, 1H), 4.75 (s, 1H), 3.32 (q, J = 6.6 Hz, 2H), 2.55 (t, J = 6.9 Hz, 2H), 1.85 (p, J = 6.9 Hz, 2H), 1.43 (s, 9H). LC-MS m/z: [M + Na]⁺ calcd for C₁₇H₁₉FN₂O₂Na⁺, 325.13; found, 325.17.

tert-Butyl (5-(2-(((4,5-Dihydro-1H-imidazol-2-yl)amino)-methyl)-5-(trifluoromethyl)phenyl)pentyl)carbamate (20a).

To a solution of 18a (100 mg, 0.28 mmol, 1.0 equiv) in 5 mL methanol was added Raney nickel in water (10 mg) and ammonia solution (100 μL), react overnight at room temperature in hydrogen atmosphere. Monitored by LC-MS until the reaction was completed. Filtered and the filtrate was concentrated under low pressure to yield the crude product 19a and used into the next step without further purification. LC-MS m/z: [M + H]⁺ calcd for C₁₈H₂₈F₃N₂O₂⁺, 361.21; found, 361.09.

The residue was dissolved into 5 mL MeOH/AcOH = 10/1, and methyl 2-(methylthio)-4,5-dihydro-1H-imidazole-1-carboxylate (49 mg, 0.28 mmol, 1.0 equiv) was added. React at 70 °C overnight and monitor the reaction by LC-MS until the reaction is completed. Concentrated and the residue was dissolved into 10 mL DCM and neutralized with saturated NaHCO₃ aqueous solution, separated and the organic phase was washed with brine, dried over anhydrous Na₂SO₄, concentrated to yield crude product 20a and used into next step without further purification. LC-MS m/z: [M + H]⁺ calcd for C₂₁H₃₂F₃N₄O₂⁺, 429.25; found,429.38.

tert-Butyl (5-(2-(((4,5-Dihydro-1H-imidazol-2-yl)amino)-methyl)-5 fluorophenyl)pentyl)carbamate (20b).

Compound 20b was synthesized in a similar way to 20a. except for using 18b instead of 18a as the start material. LC-MS m/z: [M + H]⁺ calcd for C₂₀H₃₂FN₄O₂⁺, 379.25; found, 379.38.

4-(2-(5-Aminopentyl)-4-(trifluoromethyl)benzyl)-7-benzyl-2,4,6,7,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-5(1H)-one (1).

To a solution of 20a in 2 mL methanol was added methyl 1-benzyl-4-oxopiperidine-3-carboxylate 21c (80 mg, 0.28 mmol, 1.0 equiv) and MeONa (25 wt % in methanol) (0.19 mL, 0.84 mmol, 3.0 equiv). React at 60 °C overnight and monitor the reaction by LC-MS until the reaction is completed. Concentrated and the residue was purified by reverse chromatography column (C18, 10–100% methanol/0.1% TFA in H₂O). The above product was dissolved into 2 mL TFA/DCM = 1/1, the reaction mixture was stirred at room temperature for 30 min. Concentrated and the residue was purified by reverse chromatography column (C18, 10–100% methanol/0.1% TFA in H₂O) to afford 1, White solid, 0.13 g, yield: 63%. ¹H NMR (400 MHz, MeOD) δ 7.57 (s, 1H), 7.54–7.42 (m, 6H), 7.18 (d, J = 8.1 Hz, 1H), 5.22 (s, 2H), 4.45 (dd, J = 10.8, 7.8 Hz, 2H), 4.33 (s, 2H), 4.06 (dd, J = 10.6, 7.9 Hz, 2H), 3.83 (s, 2H), 3.45 (t, J = 6.1 Hz, 2H), 3.06 (t, J = 6.2 Hz, 2H), 2.95 (t, J = 7.6 Hz, 2H), 2.84 (t, J = 7.9 Hz, 2H), 1.80–1.68 (m, 4H), 1.58–1..49 (m, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₉H₃₅F₃N₅O⁺, 526.2788; found, 526.2788.

3-((4-(2-(5-Aminopentyl)-4-(trifluoromethyl)benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (2).

Compound 2 was synthesized in a similar way to 1, except for using 20a and 21d as the starting materials. White solid, 85 mg, yield: 39%. ¹H NMR (400 MHz, MeOD) δ 7.83 (s, 1H), 7.79–7.71 (m, 2H), 7.59 (d, J = 7.8 Hz, 1H), 7.58–7.56 (m, 1H), 7.47 (d, J = 8.1 Hz, 1H), 7.18 (d, J = 8.0 Hz, 1H), 5.23 (s, 2H), 4.53–4.44 (m, 2H), 4.12–4.03 (m, 4H), 3.54 (s, 2H), 3.18–3.08 (m, 2H), 2.99–2.89 (m, 4H), 2.84 (t, J = 7.9 Hz, 2H), 1.80–1.68 (m, 4H), 1.58–1..49 (m, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₀H₃₄F₃N₆O⁺, 551.2741; found, 551.2716.

4-(2-(5-Aminopentyl)-4-fluorobenzyl)-7-benzyl-2,4,6,7,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-5(1H)-one (3).

Compound 3 was synthesized in a similar way to 1, except for using 20b and 21c as the starting materials. White solid, 98 mg, yield: 42%. ¹H NMR (400 MHz, MeOD) δ 7.55–7.50 (m, 2H), 7.50–7.45 (m, 3H), 7.04 (dd, J = 9.8, 2.7 Hz, 1H), 6.99 (dd, J = 8.6, 5.5 Hz, 1H), 6.88 (td, J = 8.4, 2.7 Hz, 1H), 5.13 (s, 2H), 4.46 (dd, J = 10.8, 7.9 Hz, 2H), 4.36 (s, 2H), 4.07 (dd, J = 10.6, 7.9 Hz, 2H), 3.87 (s, 2H), 3.49 (t, J = 6.1 Hz, 2H), 3.07 (t, J = 6.2 Hz, 2H), 2.94 (t, J = 7.7 Hz, 2H), 2.77 (t, J = 7.8 Hz, 2H), 1.79–1.68 (m, 4H), 1.57–1.46 (m, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₈H₃₅FN₅O⁺, 476.2820; found, 476.2885.

3-((4-(2-(5-Aminopentyl)-4-fluorobenzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-7(6H)-yl)-methyl)benzonitrile (4).

Compound 4 was synthesized in a similar way to 1, except for using 20b and 21d as the starting materials. White solid, 96 mg, yield: 40%. ¹H NMR (400 MHz, MeOD) δ 7.87 (s, 1H), 7.82–7.74 (m, 2H), 7.61 (t, J = 7.8 Hz, 1H), 7.04 (dd, J = 9.8, 2.7 Hz, 1H), 6.99 (dd, J = 8.6, 5.5 Hz, 1H), 6.89 (td, J = 8.4, 2.7 Hz, 1H), 5.13 (s, 2H), 4.55–4.43 (m, 2H), 4.18 (s, 2H), 4.07 (dd, J = 10.6, 8.2 Hz, 2H), 3.65 (s, 2H), 3.25 (t, J = 5.9 Hz, 2H), 3.01–2.90 (m, 4H), 2.77 (t, J= 7.8 Hz, 2H), 1.78–1.68 (m, 4H), 1.57–1.46 (m, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₉H₃₄FN₆O⁺, 501.2773; found, 501.2803.

N-(5-(2-((7-Benzyl-5-oxo-1,2,6,7,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-4(5H)-yl)methyl)-5-(trifluoromethyl)-phenyl)pentyl)acetamide (5).

To a solution of compound 1 (20 mg, 0.038 mmol, 1.0 equiv) in 2 mL DCM were added triethylamine (26 μL, 0.19 mmol, 5.0 equiv) and acetic anhydride (11 μL, 0.114 mmol, 3.0 equiv). The reaction mixture was stirred at room temperature for 2 h. Concentrated and purified by prepared HPLC. White solid, 22 mg, yield: 86%. ¹H NMR (400 MHz, MeOD) δ 7.58–7.40 (m, 7H), 7.18 (d, J = 8.1 Hz, 1H), 5.24 (s, 2H), 4.49 (dd, J = 10.5, 8.1 Hz, 2H), 4.39 (s, 2H), 4.09 (dd, J = 11.0, 7.6 Hz, 2H), 3.94–3.88 (m, 2H), 3.52 (t, J = 6.1 Hz, 2H), 3.19 (t, J = 6.9 Hz, 2H), 3.10 (t, J = 6.1 Hz, 2H), 2.82 (t, J= 7.8 Hz, 2H), 1.86 (s, 3H), 1.72 (p, J = 8.1 Hz, 2H), 1.58 (p, J = 7.1 Hz, 2H), 1.44 (q, J = 7.8 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₁H₃₇F₃N₅O₂⁺, 568.2894; found, 568.2876.

N-(5-(2-((7-(3-Cyanobenzyl)-5-oxo-1,2,6,7,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-4(5H)-yl)-methyl)-5-(trifluoromethyl)phenyl)pentyl)acetamide (6).

Compound 6 was synthesized in a similar way to 5, except for using compound 2 as the starting material. White solid, 25 mg, yield: 92%. ¹H NMR (400 MHz, MeOD) δ 7.85 (s, 1H), 7.76 (t, J = 8.3 Hz, 2H), 7.59 (t, J = 7.8 Hz, 1H), 7.55 (s, 1H), 7.45 (d, J = 8.1 Hz, 1H), 7.18 (d, J = 8.1 Hz, 1H), 5.24 (s, 2H), 4.49 (dd, J = 11.2, 7.4 Hz, 2H), 4.14–4.04 (m, 4H), 3.60 (s, 2H), 3.18 (q, J = 6.4 Hz, 4H), 2.96 (t, J = 5.9 Hz, 2H), 2.82 (t, J = 7.8 Hz, 2H), 1.86 (s, 3H), 1.72 (p, J = 7.6 Hz, 2H), 1.58 (p, J = 7.1 Hz, 2H), 1.44 (q, J = 7.8 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₂H₃₆F₃N₆O₂⁺, 593.2846; found, 593.2882.

3-((4-(2-(5-(Dimethylamino)pentyl)-4-(trifluoromethyl)-benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (7).

To a solution of compound 2 (10 mg, 0.018 mmol, 1.0 equiv) in 1 mL methanol was added 37 wt % formaldehyde aqueous solution (10 μL) and NaBH₃CN (2.3 mg, 0.036 mmol, 2.0 equiv). The reaction mixture was stirred at room temperature overnight and purified by prepared HPLC. White solid, 13 mg, yield: 90%. ¹H NMR (400 MHz, MeOD) δ 7.95 (s, 1H), 7.86 (t, J = 8.8 Hz, 2H), 7.67 (t, J = 7.5 Hz, 1H), 7.58 (s, 1H), 7.44 (d, J = 8.1 Hz, 1H), 7.18 (d, J = 8.1 Hz, 1H), 5.25 (s, 2H), 4.56–4.45 (m, 4H), 4.15–4.05 (m, 2H), 3.93 (s, 2H), 3.56 (t, J = 6.1 Hz, 2H), 3.19–3.09 (m, 4H), 2.88 (s, 6H), 2.84 (t, J = 7.9 Hz, 2H), 1.79 (h, J = 7.4 Hz, 4H), 1.52 (p, J= 7.8 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₂H₃₈F₃N₆O⁺, 579.3054; found, 579.3028.

2-((5-(2-((7-(3-Cyanobenzyl)-5-oxo-1,2,6,7,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-4(5H)-yl)-methyl)-5-(trifluoromethyl)phenyl)pentyl)(methyl)amino)-N-methylacetamide (8).

To a solution of 2 (25 mg, 0.045 mmol, 1.0 equiv) in 1 mL methanol were added ethyl 2-oxoacetate (50% toluene) (9 μL, 0.045 mmol, 1.0 equiv), 10 μL AcOH and NaBH₃CN (11 mg, 0.090 mmol, 2.0 equiv). The reaction mixture was stirred at room temperature overnight and purified by prepared HPLC.

The above product was dissolved into 1 mL methanol, 0.2 mL saturated NaHCO₃ aqueous solution and Boc₂O (21 μL, 0.090 mmol, 2.0 equiv) were added to the solution. The reaction mixture was stirred at room temperature for 2 h. The reaction mixture was diluted with 5 mL DCM and washed with brine, the organic phase was dried over anhydrous Na₂SO₄. Concentrate and the residue was used without purification. LC-MS m/z: [M + H]⁺; calcd for C₃₉H₄₈F₃N₆O₅⁺, 737.36 found, 736.65.

The above compound was dissolved into 1 mL methanol and 0.20 mL 1 M NaOH aqueous solution, the reaction mixture was stirred at room temperature overnight. The reaction mixture was purified by prepared HPLC.

To a solution of above product into 1 mL DMF were added 2 M methylamine THF solution (63 μL, 0.126 mmol, 3.0 equiv), EDCI (24 mg, 0.126 mmol, 3.0 equiv), HOAT (11.4 mg, 0.084 mmol, 2.0 equiv) and 50 μL NMM. The reaction mixture was stirred at room temperature overnight. Dilute the solution with 10 mL DCM and wash the solution with brine for 3 times. Dried over anhydrous Na₂SO₄ and concentrated. The residue was dissolved into 1 mL DCM/TFA = 1/1, stirred at room temperature for 30 min, concentrated and the residue was purified by prepared HPLC, compound 22 was obtained, yield: 40%. LC-MS m/z: [M + H]⁺ calcd for C₃₃H₃₉F₃N₇O₂⁺ 622.31; found, 622.31.

To a solution of compound 22 (5 mg, 0.008 mmol, 1.0 equiv) in 1 mL methanol were added 37 wt % formaldehyde aqueous solution (5 μL) and NaBH₃CN (1.0 mg, 0.016 mmol, 2.0 equiv). The reaction mixture was stirred at room temperature overnight and purified by prepared HPLC. White solid, 4.5 mg, yield: 65%. ¹H NMR (400 MHz, MeOD) δ 7.87 (s, 1H), 7.78 (t, J = 7.9 Hz, 2H), 7.61 (t, J = 6.9 Hz, 1H), 7.58 (s, 1H), 7.46 (d, J = 8.3 Hz, 1H), 7.18 (d, J = 8.1 Hz, 1H), 5.24 (s, 2H), 4.49 (t, J = 9.4 Hz, 2H), 4.20 (s, 2H), 4.07 (t, J = 9.4 Hz, 2H), 3.92 (s, 2H), 3.67 (s, 2H), 3.27 (d, J = 6.1 Hz, 2H), 3.18 (s, 2H), 3.00 (t, J = 6.1 Hz, 2H), 2.91 (s, 3H), 2.84 (t, J = 8.1 Hz, 2H), 2.80 (s, 3H), 1.80 (hept, J = 7.9 Hz, 4H), 1.53 (q, J = 7.6 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₄H₄₁F₃N₇O₂⁺, 636.3268; found, 636.3250.

tert-Butyl (2-(2-Cyano-5-(trifluoromethyl)phenoxy)ethyl)-carbamate (23a).

To a solution of 2-hydroxy-4-(trifluoromethyl)-benzonitrile (2.0 g, 0.011 mol, 1.0 equiv) in 25 mL CH₃CN were added K₂CO₃ (3.0 g, 0.022 mol, 2.0 equiv) and tert-butyl (2-bromoethyl)carbamate (3.1 g, 0.016 mol, 1.5 equiv). The reaction mixture was stirred at 80 °C overnight. M = 1.2 g, w% = 34%. ¹H NMR (400 MHz, Chloroform-d) δ 7.70 (d, J = 8.0 Hz, 1H), 7.30 (d, J = 8.1 Hz, 1H), 7.19 (s, 1H), 5.05 (s, 1H), 4.20 (t, J = 5.4 Hz, 2H), 3.63 (q, J = 5.6 Hz, 2H), 1.45 (s, 9H).

2-(2,2-Dimethoxyethoxy)-4-(trifluoromethyl)benzonitrile (23b).

To a solution of 2-hydroxy-4-(trifluoromethyl)benzonitrile (0.5 g, 2.67 mmol, 1.0 equiv) in 10 mL DMF were added K₂CO₃ (0.37 g, 2.67 mmol, 1.0 equiv), NaI (0.40 g, 2.67 mmol, 1.0 equiv) and 2-bromo-1,1-dimethoxyethane (0.90 g, 5.34 mmol, 2.0 equiv). The reaction mixture was stirred at 100 °C overnight. m = 0.34 g, w% = 46%. ¹H NMR (400 MHz, Chloroform-d) δ 7.69 (d, J = 6.6 Hz, 1H), 7.30 (d, J = 7.9 Hz, 1H), 7.22 (s, 1H), 4.75 (t, J = 5.2 Hz, 1H), 4.16 (t, J = 5.3 Hz, 2H), 3.53 (s, 6H).

3-((4-(2-(2-Aminoethoxy)-4-(trifluoromethyl)benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (9).

Compound 9 was synthesized in a similar way as compound 1, except for using 23a (0.30 g, 0.91 mol) and 21d (0.25 g, 0.91 mol) as the starting materials. White solid, 0.37 g, yield: 54%. ¹H NMR (400 MHz, MeOD) δ 7.94 (s, 1H), 7.85 (dd, J = 11.8, 7.9 Hz, 2H), 7.66 (t, J = 7.8 Hz, 1H), 7.32 (s, 1H), 7.29–7.22 (m, 2H), 5.30 (s, 2H), 4.53–4.38 (m, 6H), 4.07 (dd, J = 11.0, 7.6 Hz, 2H), 3.90 (s, 2H), 3.55–3.44 (m, 4H), 3.09 (t, J = 6.1 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₇H₂₈F₃N₆O₂⁺, 525.2220; found, 525.2250.

3-((4-(2-(2,2-Dimethoxyethoxy)-4-(trifluoromethyl)benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]-pyrimidin-7(6H)-yl)methyl)benzonitrile (26).

Compound 26 was synthesized in similar way to compound 1 without deprotection at TFA/DCM at the final step, except for using 23b (0.33 g, 1.20 mol) and 21d (0.36 g, 1.32 mol) as the starting materials. Yellow solid, 0.36 g, yield: 53%. ¹H NMR (400 MHz, MeOD) δ 7.75 (s, 1H), 7.69 (d, J = 7.8 Hz, 1H), 7.65 (d, J = 7.7 Hz, 1H), 7.53 (t, J = 7.7 Hz, 1H), 7.25 (s, 1H), 7.21 (s, 2H), 5.09 (s, 2H), 4.78 (t, J = 5.0 Hz, 1H), 4.19 (t, J = 9.1 Hz, 2H), 4.14 (d, J = 5.0 Hz, 2H), 3.89 (t, J = 9.3 Hz, 2H), 3.77 (s, 2H), 3.49–3.39 (m, 8H), 3.24 (s, 2H), 2.80 (t, J = 5.7 Hz, 2H), 2.70 (t, J = 5.8 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₉H₃₁F₃N₅O₄⁺, 570.2323; found, 570.59.

N-(2-(2-((7-(3-Cyanobenzyl)-5-oxo-1,2,6,7,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-4(5H)-yl)-methyl)-5-(trifluoromethyl)phenoxy)ethyl)acetamide (10).

Compound 10 was synthesized in similar way as 5, except for using compound 9 as the starting material. White solid, yield: 92%. ¹H NMR (400 MHz, MeOD) δ 7.88 (s, 1H), 7.79 (t, J = 8.0 Hz, 2H), 7.62 (t, J = 7.8 Hz, 1H), 7.34 (d, J = 8.0 Hz, 1H), 7.30 (s, 1H), 7.23 (d, J = 8.0 Hz, 1H), 5.15 (s, 2H), 4.46 (dd, J = 11.2, 7.4 Hz, 2H), 4.22 (d, J = 5.5 Hz, 4H), 4.09 (dd, J = 11.0, 7.6 Hz, 2H), 3.71 (s, 2H), 3.61 (t, J = 5.3 Hz, 2H), 3.29 (t, J = 5.1 Hz, 2H), 2.99 (t, J = 6.0 Hz, 2H), 1.94 (s, 3H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₉H₃₀F₃N₆O₃⁺, 567.2326; found, 567.2316.

3-((4-(2-(2-(Dimethylamino)ethoxy)-4-(trifluoromethyl)-benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (11).

To a solution of compound 9 (25 mg, 0.033 mmol, 1.0 equiv) was added 37 wt % formaldehyde aqueous solution (25 μL) and NaBH₃CN (6.2 mg, 0.10 mmol, 3.0 equiv). The reaction mixture was stirred at room temperature overnight and purified by prepared HPLC. White solid, 23 mg, yield: 92%. ¹H NMR (400 MHz, MeOD) δ 7.91 (s, 1H), 7.86–7.78 (m, 2H), 7.64 (t, J = 7.8 Hz, 1H), 7.37 (s, 1H), 7.27 (q,J = 7.4 Hz, 2H), 5.28 (s, 2H), 4.57–4.51 (m, 2H), 4.47 (dd, J = 11.3, 7.5 Hz, 2H), 4.32 (s, 2H), 4.07 (dd, J = 10.5, 8.3 Hz, 2H), 3.80 (s, 2H), 3.73–3.67 (m, 2H), 3.40 (t, J = 6.0 Hz, 2H), 3.08–3.02 (m, 8H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₂₉H₃₂F₃N₆O₂⁺, 553.2533; found, 553.2539.

3-((4-(2-(2-Morpholinoethoxy)-4-(trifluoromethyl)benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo[1,2-a]pyrido[3,4-e]-pyrimidin-7(6H)-yl)methyl)benzonitrile (12).

To a solution of 26 (50 mg, 0.088 mmol, 1.0 equiv) in 4 mL acetone/H₂O = 2/1 was added TsOH·H₂O (50 mg, 0.26 mmol, 3.0 equiv). The reaction mixture was refluxed for 4 h, the reaction was completed. Diluted with 20 mL DCM and washed with saturated NaHCO₃ aqueous solution. Dried over anhydrous Na₂SO₄, and concentrated to yield the aldehyde intermediate, and used into next step without purification.

To a solution of above crude product (0.044 mmol) in 1 mL methanol was added 10 μL morpholine, 10 μL acetic acid and 10 mg NaBH₃CN. The reaction mixture was stirred at room temperature overnight. The reaction mixture was purified by prepared HPLC. White solid, 25 mg, yield: 69%. ¹H NMR (400 MHz, MeOD) δ 7.88 (s, 1H), 7.79 (t, J = 8.1 Hz, 2H), 7.62 (t, J = 7.8 Hz, 1H), 7.36 (s, 1H), 7.27 (q, J = 8.3 Hz, 2H), 5.25 (s, 2H), 4.60–4.53 (m, 2H), 4.47 (dd, J = 11.2, 7.6 Hz, 2H), 4.22 (s, 2H), 4.08 (dd, J = 11.1, 7.6 Hz, 2H), 4.02–3.87 (m, 4H), 3.77–3.71 (m, 2H), 3.68 (s, 2H), 3.57–3.40 (m, 4H), 3.28 (d, J = 5.8 Hz, 2H), 3.00 (t, J = 5.9 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₁H₃₄F₃N₆O₃⁺, 595.2639; found, 595.2625.

3-((4-(2-(2-(4-Methylpiperazin-1-yl)ethoxy)-4-(trifluoromethyl)benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo-[1,2-a]pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (13).

Compound 13 was synthesized in similar way to compound 12, except for using 1-methylpiperazine. White solid, 22 mg, yield: 53%. ¹H NMR (400 MHz, MeOD) δ 7.92 (s, 1H), 7.83 (t, J = 8.3 Hz, 2H), 7.65 (t, J = 7.8 Hz, 1H), 7.34 (s, 1H), 7.26–7.22 (m, 2H), 5.21 (s, 2H), 4.48 (dd, J = 11.3, 7.5 Hz, 2H), 4.38 (t, J = 5.3 Hz, 2H), 4.35 (s, 2H), 4.07 (dd, J = 11.1, 7.6 Hz, 2H), 3.82 (s, 2H), 3.47–3.33 (m, 6H), 3.24–2.97 (m, 8H), 2.90 (s, 3H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₂H₃₇F₃N₇O₂⁺, 608.2955; found, 608.2952.

3-((5-Oxo-4-(2-(2-(piperidin-1-yl)ethoxy)-4-(trifluoromethyl)benzyl)-1,2,4,5,8,9-hexahydroimidazo[1,2-a]-pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (14).

Compound 14 was synthesized in similar way to compound 12, except for using piperidine. White solid, 21 mg, yield: 58%. ¹H NMR (400 MHz, MeOD) δ 7.91 (s, 1H), 7.86–7.78 (m, 2H), 7.64 (t, J = 7.8 Hz, 1H), 7.34 (s, 1H), 7.30–7.22 (m, 2H), 5.25 (s, 2H), 4.57–4.51 (m, 2H), 4.51–4.44 (m, 2H), 4.32 (s, 2H), 4.08 (dd, J = 11.1, 7.7 Hz, 2H), 3.78 (s, 2H), 3.75–3.60 (m, 4H), 3.40 (t, J = 6.2 Hz, 2H), 3.13 (br, 2H), 3.05 (t, J = 6.1 Hz, 2H), 2.00–1.50 (m, 6H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₂H₃₆F₃N₆O₂⁺, 593.2846; found, 593.2801.

3-((5-Oxo-4-(2-(2-(piperazin-1-yl)ethoxy)-4-(trifluoromethyl)benzyl)-1,2,4,5,8,9-hexahydroimidazo[1,2-a]-pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (15).

Compound 15 was synthesized in a similar way to compound 12, except for using tert-butyl piperazine-1-carboxylate, and then the above product was dissolved into 1 mL TFA/DCM = 1/1, the reaction mixture was stirred at room temperature for 30 min and concentrated, the residue was purified by prepared HPLC. White solid, 24 mg, yield: 58%. ¹H NMR (400 MHz, MeOD) δ 7.93 (s, 1H), 7.87–7.80 (m, 2H), 7.66 (t, J = 7.8 Hz, 1H), 7.34 (s, 1H), 7.27–7.20 (m, 2H), 5.23 (s, 2H), 4.53–4.43 (m, 4H), 4.41 (s, 2H), 4.07 (dd, J = 11.2, 7.6 Hz, 2H), 3.88 (s, 2H), 3.52–3.42 (m, 6H), 3.40 (t, J = 5.1 Hz, 2H), 3.38–3.32 (m, 4H), 3.09 (t, J = 6.1 Hz, 2H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₁H₃₅F₃N₇O₂⁺, 594.2799; found, 594.2790.

3-((4-(2-(2-(4-Acetylpiperazin-1-yl)ethoxy)-4-(trifluoromethyl)benzyl)-5-oxo-1,2,4,5,8,9-hexahydroimidazo-[1,2-a]pyrido[3,4-e]pyrimidin-7(6H)-yl)methyl)benzonitrile (16).

Compound 16 was synthesized in similar way as compound 5, except for using compound 15 as the starting material. White solid, 16 mg, yield: 84%. ¹H NMR (400 MHz, MeOD) δ 7.88 (s, 1H), 7.83–7.75 (m, 2H), 7.62 (t, J = 7.8 Hz, 1H), 7.36 (s, 1H), 7.31–7.22 (m, 2H), 5.25 (s, 2H), 4.62–4.54 (m, 2H), 4.47 (dd, J = 11.2, 7.6 Hz, 2H), 4.22 (s, 2H), 4.07 (dd, J = 11.1, 7.6 Hz, 2H), 4.00–3.80 (m, 4H), 3.78–3.72 (m, 2H), 3.69 (s, 2H), 3.61–3.43 (m, 4H), 3.28 (d, J = 5.9 Hz, 2H), 3.00 (t, J = 6.1 Hz, 2H), 2.15 (s, 3H). HRMS (ESI-TOF) m/z: [M + H]⁺ calcd for C₃₃H₃₇F₃N₇O₃⁺, 636.2904; found, 636.2902.

Cell Culture.

All cells were cultured in DMEM/F12 + GlutaMax medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Pen Strep).

Western Blot Assay.

Western blotting was performed using ClpP (Abcam, ab124822), Actin (Millipore Sigma, MAB1501R), DAP13 (Abcam, ab192617), Caspase-3 (CST, #9661).

Thermal Shift Assay in Lysate.

Compounds were added to 10 μg of MDA-MB-468 lysates and allowed to incubate at 23 °C for 15 min. Lysates were then heated at 67 °C for 3 min in a thermocycler to denature the proteins. An unheated control and a no-drug control were included. Denatured proteins were allowed to precipitate at 23 °C for 3 min and then spun at 14,000 rpm for 20 min at 4 °C. The soluble fraction was collected and analyzed via Western blotting using anti-ClpP antibodies.

Thermal Shift Assay in Cells.

MDA-MB-468 cells were seeded a day prior to the experiment at a density of 1 million cells per well in a 6-well plate. On the day of the experiment, appropriate amounts of compounds were added to the cells and incubated for 30 min at 37 °C. Cells were washed with PBS to remove drug-containing media and collected by scraping in PBS with proteasome inhibitors. Suspended cells were heated at 67 °C for 3 min and lysed by freeze–thaw. Lysate was spun at 14,000 rpm for 20 min at 4 °C, and the soluble fraction was analyzed via Western blotting using anti-ClpP antibodies.

Expression and Purification of hClpP.

Human ClpP (hClpP) was expressed and purified as described previously.^34,65 Wild-type hClpP (without mitochondria sequences) with N-terminal, 2x(His6-thrombin)-SUMO-TEV site were cloned into pET-28(+) expression vector and expressed in E. coli BL21(DE3). Single colony of transformed BL21(DE3) cells grown overnight in LB supplemented with kanamycin (100 μg/mL) at 37 °C with shaking. The overnight culture was diluted 1:100 in LB supplemented with kanamycin (100 μg/mL) and grown to an optical density (OD600) of ∼0.6–0.8. Protein expression was induced for 4 h at 37 °C with 1 mM isopropyl-1-thio-B-D-galactopyranoside (IPTG). The cells were harvested by centrifugation and frozen at −80 °C as pellets until further purification. The cells were thawed and disrupted in 25 mM Tris-HCl, 500 mM NaCl, 10 mM imidazole and 10% glycerol (pH 7.5) supplemented with 1× EDTA-free Roche protease inhibitor cocktail. The hClpP was purified with HisPur Cobalt Resin (ThermoScientific, Waltham, MA) and eluted with 25 mM Tris-HCl, 0.1 M NaCl, 10% glycerol and 400 mM imidazole (pH 7.5). The samples were desalted and analyzed by SDS-PAGE.

Differential Scanning Fluorimetry.

Differential scanning fluorimetry (DSF or thermal shift assay) for purified human ClpP was performed on QuantStudio 3 real-time PCR system (Applied Biosystems, ThermoScientific, Waltham, MA). Purified human ClpP proteins were diluted to a final concentration of 0.5 mg/mL in 25 mM Tris-HCl buffer (pH 7.5). DSF dye, 8X SYPRO Orange (ThermoScientific, Waltham, MA) was prepared by diluting according manufacturer’s protocol. The reactions were kept at 25 °C for 2 min and then heated from 25 to 95 °C with a rate of 0.05 °C/s. The change in the fluorescence intensities of SYPRO Orange was monitored as a function of the temperature and analyzed by Protein Thermal Shift software 1.3. Each reaction was performed in 4-replicates and GraphPad Prism 8 was used in the analysis of the difference of melting temperature (ΔT_M) of different condition compared to DMSO-treated samples from the data.

Flow-Cytometry-Based Apoptosis Detection.

MDA-MB-468 cells treated with DMSO or 1 μM of ONC212, compound 9 or doxorubicin for 72h at 37 °C in 5% CO₂ incubator. Treated cells were washed twice with FACS buffer (50 mM PBS, 2 mM EDTA, and 2% FBS) and stained with fixable viability dye eFluor 780 (ThermoScientific, Waltham, MA) for 30 min at 4 °C. After two washes, cells were resuspended in FACS buffer for flow cytometry with the Attune NxT flow cytometer (ThermoScientific, Waltham, MA). The data was analyzed with the Attune NxT and FlowJo analysis software

DAP13 Degradation Assay.

Cells were treated with drug for 24 h and lysates analyzed by Western blotting using an anti- DAP13 antibody.

Viability Assay.

Cells were seeded at a density of 5,000 cells per well in a 96-well plate 24 h before drug exposure, in quadruplicate. The media was removed by pipetting, and drug-containing media was added. After 72 h of incubation, cell viability was assessed using incubation with AlamarBlue dye for 3 h. The resulting colorimetric change was measured using a Varioskan Lux plate reader. Viability data were normalized to a vehicle control.

Generation of Resistant Cells.

MDA-MB-468 cells were treated with 50 nM ONC212 continuously and media changed every 3 days to remove dead cells. Resistant cells were then allowed to expand in drug-free media and re-exposed to ONC212. This selection process was repeated over several cycles until the growth rate of the treated cells exceeded the death rate.

Seahorse Assay.

Mitochondrial respiratory function was assessed using a Seahorse XF24 extracellular flux analyzer (Seahorse Biosciences, North Billerica, MA). MDA-MB-468 cells (250,000 cells per well) were plated in Seahorse plates and treated with either ONC212 or 9 to 50 nM for 12, 24, and 48 h. One hour before measurement, the drug-containing media was replaced with Agilent Seahorse XF DMEM media (1 mM pyruvate, 2 mM glutamine, 10 mM glucose, pH 7.4). The assay was conducted using the Seahorse XF Cell Mito Stress Test Kit. The final concentrations of inhibitors used were 1 μM oligomycin, 1 μM FCCP (as an uncoupler), and 0.5 μM complex III inhibitor antimycin A. Each plate, along with its cartridge, was loaded into the XF analyzer, and the oxygen consumption rate (OCR) was measured under basal conditions and after sequential addition of oligomycin, FCCP, and rotenone.

Mouse Pharmacokinetic Study.

Three male Swiss Albino mice were administered intraperitoneally with a solution formulation of compound 9 at a 50 mg/kg dose. The following formulation vehicle was used in the study: 100% v/v normal Saline. Blood samples (approximately 60 μL) were collected from the three test mice at 0.5, 2, and 8 h. Plasma was harvested by centrifugation of blood and stored at −70 °C ± 10 °C until analysis. Plasma samples were quantified by the fit-for-purpose LC-MS/MS method. Compound concentrations in plasma at each time point are average values from 3 test mice. Error bars represent ± SEM. All animal experiments were performed following the protocols evaluated and approved by the Institutional Animal Care and Use Committee (IACUC) at the Icahn School of Medicine at Mount Sinai (Ethics Approval Number: IACCR202500000089).

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.5c01315.

Assessment of the ClpP protein level in breast cancer cell lines (Figure S1); representative SYPRO Orange melting curve plots for DSF assay analysis (Figure S2); flow cytometry analysis (Figure S3); mouse pharmacokinetic study of compound 9 (Figure S4); ¹H NMR, ¹³C NMR and LC-MS spectra of compound 9 (PDF)

Molecular formula strings for all compounds (CSV)

ABBREVIATIONS USED

AcOH

acetic acid

Ac₂O

acetic anhydride

AML

acute myeloid leukemia

ATP

adenosine triphosphate

Boc

tert-butyloxycarbonyl

CETSA

Cellular Thermal Shift Assay

ClpP

serine caseinolytic protease P

ClpX

caseinolytic protease X

DCM

dichloromethane

DMF

N,N-dimethylformamide

DAP13

13 kDa differentiation-associated protein

DSF

differential scanning fluorimetry

EDCI

1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide hydrochloride

HPLC

high-performance liquid chromatography

HOAT

1-hydroxy-7-azabenzotriazole

MeOH

methanol

NMM

N-methylmorpholine

OCR

oxygen consumption rate

OXPHOS

oxidative phosphorylation

ROS

reactive oxygen species

SAR

structure–activity relationship

rt

room temperature

S_NAr

nucleophilic aromatic substitution

TEA

triethylamine

TFA

trifluoracetic acid

THF

tetrahydrofuran

TLC

thin-layer chromatography

TRAIL

TNF-related apoptosis-inducing ligand

TSA

thermal shift assay

TsOH·H₂O

4-toluenesulfonic acid monohydrate

UHC

unheated control

WB

Western blotting