Objective: The mitochondrial pyruvate carrier (MPC) has emerged as a promising drug target for metabolic disorders, including non-alcoholic steatohepatitis and diabetes, metabolically dependent cancers and neurodegenerative diseases. A range of structurally diverse small molecule inhibitors have been proposed but the nature of their interaction with MPC is not understood. Moreover, the composition of the functional human MPC is still debated. The goal of this study was to characterize the human MPC protein in vitro, to understand the chemical features that determine binding of structurally diverse inhibitors and to develop novel higher affinity ones.

Results: We have determined that the functional unit of human MPC is a hetero-dimer. We have compared all different classes of MPC inhibitors to find that three closely arranged hydrogen bond acceptors followed by an aromatic ring are shared characteristics of all inhibitors and represent the minimal requirement for high potency. We also demonstrate that high affinity binding is not attributed to covalent bond formation with MPC cysteines, as previously proposed. Following the basic pharmacophore properties, we identify 14 new inhibitors of MPC, one outperforming compound UK5099 by tenfold.  Two of them are the commonly prescribed drugs entacapone and nitrofurantoin, suggesting an off-target mechanism associated with their adverse effects. 

Conclusion: This work defines the composition of human MPC and the essential MPC inhibitor characteristics. In combination with the functional assays we describe, this new understanding will accelerate the development of clinically relevant MPC modulators.

1. Mass determination by SEC-MALLS

SEC-MALLS analysis was performed with a Superdex 200 10/300 GL column on an ÄKTA Explorer coupled in-line with a light scattering detector (Dawn HELEOSII, Wyatt Technologies) and a refractometer (Optilab T-rREX, Wyatt Technologies). The MPC complex was applied onto the Superdex 200 10/300 GL column equilibrated with 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.005% (w/v) LMNG, 0.005 mg/ml TOCL at 0.3 ml/min. All data were recorded and analysed with ASTRA 6.03 (Wyatt Technologies). Molecular weight calculations were performed using the protein-conjugate method with the dn/dc value for protein of 0.185 ml/g and dn/dc value for LMNG-TOCL of 0.1675 ml/g. The contributions of each protein to the overall protein-detergent-lipid complex were determined from the extinction coefficients εA₂₈₀, derived from the amino-acid sequence using the ProtParam tool on the ExPaSy server.

2. Thermostability shift analysis

The assessment of ligand binding was performed via shifts in protein thermostability upon binding, using thermal denaturation on a rotary quantitative PCR (qPCR) instrument (Qiagen Rotor-Gene Q 2plex HRM, Venlo, Netherlands). In this method, cysteine residues, buried within the protein structure, become solvent exposed during denaturation in a temperature ramp and react with 7-diethylamino-3-(4-maleimidophenyl)-4-methylcoumarin (CPM) to form fluorescent-adducts. A CPM working solution was prepared by diluting the CPM stock (5 mg/ml in dimethyl sulfoxide) 50-fold into assay buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% (w/v) LMNG, 0.1 mg/ml TOCL) and incubated for 10 min at room temperature. In each analysis, performed in duplicates or triplicates, as indicated, 3 μg of purified MPC was diluted into assay buffer containing the desired concentration of inhibitors, incubated for 10 min on ice and then 5 μL of the CPM working solution were added. Samples were incubated on ice for a further 10 min and then subjected to a temperature ramp of 5.6 °C/min. The fluorescence increase was monitored with the HRM channel of the instrument (excitation at 440–480 nm, emission at 505–515 nm). Unfolding profiles were analyzed with the Rotor-Gene Q software 2.3 and the peaks of their derivatives were used to determine the apparent melting temperature.

The second method to assess ligand binding was dye-free nano differential scanning fluorimetry (nano-DSF), which monitors fluorescence changes due to altered environments of tryptophan and tyrosine residues during unfolding. Protein samples in buffer containing 20 mM Tris-HCl pH 7.4, 150 mM NaCl, 0.1% (w/v) LMNG, 0.1 mg/mL TOCL, in the presence or absence of the indicated concentrations of small molecule inhibitors, were loaded into capillary tubes, and a temperature ramp of 5 °C/min was applied. The intrinsic fluorescence was measured using the NanoTemper Prometheus NT.48 (NanoTemper, München, Germany).

3. Pyruvate transport assays 

Transport was initiated at room temperature by diluting proteoliposomes 200-fold into external buffer, 20 mM MES, pH 6.4, 50 mM NaCl, containing 50 μM [¹⁴C]-pyruvate (500,000 dpm, Perkin Elmer, MA, USA). The reaction (0–30 s) was terminated by rapid dilution into 8 volumes of ice-cold buffer (20 mM Tris-HCl, pH 8.0, 50 mM NaCl), internal to proteoliposomes, followed by rapid filtration through cellulose nitrate 0.45 μm filters (Millipore, Gillingham, UK) and washing with an additional 8 volumes of buffer. The filters were dissolved in Ultima Gold scintillation liquid (Perkin Elmer, MA, USA) and the radioactivity was counted with a Perkin Elmer Tri-Carb 2800 RT liquid scintillation counter. For inhibition of pyruvate transport, various concentrations of compounds were added to the liposomes simultaneously with 50 μM radioactive substrate. The data analysis was performed with non-linear regression fittings using GraphPad Prism 7.0d ([Inhibitor] vs response, variable slope). The specific uptake rates were calculated based on the amount of protein used in reconstitutions, as estimated from bicinchoninic acid assay. The biological repeats represent independent proteoliposome preparations using protein from independent purifications.