A High-Throughput Screening Identifies MICU1 Targeting Compounds

Abstract

Mitochondrial Ca²⁺ uptake depends on the mitochondrial calcium uniporter (MCU) complex, a highly selective channel of the inner mitochondrial membrane (IMM). Here, we screen a library of 44,000 non-proprietary compounds for their ability to modulate mitochondrial Ca²⁺ uptake. Two of them, named MCU-i4 and MCU-i11, are confirmed to reliably decrease mitochondrial Ca²⁺ influx. Docking simulations reveal that these molecules directly bind a specific cleft in MICU1, a key element of the MCU complex that controls channel gating. Accordingly, in MICU1-silenced or deleted cells, the inhibitory effect of the two compounds is lost. Moreover, MCU-i4 and MCU-i11 fail to inhibit mitochondrial Ca²⁺ uptake in cells expressing a MICU1 mutated in the critical amino acids that forge the predicted binding cleft. Finally, these compounds are tested ex vivo, revealing a primary role for mitochondrial Ca²⁺ uptake in muscle growth. Overall, MCU-i4 and MCU-i11 represent leading molecules for the development of MICU1-targeting drugs.

Keywords: HTS; MCU; MICU1; active compounds; high-throughput screening; mitochondrial calcium uniporter; mitochondrial calcium uptake; molecular modeling; small molecules.

Graphical abstract

Figure 1

A High-Throughput Screening Identifies MCU Modulators (A) Screening flow chart to identify MCU modulators. (B) Activity distribution of 44K compounds in the primary mitochondrial Ca²⁺ uptake screen. The aequorin fluorescence signals for each compound were normalized to the activity of an inhibition control (−100% full inhibition) and of a neutral control (stimulation by histamine; 0% inhibition), which were included in each 384-well screening plate. (C) Summary table of primary and confirmation screen results. (D) Chemical structure and name of the two selected compounds.

Figure 2

MCU-i4 and MCU-i11 Negatively Modulate MCU Activity in a Small-Scale Validation Assay (A) Agonist-induced mitochondrial calcium uptake in intact HeLa cells. Cells were stimulated with 100 μM histamine and treated with 10 μM of each compound in 0.1% DMSO. Left: mean [Ca²⁺]mt peaks are shown. Right: representative traces of mitochondrial calcium uptake are shown. (B) Agonist-induced cytosolic calcium transients in intact HeLa cells. Cells were stimulated with 100 μM histamine and treated with 10 μM of each compound. Left: mean cytosolic calcium peaks are shown. Right: representative traces of cytosolic calcium transients are shown. (C) Mitochondrial calcium uptake in permeabilized HeLa cells. A buffer mimicking the cytosolic ionic composition (IB) supplemented with either 100 μM EGTA (IB/EGTA) or 3 μM [Ca²⁺] was used. HeLa cells were permeabilized by a 1-min perfusion with 100 μM digitonin (in IB/EGTA) during luminescence measurements. Left: mean mitochondrial [Ca²⁺] speed is shown. Right: representative traces of mitochondrial calcium uptake are shown. (D) Mitochondrial calcium uptake in intact HeLa cells upon long-term compound incubation. (E) Agonist-induced mitochondrial calcium uptake in intact MEFs. Cells were stimulated with 100 μM ATP and treated with 10 μM compound. (F) Agonist-induced mitochondrial calcium uptake in MDA-MB-231 cells upon treatment with 10 μM of each compound. Cells were stimulated with 100 μM ATP. (G) Agonist-induced mitochondrial calcium uptake in HEK293T cells upon treatment with 10 μM of each compound. Cells were stimulated with 100 μM ATP. (H) Δψ measurements in HeLa cells upon treatment with 10 μM of each compound at different time points. Data are presented as mean ± SD. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; one-way ANOVA except Kruskal-Wallis for (C). See also Figure S1.

Figure 3

MICU1 Is Required for MCU-i4 and MCU-i11 Activity (A) Agonist-induced mitochondrial calcium uptake peaks in intact HeLa cells in which MICU1 or MICU2 were transiently silenced. Cells were stimulated with 100 μM histamine and treated with 10 μM of either compound. (B) Mitochondrial calcium uptake peaks in control or shMICU1 HeLa stable clones. Cell were treated with 10 μM compound and stimulated with 100 μM histamine. (C) Agonist-induced mitochondrial calcium uptake peaks in intact Micu1^−/− MEFs upon expression of Micu1 or mock plasmids. Cells were treated with 10 μM compound and stimulated with 100 μM ATP. (D) Agonist-induced mitochondrial calcium uptake peaks in intact MICU1^−/− HEK293T cells upon expression of Micu1 or mock plasmids. Cells were treated with 10 μM compound and stimulated with 100 μM ATP. Data are presented as mean ± SD. ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; two-way ANOVA. See also Figure S2.

Figure 4

MCU-i4 and MCU-i11 Bind MICU1 (A and B) Predicted binding mode of MCU-i4 (A) and MCU-i11 (B). The pose is reported in the upper representation: the protein surface, as well as the ribbon, is reported in gray color. The ligands MCU-i4 and MCU-i11 are reported in magenta and cyan, respectively. The atomic distances of crucial interactions are reported in green. The three Micu1 residues selected for mutagenesis are shown in gray. Below each pose, the ligand interaction diagram reports the residues forming the pocket and the interactions observed. (C) Agonist-induced mitochondrial calcium uptake peaks in intact Micu1^−/− MEFs transfected with Micu1 or Micu1^{Q302A, Q306A, L443A}. Cells were stimulated with 100 μM ATP and treated with 10 μM of MCU-i4. (D) Agonist-induced mitochondrial calcium uptake peaks in intact Micu1^−/− MEFs transfected with Micu1 or Micu1^{Q302A, Q306A, L443A}. Cells were stimulated with 100 μM ATP and treated with 10 μM of MCU-i11. Data are presented as mean ± SD. ^∗∗p < 0.01; ^∗∗∗p < 0.001; two-way ANOVA. See also Figures S3 and S4 and Table S1.

Figure 5

MCU-i4 and MCU-i11 Reduce Mitochondrial Ca²⁺ Uptake in Skeletal Muscle Fibers and Impair Muscle Cell Growth (A) Representative scheme of the experimental design. (B) Resting mitochondrial Ca²⁺ levels of single isolated FDB fibers treated with either compound. (C) Mitochondrial Ca²⁺ uptake in single isolated FDB fibers transfected with 4mtGCaMP6f. Fibers were treated with 10 μM of MCU-i4 and MCU-i11, respectively. 6 min later, cells were stimulated with 40 mM caffeine. Left: mean mt Ca²⁺ peaks are shown. Right: representative traces of mitochondrial calcium uptake are shown. (D) Left: representative scheme of the experimental design. Right: measurements of myotubes width upon compound treatment are shown. Data are presented as mean ± SD. ^∗p ˂ 0.05; ^∗∗p < 0.01; ^∗∗∗p ˂ 0.001; one-way ANOVA.