Zc3h10 is a novel mitochondrial regulator

Abstract

Mitochondria are the energy-generating hubs of the cell. In spite of considerable advances, our understanding of the factors that regulate the molecular circuits that govern mitochondrial function remains incomplete. Using a genome-wide functional screen, we identify the poorly characterized protein Zinc finger CCCH-type containing 10 (Zc3h10) as regulator of mitochondrial physiology. We show that Zc3h10 is upregulated during physiological mitochondriogenesis as it occurs during the differentiation of myoblasts into myotubes. Zc3h10 overexpression boosts mitochondrial function and promotes myoblast differentiation, while the depletion of Zc3h10 results in impaired myoblast differentiation, mitochondrial dysfunction, reduced expression of electron transport chain (ETC) subunits, and blunted TCA cycle flux. Notably, we have identified a loss-of-function mutation of Zc3h10 in humans (Tyr105 to Cys105) that is associated with increased body mass index, fat mass, fasting glucose, and triglycerides. Isolated peripheral blood mononuclear cells from individuals homozygotic for Cys105 display reduced oxygen consumption rate, diminished expression of some ETC subunits, and decreased levels of some TCA cycle metabolites, which all together derive in mitochondrial dysfunction. Taken together, our study identifies Zc3h10 as a novel mitochondrial regulator.

Keywords: Zc3h10; functional screens; metabolism; mitochondria.

Figure 1. High‐throughput screen to identify a new regulator of mitochondrial function

1. A genome‐wide primary screen testing the ability of individual cDNAs to control Tfam promoter activity was performed in HEK293 cells. Top left inset shows performance of controls. Hits were filtered in a secondary screen measuring mitochondrial density and function in HEK293 cells. Validated positive hits were further evaluated for their ability to increase mitochondrial respiration and mtDNA levels in C2C12 myoblasts.

2. Scatter plot of primary screen data in HEK293 cells. Hits were selected based on their inhibitory or stimulatory effect on Tfam promoter activity relative to known negative (Mybbp1a) and positive (Pgc‐1α) Tfam regulators.

3. Scatter plot of secondary screen results in HEK293 cells. 126 positive and 32 negative hits controlled mitochondrial function to a similar or greater extent than the Pgc‐1α and Mybbp1a controls.

4. Scatter plot of tertiary screen data in C2C12 myoblasts. Confirmed positive hits (21) from the secondary screen were transiently overexpressed in C2C12 myoblasts, and mtDNA levels and basal cell respiration were measured. Zc3h10 surfaced as a prime candidate for follow up studies.

5. C2C12 myoblasts were co‐transfected with indicated plasmids, and luciferase activity was evaluated 48 h later. n = 3. Statistical analysis was performed by one‐way ANOVA with Dunnett's post‐test vs. Tfam‐Luc control. ***P < 0.001.

6. Tfam mRNA levels 24 h after Pgc‐1α and Zc3h10 overexpression in C2C12 myoblasts. n = 4. Statistical analysis was performed by one‐way ANOVA with Dunnett's post‐test vs. control. **P < 0.01.

7. Western blot analysis of indicated Oxphos subunits and Tfam 24 h after Pgc‐1α and Zc3h10 overexpression in C2C12 myoblasts.

8. Basal, uncoupled (Oligo), and maximal uncoupled (CCCP) respiration in control, Pgc‐1α, and Zc3h10 overexpressing C2C12 myoblasts. n = 3. Statistical analysis was performed by one‐way ANOVA with Dunnett's post‐test. **P < 0.01, ***P < 0.001 vs. basal control; ^$$ P < 0.01 vs. oligomycin control; ^# P < 0.05 vs. CCCP control.

9. Cytosolic, mitochondrial, and total ATP levels 24 h after Pgc‐1α and Zc3h10 transfection in C2C12 myoblasts. n = 5. Statistical analysis was performed by one‐way ANOVA with Dunnett's post‐test. ***P < 0.001 vs. mitochondrial control; ^$ P < 0.05, ^$$ P < 0.01 vs. total control.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure EV1. Bioinformatics analysis to prioritize screen hits and Zc3h10 expression in mouse tissues and human quadriceps

1. Tissue set enrichment analysis (TSEA) of secondary screen positive hits. Significantly enriched bar (muscle) is highlighted in red. P‐values were calculated by Fisher's exact test with Benjamini‐Hochberg correction.

2. Scatter plot of expression levels of secondary screen hits in C2C12 myotubes and mouse skeletal muscle according to the BioGPS database.

3. Protein classes of secondary screen hits according to gene ontology classification.

4. Table representing basal respiration and mtDNA fold change after overexpression of 21 hits in the tertiary screen in C2C12 myoblasts.

5. Zc3h10 expression levels in different mouse tissues and human quadriceps. Absolute quantification of mRNA concentration was obtained by comparing sample threshold cycles to a standard curve set up with Zc3h10 cDNA at known concentrations. Blue and red bars represent Zc3h10 RNA levels in adult human male and female quadriceps, respectively. Numbers within bars indicate qRT‐PCR threshold cycles.

6. Zc3h10 expression levels in different human tissues. Data have been downloaded from Genotype‐Tissue Expression (GTEx) Portal on July 27, 2017. Box plots are showing median and 25^th and 75^th percentiles. n ≥ 70. Further information about statistics can be found at https://www.gtexportal.org/home/tissueSummaryPage#sampleCountsPerTissue.

7. Schematic representation of the experimental plan for myoblasts experiments and Western blot to detect Zc3h10 in control and Zc3h10‐transfected cells.

8. Relative quantification of ROS production and apoptosis after Pgc‐1α or Zc3h10 overexpression in myoblasts. Cells were exposed to H₂O₂ and CCCP as positive controls for ROS production and apoptosis, respectively. ROS, n = 3; apoptosis, n = 3. Statistical analysis was performed by one‐way ANOVA with Dunnet post‐test. **P < 0.01 vs. control.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure 2. Zc3h10 is a nuclear protein upregulated during C2C12 differentiation

1. Zc3h10 mRNA levels, Zc3h10, MyHC, and Tfam protein levels during C2C12 differentiation. Hsp90 and β‐actin were used as loading controls. mRNA/n = 3. Statistical analysis was performed by one‐way ANOVA with Dunnett's post‐test. *P < 0.05, **P < 0.01 vs. day −1.

2. Western blot analysis of indicated Oxphos subunits and Tfam during C2C12 differentiation.

3. Western blot analysis to detect Zc3h10, α‐tubulin, histone H3, and β‐actin at different time points of C2C12 differentiation in both cytoplasm and nucleus. α‐tubulin, histone H3, and β‐actin were used as cytoplasmic, nuclear, and loading controls, respectively.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure EV2. Effect of Zc3h10 overexpression in myoblasts and myotubes

1. mtDNA levels, Tfam and myogenesis markers MyoD, MyHC7, TnnI1, and Mef2c RNA expression levels during C2C12 differentiation. n = 3. Statistical analysis was performed by one‐way ANOVA with Dunnet post‐test. *P < 0.01, **P < 0.01, ***P < 0.001 vs. day −1.

2. Scheme of the experimental plan followed up for Zc3h10 overexpression with adenoviral vectors in proliferating myoblasts.

3. Zc3h10 expression levels in control and Flag‐Zc3h10 myotubes at 24, 48, and 72 h from differentiation. n = 3. Statistical analysis was performed by Student's t‐test. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control.

4. Western blot to detect Flag‐tagged Zc3h10 at 24, 48, and 72 h after adenoviral overexpression in differentiated myotubes.

5. mRNA expression levels of mitochondria and myogenesis markers in Zc3h10 overexpressing C2C12 myotubes 48 h after induction of differentiation. n = 3. Statistical analysis was performed by Student's t‐test. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control.

6. C2C12 myotube fusion index 72 h after induction of differentiation with relative quantification and C2C12 myoblast proliferation rate of control and flag‐Zc3h10 cells. Fusion index, n = 4. Student's t‐test, ***P < 0.001 vs. control. Proliferation, n = 3. Statistical analysis was performed by two‐way ANOVA with Sidak post‐test vs. control.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure 3. Zc3h10 overexpression boosts mitochondrial function and density in myotubes

1. Basal oxygen consumption and mtDNA content in control and Flag‐Zc3h10‐infected C2C12 myotubes 24, 48, and 72 h from differentiation induction. n = 3. Statistical analysis was performed by two‐way ANOVA with Sidak post‐test. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control.

2. Basal, uncoupled, and maximal uncoupled respiration and complex I, II, and IV activity evaluation 48 h from differentiation induction in control and Flag‐Zc3h10 myotubes. n = 3. Statistical analysis was performed by Student's t‐test. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure EV3. Zc3h10 downregulation impairs mitochondrial function and C2C12 differentiation

1. A

   Scheme of the experimental plan followed up for Zc3h10 downregulation with adenoviral vectors in proliferating myoblasts.

2. B, C

   Zc3h10 mRNA (B) and protein expression levels (C) with relative quantification in C2C12 myoblasts expressing scramble or Zc3h10 shRNA. n = 3. Statistical analysis was performed by Student's t‐test; *P < 0.05, ***P < 0.001 vs. scramble.

3. D

   Western blot of selected Oxphos subunits in isolated mitochondria of C2C12 myoblasts expressing scramble or Zc3h10 shRNA.

4. E

   mtDNA content in C2C12 myoblasts. n = 3.

5. F, G

   Zc3h10 mRNA (F) and protein levels (G) in differentiating C2C12 myotubes at different time points in cells expressing scramble or Zc3h10 shRNA. n = 3. Statistical analysis was performed by Student's t‐test. ***P < 0.001 vs. scramble.

6. H

   mtDNA content in C2C12 myotubes at different time points in cells expressing scramble or Zc3h10 shRNA. n = 3.

7. I

   Gene set enrichment analysis (GSEA) of C2C12 myotubes expressing scramble or Zc3h10 shRNA 48 h after induction of differentiation. n = 5.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure 4. Impaired mitochondrial function in Zc3h10‐silenced C2C12 myoblasts and myotubes

1. A

   Western blot analysis to assess expression levels of indicated OXPHOS subunits in C2C12 myoblasts expressing scramble or Zc3h10 shRNA.

2. B

   Basal, uncoupled, and maximal uncoupled respiration in C2C12 myoblasts 48 h after infection with scramble or Zc3h10 shRNA. n = 3. Statistical analysis was performed by Student's t‐test; ***P < 0.001 vs. scramble.

3. C, D

   Basal, maximal coupled, uncoupled, and maximal uncoupled respiration (C), and complex I, II, and IV activities (D) in mitochondria isolated from C2C12 myoblasts expressing scramble or Zc3h10 shRNA. n = 3. Statistical analysis was performed by Student's t‐test. *P < 0.05, **P < 0.01, ***P < 0.001 vs. scramble.

4. E–G

   Tfam (E) and indicated Oxphos subunit (F) protein levels and (G) basal oxygen consumption in scramble and shZc3h10 myotubes 24, 48, and 72 h after induction of differentiation. n = 3. Statistical analysis was performed by two‐way ANOVA with Sidak post‐test. *P < 0.05, **P < 0.01 vs. scramble.

5. H–J

   Basal, uncoupled, and maximal uncoupled respiration (H); complex I, II, and IV activities (I); and cytosolic, mitochondrial, and total ATP levels (J) in scramble and Zc3h10 shRNA C2C12 myoblasts. (H) and (I) n = 3, (J) n = 4. Statistical analysis was performed by Student's t‐test. ***P < 0.001 vs. scramble.

6. K

   Myotube fusion index 72 h after induction of differentiation with relative quantification, and myoblast proliferation rate of scramble and shZc3h10 expressing C2C12 myotubes. Fusion index, n = 4; proliferation, n = 3. Statistical analysis was performed by Student's t‐test. **P < 0.01 vs. scramble.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure EV4. System approaches to analyze the effects of Zc3h10 knockdown

1. Western blot analysis to detect Zc3h10 in SILAC light and heavy C2C12 myoblasts expressing scramble or Zc3h10 shRNA, respectively.

2. Volcano plot of 3,712 proteins detected by SILAC analysis. Dark blue and red dots indicate downregulated and upregulated proteins, respectively, in shZc3h10 compared to scramble C2C12 myoblasts (P < 0.001). Blue and orange dots indicate downregulated and upregulated proteins, respectively (P < 0.01). Light blue and yellow dots indicate downregulated and upregulated proteins, respectively (P < 0.05). n = 1. P‐values by means of significance B were calculated by Perseus software [50,52].

3. Gene ontology analysis of significantly downregulated protein clusters (P < 0.05) in myoblasts with Zc3h10 knockdown. P‐values are indicated as EASE score: modified Fisher exact P‐value.

4. SILAC Oxphos subunit protein expression levels fold change in C2C12 myoblasts with Zc3h10 knockdown relative to scramble. Student's t‐test vs. scramble. *P < 0.05. n = 1.

5. SILAC TCA cycle enzymes protein expression levels fold change in C2C12 myoblasts with Zc3h10 knockdown relative to scramble. Student's t‐test vs. scramble. n = 1.

Data information: n indicates the number of biological replicates. P‐values were calculated from the indicated n independent experiments according to 44.

Figure 5. Zc3h10 depletion alters the ETC and rewires energy metabolism

1. A

   Volcano plot of differentially regulated metabolites in myoblasts 48 h after infection with scramble or Zc3h10 shRNA. Analysis was performed by targeted LC‐MS/MS. n = 5. Statistical analysis was performed by Student's t‐test vs. scramble. Statistical significance is indicated in the plot.

2. B–G

   Mole percent enrichment (MPE) and mass isotopomer distribution (MID) of indicated metabolites labeled with (B, E) [U‐¹³C₆] glucose (red), (C, F) [U‐¹³C₁₆] palmitate (blue), and (D, G) [U‐¹³C₅] glutamine (green) in scramble and ShZc3h10 myoblasts. Light green bars indicate reductive glutamine metabolism. n = 3. Statistical analysis was performed by Student's t‐test. *P < 0.05, **P < 0.01, ***P < 0.001 vs. scramble.

3. H

   Schematic representation of nutrient carbons flow through the TCA cycle. Large and small circles represent increased and decreased metabolites, respectively. Carbons derived from [U‐¹³C₆] glucose are indicated as red circles, [U‐¹³C₁₆] palmitate as blue circles, and [U‐¹³C₅] glutamine as green and light green circles indicating oxidative and reductive glutamine metabolism, respectively.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure 6. Tyr105 to Cys105 point mutation in human ZC3H10 is associated with altered metabolic phenotype

1. Genome browser representation of Hg38 ZC3H10 locus (upper part) and sequence of the tyrosine 105 (red) neighborhood in ZC3H10 coding sequence (lower part).

2. Body mass index (BMI), waist circumference, total fat mass, waist/hip ratio, fasting glucose and triglyceride (TG) in Tyr105 (black), Tyr105Cys (red), and Cys105 (blue) subjects. Tyr105, n = 697. Tyr105Cys, n = 893. 105Cys, n = 4. Data are expressed as mean ± SD for BMI, waist circumference, total fat mass, and waist/hip ratio. For fasting glucose and TG, data are expressed as median ± interquartile range (Tyr105 and Tyr105 Cys) and mean ± SD for Cys105. Statistical analysis was performed by Kruskal–Wallis (Dunn's post hoc analysis). *P < 0.05 vs. Tyr105 and ^# P < 0.05 vs. Tyr105Cys.

Figure EV5. Location and conservation of a missense mutation in human ZC3H10

1. PolyPhen‐2 prediction of functional effects of rs61732294 SNPs. Black bar indicates damage probability in a range from 0 (green) to 1 (red).

2. Alignment of ZC3H10 loci (upper part) and the tyrosine 105 (red) neighborhood in ˜30 vertebrates. Green bars indicate the degree of sequence conservation among species.

3. Western blot to detect Zc3h10 in C2C12 myoblasts transfected with vector control, wild‐type (Tyr105), or mutant (Cys105) Zc3h10 expression constructs. n = 3. One‐way ANOVA with Dunnett's post‐test; **P < 0.01, ***P < 0.001 vs. control (white bar). n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.

Figure 7. Loss‐of‐function mutation in Zc3h10 is associated with mitochondrial dysfunction in humans

1. Myoblasts were co‐transfected with control, wild‐type mouse Zc3h10, or Cys105 mouse Zc3h10 mutant and the indicated reporter constructs. Luciferase activity was measured 48 h later. n = 4. Data are expressed as mean ± SD. Statistical analysis was performed by one‐way ANOVA with Dunnett's post‐test. ***P < 0.001 vs. Tfam‐Luc control and ^### P < 0.001 vs. Tfam‐Luc Tyr105 wild‐type Zc3h10.

2. Basal, maximal coupled, uncoupled, maximal uncoupled respiration, and ETC complex I, II, and IV activities in C2C12 myoblasts transfected with wild‐type or Cys105 Zc3h10. n = 3. Statistical analysis was performed by one‐way ANOVA with Tukey's post‐test. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control and ^# P < 0.05, ^## P < 0.01, ^### P < 0.001 vs. wild‐type Zc3h10.

3. Basal, maximal coupled, uncoupled, and maximal uncoupled respiration, and complex I, II, and IV activities in isolated human peripheral blood mononuclear cells (PBMCs) of controls and Zc3h10 Cys105 homozygotes.

4. Tfam, Atp5a1, Uqcrc2, Sdhb, and Ndufb8 mRNA levels in PBMCs isolated from Tyr105 and Cys105 homozygote subjects.

5. Western blot showing Atp5a1, Uqcrc2, Mt‐CO1, Sdhb, and Ndufb8 protein expression levels in PBMCs isolated from indicated subjects.

6. Heatmap of targeted metabolomics in PBMCs isolated from Tyr105 and Cys105 homozygote subjects. Blue and orange squares show increased and decreased metabolites, respectively.

Data information: n indicates the number of biological replicates. Data are expressed as mean ± SD, and P‐values were calculated from the indicated n independent experiments according to 44.