CHCHD2 mutant mice display mitochondrial protein accumulation and disrupted energy metabolism

Abstract

Mutations in the mitochondrial cristae protein CHCHD2 lead to a late-onset autosomal dominant form of Parkinson's disease (PD) which closely resembles idiopathic PD, providing the opportunity to gain new insights into the mechanisms of mitochondrial dysfunction contributing to PD. To begin to address this, we used CRISPR genome-editing to generate CHCHD2 T61I point mutant mice. CHCHD2 T61I mice had normal viability, and had only subtle motor deficits with no signs of premature dopaminergic (DA) neuron degeneration. Nonetheless, CHCHD2 T61I mice exhibited robust molecular changes in the brain including increased CHCHD2 insolubility, accumulation of CHCHD2 protein preferentially in the substantia nigra (SN), and elevated levels of α-synuclein. Metabolic analyses revealed an increase in glucose metabolism through glycolysis relative to the TCA cycle with increased respiratory exchange ratio, and immune-electron microscopy revelated disrupted mitochondria in DA neurons. Moreover, spatial genomics revealed decreased expression of mitochondrial complex I and III respiratory chain proteins, while proteomics revealed increased respiratory chain and other mitochondrial protein-protein interactions. As such, the CHCHD2 T61I point-mutation mice exhibit robust mitochondrial disruption and a consequent metabolic shift towards glycolysis. These findings thus establish CHCHD2 T61I mice as a new model for mitochondrial-based PD, and implicate disrupted respiratory chain function as a likely causative driver.

Figure 1:. CHCHD2 T61I mice exhibit normal survival and subtle motor deficits.

(A) Insertion of CHCHD2 T61I point mutation by CRISPR/Cas9 gene editing and homologous recombination. (B) CHCHD2 T61I mouse SNP genotyping by qPCR. 3 clusters were autogenerated by allelic discrimination in the analysis software, including WT (black), T61I HET (pink) and T61I HOM (red). (C) WT peptide (MAQMATTAAGVAVGSAVGHTLGHAITGGFSGGGSAEPAKPDITYQEPQGAQL) and mutant peptide (MAQMAITAAGVAVGSAVGHTLGHAITGGF) were searched in brain lysates of 7 mice of different genotypes (1 WT, 2 T61I HET, 2 T61I HOM and 2 CHCHD2 KO). WT peptide was found in WT and HET mice, while mutant was found in HET and HOM mice. Neither peptide was found in KO mice. (D) CHCHD2 T61I mice showed similar body weights compared with WT through 13 months of age. Data represent mean ± SEM. N = 5–9 mice/genotype/gender. (E) T61I mice showed normal survival compared with WT in Kaplan-Meier survival curve through 16 months of age. Left, females (n = 30 WT, 18 HET and 17 HOM); right, males (n = 21 WT, 20 HET, 34 HOM). (F) Total movements of T61I mice were normal in open field test. (G) Male T61I HOM mice showed a significant decrease in latency to fall in inverted grid hang (WT vs HOM P < 0.05). (H) Female T61I HOM mice showed a trend toward a decrease in latency to fall in the accelerating rotarod test at (WT vs HOM P = 0.1). (I) Female HOM showed significantly lower time spent in the open arms (WT vs HOM P < 0.001). In all behavioral studies (F-I), data represent mean ± SEM. N = 16 WT, 16 HET and 13 HOM (female), or 14 WT, 18 HET and 17 HOM (male). *P < 0.05, ***P < 0.001 by linear mixed effects analysis.

Figure 2:. CHCHD2 T61I point mutation disrupts DA autoreceptor function without DA neuron degeneration.

Recordings were made blind to genotype. (A-C) SNc dopamine neurons from CHCHD2 T61I and control mice fired spontaneously in acute brain slices in cell attached recording configuration. (A) Example action potential activity from WT and HOM mice. (B) Spontaneous firing rates in DA neurons from CHCHD2 T61I and control mice recorded in cell attached mode were similar. (C) Regularity of spontaneous firing was evaluated by measuring the coefficient of variation of 100 consecutive interspike intervals. ISI-CV was similar across genotypes. (D) Example pacemaker firing in DA neurons from CHCHD2 T61I and control mice in whole cell current clamp (I = 0 pA) configuration. (E, F) Spontaneous firing rates in whole cell current clamp configuration were similar in DA neurons from CHCHD2 T61I and control mice. (G, H) Spontaneous AP waveforms were recorded in whole cell configuration and quantified from each genotype including AP peak (G) and threshold (H). (I) The D2R selective agonist quinpirole (1 uM) was bath applied during whole cell recordings of SNc DA neurons. Quinpirole caused hyperpolarization in most WT neurons but not T61I mutant neurons. Left, time course averages of responses; right, summary of individual responses. AP, action potential. N = 5 WT and 4 HOM male mice at 23 months; each circle is a neuron. Data represent mean ± SEM or with kernel density estimations. For all data parametric assumptions were tested to choose between t-test (parametric) or permutation (non-parametric) analysis. *P < 0.05. (J) HPLC of dorsal CPu punches from flash-frozen brain tissues at 5 or 16 months showed lower DA level in T61I HOM mice. HVA, homovanillic acid; DOPAC, 3,4-dihydroxyphenylacetic acid. Data represent mean ± SEM. N = 9–10 mice/genotype at 5 months and 3 mice/genotype at 16 months. *P < 0.05 by two-way ANOVA with Sidak post hoc test. (K) Representative images of TH immunoreactivity in striatal sections of WT, T61I HET and HOM mice at 16 months. DA neuron projection areas in dorsal and ventral striatum are indicated with dotted lines. CPu, caudate putamen; NAc, nucleus accumbens; OT, olfactory tubercule. Quantification for TH immunoreactivity showed no difference in WT, HET and HOM by two-way ANOVA with Tukey’s post hoc test. Data represent mean ± SEM. N = 3 animals/group, 4 sections/animal. Scale bar indicates 300 μm. (L) Representative images of TH immunoreactivity in midbrain sections of WT, HET, and HOM mice at 16 months. DA neuron projection areas in the midbrain are indicated with dotted lines. SN, substantia nigra; VTA, ventral tegmental area. Stereology estimating the number of DA neurons showed no difference between each group by two-way ANOVA with Tukey’s post hoc test. N = 5–7 animals/group. Data represent mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, Scale bar indicates 300 μm.

Figure 3:. T61I point mutation preferentially increases accumulation of CHCHD2 and CHCHD10 in mitochondria in SN DA neurons.

(A) Representative fluorescence images of CHCHD2 (red), CHCHD10 (green), and merge in SN (top) and VTA (bottom) subregions in the midbrain of CHCHD2 T61I mice at 16 months. Punctae was semi-automatically annotated by Cellprofiler. CHCHD2 denoted in red and CHCHD10 in green. The region of cell body is indicated with dotted lines, based on TH signal. (B) Quantification of CHCHD2 and CHCHD10 intensity, area of CHCHD2 and D10 punctae, and area of co-localization between CHCHCHD2 and D10 punctae in SN (top panel) and VTA (bottom), normalized to the averaged cell area of each group. (C) Representative fluorescence images of mitochondria (PDH, red), CHCHD2 (green), and merge in SN (top) and VTA (bottom). PDH denoted in red, total CHCHD2 in green and CHCHD2 punctae in blue. The region of cell body is indicated with dotted lines, based on TH signal. (D) Quantification of cell area, mitochondrial content, CHCHD2 punctae area in the mitochondrial, and total CHCHD2 level in the mitochondria in SN (top panel) and VTA (bottom), normalized to averaged cell area of each group. In all fluorescence staining (A-D), n = 6–8 randomly selected TH-positive cells from 6 fields in each region in each mouse, 4 mice/genotype. Examiner was blinded to mouse genotype. Data represent mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA with Tukey’s post hoc test, Scale bars indicate 1μm. (E) Impact of T61I mutation on mitochondrial morphology. DA neurons were labeled with primary TH antibody followed by secondary gold particle-conjugated antibody. N = 71 – 189 mitochondria/mouse, 3 mice/genotype. Data represent mean ± SEM. **P < 0.01 by t test with Welch’s correction. Images taken at 19,000X magnification, and scale bar indicates 1 μm.

Figure 4:. CHCHD2 T61I mice show increases in α-synuclein and phosphorylated synuclein level in the midbrain.

(A) Transcriptomic levels of CHCHD2 and CHCHD10 in DA neurons of SND (dorsal tier of substantia nigra, orange) and SNV (ventral, blue) measured with GeoMx WTA show significant downregulation of CHCHD2 in the SNV of ILB and ePD cases. Data represent mean ± SEM. Brown-Forsythe and Welch ANOVA tests were applied for data covariates with age, sex, and post-mortem delay. ***P < 0.001. NC, normalized counts; Ct, control; ILBD, incidental Lewy body disease; ePD, PD with early Braak stage pathology; lPD, PD with late Braak stage pathology. (B) Moderate positive linear associations were revealed between expression levels of either CHCHD2 or CHCHD10 with SNCA in the SN by Spearman correlation. (C) CHCHD2 localization in control and PD SN DA neurons and (E) orthogonal views of the relative location of CHCHD2 in different staged αSyn aggregations. (D) CHCHD10 localization in control and PD SN DA neurons, and (F) orthogonal views of the relative location of CHCHD10 in different staged αSyn aggregations. Scale bars represent 50 μm in (C, D), and 20 μm in (E, F). (G) Representative images and quantification of TH (green) and either total α-synuclein (Syn1) or phosphorylated synuclein, P-syn (PS129, red) immunoreactivity in SN and VTA at 63X magnification in midbrain sections of WT, HET and HOM mice at 16 months. Syn1 and P-syn immunoreactivities increased in midbrain DA neurons of HOM mice by two-way ANOVA with Tukey’s post hoc test. Data represent mean ± SEM. *P < 0.05. N = 4 sections/mouse, 4 mice/genotype. Scale bar indicates 10 μm. (H) Representative images and quantifications of Syn1 (top) and P-syn (PS129, bottom), immunoreactivity in striatal sections of mice at 16 months. Syn1 and P-Syn areas in dorsal and ventral striatum are indicated with dotted lines. Significant increases in both Syn1 and p-Syn immunoreactivity were observed in HOM mice compared with WT by two-way ANOVA with Tukey’s post hoc test. Data represent mean ± SEM. *P < 0.05, **P < 0.01. N = 4 sections/mouse, 4 mice/genotype. Scale bar indicates 300 μm.

Figure 5:. CHCHD2 mutant mouse brains exhibit a metabolic shift toward glycolysis and PPP.

(A) Female CHCHD2 T61I mice at 19 months received tail vein injections of [U-¹³C]glucose for 30 minutes before brains were harvested. Brain metabolite extract was analyzed by an ion chromatography-mass spectrometry (IC-MS) detector. Heat map shows relative total levels of metabolomics in the L column determined by non-targeted metabolomics, and fractional contribution of 13C-glucose to each metabolite by targeted metabolomics on the R. Each square shows the log value of a metabolite level or fractional contribution in a mouse normalized to the mean of the WT group. (B) Bar graph of relative amounts of metabolites along glycolysis, PPP and TCA cycle. CHCHD2 T61I heterozygotes and homozygotes showed a dose-dependent increase in distal glycolytic, PPP and TCA cycle metabolites compared with WT mice. (C) Bar graph of fractional contribution of metabolites. CHCHD2 T61I homozygous mice showed significantly higher labeling percentages in glycolytic and PPP metabolites compared with WT mice. G6P, glucose 6-phosphate; F6P, fructose 6-phosphate; F1,6BP, fructose 1,6-bisphosphate; 1,3BPG, 1,3-bisphosphoglycerate; 3PG, 3-phosphoglycerate; PEP, phosphoenolpyruvate; Pyr, pyruvate; Lac, lactate; R5P, ribose 5-phosphate; S7P, sedoheptulose 7-phosphate; Cit, citrate; iCit, isocitrate; a-KG, α-ketoglutarate; Succ, succinate; Fum, fumarate; Mal, malate. Data represent mean ± SEM. N = 5 WT, 5 HET and 4 HOM T61I female mice at 19 months. *P < 0.05, **P < 0.01, ***P < 0.001 by two-way ANOVA with Tukey’s post hoc test. (D) Respiratory exchange ratio (CO₂ production to O₂ consumption) of CHCHD2 T61I mice at 5 months was recorded for 5 days. T61I HOM mice showed significant increase in RER in the dark phase. N = 9 WT, 9 HET and 4 HOM. Data represents mean ± SEM in the bar graph. *P < 0.05 by two-way ANOVA with Dunnett’s post hoc test.

Fig. 6.. Co-fractionation mass spectrometry analysis of protein assemblies across various CHCHD2 genotypes.

(A) Biochemical fractionation of whole cell lysates (WCL) and mitochondrial (mt) lysates from mouse brain with varying CHCHD2 genotypes was performed using high-performance size exclusion chromatography (SEC-HPLC). Protein fractions were subjected to tryptic digestion and analyzed by high performance liquid chromatography-tandem mass spectrometry (LC-MS/MS) to measure peptide spectral counts. The schematic details the computational scoring pipeline from our previous work in MACP. This pipeline includes calculating protein similarity (correlation) metrics for each CHCHD2 genotype, training integrative classifiers with machine learning using the CORUM mouse complex database as a training standard, and scoring co-fractionation data to predict high-confidence interactions. These predicted interactions were clustered to define co-complex membership, and analyzed for pathobiological relevance. (B) Hierarchical clustering displaying changes in protein co-fractionation intensity profiles between mt and WCL as measured by LC-MS/MS. (C) Venn diagram illustrating the distribution of mouse protein interactions within the mt and WCL across different CHCHD2 genotypes. (D) Enrichment analysis of interacting proteins in mt mouse brain lysates from CHCHD2 heterozygous (HET) and homozygous (HOM) samples involved in annotated metabolic processes. (E) Enrichment analysis (FDR ≤ 5e-02) of protein assemblies in mt extracts from CHCHD2 point mutants, not found in WT, and also present in both whole cell and mitochondrial extracts of CHCHD2 WT but missing in mutants. These assemblies are linked to diseases such as Leigh syndrome, Parkinson’s disease, amyotrophic lateral sclerosis, and Alzheimer’s disease.

Fig. 7.. Differential analysis of predicted protein complexes in mouse brain WCL and mt lysates across different CHCHD2 genotypes.

(A, C) Complexes in WCLs (A) and mt lysates (C) exhibit differential enrichment between CHCHD2 mutants and WT. (B, D) Displays of representative complexes (indicated by dotted circles with labels) showing increased interactions in WCLs (B) and mt lysates (D) of either CHCHD2 mutants or WT. The left panels in B and D show the proportion of interactions within each node, detected across various CHCHD2 genotypes or WT, with node size representing the number of interacting proteins predicted in each complex. The right panels in B and D provide detailed views of the physically associated proteins within each cluster. (E) Complex I activity was not changed in T61I HOM mice at 5 and 11 months. N = 3 WT and 4 HOM mice at 5 months and 4 WT and 5 HOM mice at 11 months. Data represents mean ± SEM. ns, not significant by t test with Welch’s correction. (F) Summary schema of the study.