HSP60 dysfunction results in cell proliferation arrest and progressing perturbations in key mitochondrial functions. (A) Schematic representation depicting the mechanism in tetracycline-inducible HEK293 cell lines expressing wild-type HSP60 (WT) or the HSP60 dominant-negative ATPase-deficient mutant HSP60 (D423A). Inducible expressed subunits integrate with endogenously expressed subunits to form heteromeric seven-subunit HSP60 complexes. (B) Cell proliferation of WT and D423A cell lines, uninduced and induced with 50 ng/mL tetracycline for different time intervals. Represented in fold change compared to initial cell counts (statistical comparison between WT and D423A induced cell lines). (C–D) The abundance of D423A and endogenous WT HSPD1 transcript (C) and HSP60 protein (D) levels, quantified by RNA sequencing and mass spectrometry, respectively, at 0 h (uninduced), 48 h, and 72 h induced D423A cells. (E) Quantification of mitochondrial mass by MitoTracker green fluorescence in induced WT and D423A cell lines (n ≥ 3) at 0 h (uninduced), 48 h, and 72 h. (F) Percentage of cells with intact mitochondrial membrane potential in induced WT and D423A cells by JC-1 fluorescence (n ≥ 3) at 0 h (uninduced), 48 h, and 72 h induced (G) Mitochondrial superoxide levels in induced WT and D423A cells, based on MitoSOX fluorescence (n ≥ 3) at 0 h (uninduced), 48 h, and 72 h. (H) Bioenergetic profiles of induced WT and D423A cells in basal conditions, analyzed with Seahorse XFe96 analyzer at 0 h (uninduced), 48 h, and 72 h. (I) Total ATP production distinguishing the proportion of mitochondrial and glycolytic ATP production rates, analyzed using the Seahorse XFe96 analyzer at 0 h (uninduced), 48 h, and 72 h. Blue continuous line compares the total ATP production rate, and blue dashed lines compare the mitochondrial ATP production rate. Graphs represent mean ± SD (n = 3); statistical significance was determined with multiple unpaired t-tests with Holm-Sidak correction between WT and D423A induced cell lines unless otherwise stated (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). Abbreviations: ECAR, extracellular acidification rate (represents glycolysis); FCCP, Carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone; h, hours; HEK, human embryonic kidney; HSP, heat-shock protein; OCR, oxygen consumption rate (represents oxidative phosphorylation); SD, standard deviation; TET, tetracycline; tTA, reverse tetracycline transactivator; TRE, tetracycline-responsive promoter element; WT, wild-type.

HSP60 dysfunction severely regulates the mitochondrial proteome, but not the transcriptome. (A–B) Venn diagrams representing the number of quantified and differentially expressed genes in transcriptomics and proteomics. D423A cells at 0 h were compared to 48 h (A) or 72 h induced D423A cells (B). The pairing of quantified genes of transcriptomics and proteomics was performed through Ensembl gene IDs. (C-F) Volcano plots showing the comparison of induced D423A cells at 48 h and 72 h with 0 h for quantified transcriptome (C, E), and proteome (D, F). Cut-off values are indicated with dashed lines. As indicated in Figure 2 C, Mitochondrial proteins based on MitoCarta3.0 (MC3.0; [ 2 ]) are highlighted with red points. (G, H) Comparison plot of log2fold changes of Proteomics versus Transcriptomics at 48 h (G) and 72 h (H). Cut-off values are indicated with dashed lines. Mitochondrial proteins are highlighted in red. Genes significantly up- or down-regulated at both transcript and protein levels are labeled. (I-K) Violin plots distinguishing proteins from major mitochondrial pathways (I), mitochondrial sub-compartments (J), and HSP60 interactors (K), Induced D423A cells at 0 h in comparison to 48 h (light red) or 72 h induced D423A cells (dark red), respectively. The violin plots reflect all quantitated proteins in the respective pathways, compartments, and groups; violin widths are kept constant in the plots. Cut off values for transcriptomics: Benjamini-Hochberg adjusted p-value <0.05 and |log2FC|>1. Cut off values for proteomics: p-value <0.05 and |log2FC|>0.26 Abbreviations: DE, differentially expressed; FC, fold change; IMS, intermembrane space; MIM, mitochondrial inner membrane; MOM, mitochondrial outer membrane (See also the legend of Figure 1 ).

Proteomics analysis of patient fibroblasts carrying HSP60 disease-associated variants (A–B) Volcano plots showing the quantified proteome of human dermal fibroblasts of the Patient 1 (heterozygous HSP60-p.Leu47Val) (A) and of human dermal fibroblasts of the Patient 2 (homozygous HSP60-p.Asp29Gly) (B) compared to their respective human dermal fibroblast control cells (n = 3). P values were calculated using TDIST function in Excel. Cut-off values are indicated with dashed lines (p-value <0.05 and |log2FC|>0.26). Mitochondrial proteins based on MitoCarta3.0 (MC3.0; [ 2 ]) are highlighted with red points. (C) Violin plot distinguishing the mitochondrial sub-compartment protein Log2fold changes from dermal fibroblasts from Patient 1 (light red) and Patient 2 (dark red) in comparison to their respective controls. (D) Correlation of the protein Log2fold changes of Patient 1 (horizontal axis) with Patient 2 (vertical axis) in comparison to their respective controls. Points for significantly changed genes with p-values ≤ 0.05 are displayed as blue points. The correlation line is based on a linear mode; Pearson R and p values are given. (E) Functional categories of differentially expressed proteins in fibroblasts of Patient1 and Patient2 as compared to control fibroblasts illustrated by using Proteomaps [ 61 ] ( https://proteomaps.net ). Polygons correspond to KEGG pathway groups, and the sizes correlates are scaled according to the absolute values of average log(2) fold changes. (F) Enriched GO terms for significantly up-regulated fibroblast proteins in Patient 1 and Patient 2 is analyzed using the STRING database [ 62 ] ( https://string-db.org ). Proteomaps and Enrichment analyses were performed on False Discovery Rate (FDR)-corrected (FDR<0.05) values with |log2FC|>0.26.

Zebrafish Hsp60 knockout results in early larval developmental abnormalities. (A) Hsp60 protein levels of hspd1+/+, hspd1+/−, and hspd1−/− larvae at 3 dpf, and hspd1−/− larvae at 3, 5, and 7 dpf. Assessed by western blot (n = 6). (B) Lateral imaging of hspd1−/− larvae at 3, 5, and 7 dpf indicated developmental delay at 5 and 7 dpf, but hspd1+/− larvae showed no significant phenotypic differences compared to hspd1+/+ larvae. (C) Quantification of WT and KO hspd1 transcript levels in hspd1+/+ and hspd1−/− larvae at 5 dpf, quantified by RNA sequencing. (D–F) Phenotypic quantification of body length (D), eye size (E), and swim bladder inflation (F) in hspd1+/+ (n = 28), hspd1−/− (n = 31) larvae at 3, 5, and 7 dpf. (G–H) Volcano plots showing analysis of transcriptome (G), and proteome (H) changes between hspd1+/+ and hspd1−/− larvae. Cut-off values are indicated with dashed lines. Mitochondrial proteins based on MitoCarta3.0 (MC3.0; [ 2 ]) are highlighted with red points. (I) Comparison plot of log2fold changes of Proteomics versus Transcriptomics in hspd1−/− larvae. Cut-off values are indicated with dashed lines. Mitochondrial proteins are highlighted in red. Genes significantly up- or down-regulated at both transcript and protein level are labeled. (J) Violin plots distinguishing proteins from mitochondrial sub-compartments of hspd1−/− larvae in comparison to hspd1+/+ larvae. Statistical significance was determined with one-way ANOVA: ∗p < 0.05, ∗∗∗∗p < 0.0001. Abbreviations: dpf, days post-fertilization; KO, knockout; sgRNA, single guide RNAs; WT, wild type.

ISRmt is upregulated in D423A cells and hspd−/− larvae. (A–B) Enrichment analysis of differentially expressed proteins of D423A cells at 0 h in comparison to 72 h induced D423A cells (A) and in 5 dpf hspd1−/− larvae in comparison to hspd1+/+ larvae (B). Performed on EnrichR at gene name level for proteomics GO biological process and KEGG terms are included. The enriched terms are shown as a negative log10 of adjusted p-values. Adjusted p-values <0.05 were considered significant. (C–D) Volcano plots showing the transcriptomic regulation of mitochondrial integrated stress response (ISRmt) transcripts. ISRmt genes were taken from [ 27 ] and are displayed as blue points. (C) HEK D423A cells at 0 h in comparison to 72 h induction (left panel) and ISRmt transcript fold changes at 48 h and 72 h as compared to uninduced cells (right panel), (D) 5 dpf hspd1−/− larvae in comparison to hspd1+/+ larvae. (E–F) Amino acids and TCA cycle intermediate metabolites were analyzed by targeted metabolomics, and mean-centered abundance values were further normalized to total ion intensity. (E) In the D423A cells, significance was calculated against uninduced cells (mean ± SD, n = 3, multiple unpaired t-tests with Holm-Sidak correction). (F) 5 dpf hspd1−/− larvae in comparison to hspd1+/+ larvae (mean ± SD, n = 3, two-way ANOVA with Sidak's multiple comparisons correction). Abbreviations: GO, gene ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; See also Figure 1 , Figure 4 .

Zebrafish Hsp60 knockout larvae have altered lipid metabolism and oligodendrocyte counts (A–B) Enrichment analysis of differentially expressed genes of D423A cells at 0 h in comparison to 72 h (A) and in 5 dpf hspd1−/− larvae in comparison to hspd1+/+ larvae (B). Performed on EnrichR at gene name level for proteomics and transcriptomics, GO biological process and KEGG terms are included. The enriched terms are shown as a negative log10 of adjusted p-values. Adjusted p-values <0.05 were considered significant. (C–D) Volcano plots showing the transcriptomic regulation of genes involved in cholesterol synthesis pathways in D423A cells induced for 72 h (C) and in 5 dpf hspd1−/− larvae (D). Cholesterol synthesis genes and regulators are displayed as blue points; the labels for the regulating system genes are shaded in blue. (E) Whole-mount Oil Red O staining of 8 dpf zebrafish larvae shows lipid accumulation and altered lipid composition in the yolk of hspd1−/− larvae compared to hspd1+/+ larvae. (F) The length of Dendra2-CaaX(+) signal was measured caudal to yolk sac extension at 78 hpf, 4 dpf, 7 dpf, and 10 dpf. Larvae were generated from hspd1+/− to Tg(mbp:Dendra2-CaaX) mating. (G) The number of mature oligodendrocytes was quantified as caudal to yolk sac extension Dendra2-positive cells at 78 hpf, 4 dpf, and 7 dpf. Larvae were generated from hspd1+/− to Tg(mbp:Dendra2) mating. Abbreviations: TCA, tricarboxylic acid; GO, gene ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes. See also Figure 1 , Figure 2 , Figure 4 .

Overview of effects of HSP60 dysfunction (A) Schematic summary of consequences of HSP60 deficiency: (a) mitochondrial matrix protein levels are decreased, (b) leading to changes in cellular metabolism, (c) altering the pool of acetyl-CoA that alters the lipid metabolism (d), and finally activating ISRmt. Created with Biorender.com . (B) Simplified overview summarizing the changes in gene expression of metabolic enzymes and respiratory chain complexes at RNA and protein level affecting metabolic pathways and the respiratory chain as observed in HEK D423A cells induced for 72 h compared to uninduced cells.