Biallelic PPA2 Mutations Cause Sudden Unexpected Cardiac Arrest in Infancy

Abstract

Sudden unexpected death in infancy occurs in apparently healthy infants and remains largely unexplained despite thorough investigation. The vast majority of cases are sporadic. Here we report seven individuals from three families affected by sudden and unexpected cardiac arrest between 4 and 20 months of age. Whole-exome sequencing revealed compound heterozygous missense mutations in PPA2 in affected infants of each family. PPA2 encodes the mitochondrial pyrophosphatase, which hydrolyzes inorganic pyrophosphate into two phosphates. This is an essential activity for many biosynthetic reactions and for energy metabolism of the cell. We show that deletion of the orthologous gene in yeast (ppa2Δ) compromises cell viability due to the loss of mitochondria. Expression of wild-type human PPA2, but not PPA2 containing the mutations identified in affected individuals, preserves mitochondrial function in ppa2Δ yeast. Using a regulatable (doxycycline-repressible) gene expression system, we found that the pathogenic PPA2 mutations rapidly inactivate the mitochondrial energy transducing system and prevent the maintenance of a sufficient electrical potential across the inner membrane, which explains the subsequent disappearance of mitochondria from the mutant yeast cells. Altogether these data demonstrate that PPA2 is an essential gene in yeast and that biallelic mutations in PPA2 cause a mitochondrial disease leading to sudden cardiac arrest in infants.

Figure 1

PPA2 Mutations Identified in Affected Individuals from Three Families (A) Pedigrees of families F1, F2, and F3. PPA2 variants are reported according to GenBank: NM_176869.2. Individuals submitted to exome sequencing are indicated in red. (B) Cardiac muscle histology (HES staining) of affected individuals. Top: F2.II:1. Magnification ×10. Arrow shows focal infiltrate of inflammatory cells and cardiomyocyte necrosis. Bottom: F2.II:2. Magnification ×2.5. Arrow shows fibrosis and inflammatory cells. (C) Dimeric hPPA2 protein structure homology model. Views of the dimer drawn as a ribbon for each family (F1, F2, and F3). Each mutated residue is shown as a red sphere and is located on different monomers colored in green and pink. The phosphate groups (orange and red) and the four Mg²⁺ ions (purple) are drawn in ball and stick mode in the active sites. The homology model of dimeric hPPA2 is based on the 2.3 Å resolution structure of the inorganic pyrophosphatase from Schistosoma japonicum (PDB: 4QLZ) that shares 55% identity with hPPA2. It was built using Phyre2 and drawn using Pymol (DeLano WL, DeLano Scientific). (D) hPPA2 protein steady-state level is decreased in affected individuals’ fibroblasts compared to control subjects. Western blotting analysis of total protein extracts from cultured skin fibroblasts of three different age-matched control subjects (C1, C2, and C3) and affected individuals, using antibodies directed against PPA2 (Abcam cat# ab177935, 1/2,000), SOD2 (a mitochondrial matrix protein, Abcam cat#125702, 1/100,000), and β Actin (Santa Cruz Biotechnology cat# sc-81178; RRID: AB_2223230) as loading control.

Figure 2

Functional Complementation Assays and Primary Consequences of the PPA2 Mutations in Yeast (A–C) Functional complementation assay of a yeast ppa2Δ-null mutant. (A) Growth of ppa2Δ was tested either on glucose (YPD) or glycerol + ethanol medium (YPGE). The ppa2Δ strain containing a plasmid expressing wild-type yPPA2 from its own promoter was transformed with either wild-type hPPA2 (WT), hPPA2 mutant cDNAs, or the corresponding empty plasmid (−), under the control of the constitutive TPI promoter. After 5-FOA treatment (as previously described¹⁴) to eliminate the plasmid encoding yPPA2, cells expressing only wild-type (WT) or mutant hPPA2 were spotted onto YPD or YPGE plates. Drop dilution growth tests were performed at 1/10 dilution steps and yeast were incubated on YPD or YPGE plates for 2 days at 28°C. (B) DAPI staining of ppa2Δ cells transformed either with yPPA2 or the corresponding empty plasmid. (C) Western blots demonstrating expression of Por1 (mitochondrial outer membrane protein, monoclonal anti-porine [Invitrogen]), Aac2 (mitochondrial inner membrane ADP/ATP translocator, polyclonal anti-Aac2), Abf2 (mitochondrial matrix DNA-binding protein, polyclonal anti-Abf2), and the cytosolic protein Pgk1 (cytosolic phosphoglycerate kinase, monoclonal anti-Pgk1 [Invitrogen]) in total extracts of ppa2Δ cells transformed with empty plasmid (−) or wild-type (WT) or mutant hPPA2 cDNA constructs. Protein extraction and western blotting were performed as previously described. (D–F) Bioenergetic investigations. To overcome the problem of mitochondrial DNA (mtDNA) instability in ppa2Δ yeast (B), we constructed ppa2Δ yeast cells co-expressing wild-type hPPA2 under the control of a doxycycline-repressible (Tet-Off) promoter (pCM189) and one of the mutated forms: p.Ser61Phe or p.Glu172Lys under the control of the constitutive promoter TPI (pYX142 plasmid). After blocking the expression of the wild-type hPPA2 with doxycycline, only the mutated form is expressed. The strain referred to as WT contains only the regulatable wild-type hPPA2. The consequences of the mutations on mitochondrial function can then be assessed while the cells still contain functional mtDNA (Table S3). The cells were grown in 2 L of rich glycerol/ethanol medium until OD = 1 and then supplemented with 5 mM doxycycline. 12 hr later, by which time the source of newly synthesized PPA2 was totally drained or supplied by only one of the hPPA2 mutants, the cells were harvested and their mitochondria isolated. (D) State 3 respiration (in the presence of an excess of external ADP, phosphorylating conditions) and ATP synthesis, expressed as percent of the WT (see Figure S4 and Table S3 for details). Values represent mean ± SD. (E) BN-PAGE analyses. Experiments were performed as previously described. Mitochondrial proteins were solubilized with digitonin (2 g/g protein), separated in a non-denaturing 3%–13% gradient polyacrylamide gel, and stained with Coomassie blue. The figure shows the results obtained with mitochondria extracted from cells grown in the absence (−) or presence (+) of 5 mM doxycycline (DOX). (F) Mitochondrial membrane potential (ΔΨ) analyses. Energization of the mitochondrial inner membrane was monitored by rhodamine 123 fluorescence quenching with intact mitochondria using a SAFAS (Monte Carlo, Monaco) fluorescence spectrophotometer. The additions were 0.5 μg/mL rhodamine 123, 0.15 mg/mL mitochondrial proteins (Mito), 1% (V/V) of ethanol (EtOH) (electron donor), 6 μg/mL oligomycin (oligo) (ATP synthase proton channel inhibitor), 0.2 mM potassium cyanide (KCN) (complex IV inhibitor), 50 μM ADP, 1 mM ATP, and 3 μM CCCP (carbonyl cyanide m-chlorophenylhydrazone, membrane uncoupler).