ELAC2 mutations cause a mitochondrial RNA processing defect associated with hypertrophic cardiomyopathy

Abstract

The human mitochondrial genome encodes RNA components of its own translational machinery to produce the 13 mitochondrial-encoded subunits of the respiratory chain. Nuclear-encoded gene products are essential for all processes within the organelle, including RNA processing. Transcription of the mitochondrial genome generates large polycistronic transcripts punctuated by the 22 mitochondrial (mt) tRNAs that are conventionally cleaved by the RNase P-complex and the RNase Z activity of ELAC2 at 5' and 3' ends, respectively. We report the identification of mutations in ELAC2 in five individuals with infantile hypertrophic cardiomyopathy and complex I deficiency. We observed accumulated mtRNA precursors in affected individuals muscle and fibroblasts. Although mature mt-tRNA, mt-mRNA, and mt-rRNA levels were not decreased in fibroblasts, the processing defect was associated with impaired mitochondrial translation. Complementation experiments in mutant cell lines restored RNA processing and a yeast model provided additional evidence for the disease-causal role of defective ELAC2, thereby linking mtRNA processing to human disease.

Figure 1

ELAC2 Mutation Status and Gene Structure (A) Pedigrees of three families with mutations in ELAC2. Individual #57415 (F1: II-3) from unrelated parents was found to have a compound heterozygous mutation. Individuals #61982 (F2: II-5) and #65937 (F3: II-2) born to consanguineous parents (first-degree cousins) carry homozygous missense mutations in ELAC2. (B) Gene structure of ELAC2 with known protein domains of the gene product and localization and conservation of amino acid residues affected by mutations. Intronic regions are not drawn to scale.

Figure 2

Quantification and Rescue of the Mitochondrial mRNA Processing Defect by qPCR (A) Map of the mitochondrial genome with the annotation of rRNA, tRNA, and mRNA genes among the heavy and light strand. Numbering 1–10 indicates the RNA cleavage sites located 5′ terminal of rRNAs or mRNAs. As an example for the primer design, a scheme of the tRNA^Arg-ND4L junction is given. (B and C) Quantification of the ten unprocessed mitochondrial tRNA-rRNA or tRNA-mRNA junctions analyzed by qPCR in (C) muscle and (D) fibroblasts before (−) and after (+) transduction with a wild-type ELAC2 expression construct. Values are given as fold change to the average of three different controls. Error bars indicate ±1 standard deviation (SD) from the mean of at least three technical (B) or biological (C) replicates.

Figure 3

Steady-State Levels Mitochondrial RNAs and RNA-Seq in Mutant Fibroblasts (A and B) Quantification of steady-state levels of (A) 14 mt-tRNAs and (B) 3 mt-mRNAs and 16S mt-rRNA in control and affected individual’s (#57415) fibroblasts by RNA blot analysis. 18S rRNA was used as loading control. (C) Mitochondrial transcriptome analysis by next-generation sequencing (RNA-seq) of fibroblasts RNA from ELAC2 individuals and three controls. Relative sequencing coverage is plotted against the mitochondrial genome (numbering according to RefSeq accession number J01415). Colored lines represent individuals #57415 (blue), #61982 (red), and #65937 (green). The black dotted line indicates the average of three controls. The ten RNase Z cleavage sites analyzed by qPCR (Figure 2) are shown with higher magnification. (D) RNA blot analysis decorated with tRNA^Leu(UUR)- and tRNA^Val-specific probes to detect unprocessed mt-tRNA-mRNA intermediates.

Figure 4

Mitochondrial Translation and Steady-State Levels of Respiratory Chain Complexes in Mutant Fibroblasts (A and B) Analysis of mitochondrial translation products in fibroblasts from individual #57415 by [³⁵S]methionine pulse labeling (A) and quantification of three experiments (B). Error bars indicate ±1 SD from the mean of three replicates. mtDNA-encoded structural subunits of complex I (ND1, ND2, ND3, ND4, ND4L, ND5, ND6), complex III (cytb), complex IV (MT-COI, MT-COII, MT-COIII), and complex V (ATP6, ATP8) are shown. (C and D) Immunoblot analysis of respiratory chain complex subunits in crude mitochondrial extractions from individual #57415 and control (C) fibroblasts as well as (D) muscle. Porin antibody was used as a loading control in muscle.

Figure 5

Model of Potential ELAC2 Pathomechanism Suggested model of normal and impaired ELAC2 activity. Under normal conditions minor amounts of unprocessed ELAC2 substrates can be degraded by the RNA surveillance machinery. Reduced ELAC2 activity results in an accumulation of mitochondrial precursor mRNAs, which impair mitochondrial translation.