Mutations in the MRPS28 gene encoding the small mitoribosomal subunit protein bS1m in a patient with intrauterine growth retardation, craniofacial dysmorphism and multisystemic involvement

Abstract

Mitochondria contain a dedicated translation system, which is responsible for the intramitochondrial synthesis of 13 mitochondrial DNA (mtDNA)-encoded polypeptides essential for the biogenesis of oxidative phosphorylation (OXPHOS) complexes I and III–V. Mutations in nuclear genes encoding factors involved in mitochondrial translation result in isolated or multiple OXPHOS deficiencies and mitochondrial disease. Here, we report the identification of disease-causing variants in the MRPS28 gene, encoding the small mitoribosomal subunit (mtSSU) protein bS1m in a patient with intrauterine growth retardation, craniofacial dysmorphism and developmental delay. Whole exome sequencing helped identify a seemingly homozygous missense variant NM_014018.2:c.356A>G, p.(Lys119Arg) which affected a highly conserved lysine residue. The variant was present in the mother in a heterozygous state, but not in the father who likely carried a large deletion spanning exon 2 and parts of introns 1 and 2 that could account for the apparent homozygosity of the patient. Polymerase chain reaction (PCR) amplification and Sanger sequencing of MRPS28 cDNA from patient fibroblasts revealed the presence of a truncated MRPS28 transcript, which lacked exon 2. Molecular and biochemical characterization of patient fibroblasts revealed a decrease in the abundance of the bS1m protein, decreased abundance of assembled mtSSU and inhibited mitochondrial translation. Consequently, OXPHOS biogenesis and cellular respiration were compromised in these cells. Expression of wild-type MRPS28 restored mitoribosomal assembly, mitochondrial translation and OXPHOS biogenesis, thereby demonstrating the deleterious nature of the identified MRPS28 variants. Thus, MRPS28 joins the increasing number of nuclear genes encoding mitoribosomal structural proteins linked to mitochondrial disease.

Introduction

Mitochondria contain a dedicated translation machinery that includes mtDNA-encoded tRNAs (mt-tRNA), mostly unique nuclear-encoded tRNA synthetases and translation factors, as well as mitochondria-specific ribosomes (mitoribosome). The mitoribosome (55S) is built up by a small mitoribosomal subunit (mtSSU; 28S) and a large (mtLSU; 39S) mitoribosomal particles, which are ribonucleoprotein complexes composed of mtDNA-encoded 12S and 16S rRNAs, respectively, and altogether ~80 nuclear-encoded ribosomal proteins. In contrast to its bacterial and cytosolic counterparts, the mitoribosome features reduced rRNA content and expanded functional and structural roles of its constituent proteins some of which are unique to mitochondria whereas others, though retaining some conservation with their bacterial orthologs, have acquired mitochondria-specific N- or C-terminal extensions (1,2). Notably, 5S rRNA typical of bacterial ribosomes has been substituted in mitoribosomes by mt-tRNA^Val or mt-tRNA^Phe, both seemingly interchangeable in different species, tissues and cell types (3). The mitoribosome synthesizes 13 mtDNA-encoded polypeptides that are essential structural components of four of the five, ATP-generating, oxidative phosphorylation (OXPHOS) complexes: complex I (CI), complex III (CIII), complex IV (CIV) and complex V (CV). Only complex II (CII) is built up exclusively by nuclear-encoded subunits. Mutations in mt-tRNA-coding genes and in nuclear genes encoding proteins involved in mitoribosomal biogenesis and function have been shown to impair mitochondrial translation thus inhibiting OXPHOS biogenesis and leading to mitochondrial diseases (4).

To date, mutations in 10 genes encoding 7 proteins from mtSSU—uS2m (MRPS2, MIM: 611971) (5), uS7m (MRPS7, MIM: 611974) (6), uS14m (MRPS14, MIM: 611978) (7), bS16m (MRPS16, MIM: 609204) (8), mS22 (MRPS22, MIM: 605810) (9–11), mS23 (MRPS23, MIM: 611985) (12) and mS34 (MRPS34, MIM: 611994) (13)—and 3 proteins from mtLSU—uL3m (MRPL3, MIM: 607118), bL12m (MRPL12, MIM: 602375) and mL44 (MRPL44, MIM: 611849)—have been linked to heterogeneous, multisystemic presentations affecting predominantly the brain, the heart and the liver (Table 1 and Discussion). In these cases, age of onset was in the antenatal or neonatal period whereas severity of the clinical presentation and mortality varied considerably between genes and even between patients with disease-causing variants in the same gene.

Table 1

Clinical and biochemical presentations of disease-causing variants in MRPS and MRPL genes

Gene (Protein)	Total number of patients	Patient nr. in the reference	Variant(s)	Onset (age if reported)	Age of death / age at last follow-up	Clinical presentation	Laboratory findings			Reference
							Biochemistry	Histology	OXPHOS
MRPS2 (uS2m)	2	Pt1	NM_016034.4: c.328C>T, p.(Arg110Cys) c.340G>A, p.(Asp114Asn)	Infantile (1 year)	Alive at 11 years	Sensorineural deafness; minor dysmorphic features; skin wrinkling prominent on abdomen and hands; FTT; DD; strabismus of left eye	Elevated liver enzymes and serum alanine; hypoglycemia; hyperlactatemia; increased KG in urine	N/A	M: ↓CIII and CIV; L: ↓CI, CIII and CIV; Fb: ↓CIV	(5)
		Pt2	NM_016034.4:c.413G>A, p.(Arg138His)	Infantile (6 years)	Alive at 11 years	Sensorineural deafness, DD acute hypoglycemic episodes; exercise intolerance;	Hypoglycemia; hyperlactatemia and hyperlactaturia; Increased KG in urine.	N/A	M: ↓CIV; L: ↓CI, CII, CIII and CIV; Fb: ↓CIII and CIV
MRPS7 (uS7m)	2S	Pt1	NM_015971.3:c.550A>G, p.(Met184Val)	Neonatal (~6 months for Pt1)	Died at 141/2 years	Bilateral sensorineural deafness; liver involvement: initial hepatomegaly, liver failure at 14 years of age; renal dysfunction at 14 years; FTT	Hypoglycemia; hyperlactatemia; increase in 3-hydroxybutirate and C6-C10 dicarboxylic acids	L: steatosis, mild fibrosis.M: increased size of lipid droplets	M: ↓CIII; L: ↓ CI, CIII and CIV; Fb: ↓CI and CIV	(6)
		Pt2			Alive at 171/2 years	Bilateral sensorineural deafness; mild learning difficulties; Primary hypogonadism and adrenal failure	Intermittent hyperlactatemia and hypoglycemia; Mild TSH elevation; Elevated serum urea and creatinine	M: predominance of type 1 fibers. L: mild increase of spaces of Disse; swollen hepatocytes; mild steatosis K: atrophic renal tubes with giant mitochondria	M: unchanged L: ↓CI, CIII and CIV
MRPS14 (uS14m)	1	Pt1	NM_022100.2: c.322C>T, p.(Arg108Cys)	Antenatal	Alive at 5 years	Stable HCM (Wolf–Parkinson White syndrome) completely receded at 5 years of age; DD; FTT; muscle hypotonia; dysmorphic features	Hyperlactatemia	N/A	M: ↓ CIV	(7)
MRPS16 (bS16m)	1	Pt1	NM_016065.3: c.331C>T, p.(Arg111*)	Antenatal/neonatal	Died at 3 days	Dysmorphism; redundant skin on neck. On FUS: Dilation of cerebral ventricles, agenesis of corpus callosum. CM; FTT; hypotonia	Increased liver enzymes; Hyperlactatemia	N/A	M: ↓CI, CII+CIII, CIV and CV; L: ↓CI, CII+CIII and CIV; Fb: ↓CII+CIII and CIV	(8)
MRPS22 (mS22)	10	Pt1S Pt2S Pt3S*	NM_020191.2: c.509G>A, p.(Arg170His)	Antenatal/neonatal	Died at 2–22 days	HCM (Pt1 and Pt2), Generalized antenatal edema, ascites and subcutaneous edema at birth; Renal tubulopathy; Muscle hypotonia	Hyperammonemia, hyperlactatemia	N/A	M (P1): ↓CI, CIII, CIV and CV; Fb (P1): ↓CIV; LC (P2): ↓CIV	(9)
		Pt1	NM_020191.2: c.644T>C, p.(Leu215Pro)	Antenatal/neonatal	Alive at 5.5 years	Dysmorphic features: antenatal microcephaly, Cornelia de Lange-like features, redundant skin on neck; Stable biventricular HCM (Wolf–Parkinson-White syndrome); Corpus callosum hypoplasia and dilation of third ventricle, leukoencephalopathy; Muscle hypotonia; FTT; transient seizures and EEG abnormalities; tetraspasticity	Severe metabolic acidosis with hyperlactatemia; Elevated urinary TCA-cycle intermediates	N/A	Fb: ↓CI, CIII and CIV	(10)
		Pt1	NM_020191.2: c.1032_1035dup, p.(Leu346Asnfs*21)	Neonatal (2 days)	Died at 3 days	Agenesis of corpus callosum, periventricular cysts, hemorrhagic lesions or calcifications; CM (atrial and ventricular septal defects, coronary artery fistula)	Lethal hyperlactatemia; moderate hyperammonemia; massive lactaturia	N/A	Fb: ↓CI, CIII and CIV	(11)
		Pt1	NM_020191.2: c.339+5G>A	Infantile (4 months)	Alive at 4 years	Dysmorphic features; Brain MRI at 15 months: Leigh-like lesions affecting medulla oblongata and brain stem, but normal basal ganglia, cerebellar and cerebral atrophy. Hypotonia; Severe DD, oxygen dependence, tetraspasticity, unresponsiveness to audiovisual stimuli. Mosaic Down syndrome	Elevated urine organic acids, lactate and pyruvate	N/A	N/A	(36)
		Pt1S,C Pt3S,C Pt2 C	NM_020191.2: c.605G>A, p.(Arg202His)	Childhood–adolescence (9–19 years)	Alive at time of report	Delayed puberty; Primary ovarian insufficiency; Hypergonadotropic hypogonadism	High basal gonadotropin levels and undetectable estradiol	Pt3: Fibrotic ovaries without follicles	N/A	(19)
		Pt4	NM_020191.2: c.404G>A, p.(Arg135Gln)	14 years	Alive at time of report	Small for gestational age; mild facial dysmorphism; developmental delay; Bilateral axonal polyneuropathy; Hypergonadotropic hypogonadism Amenorrhea Primary ovarian insufficiency	Slightly elevated blood lactate Very low ketosteroids. Elevated methemoglobin	N/A	N/A
MRPS23 (mS23)	1	Pt276	NM_016070: c.119C>G, p.(Pro40Arg)	Infantile (11/2 years)	N/A	Liver disease	Hypoglycemia	N/A	Fb: ↓CI and CIV	(12)
MRPS28(bS1m)	1	Pt1	NM_014018.2:c.356A>G, p.(Lys119Arg)c.214_395del, p.(Gly72Glufs*16)	Antenatal /infantile	Alive at 30 years	Severe intrauterine growth retardation, hypotrophy and microcephaly;Facial dysmorphism;Sensorineural deafness;Hepatomegaly, FTT, DD, altered ERGAt 24 years of age: bilateral lesions in globus pallidus and cerebellar atrophy	Hyperlactatemia, high alanine;Hypoglycemia;Elevated liver enzymes and urinary TCA-cycle metabolites.	N/A	M: ↓CV;L: ↓CIV;Fb: ↓CIV;	Thisstudyand (14)
MRPS34 (mS34)	6	Pt1	NM_023936.1: c.(321+1G>T), p.(Val100_Gln107del)	Infantile (4 months)	Died at 9 months	Leigh syndrome; Microcephaly; DD	Hyperlactatemia; Increased alanine levels in urine;	M: Atrophy of type 2 fibers; COX deficiency	M: ↓CI, CIII and CV; L: ↓CI and CIV; Fb: ↓CI and CIV	(13)
		Pt2a**	c.322-10G>A, p.(Asn108Leufs*12), p.(Asn108Glyfs*50), =	Infantile (6 months)	Alive at 17 years	Leigh-like syndrome; Hypotonia; DD; strabismus, ptosis; Developmental regression at 12 years; Tremors, spasticity; Horseshoe kidney; Microcephaly; coarse features	Mild hyperlactatemia	N/A	M(12 years): ↓CI, CIII and CV
		Pt2b**		Infantile (6 months)	Alive at 14 years	Similar to Pt2a; MRI at 9 years: mild cerebellar atrophy	Mild hyperlactatemia	N/A	N/A
		Pt3a**	c.322-10G>A, p.(Asn108Leufs*12), p.(Asn108Glyfs*50), =	Infantile (6 months)	Alive at 7 years	Leigh syndrome; Hypotonia; Developmental regression; Optic atrophy; Dysmorphic facies; Premature puberty	Mild hyperlactatemia	M: atrophy of type II and I fibers; increased COX staining; increase in subsarcolemmal mitochondria	M: ↓CI+CIII, CII, CII+CIII and CV
		Pt3b**		Infantile (6 months)	Alive at 2 years	Leigh syndrome; DD	Mild hyperlactatemia	N/A	N/A
		Pt4	c.37G>A, p.(Glu13Lys); c.94C>T, p.(Gln32*)	Infantile (~10 days)	Died at 81/2 months	Leigh syndrome; Developmental regression; Dystonia; Pyramidal syndrome of lower limbs; Tubulopathy	Transient metabolic acidosis	M: lipidosis	M: ↓CIV; Fb: ↓CIV
MRPL3 (uL3m)	4S	Pt1	c.950C>G, p.(Pro317Arg)	Infantile (5–6 months)	Died at 17 months	HCM (stable in P3 and P4), Hepatomegaly, liver dysfunction, FTT, Hypergyria of the cortical brain (P3 and P4)	Hyperlactatemia, elevated liver enzymes	N/A	M: ↓CIII and CIV; Fb: ↓CIV	(37)
		Pt2			Died at 15 months				Fb: ↓CIV
		Pt3T Pt4T			Alive at 3 years				M(P3): ↓CIV Fb(P3): ↓CI and CIV Fb(P4): ↓CI and CIV
MRPL12 (bL12m)	3S	Pt1	c.542C>T, p.(Ala181Val)	Antenatal/neonatal	Died at 2 years	Antenatal hypotrophy (Dwarfism excluded); DD, FTT At 2 years: tonic seizures, cerebellar ataxia and tremor; mild hepatomegaly; mild dysmorphism; MRI: lesions in white matter and basal ganglia	Hyperlactatemia; Very low plasma IGF1	N/A	M: ↓CI+CIII and CIV; L: ↓CI+CIII and CIV; Fb: ↓CI and CIV	(38)
		Pt2T, Pt3T		Antenatal	Pregnancy terminated	N/A	N/A		AC: ↓CIV
MRPL44 (mL4)	4	Pt1S	c.467T>G, p.(Leu156Arg)	Infantile (3–6 months)	Died at 6 months	HCM Liver steatosis	N/A	H: Hypertrophied and pale. Diffuse microvesicular steatosis; L: Fatty degeneration.	H: ↓CI and CIV M: ↓CIV	(15)
		Pt2S			Alive at 14 years	Stable and asymptomatic HCM. Liver steatosis Mild granular pigmentation of retina	Transiently elevated ALAT and ASAT. Hyperlactatemia	Fatty degeneration in muscle and liver Decreased CIV activity	M: ↓CIV
		Pt1	c.233G>A, p.(Arg78Gln); c.467T>G, p.(Leu156Arg)	Infantile (6 months)	Alive at 8 years of age	Severe non-obstructive HCM Mild hepatopathy. Moderate exercise intolerance and muscle weakness	Hyperlactatemia; decreased free carnitine Transiently elevated ASAT and GGT enzymes	Mitochondrial proliferation in cardiac muscle	Fb: ↓CIV	(39)
		Pt2	c.467C>T, p.(Leu156Arg)	Neonatal (2 days)	Alive at 26 years	Non-obstructive HCM; Learning difficulties, memory and concentration problems Hemiplegic migraine. Tapetoretinal degeneration with pigment deposits in retina. White matter abnormalities, basal ganglia and midbrain hyperintensities at 24 years.	Metabolic acidosis	Hypertrophy of the heart muscle, myocytolysis, perivascular and interstitial fibrosis	H: ↓CI and CIV

^Ssiblings;

^**a and b used to indicate sibship in original report;

T

^Ttwins;

^Ccousins; ↓, indicates a decrease in the activities of the complexes indicated after the arrow;

⁺indicates a combined measurements of the OXPHOS complex activity; COX, cytochrome oxidase; CI to CV, OXPHOS complexes I to V; ERG, electroretinogram; KG, a-ketoglutaric (oxoglutaric) acid; TSH, thyroid-stimulating hormone; ALAT, alanine aminotransferase; ASAT, aspartate aminotransferase; GGT, gamma-glutamyl transferase; DD, developmental delay; N/A, not available; M, muscle; L, liver; K, kidney; Fb, fibroblasts; LA, lactic acidosis; FUS, fetal ultrasound; HCM, hypertrophic cardiomyopathy; CM, cardiomyopathy; FTT, failure to thrive; LC, lymphocytes. MRI, magnetic resonance imaging

^*The clinical presentation of this patient was not reported in detail.

Here we report the identification and characterization of disease-causing variants in the MRPS28 gene encoding the mtSSU protein bS1m located at the mRNA exit channel of the mitoribosome. Although the patient was previously described (14), the disease-causing gene was unknown until now. Molecular and biochemical analysis of patient fibroblasts revealed an impaired mtSSU biogenesis and a dramatically decreased mitochondrial translation, which led to assembly defects of all OXPHOS complexes containing mtDNA-encoded proteins. Expression of wild-type MRPS28 cDNA in patient fibroblasts restored mitochondrial translation and OXPHOS biogenesis confirming the role of the identified MRPS28 variants in the disease pathomechanism in this patient.

Results

Clinical presentation and identification of MRPS28 variants by exome sequencing

A detailed clinical description of the patient was first reported by (14). Below we present a brief and updated summary of important clinical features of the patient that are relevant to the discussion and the identity of the disease gene. The patient was born to healthy parents of French origin at 34 weeks of pregnancy by cesarean section because of an intrauterine growth retardation (birthweight: 1670 g, occipital frontal circumference (OFC): 28.5 cm). He had two healthy siblings, but an elder sister died at 27 months of age due to hepatic failure—her clinical presentation included failure to thrive (FTT), sensorineural deafness, liver enlargement, facial dysmorphism and iterative episodes of hypoglycemia and lactic acidosis. The patient also presented with a FTT and had iterative episodes of dehydration, hypoglycemia and major swallowing difficulties, which required gastrostomy for enteral nutrition. His facial dysmorphism, closely similar to that of his diseased sister, included round face, short neck, lid ptosis, long convex philtrum, low set, posteriorly rotated ears, sensorineural deafness and cataract. X-rays showed flat vertebrae and cone-shaped terminal phalanges. His height, weight, OFC gain (--4SD) and psychomotor development were markedly delayed, with no speech, but his personality was jovial and outgoing. Metabolic workup consistently showed high plasma lactate, high alanine and lactate/pyruvate ratios, elevated liver aminotransferases [aspartate aminotransferase (ASAT) and alanine aminotransferase (ALAT)] and urinary excretion of Krebs cycle intermediates. Heart ultrasound and brain CT scan were normal, but electroretinograms (ERGs) were markedly altered. However, computer tomography at 24 years of age revealed bilateral lesions in globus pallidus and cerebellar atrophy (Fig. 1). The patient is now a 30-year-old adult with little autonomy and needs continuous assistance.

Enzymatic analysis of respiratory activities in liver and muscle biopsies, as well as lymphocytes (LCs) and skin fibroblasts, revealed a strong inhibition of CIV activity, which together with the clinical presentation was consistent with a diagnosis of mitochondrial disease (14). However, the disease-causing gene was not identified at that time. Therefore, we recently carried out whole-exome sequencing on genomic DNA from this patient. To identify putative disease-causing variants, we excluded all non-coding, intronic (except for splice sites) and synonymous variants, as well as those with a frequency >0.1% in dbSNP, 1000 Genomes, Exome Variant Server or our in-house database. Thus, we could identify a homozygous variant NM_014018.2:c.356A>G, p.(Lys119Arg) in the MRPS28 gene encoding the mtSSU protein bS1m. In silico analysis suggested that the variant had a strong pathogenic probability (Table 2). Sanger sequencing of genomic DNA from the patient and his parents confirmed the presence of c.356A>G in the patient and also in his mother who was heterozygous for this variant (Fig. 2A). However, the father was apparently homozygous for the wild-type allele, which prompted us to hypothesize that instead he carried a deletion spanning exon 2 of the MRPS28 gene. Two lines of evidence supported this hypothesis: (i) analysis of the exome data showed that the patient was heterozygous for the single nucleotide variant (SNV) rs10957952 located in intron 1 of the MRPS28 gene (c.213+274A>G) suggesting that loss of heterozygosity in exon 2 could be due to a large deletion while the presence of the heterozygous SNV rs34059899 in the STMN2 gene upstream of MRPS28 set the maximum size of the deletion to 365055 bp and (ii) amplification of the MRPS28 cDNA using 5′ and 3′ end primers revealed two amplification products in both control and patient cDNA samples---a longer amplicon corresponding to the full length MRPS28 cDNA, and a shorter amplicon, which had an increased abundance in the patient sample (Fig. 2B). Sanger sequencing of the two amplification products from patient cDNA showed that while the longer amplicon contained the missense c.356A>G variant, the shorter amplicon lacked exon 2 (Fig. 2C). In silico analysis predicted that loss of exon 2 results in a frameshift and a truncated polypeptide of only 86 amino acids, p.(Gly72Glufs^*16) (Fig. 2D). Thus, we conclude that the patient is compound heterozygous for a missense variant c.356A>G and a deletion of unknown length spanning exon 2 and parts of introns 1 and 2 of the MRPS28 gene.

Figure 1

Brain magnetic resonance imaging (MRI) of the patient revealing previously absent structural brain abnormalities. Axial T2 (left) and sagittal T1 (right) weighted MRI scans of the patient at 24 years of age showing hyperintensities in the bilateral pallidi and cerebellar atrophy.

Table 2

In silico pathogenicity prediction for the c.356A>G variant

cDNA1	c.356A>G
Amino acid2	p.Lys119Arg
SIFT	Deleterious (score 0)
Mutation Taster	Disease causing (probability 1)
PolyPhen-2	Probably damaging (score: 1.000; sensitivity: 0.00; specificity: 1.00)
Parental origin	Maternal

¹accession number NM_014018.2;

²accession number NP_054737.1; SIFT, Mutation Taster and PolyPhen-2 v2.2.2r398 were used at default settings. The prediction cut-off for SIFT is at 0.05. Values higher than the cut-off indicate tolerated substitutions (28). Mutation Taster returns a probability value in the range between 0 and 1 with a cut-off at 0.5. Probabilities >0.5 are indicative of disease-causing substitutions (30). Scores of the PolyPhen’s HumVar model range from 0 to 1 with score closer to 1 indicating an increasing probability that a variant is pathogenic (29).

Figure 2

Segregation of the identified MRPS28 variants and their effect on protein structure. (A) Pedigree analysis and inheritance of the identified variants in the MRPS28 gene. The missense c.356A>G variant was inherited from the mother whereas the father likely carried a deletion of exon 2, which can explain the apparent homozygosity of the patient. A female sibling of the patient died in infancy but did not have a formal genetic diagnosis. (B) Polymerase chain reaction (PCR) amplification of MRPS28 cDNA from control (C1 and C2) and patient (Pt) samples reveals the accumulation of a second, faster migrating amplicon (amplicon 2) in patient samples. Amplicon 1 corresponds to the full-length MRPS28 cDNA. (C) Sanger sequencing of amplicons 1 and 2 from the patient sample. Amplicon 1 was found to carry the c.356A>G variant whereas amplicon 2 carried a deletion of exon 2. (D) Predicted protein sequence of the transcript represented by amplicon 2 in (C). The amino acid sequence common to wild-type bS1m and the putative truncated polypeptide is indicated in black; the differences in the sequences of the two polypeptides after Lys71 are emphasized by differential coloring: the wild-type bS1m sequence is shown in blue whereas the sequence of the truncated polypeptide is shown in brown. (E) Multiple sequence alignment of bS1m homologs in different organisms. Only the region containing the highly conserved Lys119 (indicated by an arrow) is shown. (F) Localization and interactions of lysine (shown in red on the left-hand panel) and arginine (shown in red on the right-hand panel) at position 119 in the mitoribosomal structural model 3jd5. While Lys119 does not seem to form any H-bonds with surrounding residues, Arg119 has the potential to interact with His92 (black) from uS2m (grey) and same-chain Phe120 (yellow). H-bonds are shown in green and clashes are in red. BS1m polypeptide is shown in yellow.

The p.(Lys119Arg) substitution destabilizes bS1m resulting in impaired mtSSU assembly and inhibited mitochondrial translation

The c.356A>G variant results in the substitution of a highly conserved lysine residue with arginine at position 119 in the bS1m protein (Fig. 2E). Using the available structural model for the mammalian mitochondrial ribosome, we examined the interaction of Lys119 and Arg119 with their surrounding amino acids and the effect of p.(Lys119Arg) on the structure of bS1m. This analysis revealed that Lys119 does not appear to form intra- or inter-chain H-bonds with residues from bS1m or the neighboring uS2m protein whereas arginine at the same position appears to clash with the aromatic ring of its same-chain neighbor Phe120 and is also within H-bond formation distance with the side chain of His92 from uS2m (Fig. 2F). This may result in the structural destabilization of the bS1m protein and impairment of its ability to assemble into mtSSU. Consistent with the possible destabilizing effect of p.(Lys119Arg), abundance of bS1m was dramatically decreased in mitochondrial extracts from patient fibroblasts (Fig. 3A) as were other mtSSU proteins whereas abundance of mtLSU proteins was comparable to that in control cells (Fig. 3B). Different studies have shown that loss of individual mtSSU or mtLSU proteins can lead to a decreased abundance of 12S or 16S rRNA (5,6,10,15). In agreement with this, 12S rRNA was specifically decreased in patient fibroblasts whereas ND1 mRNA was unaffected; 16S rRNA was significantly upregulated perhaps as a compensatory response to impaired mtSSU assembly (Fig. 3C). The observation that stability of mtLSU is seemingly unaffected by the absence of mtSSU is consistent with previous studies in patient fibroblasts carrying MRPS/MRPL mutations showing that the two mitoribosomal particles can assemble independently of each other (5,13,15,16). These results suggest that mtSSU assembly is specifically impaired in patient cells. Indeed, linear density sucrose gradient ultracentrifugation analysis of mtSSU and mtLSU revealed that hardly any assembled mtSSU could be detected in patient fibroblasts (Fig. 3D).

Figure 3

Decreased abundance of mutant bS1m and other mtSSU proteins, impaired mtSSU assembly and inhibition of mitochondrial translation in patient fibroblasts. (A) and (B) Immunoblot analysis of mitochondrial extracts from control (C1 and C2) and patient (Pt) fibroblasts. Abundance of bS1m and the mtSSU proteins uS5m, mS40, mS34, uS15m and mS31 was decreased in patient fibroblasts whereas abundance of mtLSU proteins uL11m, mL37 and mL44 was comparable to that in controls. Porin was used as a loading control. Representative of three independent experiments is shown. (C) Relative expression levels of 12S rRNA, 16S rRNA and ND1 mRNA in control and patient fibroblasts determined by quantitative RT-PCR. This analysis demonstrated the specific decrease of 12S rRNA in patient fibroblasts; ND1 mRNA was unchanged whereas 16S rRNA was upregulated. 18S rRNA was used for normalization. Values for the two controls were averaged and set as 100% (C^*). Error bars represent standard error of the mean; C^*, n = 6; Pt, n = 3. Student’s t-test: ^*P < 0.05; ^**P < 0.01. (D) Linear density sucrose gradient sedimentation analysis of mitoribosomal assembly in control and patient mitochondrial extracts. The sedimentation profiles of mtSSU (28S), mtLSU (39S) and the monosome (55S) were determined by immunoblotting of the gradient fractions with antisera against mS40 and mL37 from mtSSU and mtLSU, respectively. Abundance of assembled mtSSU is decreased in patient fibroblasts. A representative of two independent experiments is shown. (E) Pulse-labeling of mitochondrial translation products in control and patient fibroblasts demonstrating that mitochondrial translation was severely inhibited in patient fibroblasts. Coomassie staining was used to assess loading. A representative of four independent experiments is shown.

Consistent with the finding that mitoribosomal assembly was compromised in patient fibroblasts, in vitro pulse labeling of the 13 mtDNA-encoded polypeptides revealed that mitochondrial translation was strongly inhibited in these cells (Fig. 3E).

Patient fibroblasts have reduced abundance of assembled OXPHOS complexes and decreased respiration rates

The 13 mtDNA-encoded polypeptides are core structural components of OXPHOS CI, CIII, CIV and CV. Therefore, inhibition of mitochondrial translation inevitably leads to impaired biogenesis of these complexes, but not OXPHOS CII, which does not contain any mtDNA-encoded subunits. This biogenesis defect can manifest as an isolated or a multiple OXPHOS deficiency in enzymological analyses. Consistent with this, diagnostic enzymology of muscle biopsies and patient fibroblasts previously revealed a dramatic CIV deficiency (14). Western immunoblotting of patient fibroblasts’ mitochondrial extracts carried out by us revealed decreased abundance not only of mtDNA-encoded but also some nuclear-encoded OXPHOS proteins suggesting an impaired assembly of all OXPHOS complexes containing mtDNA-encoded subunits (Fig. 4A). In the absence of mtDNA-encoded proteins, nuclear-encoded OXPHOS subunits continue to be synthesized and are imported into mitochondria but fail to assemble into OXPHOS complexes. Some of these unassembled subunits are unstable and are rapidly degraded by the mitochondrial proteolytic machinery, which explains the decrease in their abundance observed by immunoblotting. Therefore, to further assess the assembly status and function of the OXPHOS system, we carried out BN-PAGE and respirometry analyses on control and patient fibroblasts. BN-PAGE revealed a dramatic decrease in the abundance of CI and CIV, and to a lesser extent of CIII (Fig. 4B). An accumulation of the stable CV subcomplex, frequently observed when assembly of the ATPase F1 stalk region containing mtDNA encoded ATP6 and ATP8 is compromised, could also be observed in these analyses. Finally, high-resolution respirometry analysis of patient and control fibroblasts revealed a significant decrease of mitochondrial respiration in patient cells consistent with an impaired OXPHOS function in these cells (Fig. 4C).

Figure 4

OXPHOS biogenesis and function are severely compromised in patient fibroblasts. (A) Immunoblotting of control (C1 and C2) and patient (Pt) mitochondrial extracts against mtDNA- and nuclear-encoded proteins from the five OXPHOS complexes (indicated in brackets). A representative of three independent experiments is shown. Porin is used as a loading control. (B) BN-PAGE analysis of mitoplast extracts in 2% DDM from control and patient fibroblasts demonstrating a reduction in the abundance of the OXPHOS complexes containing mtDNA-encoded polypeptides. The complexes were detected by immunoblotting with antisera against a single protein from each complex indicated in brackets. CII, containing exclusively nuclear-encoded subunits, was detected using antibody against SDHA and was used as a loading control. A representative of three independent experiments is shown. (C) High-resolution respirometry of control and patient fibroblasts demonstrating severely decreased respiration of patient cells. Respiration was assessed in the presence of the following different substrates and uncouplers: PGM, pyruvate, glutamate and malate; dig, digitonin; ADP, adenosine diphosphate; suc, succinate; CCCP, carbonyl cyanide m-chlorophenylhydrazone; AMA, antimycin A; TMPD, tetramethyl- p-phenylenediamine; asc, ascorbate. Respiration states are indicated: R, routine respiration ; L, leak state; ETS, maximum noncoupled flux (capacity of the electron transfer system); ROX, residual oxygen consumption; CI + II, CIV, respiration at the level of OXPHOS complexes I + II and IV. Data were normalized to cell number. The oxygen consumption was expressed in pmol oxygen/s/cell number. C, n = 3; Pt, n = 3. Student’s t-test: ^*P < 0.05; ^***P < 0.001.

Expression of wild-type MRPS28 restores mitochondrial translation and OXPHOS biogenesis in patient fibroblasts

To demonstrate that mutations in the MRPS28 gene were responsible for the observed molecular phenotype of patient fibroblasts and therefore are likely disease-causing in nature, we carried out complementation analyses in immortalized control and patient fibroblasts. Stable transduction with lentiviral particles expressing wild-type MRPS28 cDNA restored the steady-state levels of bS1m and other mtSSU proteins (Fig. 5A) leading to restoration of mitochondrial translation (Fig. 5B). Abundance of individual OXPHOS proteins was increased in complemented patient fibroblasts (Fig. 5C) as was the abundance of all previously diminished OXPHOS complexes attesting to the restoration of OXPHOS biogenesis in these cells (Fig. 5D) and confirming the pathogenic nature of the identified MRPS28 variants.

Figure 5

Expression of wild-type MRPS28 cDNA restores the abundance of mtSSU and OXPHOS proteins, mitochondrial translation and OXPHOS biogenesis. (A) Immunoblotting of control (C1 and C2) and patient (Pt) fibroblast mitochondria demonstrating the increased abundance of bS1m and other mtSSU proteins in patient cells stably expressing wild-type MRPS28 cDNA. Detections were performed as in Figure 3A and B. A representative immunoblot of three independent experiments is shown. (B) Pulse-labeling of mitochondrial translation products in control and patient fibroblasts stably transduced with GFP-expressing vector and patient fibroblasts stably expressing wild-type MRPS28 cDNA. Mitochondrial translation is restored in the latter, but not in patient cells expressing GFP. Coomassie was used to demonstrate equal loading. A representative of two independent experiments is shown. (C) Immunoblotting of mitochondrial extracts from control and patient fibroblasts stably expressing either GFP or bS1m performed as in Figure 4A. Abundance of mtDNA- and nuclear-encoded OXPHOS proteins is restored upon expression of wild-type bS1m consistent with increased OXPHOS biogenesis. A representative of three independent experiments is shown. (D) BN-PAGE analysis of the abundance of individual OXPHOS complexes in mitoplast extracts from control and patient fibroblasts expressing either GFP or bS1m performed as in Figure 4B. Expression of wild-type bS1m restored OXPHOS assembly in patient fibroblasts. Representative of three independent experiments is shown.

Discussion

We identified the MRPS28 gene as a new mitochondrial disease gene whose variants cause impaired mitoribosomal biogenesis, inhibition of mitochondrial translation and multiple OXPHOS assembly defects in patient fibroblasts. The patient in this report had a multisystemic disease with dysmorphism, as well as brain and liver involvement. He had sensorineural deafness and severe psychomotor developmental delay (DD), and developed structural brain abnormalities, which were not present during infancy. Consistent with the fact that neurological involvement is quite common in mitochondrial disease (17), patients with disease-causing MRPL/MRPS variants frequently presented with variably severe neurological involvement and, aside from the patients who died in infancy, the patient discussed here was on the severe end of the clinical spectrum (Table 1). On the other hand, his liver involvement was relatively mild, included hepatomegaly and elevated liver enzymes, but did not progress to liver dysfunction seen in other patients with MRPL/MRPS variants. Neonatal liver involvement was also observed in approximately half of the patients with pathogenic MRPL/MRPS variants and ranged from an elevation of liver enzymes, hepatomegaly, liver steatosis to liver failure and death (Table 1). The patient also presented with intrauterine growth retardation and dysmorphic features. Dysmorphism is an uncommon presentation in mitochondrial disease (18), but it seems that more than half of the patients with MRPL/MRPS variants have dysmorphic features. It is tempting to speculate that these features may be an important aspect of the clinical presentation of MRPL/MRPS-linked disease and reflects a possible role of these genes in embryofetal development (7,15). Finally, the patient in our report also presented with one of the longest documented survivals for patients carrying MRPL/MRPS disease variants. Survival beyond infancy, well into childhood or adolescence and even into early adulthood, was documented for individuals with pathogenic variants in most of the previously reported MRPS or MRPL disease genes. In few of these cases, the disease phenotype stabilized or even improved later in life. Such was the case of a recently reported patient with a pathogenic MRPS14 variant who presented with a stable hypertrophic cardiomyopathy (HCM) that completely receded by the age of 5 years and the patient showed a significant improvement in her general condition (Table 1) (7). The mechanism behind this improvement is unclear, but it was proposed that a cardiac-specific compensatory response may have contributed to this process (7). More recently, the clinical spectrum of mitoribosomal deficiencies was further expanded with the report of pathogenic variants in MRPS22 linked to primary ovarian insufficiency (19). Unlike most previous reports, in these cases OXPHOS function in patient fibroblasts was not compromised. As most of the reported surviving patients with MRPS/MRPL variants were relatively young at the time of the last reported follow-up, it would be interesting to know how the clinical course has since evolved because this could provide important information that can help understand the pathophysiology of MRPL/MRPS-linked mitochondrial diseases.

Clinical presentations of MRPL/MRPS disease variants are heterogeneous and most often multisystemic affecting the brain, heart and/or liver with severity of presentations varying from neonatal lethality to survival into adulthood. This heterogeneity was seen not only for patients with mutations in different mitoribosomal proteins but also between patients with mutations in the same gene and even siblings, carrying identical mutations, showed differences in the severity of their clinical presentations and the clinical course of the disease (Table 1). This is difficult to explain in view of the ubiquitous expression of these genes and the important roles of the proteins that they encode in OXPHOS biogenesis. Obviously, the clinical presentation results from the complex interaction of multiple factors including the specific effects of the mutations on the stability and function of the mutant proteins, the expression levels of the target protein and its interaction partners, the specific responses of each tissue to mitochondrial dysfunction, the effect of modifier genes or the genetic background, the environment and possibly other factors. Most of the affected proteins have essential structural roles for mitoribosomal assembly, which can be deduced from the observation that their depletion results in impaired mtSSU or mtLSU assembly in patient fibroblasts and from the fact that all of them, except uS14, assemble early during mtSSU or mtLSU biogenesis (20). In addition to being structural components of the mitoribosome, these proteins may also play roles in the regulation of mitoribosomal function at different stages of the translation process, which can be impacted by their mutations. Thus, each mutation can differently affect (i) the stability and residual levels of the mutant protein in different tissues, (ii) the ability of the mutant protein to support the assembly of (semi)functional mitoribosomes and (iii) the effect of the mutation on the function of the impacted protein. The residual levels of the mutant protein will be influenced by the effect of the mutation on its turnover rate and its tissue-specific expression levels. The latter can vary naturally as was shown for wild-type mL44, which is less abundant in heart than in skeletal muscle and even though mutant mL44 was decreased in both tissues, significant levels of the protein could still be detected in skeletal muscle (15). However, it is unclear how this correlates with the clinical presentation as patients with MRPL44 variants present with CM accompanied by liver and/or neurological disease, but the relative expression levels of mL44 in brain and liver were not examined. Because the molecular characterization of most MRPL/MRPS disease variants were carried out in patient fibroblasts, it is impossible to assess the contribution of the abovementioned factors in the clinical presentation of these variants. Finally, the role of genetic and environmental modifiers in mitochondrial diseases is poorly understood and rather difficult to assess experimentally. Such modifiers were hypothesized as a possible determinant of the different clinical outcomes for two siblings with a homozygous MRPS7 disease variant one of which died from liver failure at the age of 14 years while the other was alive at 17 years of age with normal liver function (6). These factors may also, at least in part, account for the different clinical outcome of the patient discussed in this report and his sister.

Due to its lack of significant sequence homology to any bacterial ribosomal protein, bS1m was initially suggested to be specific to mitochondria (21). However, analysis of the available mitoribosomal structural models shows that the central region of bS1m adopts a structural conformation similar to the N-terminal oligonucleotide/oligosaccharide-binding (OB-fold) domain of bacterial ribosomal protein S1 (bS1) (2). BS1 is the largest ribosomal protein in bacteria and contains six OB-fold domains (β-barrel-shaped domains) of which the N-terminal domain (region spanning amino acids 1–93) is involved in ribosome binding whereas the remaining five domains are responsible for mRNA binding during translation initiation (22,23). BS1 has a high affinity for mRNA and its binding sites include A/U rich sequences upstream of the Shine-Dalgarno (SD) sequence, which drives translation initiation in most bacterial mRNAs. In vitro and in vivo, bS1 acts together with bS2, a homolog of another recently reported mitochondrial disease gene, in translation initiation of mRNAs carrying SD sequences (5,24–26). Structurally bS1 has no apparent role in ribosomal assembly and may be only loosely associated with the bacterial ribosome (27). Our data suggests that, in contrast to bS1, bS1m is an important structural component of the mitoribosome because mtSSU assembly is compromised in its absence. Such a role is corroborated by structural data, which shows that bS1m forms extensive interactions with uS2m and bS21m and by kinetics analysis of mitoribosomal assembly, which indicates that it assembles early during mtSSU biogenesis (2,20). Together with bS21m and mS31, bS1m forms the mRNA exit channel on mtSSU, but its ability to bind mRNA remains speculative because (i) it lacks any identifiable mRNA binding domains, (ii) mitochondrial mRNAs lack SD-like sequences and (iii) bS1 is dispensable for the translation of leaderless mRNAs. However, bS1m does contribute positively charged residues to the exit channel that may support mRNA interactions (2). Future functional characterization of bS1m should reveal its role at the mRNA exit channel and specifically if it functions in translation initiation in manner similar to bS1.

In conclusion, MRPS28 joins the ranks of genes encoding mitoribosomal proteins linked to impaired mitoribosomal biogenesis and function in mitochondrial disease. However, this is only the 11^th such gene discovered in the past 14 years and considering that the mitoribosome contains ~80 proteins, this suggests that mutations in these genes are relatively rare. Given the lack of specific clinical presentations associated with this group of genes, the possibility that impaired mitoribosomal biogenesis or function underlies the disease pathogenesis in patients with neonatal, multisystemic disease, especially in combination with dysmorphism, should be kept in mind.

Materials and Methods

This study adhered to the Declaration of Helsinki and written informed consent was obtained from the parents of the patient.

Whole exome sequencing

Whole exome sequencing and subsequent data analysis were carried out in collaboration with Commissariat à l’Énergie Atomique/Institut de Génomique/Centre National de Génotypage and were described previously (5).

Pathogenicity of the identified missense variant was analyzed in silico using the sorting intolerant from tolerant (SIFT) (28), PolyPhen 2 (29) and Mutation taster (30) algorithms included in the Alamut Visual 2.7 software (Interactive Biosoftware, Rouen, France).

Cell culture and immortalization

Control and patient fibroblasts were cultured at 37°C in a humidified atmosphere with 5% CO₂ in DMEM-GlutaMAX I medium supplemented with 10% (v/v) FBS, 100 U/μl penicillin and 100 μg/μl streptomycin (ThermoFischer Scientific, Montigny-le-Bretonneux, France).

Immortalization of control and patient fibroblasts was performed by stable transduction with a blasticidin-selectable lentivirus expressing human papilloma virus proteins E6 and E7 as we have done before (31).

Isolation of mitochondria

For mitochondrial isolation, cells were harvested at ~90% confluence and mitochondria were isolated by differential centrifugation after homogenization in isolation buffer (220 mm sucrose, 10 mm Tris-HCl pH 7.5) supplemented with protease inhibitor cocktail with or without ethylenediaminetetraacetic acid (EDTA) depending on the downstream application. Mitochondrial suspensions in mitochondrial isolation buffer were snap frozen in liquid nitrogen and stored at −80°C.

RNA isolation, cDNA synthesis and qRT-PCR

RNA was isolated using the RNAeasy mini-RNA isolation kit (Qiagen, Hilden, Germany) following the manufacturer’s protocols. First strand cDNA synthesis was carried out on 1–2 μg of RNA using the high-capacity cDNA archive kit (ThermoFischer Scientific, Montigny-le-Bretonneux, France) following manufacturer’s protocol. The resulting cDNA was either used for amplification of MRPS28 or for transcript quantification by TaqMan-based quantitative real-time PCR (ThermoFischer Scientific, Montigny-le-Bretonneux, France) essentially as described in (32). For qRT-PCR, the following probes were used: Hs02596859_g1 against 12S rRNA, Hs02596860_s1 against 16S rRNA, Hs02596873_s1 against ND1 mRNA and Hs99999901_s1 detecting 18S rRNA.

Cloning and expression of human MRPS28

MRPS28 cDNA was amplified from total cell cDNA by PCR using a forward primer with the sequence 5′- TACACGTACTTAGTCGCTGAAGCTCTTCTATGGCGGCGCTGTGTCG-3′ and a reverse primer with sequence 5′-AGGTACGAACTCGATTGACGGCTCTTCTACCTTATTTT TCATGATGTTCTTCTTTCG-3′. These primers contained the Sap I endonuclease restriction site to facilitate the cloning of the amplified PCR product into pD2109-CMV vector (Atum, Basel, Switzerland) using the Elektra cloning technology (Atum, Basel, Switzerland). The sequence complementary to MRPS28 is shown in bold. A control vector expressing green fluorescent protein (GFP) was used as a negative control. Virus production, viral titration and generation of fibroblast lines stably expressing bS1m or GFP were performed as described in (5).

SDS-PAGE, blue native PAGE and immunoblotting

For SDS-PAGE, 15 μg mitochondrial extracts in NuPAGE LDS buffer (ThermoFischer Scientific, Montigny-le-Bretonneux, France) were fractionated through AnyKD mini-protean polyacrylamide gels (Bio-Rad, Marnes-la-Coquette, France) and blotted onto low-fluorescence polyvinylidene difluoride (PVDF) membranes (Bio-Rad, Marnes-la-Coquette, France).

For blue native polyacrylamide gel electrophoresis (BN-PAGE), mitoplasts were extracted from ~1 × 10⁶ fibroblasts by incubation in 200 μl phosphate-buffered saline (PBS) containing 2 mg/ml Digitonin (Merck, Darmstadt, Germany) for 10 min on ice followed by dilution with 1 ml of PBS and centrifugation at 16000 × g for 10 min in a chilled centrifuge. The mitoplast-containing pellets were washed once with PBS and resuspended in 40 μl ACBT buffer [1.5 M aminocaproic acid and 75 mm Bis-Tris (Sigma-Aldrich, Lyon, France)] supplemented with 2% n-dodecyl β-D-maltoside (DDM, Sigma-Aldrich, Lyon, France). After incubation on ice for 10 min, samples were centrifuged in a prechilled centrifuge at 16000 × g for 30 min. Protein concentration in the extracts was determined using Bradford reagent (Sigma-Aldrich, Lyon, France) and 10–15 μg of the extract were supplemented with Native PAGE sample buffer and Coomassie Brilliant Blue G-250 (ThermoFischer Scientific, Montigny-le-Bretonneux, France) prior to their fractionation through 4–16% NativePAGE Bis-Tris gel (ThermoFischer Scientific, Montigny-le-Bretonneux, France). Separated complexes were blotted onto a hybond-P PVDF membrane (GE Healthcare, Succursale, France). OXPHOS complexes were detected using antibodies reactive against individual proteins from the following five complexes: anti-NDUFA13 (Abcam, Cambridge, UK, ab110240) for CI, anti-SDHA (Abcam, Cambridge, UK, ab14715) for CII, anti-UQCRC2 (Abcam, Cambridge, UK, ab14745) for CIII, anti-cytochrome oxidase (COX)IV (Proteintech, Manchester, UK, 11242–1-AP) for CIV and anti-ATP5A (Abcam, Cambridge, UK, ab14748) for CV.

Other antibodies used throughout this report were the following: anti-COXII (ab110258) and Porin (ab14734) antisera purchased from Abcam (Cambridge, UK); anti-ATP8 (26723-1-AP), anti-MRPS28/bS1m (16378-1-AP), anti-MRPS15/uS15m (17006-1-AP), anti-MRPS18b/mS40 (16139–1-AP), anti-MRPS31/mS31 (16288-1-AP) and anti-MRPL44/mL44 (16394-1-AP) purchased from Proteintech (Manchester, UK); anti-MRPL11/uL11m (GTX118773) and anti-MRPS5/uS5m (GTX103930) purchased from GeneTex (Irvine, CA, USA); and anti-MRPL37/mS37 (HPA025826) and anti-MRPS34/mS34 (SAB1102450) purchased from Sigma-Aldrich (Lyon, France).

Analysis of mitochondrial protein synthesis and ribosomal assembly

De novo mitochondrial translation was assessed by in vitro pulse labeling in cultured fibroblasts as described before (33).

Analysis of mitoribosomal assembly was carried out using linear density sucrose gradient as described in (34) with modifications. Mitochondria were isolated from immortalized control and patient fibroblasts grown in five 600 cm² cell culture dishes (ThermoFischer Scientific, Montigny-le-Bretonneux, France). A total of 500 μg of isolated mitochondria were lysed in lysis buffer [260 mm sucrose, 100 mm KCl, 20 mm MgCl_2, 10 mm Tris–HCl pH 7.5, 0.5% Triton X100 and complete EDTA-free protease inhibitor cocktail (Roche, Basel, Switzerland)] at a final concentration of 5 mg/ml for 20 min on ice and cleared by centrifugation for 45 min at 9300 × g in a chilled centrifuge. The lysates were fractionated through 15–30% linear density sucrose gradients for 6 h at 288 244 × g (max speed for rotor SW 41Ti, Beckman) at 4°C and 15 fractions were collected from the top of the gradients, precipitated with trichloroacetic acid (TCA) and subjected to SDS-PAGE using a 12% Criterion TGX gel (Bio-Rad, Marnes-la-Coquette, France).

High-resolution respirometry

Cellular respiration rates were determined using the Oroboros 2 k-Oxygraph (Oroboros Instruments, Innsbruck, Austria) in 2 mL of MIR05 buffer (110 mm sucrose, 60 mm K-lactobionate, 0.5 mm EGTA, 3 mm MgCl₂, 20 mm taurine, 10 mm KH₂PO₄, 20 mm HEPES, pH 7.1, 0.1% bovine serum albumin essentially fatty acid free) at 37°C with continuous stirring and chamber volume set at 2 ml. Digitonin titration (stock 8.1 mm) was used to permeabilize the cells. The following substrates and inhibitors were used at the indicated final concentrations 2 mm malate, 10 mm pyruvate, 10 mm glutamate, 5 mm ADP, 10 mM succinate, 2.5–5 μm FCCP, 0.5 μm rotenone, 2.5 μm antimycin A, 2 mm ascorbate, 0.5 mm N,N,N',N'-tetramethyl-p-phenylenediamine (TMPD) and 100 mm azide. DatLab (Oroboros Instruments, Innsbruck, Austria) was used for data acquisition and analysis, which includes calculation of the time derivative of oxygen concentration and correction for instrumental background oxygen flux.

Structural analysis

The model for the complete structure of the mammalian 55S monosome with protein databank code 5aj3 (35) was analyzed with Deep View-Swiss Pdb Viewer v.4.0.4 software to assess the structural effect of p.(Lys119Arg) mutation.

Web resources

Exome Aggregation Consortium database, https://exac.broadinstitute.org

ClinVar, https://www.ncbi.nlm.nih.gov/clinvar

GenBank, https://www.ncbi.nlm.nih.gov/genbank/

OMIM, https://www.omim.org

Acknowledgements

We acknowledge the use of bioresources of the Necker Imagine DNA biobank (BB-033-00065).

Conflict of Interest statement. None declared.

Funding

Association Française contre les Myopathies (AFM)-Téléthon (Grant 19876 to M.D.M.); Agence Nationale de la Recherche (Grant GENOMITANR-15-RAR3-0012-07 to A.R.).