Abstract
Pathogenic variants in cytochrome c oxidase assembly factor 5 (COA5), a proposed complex IV (CIV) assembly factor, have been shown to cause clinical mitochondrial disease with two siblings affected by neonatal hypertrophic cardiomyopathy manifesting a rare, homozygous COA5 missense variant (NM_001008215.3: c.157G>C, p.Ala53Pro). The most striking observation in the affected individuals was an isolated impairment in the early stage of mitochondrial CIV assembly. In this study, we report an unrelated family in who we have identified the same COA5 variant with patient-derived fibroblasts and skeletal muscle biopsies replicating an isolated CIV deficiency. A CRISPR/Cas9-edited homozygous COA5 knockout U2OS cell line with similar biochemical profile was generated to interrogate the functional role of the human COA5 protein. Mitochondrial complexome profiling pinpointed a role for COA5 in early CIV assembly, more specifically, its involvement in the stage between MTCO1 maturation and the incorporation of MTCO2. We therefore propose that the COA5 protein plays an essential role for the biogenesis of MTCO2 and its integration into the early CIV assembly intermediate for downstream assembly of the functional holocomplex.
Introduction
Mitochondria synthesise cellular energy in the form of adenosine triphosphate (ATP) via oxidative phosphorylation (OXPHOS), comprising four respiratory chain complexes and the F1FO ATP synthase. Cytochrome c oxidase (COX), also known as complex IV (CIV), is the terminal electron acceptor of the respiratory chain which catalyses the reduction of molecular oxygen to water. CIV couples this redox reaction to the translocation of protons across the inner mitochondrial membrane, thus contributing to the generation of the proton-motive force harnessed by the F1FO ATP synthase to generate ATP.
CIV is comprised of 14 protein subunits of dual genetic origin; the three core subunits (MTCO1, MTCO2 and MTCO3) are all mitochondrially-encoded while the remaining subunits are encoded by the nuclear genome (Kadenbach, 2017; Wikström et al, 2018). Remarkably, CIV possesses two redox active copper centres (binuclear CuA and mononuclear CuB centres) and two haem groups (haem a and haem a3) (Wikström et al., 2018). These redox active cofactors are crucial for electron transfer within CIV which entails: (i) the receipt of electrons from reduced cytochrome c by the CuA centre in the MTCO2 subunit, (ii) subsequent delivery of the electron by the haem a group in membrane-spanning MTCO1 subunit to (iii) the oxygen-reducing haem a3-CuB centre (Belevich et al, 2006; Kirchberg et al, 2012; Muramoto et al, 2010). These reactions result in the pumping of a total of four protons per oxygen molecule into the mitochondrial intermembrane space (IMS).
As a consequence, an intricate assembly machinery has been generally described for complex IV on a modular basis centred around the three catalytic subunits: MTCO1, MTCO2 and MTCO3 with over 20 unique assembly factors of CIV having been characterised to date (Signes & Fernandez-Vizarra, 2018; Vidoni et al, 2017; Watson & McStay, 2020). These assembly factors are not only involved in the sequential incorporation of the protein subunits but also crucial for auxiliary processes such as translational regulation, protein stabilisation as well as the insertion of cofactors and prosthetic groups (Povea-Cabello et al, 2024; Watson & McStay, 2020).
The cytochrome c oxidase assembly factor 5 (COA5) gene (RefSeq: NM_001008215.3), previously denoted as C2orf64, was first reported in humans when a homozygous missense variant (c.157G>C, p.Ala53Pro) in this gene was shown to cause mitochondrial disease. Biochemical studies of patient-derived fibroblasts revealed isolated COX deficiency, more specifically the accumulation of CIV assembly intermediates and decreased levels of fully assembled CIV holocomplexes, leading the authors to hypothesise that COA5 is involved in the early stages of CIV assembly (Huigsloot et al, 2011). However, earlier studies carried out using the yeast orthologue of human COA5 protein, Pet191, also suggested a putative role in CIV assembly but with no impact on COX translation and copper metalation of the protein (Khalimonchuk et al, 2008; McEwen et al, 1993; Tay et al, 2004). Interestingly, contradicting evidence have been published with regards to the mitochondrial localisation of the Pet191 protein despite being a member of the twin CX9C protein family that are often found to be dependent on the Mitochondrial Intermembrane space Import and Assembly (MIA) pathway (Bragoszewski et al, 2013; Khalimonchuk et al., 2008).
Here we present an unrelated family in which a clinically affected child harbours the identical COA5 missense variant, identified by trio whole exome sequencing (Longen et al). We generated a CRISPR/Cas9-mediated COA5 knockout (COA5KO) cell line to elucidate the role of COA5 in CIV assembly and its implications on mitochondrial health and disease, utilising an array of biochemical tools including mitochondrial complexome profiling.
Results
Clinical summary and genomic studies
A female neonate born to second cousin parents presented with tachypnoeic episodes on the first day of life. Blood lactate was elevated at variable levels between 2.5 to 11 mmol/L. Clinical examination detected hyperdynamic precordium with loud second heart sound and subsequent echocardiography indicated significant biventricular hypertrophic cardiomyopathy with septal hypertrophy and non-compaction appearance of the myocardium. Due to worsening respiratory distress, she was intubated and given inotropic support after birth before being discharged. Liver function tests were abnormal with increased echogenicity of liver on ultrasound. Neuroimaging of the brain was normal. Focused metabolic biochemical investigations identified elevated urinary lactate and ethylmalonic acid. Clinical care was directed to symptomatic management only when a presumptive diagnosis of mitochondrial disorder with hypertrophic cardiomyopathy was made having identified evidence of significant complex IV activity deficiency on muscle biopsy as detailed below. She was readmitted several weeks later with further deterioration of respiratory distress concurrent with a rhinovirus infection, poor cardiac function and severe lactic acidosis and sadly passed away.
Molecular genetic testing eliminated pathogenic mitochondrial DNA (mtDNA) variants following a complete analysis of the mitochondrial genome. Trio whole exome sequencing conducted at the Exeter Genomics Laboratory identified a previously reported COA5 variant (c.157G>C, p.Ala53Pro), confirming both parents to be heterozygous carriers (Fig 1A) (Chen et al, 2023; Huigsloot et al., 2011). The homozygous COA5 missense variant has been recorded on ClinVar as a pathogenic variant associated with isolated COX deficiency (https://www.ncbi.nlm.nih.gov/clinvar/variation/31087/). The c.157G>C COA5 variant causes an amino acid change from alanine to proline at position 53 of the COA5 protein, which is located within the CX9C domain that has suggested linkage to mitochondrial protein localisation (Gladyck et al, 2021) (Fig 1B). The alanine residue is not conserved across species and is only shared between human and zebrafish (Fig 1B). Whilst the application of several in silico pathogenicity prediction tools suggested the c.157G>C COA5 variant to be damaging, the REVEL meta-predictor score was below the recommended threshold for this variant, necessitating further study (Ioannidis et al, 2016).
Patient-derived muscle biopsies and fibroblasts displayed isolated complex IV deficiency
Histochemical analysis of skeletal muscle cryosections from the patient indicated a moderate CIV deficiency in all fibres (Fig 2A). This was corroborated by quadruple OXPHOS immunofluorescence assay which showed a marked loss of MTCO1 immunoreactivity (Fig 2B) while NDUFB8 (complex I subunit) protein levels were normal (Ahmed et al, 2017). The direct measurement of respiratory chain enzyme activities in patient-derived fibroblasts also indicated a severe and isolated complex IV deficiency in the COA5 patient muscle sample (Fig 2C).
Implications of the homozygous COA5 variant on the steady-state level and assembly of complex IV
To characterise the pathogenicity of the COA5 missense variant, SDS-PAGE and BN-PAGE were used to delineate its impact on OXPHOS protein steady-state levels and assembly. Decreased steady-state levels were only observed in CIV subunit, MTCO2 while other OXPHOS complex subunits were unaffected in whole cell lysates of both patient-derived fibroblasts and skeletal muscle biopsy (Figs 3A and 3C). Likewise for BN-PAGE, only complex IV assembly was impaired in patient-derived fibroblasts and skeletal muscle whereas the remaining four OXPHOS complexes (CI, CII, CIII and CV) were unaffected, similar to age-matched controls (Figs 3B and 3D). These observations were further supported by proteomic analysis of the immortalised patient fibroblasts which highlighted the isolated CIV deficiency. CIV protein subunits exhibited a statistically significant decrease in protein abundance up to almost four-fold while CI, CII, CIII and CV protein abundance remained unchanged (Fig 3E).
Functional characterisation of COA5 in a CRISPR/Cas9 knockout cell line
Both the primary and immortalised patient fibroblasts harbouring the homozygous c.157G>C COA5 variant displayed arrested cell growth in culture, likely owing to the severe CIV deficiency and therefore making further experimentation challenging in primary fibroblasts. To enable in-depth characterisation of the functional role and the implicated pathogenicity of COA5, CRISPR/Cas9 gene editing was utilised to generate a homozygous COA5 knockout (COA5KO) in the immortalised human U2OS cell line for a more stable cellular model system. The COA5KO cell line generated contained a homozygous 7 base pair deletion in the COA5 gene (c.287_290+3del, p.Val61del) as verified by Sanger sequencing (Figure EV1). Western blot analyses of the COA5KO cell line also confirmed an isolated CIV defect in terms of protein steady-state level (Fig 4A-B) as well as OXPHOS complex assembly (Fig 4C), successfully mimicking the biochemical phenotype observed in patient-derived biopsies. Interestingly, an accumulation of the complex II subunit, SDHA protein at around 70 kDa was also observed on BN-PAGE analysis of the COA5KO cell line which was absent in the isogenic control (Fig 4C).
Mitochondrial Complexome Profiling of COA5KO cell line
To define the functional involvement of COA5 in CIV assembly, mitochondrial complexome profiling, which combines BN-PAGE and tandem mass spectrometry (LC-MS/MS), was used to determine the presence and arrangement of the OXPHOS system and related protein complexes (Alahmad et al, 2020; Alston et al, 2018; Cabrera-Orefice et al, 2022; Lobo-Jarne et al, 2020). This technique enables visualisation of the specific stage of CIV assembly impacted when COA5 is absent using the COA5KO cell line and potentially uncovering interacting partners of the COA5 protein to elucidate its functional role.
First, complexome profiling (Fig 5) confirmed our result on BN-PAGE_(Fig 4). An increased abundance of respiratory supercomplexes containing complexes I and III2 (S0: I+III2) was observed in the COA5KO cell line (Fig 5, right panel). As observed in BN-PAGE analysis of the COA5KO mitochondrial extracts (Fig 4C), accumulation of the SDHA subunit of CII was also detected by complexome profiling at a molecular size of approx 80 −100 kDa (native calibration of soluble complexes), but not in the isogenic control (highlighted in orange box in Fig EV2, right panel). Interesting, SDHAF1 and SDHAF2 comigrate at the same range indicating an assembly intermediate (Fig EV3, right panel, orange box). When assessing complex III subunits, a complete loss of the supercomplexes III2+IV (S0) was observed in the COA5KO cell line (Fig 5). Of all the OXPHOS complexes, only complex V content and assembly were unaffected (Fig 5).
Next, we had a closer look to complex IV subunits and assembly factors. We noted a complete loss of the COA5 protein as the respective 12 kDa protein was not detected in the knockout cell line but clearly present in the isogenic control as highlighted (Fig 6A, highlighted in yellow). More importantly, the accumulation of early CIV assembly intermediates (Fig 6A, right panel, green boxed subunits and assembly factors) and the loss of fully assembled CIV holocomplexes were observed in contrast to the isogenic control cell line (Fig 6A, left panel). Interestingly, subunits of the MTCO1 and MTCO2 modules accumulate in this intermediate, suggesting that despite the lack of subunits in the individual modules, further assembly is already taking place. This provides supporting evidence to the putative role of COA5 as an assembly factor in early stages of complex IV biogenesis and the isolated CIV defect resulting from the loss of COA5 protein as observed in biochemical studies of patient-derived cells and tissues (Fig 3).
Discussion
This study identified the previously reported c.157G>C, p.Ala53Pro COA5 variant in another family. The isolated CIV deficiency associated with the COA5 variant was observed in the patient fibroblasts and skeletal muscle biopsy, firmly supported by unbiased proteomic profiling of the immortalised COA5 patient fibroblasts as the biochemical signature of COA5 deficiency. Most importantly, a CRISPR/Cas9 COA5KO cell line enabled further interrogation via complexome profiling, narrowing down the involvement of COA5 protein to a specific stage of CIV biogenesis involving MTCO2 stabilisation and its incorporation into the MTCO1-containing subcomplex. This study pinpoints specific questions that arise with regards to the functional role of the COA5 protein as a CIV assembly factor, which we elaborate on further below.
Given that the rare c.157G>C, p.Ala53Pro variant was found only in two incidences where both patients were of identical ethnicity, this could point towards the likelihood of the variant being a founder pathogenic variant within this population. However, this would require further verification through haplotype analysis, necessitating access to patient-derived samples from the previous case. Importantly, overlapping clinical and biochemical phenotypes were observed between the proband in this study and the previously reported patient by Huigsloot and colleagues back in 2019 (Huigsloot et al., 2011), further strengthening the claim that the c.157G>C (p.Ala53Pro) COA5 variant is pathogenic and causative of an isolated mitochondrial complex IV deficiency, associated with loss of steady-state COX proteins and a COX assembly defect in isolation (Fig 2A-C and Fig 3).
To further characterise the functional impact of COA5, we generated a CRISPR/Cas9-mediated knockout. The COA5KO cells successfully replicated the isolated complex IV deficiency observed in the reported patients by indicating a loss of MTCO2 protein but not MTCO1 detected at comparable levels to the isogenic control on SDS-PAGE (Fig 4A). While spontaneous degradation of MTCO1 proteins which are unable to be associated with MTCO2 is normally expected, the unaffected steady-state levels of MTCO1 suggest the presence of stabilised MTCO1 subunits despite not being assembled into functional holocomplexes. The loss of the fully assembled complex IV in COA5KO cells were not only observed in BN-PAGE analysis but also shown in complexome profiling of the COA5KO cells. Most strikingly, this has also been similarly observed in a two-dimensional BN-PAGE of patient fibroblast cell line in the previous report which demonstrated elevated levels of MTCO1 subcomplex but a marked decrease in complex IV holocomplex (Huigsloot et al., 2011). Despite this, the remaining OXPHOS complexes (complex I, II, III and V) were unaffected except the unusual accumulation of a protein complex of about 70 kDa detected with SDHA which corresponds to the size of the individual SDHA protein in COA5KO cell line.
To further infer the role of COA5 protein and its interacting partner, mitochondrial complexome profiling was employed to dissect co-migrating mitochondrial OXPHOS complexes in the COA5KO cell line in comparison to its isogenic control. Notably, the knockout of COA5 protein was confirmed using this method despite the unavailability of a robust antibody for western blotting. Most strikingly, complexome profiling of the COA5KO cell line presented evidence of the implication of COA5 in early complex IV assembly, pinpointing its involvement to the stage between MTCO1 maturation and incorporation of MTCO2 into the early assembly intermediate (Fig 6B). This is supported by the following observations: the accumulation of MTCO1-containing subcomplexes (ranging from approximately 100 to 300 kDa) which are absent in the isogenic control, (ii) the apparent absence of MTCO2 subunit and (iii) the loss of fully assembled complex IV (Fig 6A). Notably, higher molecular mass complexes were shown containing complex IV subunits in COA5KO and could reflect assembly of intermediates with the supercomplex S0 scaffold (Fernández-Vizarra & Ugalde, 2022). However, these are more likely to be prohibitin complexes, which direct the free complex IV subunits for degradation rather than supercomplex formation since complex IV biogenesis was completely obstructed (Fig 6A) (Back et al, 2002; Kohler et al, 2023; Steglich et al, 1999).
Furthermore, the higher abundance of complex I subunits, which was also observed in the SDS-PAGE analysis of COA5KO cell lysates, suggest an accumulation of supercomplex S0 (CI+III2) which could not form a full respirasome in the absence of CIV holocomplexes. Additionally, the complete loss of supercomplex III2+IV reinforced the impact of COA5KO resulting in complex IV loss, hence favouring the formation of supercomplexes only containing complexes I and III2 to mitigate the OXPHOS defect due to the absence of complex IV. Lastly, the accumulation of free SDHA was not seen in the isogenic control. This may either indicate an upregulation of SDHA as a compensatory mechanism due to the OXPHOS defect or an accumulation of a co-migrating complex of SDHA with CII assembly factors, such as SDHAF1 or SDHAF2 (Martínez-Reyes & Chandel, 2020).
Nevertheless, the question remains as to how the c.157G>C (p.Ala53Pro) COA5 variant affects mitochondrial CIV assembly. The variant likely affects the CX9C motifs of the encoded protein which has been associated with the MIA pathway for IMS localisation despite contradicting theories for its yeast counterpart (Bragoszewski et al., 2013; Gladyck et al., 2021; Herrmann & Köhl, 2007; Khalimonchuk et al., 2008; Longen et al., 2009) (Fig 1B). The replacement of alanine by proline could potentially introduce a kink to the alpha helices of the protein and consequently hampering correct folding of the protein to facilitate its import (von Heijne, 1991). Consequently, verifying the submitochondrial localisation of COA5 protein would be fundamental in further investigation to inform its functional role associated with an early stage of CIV assembly. Nevertheless, structural studies of the impact of the single amino acid change in COA5-encoded protein could also be informative for its potential impact on protein folding and subsequent effect onto its correct localisation within the mitochondria.
If the COA5 protein is localised to the IMS, as of most mitochondrial proteins with twin CX9C motifs, it will most likely be involved in the delivery or incorporation of copper atoms into complex IV which takes place in IMS. According to STRING (‘Search Tool for Retrieval of Interacting Genes/Proteins’; https://string-db.org/), COA5 protein is predicted to interact with protein partners involved in mitochondrial protein import into IMS and COX copper centre integration, though mostly based on text mining rather than experimental data. Besides, the twin CX9C motifs have also been linked to Cu(I) binding activity through the reduction of disulphide bonding between the cysteine residues, especially for characterised protein involved in COX maturation (Horng et al, 2004; Horng et al, 2005; Stroud et al, 2015). Interestingly, Nývltová and colleagues also suggested a role for COA5 in the stabilisation of what the authors termed ‘metallochaperone complexes’, comprising COX-specific copper chaperones and haem biosynthesis enzymes required for the maturation and assembly of COX subunits (Nývltová et al, 2022). Although beyond the scope of this manuscript, future experiments should focus on interrogating a direct or indirect role of the COA5 protein in the copper metalation of the MTCO2 subunit which, when affected, leads to destabilisation and eventual degradation of the MTCO2 subunit.
Nevertheless, the possibility for the role of COA5 in the mitochondrial matrix should not be eliminated before its submitochondrial localisation is confirmed. Recently, Peker and colleagues demonstrated a two-step mitochondrial import pathway where substrates of the disulphide relay system, including twin CX9C protein family, rely on oxidative folding through the MIA pathway in the IMS prior to their transport into the matrix (Peker et al, 2023). Matrix localisation of the protein, however, would point towards an involvement in complex IV biogenesis at the RNA or protein levels. A potential hypothesis is that COA5 could act as a translation factor for MTCO2 expression, similar to the involvement of TACO1 and MITRAC12 (otherwise known as COA3) proteins in the translational regulation of MTCO1 protein synthesis (Mick et al, 2012; Richman et al, 2016; Weraarpachai et al, 2009). Moreover, COA5 could also represent a chaperone or stabilising factor for the MTCO2 subunit to facilitate its incorporation into MTCO1-containing subcomplex and therefore impacting synthesis and stability of the MTCO2 protein specifically.
In conclusion, this study provides insights into a distinct mode of pathological complex IV assembly caused by the assembly factor COA5 which specifically disrupts the transitional stage between MTCO1 maturation and MTCO2 incorporation. We present functional evidence to support a role for human COA5 protein in the early stage of complex IV assembly, corroborating its pathogenicity that can contribute to isolated complex IV deficiency. This warrants further investigation to uncover the structural and functional role of COA5 with new insights into understanding its specific involvement in complex IV biogenesis.
Materials and Methods
Whole Exome Sequencing
Trio WES (https://www.exeterlaboratory.com/genetics/genome-sequencing/) was conducted at the Exeter Genomics Laboratory as previously described (Chen et al., 2023).
Histopathological and Biochemical Analyses
10 µm of frozen skeletal muscle sections were used in each assessment. For histopathological studies, H&E staining was employed to determine muscle morphology while sequential COX and SDH histochemistry was used to assess COX activity in muscle fibres. Spectrophotometric measurements of OXPHOS enzyme (complexes I-IV) activities were conducted as described in (Taylor et al) relative to citrate synthase activity. Quadruple immunofluorescence assays were carried out by measuring NDUFB8 (CI) and MTCO1 (Reddy et al) protein abundance against mitochondrial mass marker, porin using in-house analysis software as outlined in (Taylor et al.).
Cell Culture
Patient and age-matched control fibroblasts, as well as U2OS cells, were cultured in High Glucose Dulbecco’s Modified Eagle Media supplemented with 10% fetal bovine serum, 1X non-essential amino acids, 50 µg/ml penicillin, 50 µg/ml streptomycin and 50 µg/ml uridine.
SDS-PAGE
Cell pellets were harvested when reaching 80-90% confluency and resuspended in lysis buffer (50 mM Tris-HCl (pH 7.5), 130 mM NaCl, 2 mM MgCl2, 1 mM phenylmethanesulphonyl fluoride, 1% (v/v) NonidetTM P-40 and 1X EDTA free protease inhibitor cocktail). The resuspended cell pellets were incubated on ice for 10 minutes, before harvesting the resultant supernatant from centrifugation at 500 g at 4°C for 5 minutes and the protein concentrations were determined using Bradford assay.
Skeletal muscle homogenates were prepared by grinding 20 mg of frozen muscle section into powder using pestle and mortar in liquid nitrogen and resuspended in radioimmunoprecipitation (RIPA) buffer containing 1% Igepal, 1.5% Triton-X-100, 0.5% sodium deoxycholate, 10mM β-mercaptoethanol, 0.1% SDS, 1 mM PMSF and 1X EDTA-free protease inhibitor cocktail. The resuspensions were subjected to a 45-minute incubation on ice followed by three rounds of 15-second homogenisation. The soluble proteins were extracted by centrifugation at 14 000g at 4°C for 10 minutes and protein concentrations were estimated using PierceTM BCA Protein Assay Kit.
40 µg of protein extracts were resuspended in 1X Laemmli Sample Buffer and denatured at either 95°C for 5 minutes or 37°C for 15 minutes. The samples were then subjected to 12% SDS-PAGE with the Mini-Protean Tetra Cell system and transferred onto a methanol-activated Immobilon-P Polyvinylidene Fluoride (PVDF) membrane using the Mini Trans-Blot Cell system.
BN-PAGE
Cells were pelleted and resuspended in cell homogenisation buffer comprising of 0.6 M mannitol, 1 mM EGTA, 10 mM Tris-HCL pH 7.4, 1 mM PMSF and 0.1 % (v/v) bovine serum albumin (BSA). The cell suspensions were subjected to three rounds of 15x homogenisation in Teflon glass homogenisers at 4°C with intermittent differential centrifugation at 400 g for 10 minutes at 4°C to separate cytosolic protein fraction. The mitochondrial fractions were pelleted at 11 000 g for 10 minutes at 4°C and washed in cell homogenisation buffer without BSA.
Approximately 40 mg of skeletal muscle sections were processed and homogenised in Muscle Homogenisation Buffer (250 mM sucrose, 20 mM imidazole hydrochloride and 100 mM PMSF) and homogenised in Teflon glass Dounce homogeniser for 15 to 20 rounds at 4°C. The muscle homogenates were then pelleted at 20 000 g for 10 minutes at 4°C and washed twice with Muscle Homogenisation Buffer before pelleting at 20 000 g for 5 minutes at 4°C.
The final pellets were solubilised in 2% n-dodecyl β-D-maltoside (DDM) and subjected to ultracentrifugation at 100 000 g for 15 minutes at 4°C and the supernatants were extracted for protein concentration determination using PierceTM BCA Protein Assay Kit. About 10 µg of mitochondrial protein complexes were loaded and separated in the precast Native PAGE™ 4-16% Bis-Tris 1.0 mm Mini Protein Gel in the XCell SureLock Mini-Cell Electrophoresis System based on manufacturer’s protocol. The protein complexes were then immobilised onto an Immobilon-P PVDF membrane using the MiniTrans-Blot Cell system.
Immunoblotting Analysis
The membranes were blocked in 5% milk for an hour at room temperature before immunoblotting with specific primary antibodies and corresponding HRP-conjugated secondary antibodies as listed: OXPHOS cocktail (ab110411, Abcam), MTCO1 (ab14705, Abcam), MTCO2 (ab110258, Abcam), COXIV (ab14744, Abcam), NDUFB8 (ab110242, Abcam), SDHA (ab14715, Abcam), UQCRC2 (ab14745, Abcam), ATP5A (ab14748, Abcam), GAPDH (600004, ProteinTech), VDAC1 (ab14734, Abcam), Polyclonal Rabbit Anti-Mouse Ig/HRP (P0161, Dako) and Polyclonal Swine Anti-Rabbit Ig/HRP (P0399, Dako).
Finally, resultant signal was detected using SuperSignal™ West Pico PLUS Chemiluminescent Substrate (Thermo Scientific) and analysed with ChemiDoc® XRS+ Imaging Systems and Image Lab Software (BioRad).
On-beads precipitation and protein digestion
Cell pellets were lysed with 120µl RIPA buffer containing protease and phosphates inhibitors. 10µg of protein from all cell lysates were precipitated with 70% acetonitrile onto magnetic beads (MagReSynAmine, Resyn Biosciences). The proteins were washed on the beads with 100% acetonitrile, 70% ethanol and then resuspended in 50µl 50mM ammoniumbicarbonate containing 10mM DTT for reduction of cysteines. Samples were incubated at 37LC for 40min. Then, to alkylate proteins, 50µl of 30mM IAA in 50mM ammonium bicarbonate was added and samples were incubated at RT in the dark for 30 minutes. 0.5µg trypsin was added to each sample for overnight on-beads protein digestion at 37°C. The resulting peptides were concentrated and desalted on EVOTIPS for mass spectrometry analysis according to the standard protocol from EVOSEP.
LC-MS/MS analysis
LC-MS/MS analysis was carried out using an EVOSEP one LC system (EVOSEP Biosystems, Denmark) coupled to a timsTOF Pro2 mass spectrometer, using a CaptiveSpray nano electrospray ion source (Bruker Corporation, Germany).
200 ng of digested peptides were loaded onto a capillary C18 column (15 cm length, 150 μm inner diameter, 1.5 μm particle size, EVOSEP, Odense Denmark). Peptides were separated at 40 °C using the standard 30 sample/day method from EVOSEP.
The timsTOF Pro2 mass spectrometer was operated in DIA-PASEF mode. Mass spectra for MS were recorded between m/z 100 and 1700. Ion mobility resolution was set to 0.85–1.30 V·s/cm over a ramp time of 100 ms. The MS/MS mass range was limited to m/z 475-1000 and ion mobility resolution to 0.85-1.27 V s/cm to exclude singly changed ions. The estimated cycle time was 0.95 s with 8 cylces using DIA windows of 25 Da. Collisional energy was ramped from 20eV at 0.60 V s/cm to 59eV at 1.60 V s/cm.
Raw data files from LC-MS/MS analyses were submitted to DIA-NN (version 1.8.1) for protein identification and label-free quantification using the library-free function. The UniProt human database (UniProt consortium, European Bioinformatics Institute, EMBL-EBI, UK) was used to generate library in silico from a human FASTA file. Carbamidomethyl (C) was set as a fixed modification. Trypsin without proline restriction enzyme option was used, with one allowed miscleavage and peptide length range was set to 7-30 amino acids. The mass accuracy was set to 15ppm and precursor false discovery rate (FDR) allowed was 0.01 (1%).
LC-MS/MS data quality evaluation and statistical analysis was done using software Perseus ver 1.6.15.0CRISPR/Cas9 Gene Knockout
Wildtype U2OS cells were resuspended in room-temperature Nucleofector Solution added with Supplement from the Cell Line NucleofectorTM Kit V (Lonza) at a density of 1 × 106 cells per nucleofection reaction. COA5-targeting sgRNA (Sigma) was incubated at room temperature with HiFi Cas9 nuclease at a 1:1.2 molar ratio to form ribonucleoprotein complexes. The sgRNA sequence designed to mediate CRISPR/Cas9 knockout of the COA5 gene is as follows: 5’-TTTTGAGTGTAAAAGATCAG-3’. The sgRNA-Cas9 RNP complexes were nucleofected into wildtype U2OS cells on the NucleofectorTM 2b Device (Lonza). Nucleofected cells were resuspended in growth media as outlined in Section 2.6.1 before transferring to a 6-well plate. The cells were incubated at 37°C, 5% CO2 for 48 hours before isolating into single cell clones using FACSAria™ Fusion Flow Cytometer (BD Biosciences) in four 96-well plates. Sanger sequencing chromatographs of the selected single cell clones were analysed using Inference of CRISPR Edit (ICE) analysis (Synthego; https://ice.synthego.com/) to identify isogenic controls and COA5 knockout cell line.
Complexome Profiling (BN-PAGE, Trypsin Digestion, LC-MS/MS, Quantification)
Enriched mitochondrial proteins were extracted from COA5KO and isogenic control cell lines and solubilised with digitonin as described in (Giese et al., 2021). Equal amount of the solubilised mitochondrial protein extracts were subjected to a 3 to 18% acrylamide gradient gel (14cm ×14cm) for BN-PAGE as outlined in (Wittig, Braun and Schägger, 2006). The gel was then stained with Coomassie blue and cut into equal fractions then transferred to 96-well filter plates. The gel fractions were then destained in 50 mM ammoniumbicarbonate (ABC) followed by protein reduction using 10 mM DTT and alkylation in 20 mM iodoacetamide. Protein digestion was carried out in digestion solution (5 ng trypsin/μl in 50 mM ABC, 10% acetonitrile (ACN), 0.01% (w/v) ProteaseMAX surfactant (Promega), 1 mM CaCl2) at 37°C for at least 12 hours. After the recovery in the new 96-well plate, the peptides were dried in a SpeedVac (ThermoFisher) and finally resuspended in 1% ACN and 0.5% formic acid. Nano liquid chromatography and mass spectrometry (nanoLC/MS) was carried out on Thermo Scientific™ Q Exactive Plus equipped with ultrahigh performance liquid chromatography unit Dionex Ultimate 3000 (ThermoFisher) and a Nanospray Flex Ion-Source (ThermoFisher). The MS data was analysed using MaxQuant software at default settings and the recorded intensity-based absolute quantifications (iBAQ) values were normalised to isogenic control cell line.
Data Availability
The mass spectrometry proteomics data for label-free whole cell proteomics and complexome profiling produced in this study have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository (Perez-Riverol et al, 2022) and assigned the dataset identifier PXD050891 and PXD053461 respectively.
Author Contributions
Conceptualization: AP, MW, IW and RWT
Data curation and formal analysis: JXT, ACO, JM, LST, GM, SKS, AS, KS, LH, SH, TAN
Clinical care of the family: JD
Supervision: AP, MW, CBJ, IW and RWT
Drafting the manuscript: JXT, AP, MW and RWT
Critical revision and editing of the manuscript: all authors
Funding acquisition: AP, MW, IW and RWT
Disclosure and competing interests statement: The authors declare no conflicts of interest.
The Paper Explained
PROBLEM
A homozygous COA5 variant was previously reported to cause a mitochondrial cardiomyopathy phenotype but the role of the COA5 protein is poorly defined despite being proposed to be an assembly factor for cytochrome c oxidase (COX), an essential protein of the mitochondrial respiratory chain.
RESULTS
Through a series of biochemical experiments including LC-MS/MS-based complexome profiling, our study showed that the human COA5 protein is essential for the biogenesis and incorporation of the MTCO2 subunit, specifically, to form functional COX holocomplexes.
Additionally, we present an additional case of COA5-related mitochondrial disease, confirming the pathogenicity of the reported homozygous COA5 variant that leads to mitochondrial dysfunction.
IMPACT
We have provided additional insights into the functional role of human COA5 protein and its specific implications on early-stage COX assembly, correlating to isolated COX deficiencies observed in patients harbouring a pathogenic, homozygous COA5 variant.
Acknowledgements
RWT is supported by the Wellcome Centre for Mitochondrial Research (203105/Z/16/Z), the Medical Research Council (MRC) International Centre for Genomic Medicine in Neuromuscular Diseases (MR/S005021/1), the UK NIHR Biomedical Research Centre in Age and Age-Related Diseases award to the Newcastle upon Tyne Hospitals NHS Foundation, the Lily Foundation, LifeArc and the UK NHS Highly Specialised Service for Rare Mitochondrial Disorders. RWT and MW are supported by the Pathology Society; RWT, MW and AP are supported by Mito Foundation. JXT was supported by a PhD studentship from the Lily Foundation and a Newcastle University Overseas Research Studentship award. CBJ is supported by funding from the Academy of Finland (decision #336455), the Magnus Ehrnroot Foundation and the Jane and Aatos Erkko Foundation (#230004). The Proteomics Core Facility, University of Oslo/Oslo University Hospital is supported by the Core Facilities program of the South-Eastern Norway Regional Health Authority, and is a member of the National Network of Advanced Proteomics Infrastructure (NAPI) which is funded by the Research Council of Norway INFRASTRUKTUR-program (project number: 295910). IW is supported by the Deutsche Forschungsgemeinschaft (DFG): SFB1531-S01, project number 456687919 and WI 3728/3-1, project number 515944830.