Recessive PPTC7 deficiency triggers excessive mitophagy to cause a severe inborn error of metabolism with hypomyelinating leukodystrophy
Keri-Lyn Kozul, Ali AlAsmari, Essa Alharby, Reem Zakzouk, Youmian Yan, Aziza Mushiba, Anwar Alhamad, Emily Harrelson, Maya Ayach, Kevin Cho, Heba Zahid, Francisca De Luna Vitorino, Richard Searfoss, Xingyu Liu, Mohammed A. Saleh, Muhammad Latif, Lianjie Wei, Ali Aldawood
TL;DR
A genetic deficiency in PPTC7 causes severe metabolic and neurological issues in humans by triggering excessive mitophagy.
Contribution
This study identifies the first known human inborn error of mitophagy caused by PPTC7 deficiency.
Findings
PPTC7 deficiency in patients leads to excessive BNIP3- and NIX-mediated mitophagy.
The p.D158N variant disrupts PPTC7's phosphatase function and regulation of mitophagy.
Exogenous PPTC7 rescues increased mitophagy in patient-derived cells.
Abstract
The mitochondrial phosphatase PPTC7 has emerged as a potent regulator of metabolism and mitophagy as its global knockout leads to perinatal lethality in mice. However, no known Mendelian diseases have been linked to PPTC7 deficiency, rendering its role in human pathophysiology unclear. Here, we identify two independent homozygous variants in PPTC7: a missense variant, p.D158N, and a duplication variant (c.*57dup) within the 3’ untranslated region (UTR). These variants were detected in three patients from two unrelated families presenting with a primary mitochondrial disease characterized by hypomyelinating leukodystrophy, recurrent metabolic and lactic acidosis, and anemia with immune dysregulation. Patient samples, including plasma and primary fibroblasts, showed robust metabolic and mitochondrial dysfunction, with substantial phenotypic overlap with Pptc7 knockout murine fibroblast…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsGenomics and Rare Diseases · Mitochondrial Function and Pathology · RNA regulation and disease
Introduction
Mitochondria serve as cellular metabolic hubs, facilitating the catabolism of nutrients and the production of energy in the form of ATP^1^. Given these essential roles, cells have evolved quality control mechanisms to ensure a healthy mitochondrial pool^2^. In some cases, energy demand or environmental stressors signal for healthy mitochondria to fuse with damaged mitochondria and compensate for the dysfunction. However, when mitochondria are beyond repair, cells engage the autophagic machinery to remove damaged mitochondria in a process termed mitophagy.
The importance of mitophagy is evidenced by the pathologies resulting from its dysregulation in humans. Genetic variants in mitophagy-associated genes are associated with Mendelian diseases such as Amyotrophic Lateral Sclerosis (ALS)^3^ and Parkinson’s disease^4,5^. For example, biallelic variants in PRKN and PINK1, encoding the Parkin and PINK1 proteins, cause autosomal recessive juvenile Parkinson disease type 2 (PARK2; Mendelian Inheritance in Men/MIM #600116)^6,7^ and early-onset Parkinson disease type 6 (PARK6; MIM #605909)^8,9^, respectively. Mono- and bi-allelic variants in the selective autophagy receptor and mitophagy-associated gene SQSTM1 (also known as p62) cause neurological and neurodegenerative phenotypes, including frontotemporal dementia and ataxia^10^, amongst others (MIM #616437, 617158, 617145, 167250)^11^. Mechanistically, these proteins act as sensors and effectors of mitochondrial quality control, with their loss leading to the accumulation of damaged and dysfunctional mitochondria. The persistence of damaged mitochondria causes numerous liabilities, including compromised cellular energetics and the accumulation of potentially cytotoxic molecules such as reactive oxygen species. Such cellular challenges are likely to manifest as pathophysiology, particularly in organs with high energy demand such as the brain, heart, and muscles.
Given the multitude of diseases associated with the loss of mitophagy, some have proposed that bolstering mitophagy holds substantial therapeutic promise to maintain or even promote healthy populations of mitochondria. Interestingly, however, recent data have demonstrated that unrestrained, excessive mitophagy also triggers human disease. Biallelic variants in FBXL4, a gene encoding the substrate-targeting subunit of an E3 ubiquitin ligase, cause a severe mitochondrial DNA (mtDNA) depletion syndrome, type 13 (MTDPS13; MIM# 615471)^12,13^. Despite the identification of these pathogenic variants a decade prior, the mechanisms underlying FBXL4-associated disease pathology remained unclear until recently, when FBXL4 was shown to interact with the mitophagy receptors BNIP3 and NIX to facilitate their ubiquitination and subsequent degradation^14–17^. Mutations in FBXL4 perturb this turnover mechanism, leading to sustained BNIP3 and NIX expression and the degradation of mitochondria at an excessive rate. This loss of mitochondrial content drives the loss of mtDNA, thus explaining the hallmark feature of MTDPS13 caused by FBXL4 mutations.
Interestingly, we and others recently identified a second regulator of BNIP3 and NIX turnover: the mitochondrial phosphatase PPTC7^18–21^. PPTC7 bridges BNIP3 and NIX to FBXL4 to allow their ubiquitination and subsequent degradation^19,20^. Consistently, PPTC7 knockout cells harbor elevated levels of BNIP3 and NIX concomitant to decreased mitochondrial content^18,22^, phenocopying FBXL4 knockout models. In parallel to these cellular models, knockout of either Fbxl4 or Pptc7 in mice promotes robust metabolic dysfunction culminating in perinatal lethality^20,22,23^, underscoring the importance of these genes in early mammalian development.
The phenotypic similarities between Fbxl4 and Pptc7 knockout animal models, along with their recently discovered molecular interplay, strongly suggest that loss of function (LoF) variants in PPTC7 may trigger a mitochondrial disorder similar to MTDPS13 in humans. To date, however, there are no reports of Mendelian diseases linked to variants in PPTC7. Furthermore, over 800K population-level genomes show no reported homozygous LoF variants in PPTC7 in the Genome Aggregation Database (gnomAD, v4.1.0)^24^, strongly suggesting that PPTC7 LoF is intolerable and that associated diseases, if discovered, will be ultra-rare, recessive, severe, and probably lethal in nature.
In this study, we identified two independent homozygous variants in PPTC7 in patients with previously undiagnosed mitochondrial diseases at the genetic level. First, we identified two siblings with a novel and clinically distinct inborn error of mitophagy caused by a homozygous missense variant in PPTC7, p.D158N. While these patients presented with a phenotype suggestive of primary mitochondrial disease, such as persistent lactic acidosis, they also manifested hypomyelinating leukodystrophy, a distinct phenotype not typically associated with mitophagy-related disorders. Metabolomic profiling of plasma revealed a surprising breadth of metabolic abnormalities in PPTC7 D158N patients, suggesting simultaneous dysregulation of branched chain amino acid catabolism, the urea cycle, and glucose metabolism. Cellular models with D158N-disrupted PPTC7, including patient primary fibroblasts, showed similar defects in mitochondrial metabolism, including decreased mtDNA and mitochondrial protein levels, lower respiratory capacity, and decreased metabolic flux. We confirmed that patient fibroblasts had elevated BNIP3 and NIX expression, leading to elevated mitophagy and a fragmented mitochondrial network, which was rescued by exogenous wild-type PPTC7 expression. Finally, biochemical studies revealed that the p.D158N variant of PPTC7 disrupted binding to its requisite cofactor manganese, which was accompanied by a complete ablation of BNIP3 and NIX binding.
In parallel, we identified another patient with a homozygous single base pair duplication in the 3’ untranslated region (3’UTR) of PPTC7 that is highly conserved amongst primates. CRISPR-Cas9 modeling of this variant revealed a moderate but reproducible increase in BNIP3 and NIX protein expression and mitophagy that was rescued by expression of exogenous PPTC7. Collectively, these studies have revealed the candidacy of the PPTC7-associated inborn error of mitophagy in humans and have linked each of these variants to excessive mitophagy. Our data suggest that PPTC7 deficiency will likely be ultrarare and cause a severe inborn error of mitophagy with unique metabolic and pathophysiological presentations.
Results
Recruitment
The two families characterized in this study were recruited through the Study of Primary Mitochondrial Diseases in The Gulf Cooperation Council (GCC) region, which has >1000 patients with a diagnosis or suspicion of primary mitochondrial diseases due to variants in known or novel candidate genes (Alharby et al, manuscript in preparation).
Clinical, neurological, and biochemical evaluation
Case 1 (III.1):
An 18-month-old female child, born to consanguineous parents (first cousins), was referred to the neurology clinic for evaluation of global developmental delay and hypotonia. She was initially noted to have significant delay across all domains at the age of 7 months (Table 1). Antenatal and perinatal histories were unremarkable. However, at one week of age, she began to experience recurrent episodes of mild-to-moderate metabolic acidosis, with elevated lactate levels ranging from 6 to 12 mmol/L associated with recurrent chest infections. These episodes continued intermittently, leading to repeated hospitalizations. Upon the clinical genetic evaluation, the child had a height and weight below the 3^rd^ percentile as well as microcephaly. She had dysmorphic features including thick lips, low-set ears, and a flat nasal bridge. Neurological examination revealed axial and appendicular hypotonia. Reflexes were absent, and pupillary reflexes were sluggish but equally reactive bilaterally. The child was able to move all limbs against gravity, indicating some motor function, but significant weakness was apparent.
Brain MRI at 1 year and 3 years of age revealed reduction in volume of cerebral white matter and markedly diminished myelination (Fig. 1A–F). The cerebellar white matter was less affected. Myelin appeared on MRI scans after the age of 3 years with a pattern of hypomyelination. A scan was done during an acute crisis and showed an abnormal signal in the right caudate and right putamen with reduced diffusivity also in the corpus callosum, compatible with acute/subacute injury. The volume of white matter and the thickness of the corpus callosum remained diminished. No cerebellar atrophy was noted. There was no enhancement after intravenous administration of contrast. The diffusion weighted images show reduced diffusivity in the affected white matter. The spinal cord was normal.
Laboratory investigations including Complete Blood Count (CBC) showed neutropenia and mild anemia (Table 2). Basic metabolic screening, including liver function tests (LFT), ammonia, tandem mass spectrometry (MS), and creatine kinase (CK), were all within normal limits. Lactate was persistently high, falling within the range of mild-to-moderate metabolic acidosis (6–12 mmol/L). Plasma amino acids showed elevated alanine at 1709 μmol/L (reference range = 139–474 μmol/L), proline at 572.5 μmol/L (85–203 μmol/L), leucine at 207 μmol/L (48–207 μmol/L), and lysine at 228 μmol/L (49–204 μmol/L).
Currently, the patient is 4 years old, and despite early management, continues to experience recurrent hospitalizations (2–3 admissions/month) due to metabolic crises and recurrent infections, including community-acquired pneumonia, aspiration pneumonia, sepsis, and central line-associated bloodstream infections (CLABSI), all of which have been managed with intravenous antibiotics. Additionally, she has ongoing neutropenia and anemia.
Case 2 (III.2):
Her brother, a 2-year-old male, was referred to the metabolic clinic at 3 months of age for evaluation of recurrent vomiting and hypotonia. The pregnancy and delivery were uncomplicated. Three days post-birth, the patient began to experience recurrent episodes of fever, recurrent episodes of vomiting, metabolic acidosis, and elevated lactate, which required hospitalization (Table 1). Since then, he has had frequent admissions, including several pediatric intensive care unit (PICU) admissions for severe metabolic acidosis management. Later, generalized tonic-clonic seizures have been noted, particularly during metabolic crises. The patient has demonstrated developmental delay in motor and language skills, and he has not yet reached milestones appropriate for his age. Upon clinical assessment, his growth parameters showed weight below the 3^rd^ percentile as well as microcephaly. The patient showed dysmorphic features similar to those seen in his older sister. During his neurological examination, pupils were equal and bilaterally reactive with axial and appendicular hypotonia. The child was able to move all limbs symmetrically against gravity, though deep tendon reflexes were diminished.
A brain MRI at 1 month was unremarkable (Fig. 1G), however, a repeat MRI at 1 year of age showed hypomyelination and extensive cerebral white matter signal alterations with reduced diffusivity (Fig. 1H–K). Similar abnormalities were seen in the posterior brainstem white matter tracts, middle cerebral peduncles, and portions of the cerebellar white matter.
His CBC showed anemia and neutropenia (Table 2, Fig. 1L). The immunological workup including lymphocyte markers, immunoglobulin levels, and bone marrow aspiration were unremarkable (Supplemental Fig. 6A and B). Ammonia levels were elevated, in some instances reaching 300 μmol/L, although most ammonia levels were within normal limits. There were persistent lactate elevations in the moderate-to-severe range of metabolic acidosis (6–13.3 mmol/L). Plasma amino acids analysis revealed elevated levels of alanine at 1009 μmol/L (reference range = 139–474 μmol/L), serine at 487 μmol/L (69–271 μmol/L), proline at 371 μmol/L (85–203 μmol/L), and lysine at 447 μmol/L (49–204 μmol/L).
Despite ongoing therapeutic and physical therapy interventions, the patient continues to be hospitalized almost weekly due to recurrent infections including gastroenteritis, sepsis, pneumonia, urinary tract infections, and central line-associated bloodstream infection (CLABSI). These episodes are consistently associated with metabolic acidosis, elevated lactate levels, persistent anemia, and neutropenia, despite the negative hematological and immunological investigations.
PPTC7 variant identification using WES and WGS
To understand the genetic origins of this pathophysiology, duo whole exome sequencing (WES) for known and clinically relevant primary mitochondrial disease genes was performed but was negative for potential single nucleotide variants (SNVs) and copy number variation (CNV). WES analysis was expanded to include candidate variants in novel candidate genes with clinical and biological relevance, which identified a homozygous missense variant in the mitochondrial phosphatase, PPTC7 (NM_139283.2: c.472G>A, p.D158N)) in both patients (Fig. 1M). This variant is absent in the gnomAD database and affects a highly conserved residue (Fig. 1N). This variant has a high pathogenicity CADD score of 32 and is predicted to be strongly pathogenic by PrimateAI-3D and AlphaMissense, and deleterious or damaging by other in silico pathogenicity predictors like SIFT and MutationTaster. To exclude other gene variants (including SNVs and CNVs) in non-coding regions that may have been missed by WES, duo whole genome sequencing (WGS) was performed and revealed the same PPTC7 variant as the only potential cause that segregates with the disease in this family. Given the intolerance for homozygous LoF variants in PPTC7 in the gnomAD database (with a maximum probability score of being LoF intolerant (pLI) of 1)^24^ as well as the previous links of Pptc7 loss of function to severe metabolic dysfunction and perinatal lethality in mice^22^, we moved forward with the characterization of this potentially novel genetic driver of the suspected primary mitochondrial disease in this family.
Targeted and metabolomic profiling of patient plasma suggests broad mitochondrial dysfunction
A follow-up targeted urine organic acids analysis on the index patient (III.1) using gas chromatography–mass spectrometry (GC-MS) showed an increased urinary excretion of lactic acid, 3-hydroxyisovaleric acid, and fumaric acid (Fig. 2A), a biochemical pattern suggestive of mitochondrial dysfunction. To understand the extent to which the PPTC7 p.D158N variant disrupted systemic metabolism, we profiled plasma collected at two distinct timepoints from both patients, along with plasma from age-matched healthy pediatric controls (n=4). Using liquid chromatography-mass spectrometry (LC-MS) metabolomics, we identified 155 circulating polar metabolites, with over one-third of these showing significant changes in the D158N patients relative to control samples (Fig. 2B–C). Consistent with clinical metabolic profiles, plasma from the siblings carrying the PPTC7 D158N allele showed elevated levels of lactate as well as leucine, lysine, proline, and alanine (Fig. 2B), confirming the quality and reproducibility of our metabolomics assay.
Pathway analysis of altered metabolites confirmed changes consistent with primary mitochondrial disease. Multiple metabolites associated with glucose metabolism, including glycolytic intermediates glucose 6-phosphate (G6P), fructose-6-phosphate (F6P), dihydroxyacetone phosphate (DHAP), and pyruvate were elevated in the D158N patient plasma, while glucose itself was decreased (Fig. 2C). Such patterns may indicate elevated glycolytic flux in these patients—a known compensatory metabolic program in patients with mitochondrial disease which likely drives lactate accumulation^25^. Pentose phosphate pathway (PPP) intermediates, including G6P and ribose 5-phosphate (R5P), were also elevated in patient plasma (Fig. 2C), suggesting increased glucose utilization in the PPTC7 D158N patients. Other metabolite trends previously associated with mitochondrial disease were identified^26,27^, such as the accumulation of select TCA cycle intermediates (Fig. 2D) and various short-chain acylcarnitines (Fig. 2E), suggesting disrupted TCA cycle flux and fatty acid oxidation, respectively.
Interestingly, we identified multiple metabolite alterations previously mapped to human disease but typically seen only in a distinct subset of inborn errors of metabolism (IEMs). For instance, PPTC7 D158N patient plasma showed accumulation of the branched-chain amino acid (BCAA) leucine, as well as intermediates of BCAA catabolism (Fig. 2F)—hallmarks of maple syrup urine disease (MSUD)^28,29^. Additionally, PPTC7 D158N patients had increased circulating ornithine but decreased citrulline (Figure 2G), a metabolite signature typically seen in ornithine transcarbamylase (OTC) deficiency^30^, a proximal urea cycle defect. As the urea cycle plays a key role in ammonia detoxification, disruptions in this pathway likely contribute to the hyperammonemia observed in the PPTC7 D158N patients (Table 1). Finally, PPTC7 D158N patients harbor markers of select metabolic acidurias, including an accumulation of methylmalonate (seen in methylmalonic acidemia) and sarcosine (seen in sarcosinemia) (Fig. 2H). Importantly, each of these IEMs derive from the disruption of metabolic pathways in which key enzymes localize to the mitochondrial matrix (Fig. 2I). Collectively, these data suggest that the metabolic defects seen in the patients derive from broad dysfunction across multiple pathways associated with mitochondria, further supporting PPTC7 candidacy.
Isotopic flux tracing reveals substantial metabolic rewiring in fibroblasts homozygous for the PPTC7 variant
To understand the specific pathways perturbed in cells homozygous for the PPTC7 p.D158N variant, we established patient-derived fibroblasts from skin biopsies, as well as age-matched control fibroblasts. Consistent with mitochondrial dysfunction, patient fibroblasts displayed significantly reduced oxygen consumption rates in phosphorylating conditions upon addition of pyruvate/malate and succinate (Fig. 3A), suggesting deficiencies in Complex I- and Complex II-mediated respiration, respectively. To understand the extent to which mitochondrial metabolism was disrupted in these cells, we performed isotope-based metabolomics in patient and control fibroblasts. An initial characterization revealed that patient fibroblasts showed a reduction in the total abundance of many mitochondria-associated metabolites, including choline, carnitine, creatine, and proline (Supplemental Fig. 1A-E) as well as TCA cycle metabolites (Fig. 3B). To understand if these decreases in TCA cycle intermediate levels resulted from decreased flux, we labeled patient and control fibroblasts with U-^13^C_6_-glucose for four hours, harvested metabolites, and quantified isotopic labeling patterns via LC-MS. While patient fibroblasts showed robust labeling in glycolytic intermediates (Supplemental Fig. 1F), the incorporation of the ^13^C label into the TCA cycle was inefficient, labeling under 15% of citrate and aconitate with only trace labeling of other TCA cycle metabolites (Supplemental Fig. 1G). As the p.D158N variant of PPTC7 is likely inactivating, we profiled Pptc7 knockout (KO) mouse embryonic fibroblasts (MEFs) as an orthogonal model to test how genetic perturbation of Pptc7 affected TCA cycle flux. Using a similar experimental set up, we found that most TCA cycle metabolites had decreased isotopic labeling in Pptc7 KO MEFs, with concomitant increases in unlabeled (i.e., m+0) TCA cycle metabolites (Fig. 3C). Together, these data suggest that loss of PPTC7 function disrupts total TCA cycle pools as well as TCA cycle flux in cultured fibroblasts.
Our metabolomic profiling of patient plasma suggested broad metabolic defects that extend beyond the TCA cycle, prompting us to query our glucose tracing metabolomics datasets in both patient fibroblasts and Pptc7 KO MEFs to identify commonly perturbed metabolic signatures. We found significant decreases in labeled proline intermediates in both patient fibroblasts and Pptc7 KO MEFs, suggesting compromised biosynthesis (Supplemental Fig. 1H, I). As steps of proline biosynthesis occur within mitochondria and require NADPH^31^, these data could suggest alterations in organellar redox status; consistently, patient fibroblasts showed elevated levels of oxidized glutathione but no change in reduced glutathione relative to control fibroblasts (Supplemental Fig. 1J). Altered one carbon (1C) metabolism recently has been linked to mitochondrial dysfunction and mitochondrial disease (Fig. 3D)^32,33^; interestingly, glycine and serine are mildly upregulated in Pptc7 KO MEFs (Supplemental Fig. 1K) but significantly upregulated in PPTC7 D158N patient fibroblasts, along with 1C metabolites S-adenosylmethionine (SAM) and S-adenosylhomocysteine (SAH) (Fig. 3E). However, Pptc7 KO MEFs and PPTC7 D158N fibroblasts displayed lower isotopic labeling of glycine and serine from U-^13^C_6_-glucose, with PPTC7 D158N fibroblasts also showing lower synthesis of SAM and SAH (Supplemental Fig. 1L-M), suggesting alternative pathways contribute to these elevated metabolite levels. Finally, a recent study linked increased flux through the glycerol-3-phosphate (i.e., Gro3P) shuttle to mitochondrial disease pathogenesis^34^, as this pathway regenerates cytosolic NAD+ to alleviate reductive stress caused by decreased Complex I activity (Fig. 3F). Both Pptc7 KO MEFs (Fig. 3G) and PPTC7 D158N patient fibroblasts (Figure 3H) showed significantly elevated flux of U-^13^C_6_ glucose into m+3 Gro3P, consistent with this shuttle facilitating NAD+ regeneration in both cell models. Collectively, these data suggest disruption of PPTC7 triggers metabolic defects consistent with mitochondrial disease across multiple cell models.
Patient fibroblasts display decreased mitochondrial content, mitochondrial morphology defects and elevated amino acid levels
We recently demonstrated that PPTC7 enables mitochondrial metabolism^22^, which likely occurs at least in part through its potent suppression of BNIP3/NIX-mediated mitophagy^18–20^. Given the extensive metabolic defects observed in patient plasma and fibroblasts, we hypothesized that this variant caused decreased mitochondrial content via excessive mitophagy. To test this, we quantified mitochondrial DNA (mtDNA) levels and found significant reductions in patient fibroblasts relative to control cells (Fig. 4A). As mtDNA levels could decrease independent of mitochondrial mass, we employed proteomics to quantify relative mitochondrial protein levels as an orthogonal metric for total mitochondrial content. Consistent with our hypothesis, patient fibroblasts harbored a global reduction in mitochondrial proteins relative to control fibroblasts (Fig. 4B). Importantly, this included proteins from metabolic pathways compromised in PPTC7 D158N fibroblasts, such as those involved in the TCA cycle (Fig. 4C) and subunits of Complex I (Fig. 4D), consistent with the observed reduction in oxygen consumption rates in these cells (Fig. 3A). These data suggest that patient fibroblasts manifest metabolic defects concomitant to decreased mitochondrial content, similar to our previous findings in Pptc7 KO MEFs^18^ and PPTC7 KO 293T cells^35^.
One driver of decreased mitochondrial protein expression in PPTC7 loss of function systems could be diminished mitochondrial mass per cell. We posited that the isolation and enrichment of mitochondrial fractions from whole cells would allow us to normalize for and thus test mass-specific contributions to diminished mitochondrial protein expression. We performed proteomic analysis of whole cells and mitochondrial-enriched fractions from wild-type or PPTC7 KO HeLa cells and found that PPTC7 KO HeLa cells displayed a significant decrease in most mitochondrial proteins (Supplemental Fig. 2A), mirroring the proteomic changes in the PPTC7 D158N patient fibroblasts. Importantly, the effects of PPTC7 KO on mitochondrial protein expression were reduced in mitochondrial fractions relative to whole cells (Supplemental Fig. 2A-C), with select pathways such as the TCA cycle and complex I showing significantly rescued protein expression upon normalization of isolated mitochondrial fractions (Supplemental Fig. 2D, E). These data suggest PPTC7 maintains mitochondrial content at least in part through sustaining mitochondrial mass.
To further characterize mitochondrial alterations in patient fibroblasts, we assessed mitochondrial morphology through immunostaining and confocal microscopy. While control fibroblasts showed a reticular network of mitochondria (Fig. 4E), fibroblasts from patient III.2 displayed a fragmented mitochondrial network containing swollen organelles (Fig. 4F). Interestingly, these mitochondria appeared similar to those seen in fibroblasts from patients harboring variants in the E3 ubiquitin ligase FBXL4^15^. Given that FBXL4 works in conjunction with PPTC7 to suppress mitophagy by downregulating the mitophagy receptors BNIP3 and NIX^19,36^, we immunostained for NIX and found an upregulation of this mitophagy receptor in patient fibroblasts relative to control fibroblasts (Fig. 4E–F). These data suggested that the morphological changes seen in PPTC7 D158N mitochondria may result from elevated mitophagy. To test this, we pharmacologically activated BNIP3/NIX-mediated mitophagy in control fibroblasts with the iron chelator deferoxamine (DFO) and found that DFO-treated control fibroblasts developed similar patterns of fragmentation and swollen mitochondria as seen in patient fibroblasts (Supplemental Fig. 2E-G). Critically, transduction of lentiviral wild-type PPTC7 into D158N patient fibroblasts largely rescued the fragmented and swollen mitochondria, and simultaneously decreased NIX levels (Fig. 4G). Collectively, these data suggest that PPTC7 D158N patient fibroblasts harbor mitochondrial defects consistent with dysregulated mitophagy, which are rescued through restoration of wild-type PPTC7 expression.
Given the likely increase in mitophagy in PPTC7 D158N patient fibroblasts, we returned to our metabolomics and proteomics datasets to determine if other markers consistent with BNIP3- and NIX-mediated mitophagy were identified. Unfortunately, neither BNIP3 nor NIX were identified in our proteomics datasets. However, reanalysis of our metabolomics data revealed that PPTC7 D158N patient fibroblasts showed increased levels of most amino acids relative to control fibroblasts (Fig. 4H), which may result from the accelerated autophagic breakdown of mitochondrial proteins. Taken together, these data suggest that PPTC7 D158N fails to maintain mitochondrial content and morphology, likely due to an inability to suppress BNIP3- and NIX-mediated mitophagy.
PPTC7 D158N fails to suppress the mitophagy receptors BNIP3 and NIX
To test how the PPTC7 p.D158N variant affects mitophagy, we investigated the extent to which this variant, as well as a previously published catalytically inactive mutant (D78A)^21^ could suppress BNIP3 and NIX protein expression. Utilizing HeLa Flp-In T-REx cells, we expressed wild-type PPTC7, PPTC7 D78A, and the pathogenic variant PPTC7 p.D158N and probed for BNIP3 and NIX both basally and in response to pseudohypoxia—an established method for the transcriptional upregulation of these mitophagy receptors^37,38^. Consistent with our previous findings^21^, overexpression of wild-type PPTC7 decreased BNIP3 and NIX protein in both untreated and CoCl_2_-treated cells (Fig. 5A). However, the D158N mutant failed to decrease BNIP3 and NIX levels in either condition (Fig. 5A), behaving similarly to the D78A mutant, which we and others have shown cannot bind to these mitophagy receptors^19,21^.
We next used the pH-dependent mitophagy reporter mt-Keima^39^ to test whether elevated levels of BNIP3 and NIX in PPTC7 D158N cells increased mitophagic flux. Consistent with previous studies^19,20,22^, PPTC7 knockout in HeLa cells increased the ratio of acidic (excitation 566 nm) to healthy (excitation 458 nm) mitochondria relative to parental cells, revealing elevated mitophagy (Fig. 5B, C, Supplemental Fig. 3A-D). Importantly, this increased mt-Keima ratio was rescued by stable expression of wild-type PPTC7, but not the p.D158N variant, in both untreated and pseudohypoxic conditions (Fig. 5B, C, Supplemental Fig. 3A-D). Similar trends were seen for basal mitophagy, as overexpression of wild-type PPTC7 reduced mt-Keima flux in wild-type HeLa cells, but neither PPTC7 D78A nor D158N were able to do so (Supplemental Fig. 3E-F). Patient fibroblasts displayed elevated BNIP3 and NIX expression relative to control cells (Fig. 5D), which was rescued by exogenous expression of FLAG-tagged PPTC7 (Fig. 5E). Consistently, patient fibroblasts showed increased mt-Keima reported mitophagic flux that was rescued by lentiviral transduction of exogenous wild-type PPTC7-FLAG (Fig. 5F, G). Finally, electron microscopy of patient fibroblasts revealed dilated endoplasmic reticulum and multiple autophagic vacuoles at various stages, including autophagosomes with mitochondria (Fig. 5H). Taken together, these data suggest that the PPTC7 p.D158N variant diminishes suppression of BNIP3 and NIX expression, triggering excessive mitophagy.
PPTC7 D158N ablates phosphatase activity and BNIP3/NIX association through disrupted metal binding
Given the apparent loss of function of the p.D158N variant, we sought to understand the mechanism by which this mutation may inactivate PPTC7. PPTC7 belongs to the family of PP2C phosphatases, which bind divalent cations requisite for enzymatic activity^40–42^. To enable this binding, PP2Cs possess several motifs containing highly conserved aspartate residues^43^, one of which is D158 in PPTC7 (Fig. 6A). Consistently, an AlphaFold3 structural model of PPTC7 identified D158 as a key coordinating residue of manganese ions in a high-confidence prediction (Fig. 6B, Supplemental Fig. 4A). These data, together with the loss of function phenotypes seen across the patient samples, led us to hypothesize that an asparagine substitution at D158 would disrupt PPTC7 phosphatase activity by perturbing metal binding.
To test this hypothesis, we generated a recombinant PPTC7 protein that lacks a short, disordered N-terminal region of 32 amino acids (i.e., Δ32-PPTC7) but contains the full PP2C phosphatase domain (Supplemental Fig. 4B). As expected, wild-type Δ32-PPTC7 dephosphorylated the small molecule substrate para-nitrophenol phosphate (pNPP) in a metal-dependent fashion, yet matched concentrations of MnCl_2_ failed to promote activity of Δ32-PPTC7 D158N (Fig. 6C). We considered that D158N may compromise—but not fully ablate—PPTC7 phosphatase activity, and that our experimental conditions may have been below this detection threshold. To test this possibility, we performed phosphatase activity assays with increasing concentrations of wild-type and D158N recombinant Δ32-PPTC7, predicting that compromised phosphatase activity in D158N may become apparent at higher protein concentrations. While wild-type Δ32-PPTC7 showed a dose-dependent increase in phosphatase activity, increasing concentrations of Δ32-PPTC7 D158N protein by an order of magnitude failed to elicit detectable enzymatic activity (Fig. 6D). Based on these data, we concluded that the D158N mutation fully inactivates PPTC7 phosphatase activity in vitro.
We next evaluated the capacity of wild-type and D158N Δ32-PPTC7 to bind metal. We subjected wild-type Δ32-PPTC7 to a thermal shift assay in the presence and absence of MnCl_2_, which revealed a significant ~25°C rightward shift in denaturation temperature of Δ32-PPTC7 in the presence of 10 mM MnCl_2_ (Fig. 6E). Surprisingly, however, we found that Δ32-PPTC7 D158N also displayed a shift in thermal stability, albeit to a lesser extent than wild-type Δ32-PPTC7, suggesting that Δ32-PPTC7 D158N bound metal ions less efficiently than wild-type Δ32-PPTC7 (Fig 6F, G). To quantify the extent of disrupted metal binding, we performed thermal stability assays of both wild-type and D158N Δ32-PPTC7 across MnCl_2_ concentrations spanning two orders of magnitude (Fig. 6F, G). Treatment of both wild-type and D158N Δ32-PPTC7 promoted dose-dependent shifts in thermal stability, consistent with each protein binding MnCl_2_ (Fig. 6F, G, Supplemental Fig. 4C). Despite this, D158N Δ32-PPTC7 showed diminished capacity to be stabilized by MnCl_2_ at all tested concentrations (Fig. 6H, Supplemental Fig. 4C). Consistently, structural modeling of wild-type PPTC7 and the p.D158N variant revealed altered amino acid positioning and distance amongst manganese ions (Fig. 6I, Supplemental Fig. 6D). These data suggest that though the D158N mutant retains metal binding capacity, the coordination and/or affinity of this binding is likely compromised relative to wild-type PPTC7.
Our data suggest that PPTC7 D158N does not properly coordinate metal ions, ablating its phosphatase activity. We previously demonstrated that BNIP3 and NIX are hyperphosphorylated in PPTC7 KO cells, and that recombinant PPTC7 can directly dephosphorylate these mitophagy receptors in vitro^18^. However, multiple groups reported that PPTC7 facilitates BNIP3 and NIX degradation independent of its phosphatase activity^19,36^, and structural modeling coupled with mutational analysis revealed that PPTC7 binds to BNIP3 and NIX via its catalytic pocket^19^. Consistently, we and others demonstrated that mutations to catalytic aspartates in PPTC7 disrupted its physical association with BNIP3 and NIX^19,21^. Based on these data, we hypothesized that the p.D158N PPTC7 variant would have compromised binding to BNIP3/NIX. To test this, we overexpressed wild-type and D158N PPTC7 in both parental and PPTC7 KO HeLa FLP-In T-Rex cells, treating with deferoxamine (DFO) to stimulate mitophagy. Consistent with our previous data, PPTC7 D158N failed to decrease BNIP3 and NIX expression in both untreated and DFO-treated conditions (Fig. 6J). Interestingly, the p.D158N variant also displayed aberrant expression patterns, with less protein present in the upper doublet band (Figure 6J), which we and others have shown resides at the outer mitochondrial membrane (OMM)^19,21,36^. To test the association between PPTC7 D158N with BNIP3/NIX, we performed immunoprecipitation assays in parental or PPTC7 KO HeLa cells expressing FLAG-tagged wild-type or D158N PPTC7. Despite robust pulldown of wild-type and D158N PPTC7, and a clear interaction between wild-type PPTC7 and each mitophagy receptor, FLAG-tagged D158N PPTC7 failed to pull down appreciable BNIP3 or NIX in any condition (Fig. 6K). Collectively, these data suggest that the p.D158N variant of PPTC7 compromises both in vitro phosphatase activity as well as BNIP3/NIX binding, likely due to altered metal binding within its catalytic motif.
Identification of a third patient harboring a 3’UTR variant in PPTC7 that disrupts mitochondrial morphology and elevates BNIP3/NIX-mediated mitophagy
We next sought to identify additional undiagnosed cases with biallelic variants in PPTC7 by screening local and international clinical and diagnostic primary mitochondrial disease cohorts with negative WES/WGS. This approach resulted in the identification of a homozygous 3’UTR variant (c.*57dup) in PPTC7 in a case with previously negative WES and overlapping phenotypes with the siblings harboring the p.D158N variant. This 3’UTR variant affects a highly conserved nucleotide amongst primates (Supplemental Fig. 5) and is detected as heterozygous only in 6/1,540,424 chromosomes in gnomAD.
The patient was a 9-year-old girl who presented with progressive weakness, neuromuscular scoliosis, developmental regression, recurrent metabolic acidosis and chest infections with mild anemia and thrombocytopenia (Table 1 and 2; see Supplemental Clinical Synopsis for detailed clinical report). Brain MRI showed reduced white matter volume and hypomyelination (Fig. 7A). The family history was remarkable for four paternal relatives who passed away in their childhood due to similar progressive neurological disease (Fig. 7B). The patient continued to be bedridden with progressive neurological deterioration and respiratory complications secondary to neuromuscular weakness and passed away at the age of 11 years due to pneumonia and respiratory failure. Unfortunately, fibroblasts and other samples were not available from this patient or the affected relatives as they all passed away before identifying this potential genetic cause.
To characterize the molecular effects of the PPTC7 3’UTR variant, we used CRISPR-Cas9 to generate 293 human embryonic kidney cell lines harboring homozygous variant alleles. We generated two independent clones, B3E and E3E, which were sequence-verified to contain this variant (Fig. 7C). We first assessed mitochondrial morphology through immunostaining and confocal microscopy. While parental cells displayed a reticular mitochondrial network similar to that seen in control human fibroblasts, both clonal models of the PPTC7 3’UTR variant showed mitochondrial fragmentation (Fig. 7D). Importantly, stable reconstitution of exogenous wild-type PPTC7 restored the reticular mitochondrial network in both clones (Fig. 7D), suggesting that the 3’UTR variant of PPTC7 perturbs mitochondrial morphology.
Given the importance of PPTC7 in regulating BNIP3 and NIX, we hypothesized that the 3’UTR variant would disrupt native PPTC7 function, resulting in increased BNIP3 and NIX protein levels. Utilizing both clonal models of the 3’UTR PPTC7 variant, we immunoblotted for BNIP3, NIX and PPTC7 levels in both untreated and pseudohypoxic (i.e., DFO) conditions, using PPTC7 KO 293 cells as a positive control. We observed that BNIP3 and NIX levels were elevated in both PPTC7 3’UTR variant clones, but not to the same level as PPTC7 KO cells in either basal or pseudohypoxic conditions (Fig. 7E). In both 3’UTR PPTC7 variant clones, PPTC7 expression was detected, albeit with a reduction of the OMM form of PPTC7 (i.e., the upper band of the doublet, Fig. 7E). Importantly, stable reconstitution of PPTC7 rescued BNIP3 and NIX levels in both 3’UTR PPTC7 variant clones (Fig. 7F). Collectively, these data suggest that though the PPTC7 3’UTR variant may function as a hypomorphic rather than full loss-of-function variant, it is sufficient to increase BNIP3 and NIX protein expression.
We next utilized the mt-Keima system to test whether the elevation of BNIP3 and NIX protein levels in the PPTC7 3’UTR variant clones increased mitophagic flux. We observed a significant increase in the ratio of acidic (excitation = 566 nm) to healthy (excitation = 458 nm) mitochondria in both clones of the PPTC7 3’UTR variant when compared to parental cells, each of which was rescued upon stable expression of WT PPTC7 (Fig. 7G–H). Taken together, these data suggest that the PPTC7 3’UTR variant exhibits mitochondrial morphology alterations and elevated mitophagic flux due to disruptions in the negative regulation of BNIP3 and NIX protein expression. Given the similar molecular signature and overlapping clinical phenotypes manifested by patients with each variant of PPTC7, we propose that they represent the first cases with PPTC7 biallelic inactivating variants and deficiency in humans.
Discussion
Mitochondrial diseases remain challenging to diagnose due to both heterogeneous disease presentation as well as an incomplete spectrum of identified genes involved in disease pathology^44,45^. Advances in next-generation sequencing technologies have begun to address these challenges by not only accounting for some missing heritability, but also in identifying links between undiagnosed inborn errors of metabolism and candidate mitochondria-destined gene products. Beyond providing diagnostic value, recent initiatives to conglomerate human WES and WGS data have created powerful databases that assess the prevalence of loss-of-function (LoF) variants relevant to their predicted statistical frequency^24,46^. Such analyses have revealed genomic regions that are under considerable constraint, with few to no homozygous LoF variants present in tens of thousands of annotated human genomes of apparently asymptomatic individuals. Genes within such constrained regions are likely to be mutated only in severe, ultra-rare cases of disease, if they are all compatible with life at all.
In this study, we report the identification of homozygous variants in one such constrained gene: the mitochondrial phosphatase PPTC7. Examination of gnomAD reveals PPTC7 has a high probability of intolerance to LoF (i.e., pLI of 0.997; threshold intolerance >0.9) with no homozygous LoF variants^24^. An additional metric of constraint, the LoF observed/expected upper bound fraction (i.e., LOEUF) score is 0.42 for PPTC7, below the threshold of 0.6 demarcating intolerance for LoF variation. Additionally, there is only one homozygous missense PPTC7 variant (p.V197I) identified in one asymptomatic individual in gnomAD. Together, these metrics indicate substantial selection against LoF and damaging missense variants in PPTC7 in humans, supporting the recessive PPTC7 deficiency. Consistently, sifting through several thousands of exomes and genomes of patients with undiagnosed mitochondrial-like diseases, we identified only two potentially deleterious homozygous variants in PPTC7: one within the coding region (D158N) and one within the 3’UTR (c.*57dup). The D158N substitution was predicted to be highly pathogenic across multiple in silico pathogenicity predictors, as it perturbs a highly conserved aspartate residue critical for the enzymatic activity of most PP2C phosphatases. Consistently, our biochemical data revealed that PPTC7 D158N failed to function as a protein phosphatase in vitro and simultaneously ablated PPTC7 binding to, and subsequent regulation of, the mitophagy receptors BNIP3 and NIX. Our study thus demonstrates that D158N is a bona fide LoF variant disrupting both known functions of PPTC7^18,22^. CRISPR-mediated modeling of the homozygous PPTC7 duplication within the 3’UTR revealed that this variant is also sufficient to increase BNIP3 and NIX expression, albeit to levels that are less prominent than those seen in bona fide knockout systems. Nonetheless, these effects on BNIP3 and NIX are rescuable by exogenous PPTC7 overexpression, and cells harboring the 3’UTR variant show abnormal mitochondrial morphology and elevated mitophagic flux, suggesting this variant similarly disrupts PPTC7 function to promote mitochondrial disease. To our knowledge, these are the first patients with biallelic pathogenic variants in PPTC7.
Accumulating evidence suggests that PPTC7 plays a critical role in maintaining mitochondrial metabolism. In line with its likely LoF intolerance in humans, germline knockout of Pptc7 in mice causes hypoketotic hypoglycemia, lactic acidosis, and fully penetrant perinatal lethality^22^. Notably, all our patients with the PPTC7 variants presented with recurrent lactic acidosis and metabolic crises—symptoms which are consistent with mitochondrial disease and similar to phenotypes harbored by Pptc7 KO mice. Metabolomic profiling of plasma from these patients revealed elevated circulating branched-chain amino acids (BCAA) catabolites and altered metabolites associated with the urea cycle—patterns also seen in liver tissue derived from the Pptc7 global KO mice^22^. However, the patients presented with clinical phenotypes not described in the Pptc7 KO mice, including hypomyelinating leukodystrophy, anemia, and neutropenia. Whether Pptc7 KO animals might manifest severe neurological and immune phenotypes remains unknown, as tracking such phenotypes in the Pptc7 KO mice would be untenable due to their acute perinatal lethality. Further explorations of how PPTC7 loss affects myelination and immune cell function are warranted and will likely require conditional or tissue-specific Pptc7 KO models for study. However, the similarities of at least some metabolic derangements between the PPTC7 D158N patients and global Pptc7 knockout animals suggests that these mice, or cells derived therein, may serve as valuable tools for testing therapeutic interventions to address PPTC7 deficiency.
Mechanistically, PPTC7 was recently identified as a key negative regulator of receptor-mediated mitophagy^18–21^. Loss of PPTC7 causes post-transcriptional upregulation of the mitophagy receptors BNIP3 and NIX, two proteins known to be involved in mitochondrial clearance in response to a range of stimuli, including hypoxia or developmental cues^38^. Consistently, our studies demonstrate that both patient fibroblasts and HeLa PPTC7 KO cells rescued with PPTC7 D158N show increased expression of BNIP3 and NIX, along with elevated mitophagic flux. These phenotypes mirror those seen upon loss of the mitochondrial E3 ubiquitin ligase FBXL4^14–17^, which facilitates the ubiquitination and subsequent turnover of BNIP3 and NIX. Interestingly, multiple groups discovered that PPTC7 is required to present BNIP3 and NIX to FBXL4 to facilitate their degradation^19,36^, with both PPTC7 and FBXL4 functioning in tandem to limit BNIP3/NIX expression, thus fine-tuning the amplitude and duration of mitophagic responses^38^.
Given that FBXL4 and PPTC7 function within the same pathway, the loss of either gene should, in principle, have similar or overlapping pathological profiles in human disease. Much has been reported on FBXL4 deficiency, which causes mitochondrial DNA (mtDNA) depletion syndrome, type 13, or MTDPS13^12,13,47–49^. MTDPS13 is characterized by early infantile onset of encephalopathy, severe psychomotor delay, hypotonia, failure to thrive, brain atrophy, white matter abnormalities, swallowing difficulties sometimes associated with gastrointestinal dysmotility, and significant mortality in early childhood. Other variable clinical features include seizures, dysmorphic features, choreoathetoid movements, and hypospadias^50^. Rare features include anemia and neutropenia (either episodic or persistent) associated with recurrent infections^45,47–49^. Biochemically, MTDPS13 patients present shortly after birth with lactic acidosis, metabolic acidosis, hyperalaninemia, often with hyperammonemia (in ~50% of the cases), respiratory chain defects, and increased mitophagy^50–52^. Consistently, our study found that our patients with PPTC7 deficiency showed major phenotypic and biochemical overlap with MTDPS13, including encephalopathy, severe psychomotor delay, hypotonia, failure to thrive, brain atrophy, white matter abnormalities, swallowing difficulties, lactic acidosis, hyperammonemia, as well as increased mitophagy, mtDNA depletion, and respiratory chain defects. These data suggest that many of the observed clinical and biochemical features of PPTC7 deficiency are due to the impairment of its function as a cofactor for FBXL4 and likely result from unrestrained and excessive mitophagy. Furthermore, some clinical features of both FBXL4- and PPTC7-deficiency may stem from dysregulated BNIP3/NIX function at other organelles; for instance, dysmorphic features have been linked to the dysregulation of pexophagy^53^—the autophagic clearance of peroxisomes. As BNIP3 and NIX can also facilitate pexophagy^54^, it is possible that their dysregulation in response to loss of either FBXL4 or PPTC7 perturbs multiple organelles, resulting in complex disease presentation.
A cardinal clinical feature in our patients with PPTC7 deficiency, and not recurrently associated with FBXL4 deficiency or MTDPS13, is hypomyelinating leukodystrophy (HLD). Notably, the patient homozygous for the 3’UTR variant presented a relatively milder HLD compared to the D158N patients, consistent with it being a likely hypomorphic variant, clinically and biologically. Such a major clinical feature may be a unique feature of PPTC7 deficiency that occurs independent of, or in addition to, FBXL4-mediated turnover of BNIP3 and NIX. Interestingly, PPTC7 was originally identified as a protein phosphatase localized to the mitochondrial matrix – a role conserved through the budding yeast Saccharomyces cerevisiae^22,55–57^. We previously demonstrated that loss of PPTC7 causes the hyperphosphorylation of an array of mitochondrial proteins, and that dysregulation of phosphorylation on proteins within the TCA cycle^55^, the protein import machinery^22^, or metabolic programs^58^ disrupts mitochondrial and cellular function. Critically, our study demonstrated that the D158N variant ablates PPTC7 protein phosphatase activity, suggesting that this pathogenic variant not only disrupts mitophagy, but also this distinct role of PPTC7 in mitochondrial biology.
Some evidence suggests that the disruption of PPTC7 phosphatase activity could contribute to the HLD seen in the PPTC7 patients. HLDs represent a genetically heterogeneous but clinically overlapping group of genetic disorders with a wide clinical spectrum^59^. HLDs are characterized by a primary lack of myelination with most patients presenting with psychomotor delay, hypotonia, spasticity, movement disorders, seizures, and other features in infancy to early childhood^60^. Though most genetic causes of HLD disrupt genes encoding structural myelin proteins, some cases are triggered by recessive variants in genes involved in RNA translation or lysosomal and mitochondrial proteins^59^. One HLD-associated mitochondrial gene is HSPD1, with pathogenic biallelic variants known to cause recessive hypomyelinating leukodystrophy-4 (HLD4, MIM# 612233)^61,62^. HSPD1 encodes the mitochondrial chaperone HSP60, which is required for the folding and quality control of the mitochondrial proteins. Interestingly, tissues from Pptc7 KO mice^22^ and ΔPTC7 yeast^55^ displayed hyperphosphorylation of HSP60, which, if inactivating and at sufficient stoichiometry, could disrupt its function. These data not only suggest that HSP60 may be a conserved phosphosubstrate of PPTC7, but that its post-translational dysregulation could drive HLDs in PPTC7 patients. Biallelic variants in other nuclear-encoded mitochondrial genes, including AIFM1^63^ and PYCR2^64^, have been linked to HLDs. Though not annotated as candidate PPTC7 substrates, disrupted post-translational modifications on these proteins could, in theory, contribute to disease pathophysiology in a similar manner. Recent work has highlighted the extent and variability of mitochondrial phosphoproteomes across tissues^65,66^, suggesting future investigations on mitochondrial phosphorylation in neural precursors and oligodendrocytes should be prioritized to understand the extent to which these post-translational modifications could contribute to HLD onset and progression.
In addition to hypomyelination, other disease features seen in the PPTC7 patients are also not always observed in patients with FBXL4 deficiency. For instance, both PPTC7 D158N patients presented with severe and recurrent immune phenotypes, including anemia and persistent neutropenia associated with recurrent infection, with a negative immunological work up for lymphocyte markers, immunoglobulin levels, and bone marrow failure (Supplemental Fig. 6A and B). Single cell analysis across two independent datasets in immune cells showed differentially high expression of PPTC7 in neutrophils, which could explain the specific and isolated persistent neutropenia associated with PPTC7 deficiency (Supplemental Fig. 6C). Notably, NIX-mediated mitophagy is essential for erythroid and reticulocyte maturation^67,68^, with autophagy and receptor-mediated mitophagy linked to innate immunity, erythropoietin production, and protection against anemia^67–71^. These data suggest that disruptions in receptor-mediated mitophagy may contribute to immunological defects in patients, however, the differential presentation frequencies between PPTC7- and FBXL4-deficient patients may suggest that a threshold of excessive mitophagy may need to be reached before neutropenia and anemia become persistent. Such differences in the frequency of pathological phenotypes seen between patients with FBXL4 and PPTC7 deficiency may present unique opportunities to understand the molecular mechanisms and underlying consequences of dysregulating these two genes.
In conclusion, this study expands the genetic spectrum of primary mitochondrial disease, inborn errors of mitophagy, and genes associated with HLDs and immune dysregulation. Identifying and characterizing this novel PPTC7-associated disease may assist in solving undiagnosed cases with primary mitochondrial disease and overlapping genetic diseases. Due to the observed and predicted embryonic or early lethality associated with null variants of PPTC7 in mice and humans, we recommend screening intrauterine fetal demise (IUFD), neonatal lethality, and recurrent pregnancy loss cases for biallelic LoF variants and other inactivating variants in PPTC7. This approach may expand the clinical and genetic spectrum of PPTC7 deficiency and advance our understanding of the significance of mitophagy in early developmental stages.
Methods
Patient genetic testing and variant confirmation
DNA was extracted from whole blood samples drawn from the parents and affected siblings. Duo whole exome sequencing was performed on the affected siblings as described previously^72^. Briefly, genomic DNA (gDNA) was enzymatically fragmented and tagged with Illumina compatible adapter sequences. The libraries were paired-end sequenced on an Illumina platform with an average depth of 30x. The sequencing reads were then aligned to the Genome Reference Consortium Human Build 37 (GRCh37/hg19), as well as the revised Cambridge Reference Sequence (rCRS) of the Human Mitochondrial DNA (NC_012920). Sequence variants and copy number variations (CNVs) were called using DRAGEN, Manta and in-house algorithms. Variants with a minor allele frequency (MAF) of less than 1% in the gnomAD database, or disease-causing variants reported in HGMD^®^, literature, and in ClinVar were evaluated. All potential modes of inheritance were considered during the interpretation. Variants were classified according to the ACMG guidelines for variant classification.
In addition, the provided clinical information and family history were used to evaluate the identified variants with respect to their pathogenicity and disease causality. All relevant variants in known or candidate mitochondrial, autophagy, and neurological genes related to the phenotype of the patients were considered and confirmed by Sanger sequencing. Mitochondrial variants with a heteroplasmy level of 15% or higher were considered. Aside from the identified PPTC7 D158N variant, there were no additional potential variants that segregated with the disease. The WGS was performed to exclude causative noncoding SNVs and CNVs in known genes as described previously.
Skin biopsies and patient sample collection
Blood samples and skin biopsies (5 mm) were obtained from affected siblings and a healthy age-matched control participant. Biopsies were cultured immediately for 30 days to obtain primary fibroblasts in DMEM (Gibco^®^, Thermo Fisher Scientific) supplemented with 15% fetal bovine serum (FBS, Biowest). Primary fibroblasts were cryopreserved in high glucose DMEM (Gibco^®^, Thermo Fisher Scientific), with 15% FBS and 10% DMSO.
Plasmid Construction
FLAG-tagged PPTC7 WT was generated using Flippase (FLP) recombination target technology into the pcDNA5/FRT/TO-Venus-Flag-Gateway backbone (Addgene catalog #40999, a kind gift from Jonathon Pines^73^) The pcDNA5/FRT/TO-Venus-Flag-Gateway-PPTC7 D158N construct was generated by site-directed mutagenesis of the pcDNA5/FRT/TO-Venus-Flag-Gateway–PPTC7 WT construct, using customized oligonucleotide primers (Supplementary Table 2) targeting the aspartic acid (D) at amino acid 158 to asparagine (N). To generate the lentiviral pLVX-IRES-puro-FLAG – PPTC7 WT construct, the insert was amplified and then inserted into the digested backbone using Gibson assembly.
Cell culture, cell line generation, CRISPR editing, transfection and lentivirus preparation
All cells were cultured in DMEM (Gibco^®^, Thermo Fisher Scientific), supplemented with 10% heat-inactivated FBS (Biowest) and 1x penicillin-streptomycin (Thermo Fisher Scientific). Cells were grown in an incubator at 37°C with 5% CO_2_. Cell lines were regularly screened for mycoplasma contamination.
To generate patient fibroblasts stably expressing mt-Keima or PPTC7 WT, lentiviruses (mt-Keima or pLVX-IRES-puro-FLAG, respectively), were prepared by transient PEI (Fisher Scientific) transfection with viral packaging plasmids in Lenti-X 293T cells. Patient fibroblasts were transduced with virus supplemented with 10 ml/ul polybrene (Millipore Sigma) for 48 hours and then selected for protein expression by flow cytometry sorting or puromycin (Millipore Sigma) selection.
To generate the 3’UTR variant cell line, dupT was inserted into the 3’ UTR of the PPTC7 gene in the 293 cells at the Genome Engineering & Stem Cell Center (GESC@MGI) at Washington University in St. Louis. Briefly, the parental cells were transfected with recombinant Cas9 protein complexed with synthetic gRNA and single stranded oligodeoxyribonucleotide donors (ssODNs). The gRNA has the recognition sequence, 5’ GCCAGCAGGATCAGCaCACATGG (PAM is underlined, the adenosine (antisense) to be duplicated is in lower case). The dupT ssODN has the following sequence: 5’-ctagctgaggtgtcaagtcctgcctttcctttcatcatcccaaatttcccctgccatgtgTTgctgatcctgctggcaggaccacatttctttgccactgatctcaatggccagtgatgtaa. * represents a phosphorothioate bond, and both T’s at the site of duplication are in upper case. Both the sgRNA and ssODN were sourced from Integrated DNA Technologies (IDT, Coralville, IA). Each transfected pool was then single cell sorted into 96-well plates, and single cell clones were screened using NGS for harboring the respective mutations. All clones were confirmed for genotype, negative for mycoplasma and with STR profiles matching those of parental lines. To stably express mt-Keima or PPTC7-WT in the PPTC7 3’UTR variant clones, lentiviruses (mt-Keima or pLVX-puro-FLAG-PPTC7 WT, respectively) were prepared as outlined above, and transduced into clones prior to selection for protein expression by flow cytometry sorting or puromycin (Millipore Sigma) selection.
To generate HeLa Flp-In TREx and HeLa Flp-In TREx PPTC7 KO cells stably expressing mt-Keima, CRISPR-Cas9 technology was used as previously described^18^. HeLa Flp-In TREx and HeLa FLP-IN TRex PPTC7 KO cells stably expressing mt-Keima and doxycycline-inducible FLAG-tagged PPTC7 WT or D158N were generated using Flippase (FLP) recombination target technology. Briefly, constructs with the pcDNA5/FRT/TO-Venus-Flag-Gateway backbone (Addgene catalog #40999, a kind gift from Jonathon Pines^73^) were co-transfected with pOG44 at a ratio of 1:4 pcDNA5 to pOG44. Transfections were performed according to manufacturer’s directions using Fugene 6 (Promega). After 24 hours media was replenished, and after 24–48 hours cells were selected with 400 ml/ml Hygromycin B (Thermo Fisher Scientific) for approximately 7–10 days. To induce expression of constructs, cells were treated with 0.5 mg doxycycline (Millipore Sigma).
Transient transfections of plasmids into HEK293T cells were performed using PEI (Fisher Scientific). Following transfection, cells were incubated for 48 hours prior to harvesting and downstream functional assays.
Glucose Flux Tracing and Polar Metabolite Extraction
For glucose flux tracing, fibroblast cells from healthy controls and PPTC7 D158N patients were cultured in glucose-labeling media (glucose-free DMEM supplemented with U-^13^C_6_ glucose, 10% dialyzed FBS and 1% 200 mM L-glutamine) for 4 hours in an incubator at 37°C with 5% CO_2_. After 4 h, cells were washed with filtered PBS, followed by Mass Spectrometry (MS)-grade water (Fisher Scientific). Cells were quenched with ice-cold MS-grade methanol (Fisher Scientific), harvested and evaporated without heat using a Savant RVT 5105 Speed Vac. Acetonitrile:methanol:water at a 2:2:1 ratio (Fisher Scientific) was added to remaining pellets, which were then vortexed, freeze-dried in liquid nitrogen and thawed three times before sonication for 10 minutes at 25°C three times. Samples were placed at −20°C overnight, then centrifuged at full speed for 10 minutes to separate pellets and supernatants, which were transferred to fresh tubes.
Sample pellets were resuspended in 100 mM sodium hydroxide, vortexed, boiled at 95°C for 5 minutes and cooled for 2 minutes at room temperature. This step was repeated four times in total. Next, the total protein concentration in each sample was calculated by a Pierce BCA Protein Assay (Thermo Fisher Scientific). The original sample supernatants in new tubes were dried without heat using a Savant RVT 5105 Speed Vac, until all liquid was evaporated, and a clear pellet remained. Clear pellets were resuspended in acetonitrile:water (in a 2:1 ratio) based on the ug total protein in the original pellet, where 1 ml acetonitrile:water was added for every 2.5 mg (x 100 for the 100 ml NaOH added) protein. Samples were sonicated for 5 min at 25°C and vortexed for a total of two times. Samples were stored at 4°C for 1 hour, centrifuged at full speed for 10 minutes and then transferred to LC-MS compatible vials.
Metabolomic LC/MS analysis
Blood serum samples and polar metabolites extracted from healthy control and PPTC7 D158N patient fibroblasts, were assessed for metabolite profiles. Ultra-high performance liquid chromatography coupled with mass spectrometry (UHPLC/MS) analysis was performed using a Vanquish Flex UHPLC system (Thermo Fisher Scientific) interfaced with a Orbitrap ID-X mass spectrometer (Thermo Fisher Scientific). Polar metabolites were separated on a HILICON iHILIC-(P) Classic HILIC column (100 × 2.1 mm, 5 μm) equipped with a HILICON iHILIC-(P) Classic guard column (20 × 2.1 mm, 5 μm). The mobile phases consisted of solvent A (20 mM ammonium bicarbonate, 2.5 μM medronic acid, 0.1% ammonium hydroxide in 95:5 water:acetonitrile) and solvent B (95:5 acetonitrile:water). The column temperature was maintained at 45°C, and metabolites were eluted using a linear gradient at a flow rate of 250 μL/min with the following gradient profile: 0–1 min, 90% B; 12 min, 35% B; 12.5–14.5 min, 25% B; 15 min, back to 90% B. The injection volume was 4 μL for all polar experiments. Data was acquired in positive and negative ion mode with the following settings: spray voltage, 3.5 kV (positive) and −2.8 kV (negative); sheath gas, 50; auxiliary gas, 10; sweep gas, 1; ITT temperature, 300°C; vaporizer temperature, 200°C; mass range, 67–1,000 Da; resolution, 120,000. LC/MS data were processed and analyzed using Compound Discoverer and Skyline^74^ software. Natural-abundance correction of ^13^C was performed with AccuCor^75^.
High resolution respirometry analysis
To measure oxygen consumption rates in control and PPTC7 D158N patient fibroblasts, cell pellets were resuspended in Mir05 respiration buffer (0.5 mM EGTA, 3 mM MgCl_2_, 60 mM lactobionic acid, 20 mM taurine, 10 mM KH_2_PO_4_, 20 mM HEPES, 110 mM sucrose and 1g/L fatty acid-free bovine serum albumin, pH=7.1). Resuspensions were kept with continuous stirring at 37°C. Mitochondrial respiration was monitored using the computer-interfaced Clark-type electrode Oroboros Oxygraph-2k (Oroboros Instruments). To induce different mitochondrial states, 5 mM sodium pyruvate, 2 mM malate and 2.5 mM ADP were added.
mtDNA analysis
To assess mtDNA levels, gDNA was isolated from healthy control and PPTC7 D158N patient fibroblasts using the DNeasy Blood & Tissue Kit (Qiagen) according to manufacturer’s instructions. gDNA was quantified using a Take3 Plate Reader (Agilent). Primers targeting Beta-actin (forward: GGCTGTATTCCCCTCCATCG; reverse: CCAGTTGGTAACGCCATGT) as a measure of nuclear DNA, and mt-CO2 (forward: CGAGTCGTTCTGCCAATAGAA; reverse: CCTGGTCGGTTTGATGTTACT) and mt-ND4 (forward: GCCTCACATCATCACTCCTATT; reverse: GGCTATAAGTGGGAAGACCATT) as a measure of mitochondrial DNA were tested against the gDNA inputs. Quantitative real-time polymerase chain reactions (qRT-PCRs) were performed on a CFX96 Touch Real-Time PCR System (Bio-Rad). Relative quantitation was calculated using the 2^−ΔΔ^C_T_ method as previously described^76^.
Proteomics sample preparation
Sample preparation for all cell types and mitochondrial fractions were performed using the following protocol. Cells were cultured in DMEM supplemented with 10% FBS and 1x Pen/Strep, and grown to 90% confluency prior to harvesting by scraping in PBS on ice. For whole-cell proteomics, cell pellets were freeze-dried in liquid nitrogen and stored at −80°C prior to preparation. For isolated mitochondrial fractions, cell pellets were resuspended into mitochondrial isolation buffer (0.1 M Tris-MOPS, pH 7.4, 0.1 M EGTA, 1 M sucrose; in dH20) and homogenized on ice using a 7 mL dounce tissue grinder (with tight pestle). Cell homogenates were centrifuged at low speed (600 × g for 10 minutes at 4°C). Supernatants were collected and centrifuged at a higher speed (7000 × g for 10 minutes at 4°C) to separate mitochondrial-enriched fractions (pellet) from cytosol and other organelles (supernatant). Pellets were washed in mitochondrial isolation buffer, then freeze-dried in liquid nitrogen and stored at −80°C prior to preparation. Pellets were resuspended in 8M Urea in 50 mM Ammonium Bicarbonate with protease and phosphatase inhibitors (1x protease inhibitor cocktail (0.5 mg/ml pepstatin A, chymostatin, antipain, leupeptin and aprotinin) and 1x phosphatase inhibitor cocktail (0.5 mM imidazole, 0.25 mM sodium fluoride, 0.3 mM sodium molybdate, 0.25 mM sodium orthovanadate and 1 mM sodium tartate). Cells were lysed via sonication with a microtip probe sonicator. Protein concentrations were then estimated via Pierce BCA 660 nm Protein Assay (Thermo Fisher Scientific). Proteins were reduced by 10 mM TCEP for 30 minutes at 37°C then alkylated by 30 mM iodoacetamide for 30 minutes at room temperature in the dark. Urea was diluted to 2M by addition of 50 mM Ammonium Bicarbonate and trypsin digested overnight with 1:50 MS-grade trypsin (Thermo Fisher Scientific). Peptides were desalted using in-house packed stage-tips of C18 material (Thermo Fisher Scientific) and dried using a Savant RVT 5105 Speed Vac.
Proteomic LC/MS Analysis
For the fibroblast samples, 500 ng peptide was characterized using a nanoACQUITY ultrahigh-pressure liquid chromatography (UPLC) System (Waters) coupled with the ZenoTOF 7600 mass spectrometer (SCIEX). The analytes were separated on an Phenomenex Kinetex XB C18 column (2.6 μm, 0.3 × 150 mm, Phenomenex) at a flow rate of 10 μL/min, and temperature of 45°C. A solution of water containing 0.1% FA and acetonitrile containing 0.1% FA were used as solvents A and B, respectively. The chromatography gradient consisted of 2% solvent B over 0–1 min, 2–30% solvent B over 1–46 min, 30–80% solvent B over 46–47 min, 80% solvent B over 47–49 min, and finally equilibrated with 2% solvent B over 5 minutes. The ZenoTOF 7600 was equipped with an OptiFlow Turbo V ion source and operated in SWATH mode with Zeno trap activated. The source conditions were as follows: ionization voltage: 5000 V, positive polarity, temperature: 200°C, ion source gas 1:20 psi, ion source gas 2: 60 psi, curtain gas: 35 psi. The Zeno SWATH DIA method consisted of 85 variable-width SWATH DIA windows that spanned the mass range 399.5–903.5 m/z, and the acquisition settings were as follows: MS1 accumulation time: 100 ms, MS1 m/z range: 400 to 1500, MS2 accumulation time: 13 ms, MS2 m/z range: 140 to 1800. All raw data were processed with Spectronaut (v19.5) using the directDIA mode. The peptides were searched using default parameters: Trypsin/P, two missed cleavages allowed, toggle N-terminal M, fixed modification: C carbamidomethylation, variable modification: N-terminal acetylation, M oxidation, peptide length range: 7 to 52 amino acids. The proteins were mapped against the Homo sapiens UniProt canonical sequence database (UP000005640).
For the HeLa cell and mitochondrial samples, 500 ng of trypsin-digested peptides were analyzed on a Vanquish Neo UHPLC system (Thermo Fisher Scientific, CA) coupled to an Orbitrap Ascend mass spectrometer (Thermo Fisher Scientific, CA). Peptides were separated on an EASY-Spray PepMap column (2 μm particle size, 75 μm × 150 mm; Thermo Fisher Scientific) at a flow rate of 0.3 μL/min with the column temperature maintained at 40 °C. Mobile phases consisted of 0.1% formic acid (FA) in water (solvent A) and 0.1% FA in acetonitrile (solvent B). The gradient was programmed as follows: 2% B for 0–2 min, 2–32% B from 2–107 min, 32–42% B from 107–112 min, 42–95% B over 1 min, held at 95% B for 8 min, and then re-equilibrated to 2% B by the end of the 120-min run. The Orbitrap Ascend was operated with an EASY-Spray ion source. Data were acquired in data-dependent acquisition (DDA) mode with MS1 resolutionat 120k over a mass range of 350–2000 m/z, with a maximum injection time of 50 ms. Precursors with charge states 2–6 were included for MS/MS, and dynamic exclusion was set to 60 s. For MS/MS, isolation window was set to 1.6 m/z. Fragmentation was performed with higher-energy collisional dissociation (HCD) with a normalized collision energy of 30%. MS2 was analyzed with Orbitrap at 30,000 resolution with a first mass of 110 m/z. All DDA raw files were searched using FragPipe v22 against the Homo sapiens UniProt canonical protein database (UP000005640).
SDS-PAGE and Immunoblotting
Harvested cells were lysed with radioimmunoprecipitation (RIPA) buffer (0.5% wt/vol sodium deoxycholate, 150 mM sodium chloride, 1.0% vol/vol IGEPAL CA-630, 1.0% SAS, 50 mM Tris pH 8.0, 1 mM EDTA pH 8.0 in water) supplemented with 1x protease inhibitor cocktail (0.5 mg/ml pepstatin A, chymostatin, antipain, leupeptin and aprotinin) and 1x phosphatase inhibitor cocktail (0.5 mM imidazole, 0.25 mM sodium fluoride, 0.3 mM sodium molybdate, 0.25 mM sodium orthovanadate and 1 mM sodium tartate), on ice. Samples were then centrifuged at 21,100 × g at 4°C for 10 minutes. Lysates were collected and protein concentrations were estimated via Pierce BCA Protein Assay (Thermo Fisher Scientific).
Lysates (20–40 μg) were mixed with 5x sample buffer (312 mM Tris-Base, 25% wt/vol sucrose, 5% wt/vol SDS, 0.05% wt/vol bromophenol blue, 5% vol/vol b-mercaptoethanol, pH 6.8) and boiled at 95°C for 5 minutes. Samples were run on SDS-PAGE gels with Precision All-Blue Protein Standards (Bio-Rad) and were then transferred onto nitrocellulose membranes. Following transfer, membranes were incubated in blocking buffer (3% nonfat dairy milk powder in TBS-T) for 30–60 minutes. Membranes were then washed in TBS-T and incubated with primary antibodies diluted in 2% BSA in TBS-T for 24–48 hours at 4°C or 1 hour at room temperature (time periods varying between antibodies). Following incubation with primary antibodies, membranes were washed 3x with TBS-T for 5-minute intervals and incubated with secondary fluorophore-conjugated antibodies (diluted 1:10,000 in TBS-T) for 30 minutes at room temperature. Membranes were washed 3x with TBS-T for 5-minute intervals and then imaged with an OdysseyFC Imager (Li-COR) using Image Studio software (v. 5.2). The primary antibodies used were as follows: Anti-beta-Actin (mouse, 1:1000, Cell Signaling Technology (CST), 3700S), Anti-BNIP3 (rabbit, 1:1000, CST, 44060S), Anti-FLAG M2 (mouse, 1:2000, Millipore Sigma, F1804), Anti-FLAG (rabbit, 1:1000, CST, 14793), Anti-NIX (rabbit, 1:1000, CST, 12396S), Anti-PPTC7 (rabbit, 1:1000, Novus Biologicals, NBP190654).
Co-Immunoprecipitation Assays
Cells were plated on 1×10 cm^2^ dish per condition. Following harvesting, pellets were lysed in immunoprecipitation buffer (50 mM TRIS pH 7.5, 150 mM sodium chloride, 1 mM EDTA, 1 mM EGTA, 5mM magnesium chloride, 1mM b-glycerophosphate 1.0% Triton-X), supplemented with 10 mM iodoacetamide, 1x protease inhibitor cocktail (0.5 mg/ml pepstatin A, chymostatin, antipain, leupeptin and aprotinin) and 1x phosphatase inhibitor cocktail (0.5 mM imidazole, 0.25 mM sodium fluoride, 0.3 mM sodium molybdate, 0.25 mM sodium orthovanadate and 1 mM sodium tartrate), on ice for 30 min. Cell lysates were collected by centrifugation at 21,100 × g for 10 minutes at 4°C. FLAG-tagged PPTC7 WT or PPTC7 D158N was immunoprecipitated by incubating cell lysates with bead-conjugated FLAG (Sigma; #A2220) for 3 hours at 4°C. Immunoprecipitants were washed with the immunoprecipitation buffer 4 times. After final wash, supernatant was removed and samples were eluted into 2x sample buffer (125 mM Tris-Base, 10% wt/vol sucrose, 2% wt/vol SDS, 0.02% wt/vol bromophenol blue, 2% vol/vol b-mercaptoethanol, pH 6.8). Samples were then boiled at 95°C for 10 minutes, and run by gel electrophoresis, for detection by immunoblotting.
Indirect Immunofluorescence
To stain samples for immunofluorescence analysis, cells were plated on 1.5 mm glass coverslips for 24–48 hours. Prior to fixation, media was supplemented with 250 nM MitoTracker^™^ Red CMXRos (Thermo Fisher Scientific) for 30 minutes in an incubator set to 37°C and 5% CO_2_. Coverslips were then washed three times with PBS fixed with ice-cold methanol for 10 minutes at −20°C. After fixation, coverslips were washed with PBS three times. To reduce non-specific binding, fixed coverslips were blocked with 2% bovine serum albumin (BSA) in PBS for 30 minutes. Cells were then labelled with primary antibody diluted in 2% BSA in PBS for 1 hour, followed by three washes in 5-minute intervals with 2% BSA in PBS + Tween-20. Cells were then labelled with secondary antibodies diluted at 1:400 in 2% BSA in PBS for 30 minutes, followed by washes as outlined above. Finally, cells were stained with Hoechst diluted at 1:1000 in PBS for 1 minute and washed three times with PBS. Coverslips were mounted on glass microscope slides using Prolong^™^ Diamond Antifade Mountant (Thermo Fisher Scientific). Images were acquired using a Zeiss LSM980 Airyscan 2 Confocal microscope using a 63x high NA oil immersion objective. The primary antibodies used were Anti-NIX (rabbit, 1:400, CST, 12396S) and Anti-FLAG M2 (mouse; 1:400, Sigma, F1804).
Mitophagy Assays
To detect mitophagy levels, the mt-Keima assay was performed in cells stably expressing mt-Keima as previously described^18,39^. Dual-excitation images at 561 nm (mito-lysosomal signal) and 458 nm (mitochondrial signal) were acquired using a Zeiss LSM980 AiryScan 2 microscope, with a 63x high NA oil immersion objective and environmental chamber set to 37°C and 5% CO_2._ Images were taken blinded by using the tiling position feature of randomized positions in the field of vision. Quantification of the 561/458 nm ratio was performed using ImageJ2 software. Single cells were segregated into regions of interest (ROIs). The ROIs were then cropped and split into separate channels for threshold processing and separate measurements of each channel in ImageJ, which was used to calculate the mitophagy ratio. Data was normalized to a control condition and graphed using GraphPad Prism (version 10.6.1) software. To detect mt-Keima fluorescence by flow cytometry, mt-Keima cells were sorted to detect differences in the 561/458 nm ratio of fluorescence.
Transmission electron microscopy assay
To analyze healthy control and PPTC7 D158N patient fibroblasts by transmission electron microscopy, cells were fixed in 2.5% glutaraldehyde (Electron Microscopy Sciences) prepared in 0.1 M sodium cacodylate buffer (pH 7.4) at 4°C. Within 24 hours, samples were rinsed three times with fresh 0.1 M sodium cacodylate buffer (pH 7.4) at room temperature for 5 minutes. Following fixation, samples were incubated in 1% osmium tetroxide (Electron Microscopy Sciences) in 1.5% potassium ferrocyanide for 1 hour at room temperature, followed by an additional incubation in 2% aqueous osmium tetroxide for 30 minutes. After each fixation step, samples were rinsed three times with distilled water for 5 minutes at room temperature. Subsequently, samples were incubated in 1% uranyl acetate at 4°C for 12 hours, and then dehydrated through a graded ethanol series (50%, 70%, 80%, 90%, and 100%), followed by 100% acetone. Infiltration was performed using acetone/Epon resin (Electron Microscopy Sciences) mixtures in ratios of 3:1, 1:1, and 1:3. Finally, samples were embedded in 100% Epon resin and polymerized at 60°C for 72 hours. Ultrathin sections (~90 nm) were cut using an ultramicrotome (UC6, Leica Microsystems), mounted on copper grids, and stained with 3% lead citrate (Electron Microscopy Sciences) for 2 minutes. Transmission electron micrographs were acquired using either a Titan ST or a Themis microscope (Thermo Fisher Scientific) operating at 300 kV.
PPTC7 recombinant protein generation and phosphatase assays
Recombinant d32-PPTC7 and d32-PPTC7 D158N were generated and assayed for phosphatase enzyme activity as previously described^22,55,58^ with some key protocol modifications. First, each protein was cloned into pGEX-6P-1, expressing a GST-fusion that was purified with concentrations quantified as described^77^. Both recombinant proteins were assayed for activity using the generic phosphatase substrate para-nitrophenyl phosphate (pNPP) (New England Biolabs). Unless otherwise noted in the figure legends, 250 ng of phosphatase was incubated with 10 mM pNPP in the presence of 10 mM MnCl_2_ diluted in a final volume of 100 μl of 50 mM Tris (pH = 8.0). Addition of pNPP was used to initiate the reaction, and reactions lacking enzyme or MnCl_2_ served as negative controls. The dephosphorylation of pNPP to para-nitrophenol (pNP) produces a colorimetric reaction which was followed by monitoring absorbance at 405 nm on an Epoch2 plate reader (BioTek) controlled by Gen5 software (version 3.10). Reactions were run at room temperature for 30 minutes, with the maximal slope of a minimum of 5 points used to calculate relative enzyme activity.
Thermal shift assays
To test protein thermal shifts, PPTC7 WT and PPTC7 D158N recombinant proteins were generated as described in the previous section. Recombinant proteins (0.2 mg/mL) were mixed with varying concentrations of MnCl_2_ diluted in PreScission protease buffer (20 mM Tris, 150 mM NaCl, 0.5 mM EDTA, 1 mM DTT, pH = 7.4) and incubated on ice for 5 minutes. SYPRO^™^Orange Protein Gel Stain (Thermo Fisher Scientific) was then added to samples to reach a final concentration of 5x, and 25 μL sample was added to each well of a 96-well plate, prior to sealing with film. Samples were run using a CFX96 Touch Real-Time PCR System (Bio-Rad) controlled by CFX Maestro software (version 2.3) with increments of 0.5°C for 10 seconds between 10–95°C.
Statistical Analyses
For proteomic data analysis, p-values were calculated using a two-sided, unpaired t-test using the package SciPy in Python as described^78^. Statistical analyses for all other data were performed using GraphPad Prism (version 10.6.1) software. Data from three or more biologically independent replicates were used for all statistical comparisons. No statistical tests were done to pre-determine sample sizes. Similar data variances were observed between groups. T-tests or one-way ANOVA was used for comparisons between means. p-values greater than 0.05 were considered non-significant.
Supplementary Material
Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Nunnari J. & Suomalainen A. Mitochondria: In Sickness and in Health. Cell 148, 1145–1159 (2012).22424226 10.1016/j.cell.2012.02.035PMC 5381524 · doi ↗ · pubmed ↗
- 2Uoselis L., Nguyen T. N. & Lazarou M. Mitochondrial degradation: Mitophagy and beyond. Molecular Cell 83, 3404–3420 (2023).37708893 10.1016/j.molcel.2023.08.021 · doi ↗ · pubmed ↗
- 3Maruyama H. Mutations of optineurin in amyotrophic lateral sclerosis. Nature 465, 223–226 (2010).20428114 10.1038/nature 08971 · doi ↗ · pubmed ↗
- 4Valente E. M. Hereditary Early-Onset Parkinson’s Disease Caused by Mutations in PINK 1. Science 304, 1158–1160 (2004).15087508 10.1126/science.1096284 · doi ↗ · pubmed ↗
- 5Lücking C. B. Association between Early-Onset Parkinson’s Disease and Mutations in the Parkin Gene. New England Journal of Medicine 342, 1560–1567 (2000).10824074 10.1056/NEJM 200005253422103 · doi ↗ · pubmed ↗
- 6Mori H. Pathologic and biochemical studies of juvenile parkinsonism linked to chromosome 6q. Neurology 51, 890–892 (1998).9748052 10.1212/wnl.51.3.890 · doi ↗ · pubmed ↗
- 7Wang Y. Risk of Parkinson Disease in Carriers of Parkin Mutations: Estimation Using the Kin-Cohort Method. Arch Neurol 65, 467–474 (2008).18413468 10.1001/archneur.65.4.467PMC 2836931 · doi ↗ · pubmed ↗
- 8Hatano Y. PARK 6-linked autosomal recessive early-onset parkinsonism in Asian populations. Neurology 63, 1482–1485 (2004).15505170 10.1212/01.wnl.0000142258.29304.fe · doi ↗ · pubmed ↗