Engineered bacterial lipoate protein ligase A (lplA) restores lipoylation in cell models of lipoylation deficiency

Abstract

Protein lipoylation, a vital lysine post-translational modification, plays a crucial role in the function of key mitochondrial tricarboxylic acid cycle enzymatic complexes. In eukaryotes, lipoyl post-translational modification synthesis occurs exclusively through de novo pathways, relying on lipoyl synthesis/transfer enzymes, dependent upon mitochondrial fatty acid and Fe–S cluster biosynthesis. Dysregulation in any of these pathways leads to diminished cellular lipoylation. Efficient restoration of lipoylation in lipoylation deficiency cell states using either chemical or genetic approaches has been challenging because of pathway complexity and multiple upstream regulators. To address this challenge, we explored the possibility that a bacterial lipoate protein ligase A (lplA) enzyme, which can salvage free lipoic acid bypassing the dependency on de novo synthesis, could be engineered to be functional in human cells. Overexpression of the engineered lplA in lipoylation null cells restored lipoylation levels, cellular respiration, and growth in low glucose conditions. Engineered lplA restored lipoylation in all tested lipoylation null cell models, mimicking defects in mitochondrial fatty acid synthesis (MECR KO), Fe–S cluster biosynthesis (BOLA3 KO), and specific lipoylation-regulating enzymes (FDX1 [ferredoxin 1], LIAS [lipoyl synthase], and LIPT1 [lipoyl (octanoyl) transferase 1] KOs). Furthermore, we describe a patient with a homozygous c.212C>T variant LIPT1 with a previously uncharacterized syndromic congenital sideroblastic anemia. K562 erythroleukemia cells engineered to harbor this missense LIPT1 allele recapitulate the lipoylation-deficient phenotype and exhibit impaired proliferation in low glucose that is completely restored by engineered lplA. This synthetic approach offers a potential therapeutic strategy for treating lipoylation disorders.

Protein lipoylation is a post-translational modification that regulates key facets of mitochondrial metabolism, including pyruvate dehydrogenase, α-ketoglutarate dehydrogenase, branched-chain α-ketoacid dehydrogenase, and glycine decarboxylase complexes, the latter also referred to as the glycine cleavage system. This lipoylation process is evolutionarily conserved from bacteria to humans (1, 2, 3, 4, 5, 6). In humans, lipoic acid is synthesized de novo in a multistep process (Fig. 1A). First, an 8-carbon fatty acid (octanoic acid) synthesized in the mitochondria and covalently attached to an acyl-carrier protein is transferred to a conserved lysine residue within the Glycine Cleavage System Protein H protein by lipoyl (octanoyl) transferase 2 (LIPT2). Second, the lipoyl synthase (LIAS) adds two sulfurs from its auxiliary Fe–S cluster to positions C-6 and C-8 in a reaction that requires SAM (7, 8, 9, 10) and ferredoxin 1 (FDX1) (11, 12). The synthesized lipoic acid is then transferred to the target lipoylated complexes by LIPT1 resulting in functional lipoylated complexes (13).

Figure 1.

Modified bacterial lplA can rescue lipoylation defects in human cells.A, schematic of the de novo lipoylation pathway in human cells and lplA-mediated salvaging of lipoic acid (LA) in bacteria. B, lplA from Escherichia coli and lplA from Thermoplasma acidophilum (lplA_T) were codon optimized and cloned into human expression plasmids with and without a mitochondria localization signal (MLS) and V5 tag. K134R is an enzymatic dead mutant of lplA. C, WT and lipoylation null FDX1 KO HEK293T cells were treated with 100 μM LA when overexpressing either E. coli lplA or Thermoplasma acidophilum lplA (lplA_T) with and with a mitochondrial localization signal (+MLS). D, WT and lipoylation null FDX1 KO HEK293T cells were treated with 100 μM LA when overexpressing either FDX1 (used as a positive control), WT lplA (WT), MLS-tagged lplA, and the enzymatic dead K134R mutant lplA. Levels of lipoylation and indicated protein levels were analyzed by immunoblotting. HEK293T, human embryonic kidney 293T cell line.

De novo lipoic acid synthesis relies heavily on the mitochondrial lipid and Fe–S cluster biosynthesis pathways. Disruptions in these pathways lead to diminished levels of protein lipoylation, which is a characteristic feature of several debilitating human genetic disorders, often stemming from mutations in genes associated with Fe–S cluster biogenesis (14, 15, 16), mitochondrial fatty acid biosynthesis (17, 18), or directly involved in the lipoylation pathway (13, 19, 20, 21, 22, 23, 24). Each is categorized as a subclass of mitochondrial inborn errors of metabolism. The complexity of the de novo lipoic acid synthesis pathway is a significant challenge to develop therapeutic strategies aimed at restoring lipoylation in these diverse diseases.

Unlike eukaryotes, some bacteria and archaea can salvage free lipoate or lipoic acid from the environment using lipoate protein ligase A (lplA) (25, 26, 27) (Fig. 1A). Lipoate protein ligases have been described in strains of Escherichia coli, Listeria monocytogenes, Bacillus subtilis, Sand taphylococcus aureus; there are also structural analogs in Thermoplasma acidophilum. These enzymes scavenge free lipoic acid, and, in the presence of ATP, bypass the requirement of the multienzyme/step of de novo synthesis. Recently, it was shown that these enzymes can be engineered to be partially active in yeast (28). In this study, we engineered a bacterial lplA enzyme that completely restores the lipoylation state in multiple human lipoylation null genetic models (MECR, BOLA3, FDX1, LIAS, LIPT1, LIPT2 KO JHH7 cells). We further describe a patient with a homozygous, previously described mutation in LIPT1 (c.212C>T, p.Ser71Phe) with a novel syndromic sideroblastic anemia phenotype. Erythroleukemia K562 cells homozygous for the p.Ser71Phe variant exhibited loss of lipoylated proteins and impaired cell growth in metabolically restrictive conditions, phenotypes that were completely restored by exogenous overexpression of engineered lplA. These findings provide evidence that engineered bacterial enzymes could be instrumental in overcoming human metabolic disorders.

Results

Engineered E. coli lplA restores lipoylation in human cells with disrupted lipoic acid synthesis

Some strains of bacteria possess lplA enzymes that can bypass the requirement of de novo lipoic acid synthesis (Fig. 1A). We engineered two codon-optimized lplA enzymes, one from E. coli (strain K12) referred to as lplA, and one from the Archaea T. acidophilum (strain ATCC 25905) referred to as lplA_T for expression in mammalian cells. In each, we also added a V5 tag to enable immunoblot detection and introduced a mitochondrial localization signal (MLS) to determine if localization to mitochondria is essential for efficient lipoylation (Fig. 1B). We used a previously characterized lipoylation null cell model (FDX1 KO (11)) to determine the ability of these enzymes to restore lipoylation in the presence or absence of exogenous lipoic acid to the media.

As expected, supplementation of lipoic acid alone to the media had no effect on lipoylation in both WT and lipoylation null cells as determined by immunoblot analysis of lipoylated DLAT and DLST. Restoration of the de novo synthesis pathway in these cells by FDX1 reconstitution completely restored lipoylation, serving as a positive control. Overexpression of lplA or lplA_T without the MLS had no effect on the levels of lipoylation in the presence or absence of lipoic acid supplementation (Fig. 1C). However, lipoylation was restored by overexpressing lplA, and to a lesser extent lplA_T, fused to an MLS (MLS-lplA) supported by lipoic acid supplementation (Fig. 1C). Mutation of the lysine 134, required for binding of lipoic acid (UniProt), completely diminished the ability of MLS-lplA to restore cellular lipoylation even when lipoic acid was supplemented to the media (Fig. 1D). Together, these findings demonstrate that lipoylation can be fully restored in FDX1 KO cells, which are defective in lipoylation, through overexpression of engineered E. coli lplA and supplementation with lipoic acid.

To establish that restoring lipoylation with bacterial proteins is functional, we assessed the basal respiratory capacity and proliferation in low glucose conditions (2 mM glucose) in transduced cells (11). Consistent with lipoylation detected by immunoblot analysis (Fig. 1C), overexpression of MLS-lplA and to a lesser extent MLS-lplA_T restored basal respiration (Fig. 2, A and B), and the ability of lipoylation null cells to proliferate in low glucose conditions (Fig. 2, C and D) in media supplemented with lipoic acid. Interestingly, despite the lower level lipoylation in lplA_T cells, it was sufficient to restore cellular respiration and growth to WT levels. Given the superior effect observed with MLS-lplA, we focused on this enzyme.

Figure 2.

Engineered lplA restores respiration and growth in low glucose conditions in lipoylation null cells.A–D, basal oxygen consumption rate (OCR) (A and B) and growth in low glucose conditions (2 mM glucose) (C and D) were determined for WT and lipoylation null (FDX1 KO) HEK293T cells overexpressing WT lplA, MLS-lplA, or MLS-lplA_T in the absence (A and C), and in the presence of 100 μM lipoic acid supplementation to the media (C and D). OCR was measured in at least eight biological replicates and presented are three replicated measurements of the different eight biological replicates. Viability was assessed in at least three biological replicates and mean ± SD is presented. Statistical analysis was conducted one-way ANOVA with multiple comparisons to FDX1 KO (used as control) and using Dunnett method to correct for multiple comparisons. ∗∗∗∗p < 0.0001, ∗p < 0.05. HEK293T, human embryonic kidney 293T cell line.

MLS-LplA restores lipoylation in multiple lipoylation-deficient cell models

Lack of cellular lipoylation can result from disruption of multiple pathways. To test if MLS-lplA can rescue the spectrum of lipoylation defects, we generated diverse models of eukaryotic lipoylation deficiency (Fig. 3A). First, we deleted the MECR gene, which is a critical component of the mitochondrial fatty acid synthesis pathway, upstream of protein lipoylation synthesis. Multiple defects in mitochondrial fatty acid biosynthesis (including mutations in MECR) are known to result in lipoylation defects. Second, we deleted the BOLA3 gene, an Fe–S cluster chaperone required for LIAS activity, reflecting defects in upstream Fe–S cluster biosynthesis. Finally, we deleted multiple genes in the lipoylation pathway LIAS, LIPT1, LIPT2, and FDX1 (Figs. 3 and S1). Not unexpectedly, MECR, BOLA3, LIAS, LIPT1, LIPT2, and FDX1 KO cells all exhibited loss of protein lipoylation (Figs. 3B and S1, A and B). Regardless of the upstream mechanism driving the loss of lipoylation, this reduction in lipoylation was completely restored by overexpression of MLS-lplA and lipoic acid supplementation (Figs. 3B and S1C). Thus, establishing that MLS-lplA could be used to rescue cellular phenotypes associated with a variety of lipoylation deficiency disorders.

Figure 3.

Mitochondrial-targeted lplA can restore lipoylation in different lipoylation null models.A, schematic representing the different lipoylation regulating enzymes and their corresponding upstream regulation (orange-mitochondrial fatty acid synthesis regulating the synthesis of the lipoic precursor (MECR)). Green lipoylation is dependent on a functional Fe–S cluster biosynthesis pathway (BOLA3). Blue-lipoylation regulating enzymes (LIPT1, LIAS, and FDX1). B, MLS-lplA was overexpressed in JHH7 cells with CRISPR/Cas12 KO of either MECR, BOLA3, LIPT1, LIAS, FDX1, or Ch3.1 as control. The levels of lipoylated DLAT and DLST were analyzed in the presence or absence of supplemented 100 μM lipoic acid to the media.

MLS-lplA rescues lipoylation in a cellular model of sideroblastic anemia because of a homozygous LIPT1 missense mutation

We studied a 15-year-old Mexican female patient with congenital microcephaly, spastic quadriplegia, severe cerebral palsy, and a seizure disorder, subsequently diagnosed with congenital sideroblastic anemia (CSA) at age 10. Prior to the CSA diagnosis, she had received several transfusions for anemia of uncertain etiology. At diagnosis, she was treated with pyridoxine, ascorbic acid, and folic acid to which she reportedly had a response and was stable for approximately 3 years. Thereafter, she developed a transfusion-dependent anemia, with a baseline hemoglobin of 5.2 g/dl. At the time of last evaluation, 1 month following transfusion, she had normocytic anemia (hemoglobin 7.2 g/dl and mean cell volume 86.5 fl) with an increased red blood cell distribution width (16.5%) and mild aniosopoikilocytosis and notable for occasional microcytes. There was no evidence of hemolysis or parvoviral infection, and liver function was normal. The serum iron was normal (58 μg/ml), but the ferritin was markedly increased (1395 μg/L). Erythrocyte protoporphyrin was mildly increased. The bone marrow showed a decreased myeloid:erythroid ratio (∼1:1) and numerous ring sideroblasts. The patient was subsequently lost to follow up (year 2000).

Whole exome sequencing (WES) analysis demonstrated a homozygous c.212C>T (p.Ser71Phe) variant (rs767568897, gnomAD VAF = 0.000014) in LIPT1, which was present in the heterozygous state in both parents (Fig. 4A). This variant has previously been reported in a patient with a fatal early onset lactic acidosis in the compound heterozygous state (c.212C>T/c.292C>G; p.Ser71Phe/p.Arg98Gly) (23). The crystal structure of bovine LIPT1 has been resolved (29), and Thr71 (in cows, Ser71 is not conserved) directly engages with His108, which forms the hydrophobic pocket (His108, His177, Trp66, and Arg99) where the dithiolane ring of Lipoyl-AMP is buried (Fig. 4B).

Figure 4.

MLS-lplA overexpression rescues lipoylation defects in cell line models with an engineered patient mutation in LIPT1 (c.212C>T).A, whole exome sequencing analysis showed a homozygous c.212C>T (p.Ser71Phe) variant in LIPT1, which was present in the heterozygous state in both parents. B, structure of bovine LIPT1 in complex with lipoyl AMP (Protein Data Bank code: 2E5A). Residue Thr71 (the bovine Ser71) and His82 interacting with the lipoyl-AMP are annotated. C, K562 cell clones were generated with either WT LIPT1 (C/C), heterozygous (T/C), or homozygous (T/T) 212C>T mutation in LIPT1 and characterized for the levels of protein lipoylation and indicated proteins by immunoblot analysis. D, three homozygous WT (C/C WT) and mutant (T/T Mut) K562 cell line clones were analyzed for their replicative proliferation in low glucose conditions (2 mM). The data are presented as the mean ± SD of the three clones, each measured in biological triplicates. E, the levels of lipoylation and other indicated proteins are presented in K562 WT (LIPT C/C), heterozygous mutant (LIPT1 T/C), and LIPT1 homozygous mutant (LIPT1 T/T) cells with and without the overexpression of MLS-lplA in the presence and absence of 100 μM lipoic acid (LA) supplementation. F, the replicative viability of three homozygous WT (C/C WT), mutant (T/T Mut), and mutant (T/T Mut) overexpressing MLS-lplA K562 cell line clones growing in low glucose conditions (2 mM) in the presence of supplemented 100 μM LA was analyzed over time. The data are presented as the mean ± SD of the three biological replicates. The slope of the linear fit and the SD of the fit are also presented.

Using CRISPR–Cas9 editing technology, we engineered K562 erythroleukemia cells with the CSA patient-specific c.212C>T (p.Ser71Phe) mutation in LIPT1. Analysis of multiple individual clones confirmed that the lipoylation defects are present only in the homozygous (T/T) mutants (Fig. 4C), resulting in impaired proliferation in low glucose conditions (Fig. 4D). Overexpression of MLS-lplA together with lipoic acid supplementation restored protein lipoylation (Fig. 4E) and the ability to proliferate in low glucose (Fig. 4F). Quantification of the growth kinetics show that LIPT1 mutant cells' growth in low glucose was reduced to one cell division per ∼4 days (0.25 cell divisions/day) from ∼2 days (0.53 cell divisions/day) in the WT cells. MT-LPLA overexpression in LIPT1 mutant cells in the presence of lipoic acid restored the growth rate in low glucose conditions of the LIPT1 mutants to rates that were slightly faster than the WT levels (0.58 divisions/day) (Fig. 4F). Together, these findings indicate that bacterial lplA enzyme expressed in human cells and targeted to mitochondria can completely restore lipoylation in different genetic models of a human mitochondrial disease associated with a lipoylation-deficient phenotype. Given the small size of the MLS-lplA enzyme, it may present a new therapeutic modality for diseases associated with defects in lipoylation.

Discussion

A subcategory of inborn errors of metabolism disorders is attributed to reduced cellular protein lipoylation. The diverse mechanisms underlying lipoylation defects make it very challenging to propose viable therapeutic strategies to address these disorders individually. However, in this work, we provide a proof-of-concept method that can efficiently restore lipoylation defects in cells, regardless of the upstream cause. Overexpression of an engineered bacterial lplA enzyme completely restored lipoylation in multiple lipoylation null cell models by utilizing lipoic acid supplemented in the media, bypassing the need for a functional lipoic acid biosynthesis pathway.

Previous studies have described patients with biallelic LIPT1 mutations with overlapping phenotypes, including lactic acidosis, developmental delay, hypotonia, spastic quadriplegia, cortical/cerebellar atrophy/microcephaly, abnormal myelination, and aortic root dilatation (13, 21, 22, 23, 24), but none has been reported with sideroblastic anemia as described here. It is possible that the spectrum of clinical manifestations arises from residual activity of the different LIPT1 mutants. The c.212C>T (p.Ser71Phe) LIPT1 mutation was previously characterized in a patient (24), with symptoms that included bradycardia, hypertonia, dystonic movements, and pulmonary hypertension. That patient was heterozygous for the c.212C>T (p.Ser71Phe) LIPT1 mutation in trans of a different allele (c.292C>G p.Arg98Gly) (24). This suggests that the diversity of clinical phenotypes observed in different LIPT1 mutant patients is largely affected by the residual activity of the mutant LIPT1; complete loss of function is almost certainly embryonic lethal, as shown in mouse models (13). In addition, the sideroblastic phenotype observed here may indicate that other CSAs because of defects in mitochondrial Fe–S biogenesis are principally or in part the consequence of their effect on lipoylation (30).

This work establishes that lplA enzyme can be efficiently targeted to mitochondria to restore any lipoylation defects. Unlike in yeast where lplA overexpression could not restore respiration defects induced by lip3 deletion (human LIPT1) because of the failure to restore lipoylation on kgd2 (human DLST) (28), we show that mitochondrial targeting of lplA in human cells is effective. In addition, we establish that external lipoic acid supplementation enables dosage regulation. Despite this dosage regulation, we observed that the overexpression of MLS-lplA in the presence of exogenous lipoic acid was very efficient in restoring the different lipoylation defects but did not result in lipoylation levels that were profoundly higher than basal lipoylation levels. Also, overexpression of MLS-lplA did not result in any detected lipoylated proteins that were not observed in parental cells and did not exhibit a significant increase in growth kinetics in low glucose over parental cells. This specificity in lplA-mediated lipoylation suggests that the mechanism of substrate recognition is shared between bacterial lplA and human LIPT1, and that the maximal level of lipoylation, at least for the detected DLST and DLAT enzymes, is not constrained by the availability of synthesized lipoic acid.

The challenges for developing gene therapies for mitochondrial diseases include achieving optimal expression in the desired organs (delivery), ensuring the delivery of the therapeutic gene into mitochondria and achieving adequate gene dosage (31). Any successful in vivo application of this lipoylation reconstruction strategy will require efficient delivery to target tissues (32) using established nanoparticle- (33, 34, 35) or virus- (36, 37)based approaches. The feasibility of such an approach is supported by the Food and Drug Administration approval of gene therapies for neurological disorders such as spinal muscular atrophy (38) and retinal dystrophy (39) and numerous clinical trials currently ongoing at different stages (31) using gene therapy approaches to treat different neurological disorders. One example conceptually resembling lipoylation disorders is propionic acidemia, a rare, metabolic, life-threatening disorder that causes multisystemic complications including growth retardation, neurological manifestations, cardiomyopathy, arrythmias, and more. Recently, it was demonstrated in a phase 1/2 trial that a gene therapy approach, which introduced the mRNA encoding both human PCCA (728 amino acids) and PCCB (539 amino acids) proteins in a lipid nanoparticle could be beneficial as a therapeutic for propionic acidaemia. In summary, our findings demonstrate that a small and tunable enzyme is sufficient to restore the defects induced by a variety of upstream mechanisms, establishing an attractive and feasible synthetic approach for the future development of lplA-based gene therapy to treat inherited lipoylation disorders.

Experimental procedures

Cell lines and growth conditions

Parental and engineered lines of human embryonic kidney 293T (HEK293T) and JHH7 cells used in this study were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS). K562 cells were grown in RPMI media supplemented with 10% FBS. For the glucose experiments, cells were grown in either Dulbecco's modified Eagle's medium without glucose (ThermoFisher Scientific) or in RPMI media without glucose, each supplemented with 10% dialyzed FBS and 1 mM pyruvate and 2 mM fresh glutamine and indicated concentrations of glucose (2mM-10 mM). All cell lines are routinely validated using short-tandem repeat profiling and routinely tested for mycoplasma contamination using a PCR-based approach.

Generation of HEK293T LIPT2/LIAS/LIPT1/FDX1 KO cell lines

To generate HEK293T FDX1, LIPT2, LIAS, or LIPT1 KO cell lines, cells were pretransduced with previously validated gRNAs targeting LIPT2/LIAS/LIPT1 cloned into BRD016 guide only lentivector. Cells were selected for 1 week to ensure optimal gRNA expression and then subjected to CRISPR–Cas9 ribonucleoprotein (RNP) nucleofection. The RNP for nucleofection was formed using a 1:1 ratio of Alt-R CRISPR–Cas9 tracrRNA (100 μM) and Alt-R CRISPR–Cas9 predesigned crRNAs (100 uM) (Integrated DNA Technologies). The mixture was incubated at 95 °C for 5 min and then brought to room temperature to form the gRNA duplex. The gRNA duplexes (50 μM) were then combined with Alt-R Cas9 enzyme (61 μM) (Integrated DNA Technologies) at a ratio of 3:2 and incubated at room temperature for 10 to 20 min to form the RNP for the nucleofection described later.

Nucleofection of HEK293T cells was conducted according to the manufacturer’s protocol (Lonza). About 500,000 cells per sample were pelleted by centrifugation at 4 °C using a tabletop centrifuge at 3000 rpm for 5 min. Cell culture media supernatant was aspirated and discarded, and the cells were resuspended in 30 μl of a mixture containing 20 μl Lonza SF cell line solution, 5 μl of the freshly complexed Cas9-RNP, 1 μl electroporation enhancer, and 4 μl PBS (Lonza Biosciences). Each cell suspension was added to individual wells of an electroporation cuvette (Lonza Biosciences) and then nucleofected using the Lonza 4D system program A549 pulse code CM-130. Cells were then transferred from the cuvette into prewarmed media in a 6-well or 12-well culture plate and allowed to grow undisturbed until reaching 80% confluence. The bulk nucleofected cell population was validated for target KO through immunoblot analysis of loss of protein lipoylation, then used in subsequent experiments.

CRISPR–Cas12a gene modification efficiency validation using next-generation sequencing

JHH7 KOs were validated by sequencing through the MGH CCIB DNA Core CRISPR Sequencing platform. Samples were prepared for sequencing by extracting gDNA using the Qiagen DNeasy Blood and Tissue Kit and amplifying using a set of primers (described in Table S1). Resulting FastQ files were analyzed using the CRISPResso2 program from the Pinello Lab (Available in Github). Final KO efficiency % comes from the % of sequences modified of the total number of aligned sequences after setting the quantification window center to 1 (default for Cpf1) and quantification window size to 20. https://github.com/pinellolab/CRISPResso2.

Lentiviral production

Lentiviral production was performed using HEK293T cells as described on the Broad Institute Genetic Perturbation Web Portal (https://portals.broadinstitute.org/gpp/public/).

Generation of JHH7 LIPT1/MECR/BOLA3/LIAS/FDX1 KO cells

To generate the JHH7 LIAS, LIPT1, FDX1, BOLA3, or MECR KO cell lines, JHH7 cells overexpressing a previously validated CRISPR-EnAsCas12a (lentivector, pRDA_174) were transduced with sgRNAs cloned into pRDA-052 plasmid (two gRNAs for each vector, see Supplementary Table). Cells were selected with puromycin 1 μg/ml for 4 days. KOs were validated using immunoblot or sequencing-based approaches.

Cloning of bacterial lplA and lplA_T genes

lplA genes were mammalian codon optimized and cloned into lentiviral expression vectors. An N-terminal MLS and C-terminal V5 tag were added where mentioned, and a K134R mutant lplA (enzymatic dead) was cloned to be used as a control (detailed sequences provided in Table S1). lplA and lplA_T constructs were mammalian codon-optimized and synthesized by Twist Biosciences into their pTwist Lenti SFFV expression vector. To clone MLS-lplA and WT lplA_T vectors, we generated PCR fragments with 25 bp homology arms and used Gibson Assembly Master Mix (NEB) to create plasmids for transformation and sequence verification. All lplA and lplA_T constructs were synthesized into the pTwist Lenti SFFV expression vector and used in subsequent experimental assays. Synthesized MLS-lplA in a DONR vector was used to create the plix313 MLS-lplA vector with hygromycin resistance used in the JHH7 KO studies.

Establishing the K562 cells harboring both homozygous and heterozygous c.212C>T/p.Ser71Phe mutations in LIPT1

For genetic validation experiments, homozygous and heterozygous mutations in LIPT1 were introduced in K562 cells as previously described (40) with the following crRNA: 5′-ATGCCTACCAATTACAACAG and asymmetrical template of reparation:

5′-ATCAAAATCTGGCTGTGGAAGACTGGATCCATGACCATATGAATCTAGAAGGCAAACCAATTCTATTCTTTTGGCAGAATTCTCCTTTT.GTTGTAATTGGTAGGCATCAAAATCCTTGGCAGGAA (bolded: PAM site with a dot indicating cutting site, italic: crRNA: italic; underlined: introduced mutations p.70Pro = and p.Ser71Phe).

The K562 clones were genotyped twice (15 days after nucleofection, to establish the genotype, and after four passages, to confirm the genotype). In each case, DNA was prepared with QuickExtract DNA Extraction Solution (Lucigen) following the manufacturer’s directions and quantified using a Nanodrop. About 100 ng were used to amplify LIPT1 first exon (that contains the targeted locus) with target-specific primers (LIPT1F 5′-TCAAAATCTGGCTGTGGAAGAC and LIPT1-R 5′GGTAGACTGTTCCTCCTCCAC) using the NEB Q5 master mix (NEB M0492L), following the manufacturer’s protocol. PCR products were sequenced with the same primers by Genewiz from Azenta Life Sciences. Traces were analyzed with Sequencher 5.4.1. The K562 edited clones (WT, heterozygous, and homozygous) were cultured with RPMI1640 medium (11875-093; Gibco) and supplemented with 10% FBS (MilliporeSigma), 2 mM glutamine (Corning), and 100 IU/ml penicillin/100 μg/ml streptomycin (Corning), and routinely tested for mycoplasma.

Whole exome sequencing

Patient samples were obtained with informed consent. WES of leukocyte DNA at ∼50x mean exon coverage was performed on the patient and parents using Agilent SureSelect XT Human All Exon V5 library preparation and an Illumina HiSeq2000 with 100 bp paired end. Analysis was performed using the Genuity Discovery Platform (genuitysci.com, Genuity Science).

Cell proliferation assays

HEK293T FDX1 KO cells overexpressing different lplA variants were plated on 96-well plates (at seeding concentration of 50 K/ml), and rate of proliferation was analyzed by measuring the confluency using Incucyte (standard protocols) when growing in media with low glucose (2 mM) in the presence or absence of 100 μM lipoic acid. K562 WT, LIPT1 mutant, and lplA enzyme overexpressing cell viability was analyzed by plating cells in 6-well format (at least three biological triplicates) at 250 K/ml seeding concentration in indicated media (low glucose 2 mM with 100 μM lipoic acid supplementation). The replicative rate was calculated by performing a linear regression using the PRISM software (GraphPad).

Oxygen consumption rate analysis

Oxygen consumption rate of intact cells was measured using Seahorse XF96 Analyzer (Seahorse Biosciences). Each measurement was performed over 6 min after a 3 min mix and a 3 min wait period. Basal measurements were collected three times with eight biological replicates.

Immunoblot analysis

Cells were lysed using radioimmunoprecipitation assay lysis buffer (Sigma–Aldrich) with protease inhibitor cocktail tablets (Sigma–Aldrich). Protein was quantified using the bicinchoninic acid method with a bovine-specific albumin standard curve for normalization. Each protein extract was boiled in 1x LDS sample buffer with 1:10 Tris(2-carboxyethyl)phosphine solution reducing agent and then size fractionated via precast SDS-PAGE Bis–Tris 4 to 12% gels (Thermo Fisher Scientific), and transferred onto nitrocellulose membranes with the iBlot-2 system (Thermo Fisher Scientific). LICOR Odyssey blocking buffer was used for blocking and antibody mixes. Antibodies used in this study include FDX1 (abcam; ab108257), lipoic acid (Millipore, catalog no.: 437695 and abcam ab58724), LIAS (Proteintech 11577-1-AP), DLAT (Cell signaling 12362S and Thermo), DLST (Cell Signaling 5556S), vinculin (abcam ab130007), OXPHOS (abcam ab110413), and IRDye secondary antibodies (LI-COR). Membranes were imaged on a LI-COR Odyssey machine. FDX1, DLAT, DLST, lipoic acid, LIAS, and MECR antibodies were validated using gene deletion in this study and previous studies (11, 41). OXPHOS and vinculin antibodies corresponded to expected molecular weight as described by vendor.

Data availability

All data supporting the findings of this study are available within this article. Any further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Peter Tsvetkov (ptsvetko@bidmc.harvard.edu).

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by Donita C. Brady