Tailored Amino Acid Treatment for Pediatric Patients With Mitochondrial Aminoacyl‐tRNA Synthetase Deficiencies Authors: I. J. J. Muffels, G. Kok, Z. Tang, L. S. Medeiros, F. O. Poswar, W. P. Vermeij, R. Rodenburg, K. Smit, S. K. Sukumaran, S. B. Wortmann, L. Averdunk, J. Munch, J. Pak‐lam Ho, A. M. K. Kwok, E. E. S. Nieuwenhuis, P. M. van Hasselt, I. V. D. Schwartz, C. M. A. Lubout, A. M. Bosch, K. Koop, E. M. M. Hoytema van Konijnenburg, S. A. Fuchs Published: 2026 ABSTRACT ABSTRACTAminoacyl‐tRNA synthetases (ARSs) facilitate the loading of transfer RNAs with their cognate amino acids in the cytosol (ARS1), mitochondria (ARS2), or both (dual ARS). ARS deficiencies are clinically heterogeneous and can cause severe, occasionally fatal disease. Supplementation with cognate amino acids to prevent local amino acid depletion has shown promise in ARS1 deficiencies, but its potential in ARS2 deficiencies remains largely unexplored. We investigated the effects of cognate amino acid depletion on mitochondrial function in fibroblasts from seven patients with ARS2 deficiencies using Seahorse respirometry and imaging flow cytometry. During amino acid depletion, patient fibroblasts showed decreased mitochondrial membrane potential and reduced basal and maximal respiratory capacity. We additionally evaluated oral cognate amino acid supplementation in nine patients with SARS2, EARS2, LARS2, WARS2, FARS2, or VARS2 deficiencies. Supplementation was well tolerated, with no adverse effects observed. Several patients showed clinical improvements during treatment, including improved development, increased growth, and/or improved cardiac function. In two patients, treatment discontinuation coincided with severe neurological regression. Together, these findings demonstrate that assessing mitochondrial function during cognate amino acid depletion provides a novel functional assay for ARS2 deficiencies that may aid variant interpretation and supports a biological rationale for cognate amino acid supplementation. Clinical supplementation was well tolerated and showed preliminary beneficial effects. These findings support further investigation of cognate amino acid supplementation as a potential supportive treatment strategy for patients with ARS2 deficiencies. Introduction Aminoacyl‐tRNA synthetases (ARS) are essential for translating messenger RNA to proteins by coupling amino acids to their cognate transfer RNAs (tRNAs) in the cytosol (ARS1), mitochondria (ARS2) or both (Glycyl‐RS, GARS1 or Lysyl‐RS, KARS1). The clinical phenotype of patients with ARS deficiencies differs depending on the specific ARS and whether their location is cytosolic or mitochondrial [1, 2, 3, 4]. In ARS1‐deficient patients, it is generally conceptualized that aminoacylation often suffices in homeostatic conditions, but falls short during increased translational demands (rapid growth, illness) or decreased amino acid availability (starvation, vomiting) [5]. In fibroblasts of ARS1‐deficient patients, we found that low cognate amino acid concentrations were detrimental for cellular growth [5]. To prevent amino acid deficiencies, several studies explored oral supplementation for ARS1‐deficient patients with cognate amino acids, resulting in individual clinical improvements in, for example, growth, head circumference, development, gastrointestinal and pulmonary symptoms, depending on the underlying phenotype [5, 6, 7, 8, 9]. The beneficial effects of amino acid supplementation in ARS1 deficiencies prompted us to explore the effects in dual ARS and ARS2 deficiencies. Although starvation and illness are similarly triggers for clinical deterioration for mitochondrial ARS deficiencies [10], it is not obvious that amino acid deficiency would lead to similar pathophysiological defects, as the regulation of mitochondrial translation differs from the regulation of cytosolic translation [11]. Moreover, due to the versatile function of mitochondria, supplemented amino acids may be used for pathways other than mitochondrial translation, for example to fuel the TCA cycle. Recently, beneficial effects on clinical symptoms have been observed during amino acid treatment in a patient with FARS2 deficiency during a short open‐label study [12], and in two patients with YARS2 deficiency [13]. Additionally, in four patients with RARS2 deficiency, it was reported that brief administration of arginine led to EEG improvement and higher energy levels, but seizure frequency did not change [14]. In six adults with AARS2 and DARS2 deficiencies, amino acid treatment yielded no clear clinical benefits [15]. Taken together, these conflicting and preliminary findings underscore the need for systematic evaluation of amino‐acid supplementation in larger, well‐characterized pediatric cohorts. Here, we tested the effects of amino acid depletion on mitochondrial function in fibroblasts derived from seven patients with various ARS2 deficiencies. Since amino acid depletion led to mitochondrial dysfunction in patient‐derived fibroblasts, we supplemented nine ARS2‐deficient patients with cognate amino acids for a period ranging from 6 months to 5 years to prevent shortages, and report on the clinical effects observed. Methods Patient Inclusion Inclusion criteria for the in vitro section of this study were the presence of pathogenic bi‐allelic variants in ARS2 genes and availability of fibroblasts that were obtained for diagnostic purposes. For the clinical part of this study, the inclusion criteria were as follows: (1) The presence of pathogenic, bi‐allelic variants in ARS2 genes. (2) Patients < 18 years of age at the start of treatment. (3) Patients were treated for at least 6 months with ≥ 3× the recommended WHO intake of the cognate amino acid. (4) Height, development and NMDPS scores could be retrieved from patient records before and after treatment. Genetic variants were identified using Exome Sequencing [16]. Patients signed informed consent for the use of their medical data and storage and expansion of their cells in the Wilhelmina Children's Hospital metabolic biobank (TCBio 19‐489/B, https://tcbio.umcutrecht.nl). All procedures performed in studies involving human participants were in accordance with national and local institutional review boards (IRBs) of the participating centers. Fibroblast Cultures Fibroblasts were obtained from forearm skin biopsies and cultured in Ham's F‐12 supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin (PS). For amino acid sensitivity experiments, amino‐acid free DMEM/F‐12 with HEPES and NaHCO3 (US Biological) was supplemented with 1% PS, 1% GlutaMAX (Gibco), 10% dialyzed FBS (ThermoFisher) and all amino acids except the cognate amino acid in the following concentrations: L‐Alanine 4.45 mg/L, L‐Arginine 147.5 mg/L, L‐Asparagine 7.5 mg/L, L‐Aspartic Acid 6.65 mg/L, L‐Cysteine hydrochloride monohydrate 17.56 mg/L, L‐Cysteine dihydrochloride 31.29 mg/L, L‐glycine 18.75 mg/L, L‐Histidine 31.48 mg/L, L‐methionine 17.24 mg/L, L‐Isoleucine 54.47 mg/L, L‐Leucine 59.05 mg/L, L‐Lysine 91.25 mg/L, L‐Phenylalanine 35.48 mg/L, L‐Valine 52.85 mg/L, L‐Proline 17.25 mg/L, L‐Serine 26.25 mg/L, L‐Threonine 53.45 mg/L, L‐Tryptophan 9.02 mg/L, L‐Tyrosine 55.79 mg/L. GlutaMAX was replaced by L‐glutamic acid of 7.35 mg/L for P2AARS2 and P4AARS2 since GlutaMAX contains alanine. For amino acid sensitivity tests, patient‐ and healthy donor derived fibroblasts were treated with the specific amino acid, with normal blood plasma concentrations defined as 100% (L‐Valine 223 μmol/L, L‐Alanine 350 μmol/L, L‐Glutamate 550 μmol/L, L‐Lysine 129 μmol/L, L‐Tryptophan 43 umol/L, L‐leucine 114 umol/L, L‐isoleucine 57 umol/L) and in deprivation conditions as 1% of these concentrations. In both the 1% and the 100% condition, dialyzed FBS was used. Amino acids concentration in culture medium during supplementation and depletion were assessed previously by Kok et al. [17]. Seahorse Mitochondrial Stress Test Dermal fibroblasts were plated at 10 000 cells/well in Seahorse XF24 V7 Cell Culture Microplates (Agilent) and treated with the desired amino acid concentrations for 48 h. Afterwards, cells were incubated with Seahorse XF DMEM Medium (Agilent) for 1 h in a non‐CO2 incubator. Mitochondrial respiration was measured as oxygen consumption rate (OCR) using the Seahorse XFe24 Extracellular Flux Assay Kit (Mito Stress Test, Agilent) and XFe24 Seahorse Analyzer (Agilent) with 1 μM oligomycin, 1.5 μM FCCP, 1 μM Rotenone (All MedChemExpress) and 1 μM antimycinA (Sigma‐Aldrich) following the manufacturer's protocol and with three or four technical replicates per condition [18]. After analysis, nuclei were stained with Hoechst 33342 (1:1000, Thermo Fisher Scientific), visualized with Leica Thunder Live Imager (Leica Microsystem) with intensive computational clearance method, and counted using ImageJ v1.53o (NIH) for normalization. For each individual in this study, 3–4 wells were analyzed per condition. All of these wells were analyzed, unless the drop or increase in OCR after addition of Oligomycin, FCCP or Rotenone/Antimycin A was less than 20%. Limited changes to these compounds suggest that the compound was not released from the upper chamber in the cartridge. Membrane Potential Measurements Fibroblasts were incubated with Tetramethylrhodamine, methyl ester (TMRM, Sigma Aldrich) in Hank's Balanced Salt Solution (HBSS, Gibco) for 35 min. To assess TMRM background, Carbonyl cyanide‐p‐trifluoromethoxyphenylhydrazone (FCCP, Sigma Aldrich) was added during the final 5 min in 50% of the wells. Cells were harvested using TrypLE and visualized using the Amnis Imagestream Mk II (Luminex). Cell Proliferation Cell proliferation was measured using a real‐time‐cell‐analyzer (xCELLigence MP) as described previously [5]. Amino Acid Supplementation Oral amino acid supplementation protocols, including dosing strategy, safety assessment and outcome parameters are added as Table S2. Since a recent study demonstrated that many pathogenic variants impact the editing domain of ARS enzymes, thereby preventing the incorporation of non‐cognate amino acids, we focused our protocol on ensuring an adequate intake of cognate amino acids, and less so on protein supplementation (i.e., intake of non‐cognate amino acids) [17]. Statistical Analyses Imaging Flow Cytometry data analysis was performed using IDEAS software (version 6.2, Amnis). Seahorse data were analyzed using Agilent Wave software (version 2.6, Agilent). Results Mitochondrial Dysfunction Upon Amino Acid Deprivation in Mitochondrial ARS Deficiencies Sixteen patients were eligible for inclusion in this study, and we were able to obtain fibroblasts from seven of these patients (Figure 1, Note S1): AARS2 (N = 2, Figure S5), VARS2 (N = 2), EARS2 (N = 1), WARS2 (N = 1) and KARS (N = 1). All ARS2‐deficient patient fibroblast lines showed reduced basal and maximal respiration during cognate amino acid deprivation, which was not observed in healthy controls or patients with IARS1 and LARS1 deficiency during amino acid depletion (Figures 2A–D and S1A,B). For P4VARS2, we observed only a small decrease in cell count‐normalized maximum respiration compared to the large decrease in non‐normalized maximum respiration (Figure S1C), and thus the reduced mitochondrial respiration might be related to cellular growth deficits in this patient. In fibroblasts of a patient with primary mitochondrial disease due to a homoplasmic pathogenic variant (c.3243A>G) in the mitochondrial leucine tRNA (MT‐TL1) gene, valine depletion did not affect mitochondrial dysfunction, suggesting that the effect of cognate amino acid depletion is ARS2‐specific (Figure S1D). Flow chart. Showing the inclusion process of ARS2‐deficient patients for in vitro and clinical studies. ARS2 and dual ARS deficiency patients show mitochondrial dysfunction during amino acid deprivation. (A) Oxygen consumption rate (OCR; Y‐axis) for each measurement (X‐axis) with the injection of different compounds (Oligomycin, FCCP, Rotenone/Antimycin A, arrows) in VARS2 (2 patients, 3 and 4 replicates each), AARS2 (2 patients, 3 and 4 replicates), EARS2 (1 patient, 5 replicates), WARS2 (1 patient, 5 replicates) deficient fibroblast lines during treatment with normal blood plasma amino acid concentrations (indicated at 100%) and amino acid deprivation conditions (indicated as 1%). The dots represent the average of all donors and technical replicates combined. Error bars represent SEM. (B) Oxygen Consumption Rate (OCR; Y‐axis) for each measurement (X‐axis) with the injection of different compounds (Oligomycin, FCCP, Rotenone/Antimycin A, arrows) in 100% and 1% amino acid conditions for 2 days in a patient with KARS deficiency (3 replicates). The dots represent the average of all donors and technical replicates combined. Error bars represent SEM. The graph on the left has a Y‐axis ranging from 0 to 80 pmol/min, similar to Figure 2A. The graph on the right has an adjusted Y‐axis ranging from 0 to 20 pmol/min. (C) OCR of one representative healthy control fibroblast line treated with 100% alanine and 1% alanine for 2 days. The dots represent the average of three technical replicates. Error bars represent SEM. (D) Bar graphs showing the average basal and maximal OCR of healthy controls, ARS2 and dual ARS deficiency patients. The OCR values of the 1% condition were normalized to the 100% condition and depicted as percentages. These normalized values were compared between patients and healthy controls. The bars represent the mean percentage of all healthy controls/patients ±SD. Each dot corresponds to one biological replicate (patient/healthy control). The number of technical replicates is similar as described for Figure 2A. Statistics were calculated using two‐way repeated measures ANOVA. (E) Mitochondrial membrane potential during 1% and 100% conditions. The dots refer to TMRM intensity values of single cells, after normalization on FCCP‐treated samples. The black line indicates the median value; the upper and lower hinges correspond to the 25th and 75th percentile. The upper and lower whiskers extend to 1.5 × Interquartile range (IQR). Fibroblasts of all mitochondrial ARS2‐deficient patients showed decreased mitochondrial membrane potential during amino acid deprivation, except for P2EARS2 (Figure 2E). Decreased membrane potential during amino acid deprivation was not observed in healthy controls, except for healthy controls during lysine depletion, which did lead to a significant drop in membrane potential (Figure 2E). Decreased membrane potential was similarly not observed in ARS1‐deficient patients during cognate amino acid depletion. Clinical Observations in Patients Treated With Amino Acid Supplementation The mitochondrial dysfunction observed in fibroblasts of patients with ARS2 deficiencies during amino acid depletion suggested that amino acid availability correlated with mitochondrial function. Therefore, we initiated cognate amino acid supplementation in nine patients (Figure 1). We started treatment in patients with double the daily recommended WHO amino acid intake, which was increased to three times if well tolerated (Table S2). Figure 3A shows the individual treatment regimens, duration and a summary of clinical effects for each patient. Clinical characteristics of the patient cohort are summarized in Table 1. Clinical improvements of the patients are summarized in Table 2. Summary of symptoms and treatment effects in ARS2‐deficient patients. (A) Graphical representation of clinical disease course over time before and during treatment for each patient. Blue bars indicate the time during which amino acid treatment was given. Below, disease course and clinical effects of treatment are summarized. (B) Showing the Denver developmental scale for patients in whom developmental milestones were recorded before and during treatment (P3VARS2, P1SARS2, P2EARS2). AARS2‐deficient patients (aged > 8 years) had already acquired all developmental milestones before treatment initiation and were not included. P4VARS2 was on ventilator support and did not develop any developmental milestones before and during treatment. P1WARS2 her developmental progression was recapitulated in the Bayley‐III scores and not shown in these graphs. The green lines and planes indicate the normal range (50th–90th percentile, p50–p90) for developmental progression. The black dotted lines indicate the age where treatment was initiated. Representative domains are shown for each patient. (C) Showing patient height at the start and after oral amino acid supplementation. (D) Showing Newcastle Pediatric Mitochondrial Disease Scale scores (NPMDS) before and after amino acid supplementation. Two‐way ANOVA with uncorrected Fisher's LSD was used to calculate statistics for the 3‐, 6‐, and 12‐month timeframes. The 3‐month timepoint p value is shown. For the 6‐month timepoint, this was 0.054; for the 12‐month timepoint, the p value was 0.0273. TABLE 1Clinical characteristics of ARS deficient patients.Patient IDP1VARS2 P2VARS2 P4VARS2 P1SARS2 P1EARS2 P2EARS2 P1WARS2 P2WARS2 P1FARS2 P2FARS2 P1LARS2 Sibling of P2VARS2 Sibling of P1VARS2 Sibling of P3VARS2 Sibling of P2EARS2 Sibling of P1EARS2 Sibling of P2FARS2 Sibling of P1FARS2 Genetic variantNM_001167734 c.1258G>A; c.1973G>A; p.Ala420Thr; p.Arg658HisNM_001167734 c.1258G>A; c.1973G>A; p.Ala420Thr; p.Arg658HisNM_001167734 c.1940C>T; c.763A>G p.Thr647Met; Lys255GluNM_017827 c.1347G>A HomozygousNM_001083614 c.322C>T; c.920T>C; p.Arg108Trp p.Leu307SerNM_001083614 c.322C>T; c.920T>C; p.Arg108Trp p.Leu307SerNM_015836 c.938A>T; c.797delC; p.Lys313Met; p.Pro266Argfs*NM 015836.3: c.938A>T, p.Lys313Met HomozygousNM_006567.3: c.1082C>T; exon 5 deletion; p.Pro361Leu;?NM_006567.3: c.1082C>T; exon 5 deletion; p.Pro361Leu;?NM_015340.4: c.1256G>A; 2024A>G, p.Arg419Gln; Asp675GLyCurrent ageDeceased at 7 weeks of age6 years of age3 years of age6 years of ageDeceased 30 days after birth1 year of age6 years of age3 years of age7 years of age12 years of age2 years 7 monthsReceived palliative care?Yes, between 1 and 7 weeks of ageNoYes, at 9 months of ageNoYes, 15–30 days after birthYes, at 8 months of ageNoNoNoNoNoAge of symptom onset1 day after birth7 days after birthDuring pregnancy1 month of age3 weeks of age6 h after birth3 months of age6 months of age2 years of age16 months of ageAt birthPhenotypeHypertrophic cardiomyopathyHypertonia, hypertrophic cardiomyopathy, bilateral vocal cord paralysisLactic acidosis, hypertrophic cardiomyopathy, pulmonary hypertensionSpasticity and weakness of lower limbs, developmental delay, clonus, ataxia.Feeding difficulties, lactic acidosis, enterocolitis, sepsis.Seizures, feeding difficulties, axial hypotonia, limb hypertonia, hyperreflexia.Developmental delay, dystonia, tremor, sleeping difficulties and feeding difficulties necessitating PEG‐J feeding.Ataxia, tremor, developmental delay.Developmental delay, tremor, metabolic stroke, hemiparesisDevelopmental delay, swallowing difficulties, seizuresCongenital anemia, developmental delay, sensorineural hearing loss, penoscrotal hypospadia, ASDII, a small perimembranous VSD and a small muscular VSD.Treatment durationNA5 years2 years 6 months3 yearsNA6 months, currently 5 months without treatment2 years8 months2.5 years2.5 years15 monthsCurrent amino acid dosageNA100 mg/kg/day valine200 mg/kg/day valine100 mg/kg/day serineNA80 mg/kg/day glutamate60 mg/kg/day tryptophane40 mg/kg/day tryptophane60 mg/kg/day Phenylalanine50 mg/kg/day Phenylalanine95 mg/kg/day LeucineAdditional medicationNAMagnesium‐malate, 50 mg/day. Levocarnitine, 150 mg/day.3 g/kg/day protein supplementation, Sildenafil 5 mg Q8H, furosemide 0.5 mg/kg Q12H, Bisoprolol 0.05 mg/kg BD and Spironolactone 0.5 mg/kg Q12H3.5 g/kg protein supplementation, Thiamine 100 mg/dayNAPhenobarbital 5 mg/kg/day, Levetiracetam 36 mg/kg/day, Baclofen 2.5 mg PRN, Domperidone 2.1 mg Q8H, Esomeprazole 15 mg BD.Biotin 20 mg/day, Thiamine 15 mg/day, Omeprazole BD, Macrogol 4–8 mg BD, Melatonine 1 mg/day. Chloralhydrate PRN. iron supplementationNAThiamine 50 mg Q12H, Coenzyme Q10 50 mg Q12H, Levetiracetam 200 mg Q12HCoenzyme Q10 250 mg Q12H, Thiamine 150 mg Q8H, Oxcarbazepine 600 mg Q12HNAAbbreviations: BD = Bis in Die, twice a day; NA = not assessed; PEG‐J = percutaneous gastrostomy; PRN = Pro Re Nata, as needed; QxH = every x hours. Clinical characteristics of ARS deficient patients. Abbreviations: BD = Bis in Die, twice a day; NA = not assessed; PEG‐J = percutaneous gastrostomy; PRN = Pro Re Nata, as needed; QxH = every x hours. TABLE 2Clinical details of amino acid supplementation in ARS2‐deficient patients.Patient IDP1VARS2 P2VARS2 P4VARS2 P1SARS2 P1EARS2 P2EARS2 P1WARS2 P2WARS2 P1FARS2 P2FARS2 P1LARS2 Treated?NoYesYesYesNoYesYesYesYesYesYesTreatment ScheduleNAContinuously treated for 5 yearsContinuously treated for 6 months, stopped for 2 months, treated for 2 yearsContinuously treated for 3 yearsNA6 months with treatment, 5 months without treatment24 months of treatment8 months of treatmentContinuously treated for 2.5 yearsContinuously treated for 2.5 yearsContinuously treated for 17 monthsFollow‐upDeceased at 2 months of age5 years2.5 years6 yearsDeceased at 1 month of age11 months24 months8 months2.5 years2.5 years17 monthsGrowthNAWeight: −2 SDS to −0.38 SDS Height: −2 SDS to −1.44 SDS Head Circumference: −3 SDS to −0.86 SDSWeight: −4 SDS to −1 SDS Height: −4 SDS to 0 SDS Head Circumference: −3 SDS to −3 SDSWeight: 0 SDS to 0 SDS Height: 0 SDS to 0 SDS Head Circumference: −1 SDS to 0.8 SDSNAWeight: −4 SDS to −1 SDS Height: −3 SDS to −1.8 SDS Head Circumference: −3 SDS to −2 SDSWeight: −1.5 SDS to −0.3 SDS Height: −2.12 SDS to −1.92 SDS Head Circumference: −1.3 SDS to −0.88 SDSWeight: −1.8 SDS to −0.29 SDS Height: −2.03 SDS to −2 SDS Head Circumference: −1.28 SDS to −0.78 SDSWeight: −1.57 SDS to −2.04 SDS Height: −1.92 SDS to −1.84 SDS Head Circumference: −0.23 SDS to 0.17 SDSWeight: −1.84 SDS to −0.8 SDS Height: −2.97 SDS to −2.3 SDS Head Circumference: 0.1 SDS to 0.5 SDSWeight: 0.5 SDS to 1.3 SDS, Height: 0.5 SDS to 1.3 SDS, Head Circumference: 1.3 SDS to 1.3 SDSNeurologicalNALower limb hypertonia resolvedDystonia, developmental delay and seizures persist with treatment.Lower limb weakness and spasticity remained similar after serine treatment.NASeizures persisted Sleeps better. Less agitation and unrest. Shows interest in food. Weaned off levodopa without return of postural tremor. Dystonia unaltered. She had more energy and was much more active during the day. Her tremors decreased. Walking is more stable.Improved stamina. Pulls himself up more often and tries to drink independently. He plays more role‐playing games. He had learned additional words and forms sentences. Emotional participation and interaction had also improved. There was an improvement of 20 s in the 9‐Whole‐Peg Tet.Much more alert, awake, improved stamina. Less tired at school and laughs more. Increased walking time on the walker. He is more attentive. Ability to swallow saliva has improved. Since taking phenylalanine, he no longer vomits during the day. He is now able to support himself and play in a prone position.Feeling more energetic and appears stronger by parentsDevelopmentNAAttends regular kindergarten, no developmental delay.Development remains delayed, no measurements prior to treatment initiation. At 2.5 years of age, his gross motor, fine motor, speech and social skills are estimated at 6 months.Verbal development improved during serine treatment: Read and write 3‐letter words. From 1‐word sentences to 3–6‐word sentences.NADeveloped visual tracking and smiling at parentsDevelopmental age improved from 7–11 months to 14 months (Bayley III).Starts to use new words, and two‐word combinations since initiation of tryptophane.Plays more role‐playing games, forms sentences and learns additional words.Started babbling more, swearing. Plays and supports himself in a prone position.Motor milestones from 7/8th month of life to 9/10th month of life, fine motor skills from 12th month of life to 12/18th month of life, started to walk along furniture, plateau in development since then (6 months).GINAWeaned off tube feeding at the age of 2.NAHer constipation resolved within the first two weeks of treatment.No more vomiting during the day.Cardiac functionNAResolution of pulmonary hypertension resolution of cardiac hypertrophy (from 300% to 130% to 100% of normal size). NT‐proBNP dropped from 14.778 to 299 ng/L Troponin T dropped from 279 ng/L to 46 ng/L TAPSE 5.6 mm to 16.2 mm Cardiac hypertrophy: no improvements NALaboratoryNALactate values normalizedNA Lactate values remained normal during treatment ImmunologicDecreased occurrence of episodes of unexplained fever.Abbreviations: GI = gastrointestinal; NA = not assessed. Clinical details of amino acid supplementation in ARS2‐deficient patients. Sleeps better. Less agitation and unrest. Shows interest in food. Weaned off levodopa without return of postural tremor. Dystonia unaltered. NT‐proBNP dropped from 14.778 to 299 ng/L Troponin T dropped from 279 ng/L to 46 ng/L TAPSE 5.6 mm to 16.2 mm Cardiac hypertrophy: no improvements Lactate values remained normal during treatment Abbreviations: GI = gastrointestinal; NA = not assessed. Amino acid supplementation was well tolerated in all patients, since adverse effects were not reported and all laboratory values remained within normal range during treatment (Figure S2). In a subset of patients, laboratory values improved during treatment, specifically lactate values (P2VARS2 and P3VARS2), troponin and NT‐proBNP values (P3VARS2). We assessed development with Bayley‐III developmental scales or Denver developmental scales. In P1SARS2, language development was delayed at first but significantly accelerated after treatment (Figure 3B). Similarly, P1WARS2 showed improved language and social development during Tryptophan treatment, as measured using Bayley‐III. While P2VARS2 showed mild motor developmental delay during infancy (p90), her development significantly accelerated during valine treatment, as she now reached her milestones within the 50th percentile for age (Figure 3B). In contrast, the majority of reported VARS2‐deficient patients show developmental delay (86%) [2]. VARS2‐Deficient Patients P1VARS2 presented with respiratory insufficiency and hypertrophic cardiomyopathy immediately after birth. He died at the age of 7 weeks. Post‐mortem autopsy revealed pathogenic VARS2 variants. Three years later, the younger sister of P1VARS2, P2VARS2 was born. Two days after birth, P2VARS2 was diagnosed with laryngomalacia, for which she was admitted to the hospital. She developed cardiomyopathy, lactic acidosis and respiratory insufficiency, requiring mechanical ventilation and tracheotomy. The same pathogenic variants as in P1VARS2 were identified, accompanied by decreased VARS2 protein expression and complex I and complex IV deficiency in fibroblasts (Figure S4, Table S1). Brain MRI showed a thalamic infarction in the right hemisphere. At the age of 1 month, valine supplementation (25 mg/kg/day) was briefly administered but feeding difficulties limited her intake. As her condition worsened, a palliative trajectory was initiated, and valine supplementation was stopped. At the age of 3 months, her symptoms stabilized spontaneously and her intake improved. At the age of 4 months, valine supplementation (60 mg/kg/day) was reinitiated. One month later, protein fortification was added (22.5 g protein/day). Compared to 6 weeks of age, her condition has significantly improved; her pulmonary hypertension completely resolved; her hypertrophic cardiomyopathy completely resolved (heart size: 300% to 130% of normal within 1 year). Her lower limb hypertonia resolved. She never developed microcephaly or epilepsy and reached all developmental milestones at a normal age (Figure 3B). She is currently 6 years old. P3VARS2 was born prematurely at 35 weeks due to restricted fetal movement and growth deficits. Shortly after birth, pulmonary hypertension, cardiac hypertrophy and lactic acidosis were noted, for which he required mechanical ventilation. At 3 months of corrected age, a severe pulmonary hypertensive crisis and cardiopulmonary resuscitation led to ICU admission and ECMO together with Continuous Renal Replacement Therapy (CRRT). At the same time, exome sequencing revealed VARS2 variants, and valine supplementation was initiated. Two days after treatment initiation, ECMO treatment could be discontinued, and was replaced by Nitric Oxide, ventilator support and inotropy, which was slowly weaned down. His lactate values normalized during treatment (Figure S2). Additionally, troponin T and NT‐proBNP levels dropped significantly (Figure S2). All of his growth parameters improved during treatment (Figure S3). His biventricular cardiac hypertrophy persisted with no obvious improvements. While LVEF was normal before treatment and remained normal, TAPSE values improved from ~5–6 to 16.2 mm (reference range > 10 mm). At the age of 7 months, he became septic, impeding the intake of enteral valine, leading to worsened cardiomyopathy (TAPSE 10.6 mm). However, his condition stabilized and valine was tolerated again. His cardiomyopathy and pulmonary hypertension normalized again after valine intake (TAPSE 16.6 mm at 30 months). At the age of 2, he was able to eat independently. He still shows significant developmental delay at 2.5 years of age, since his gross motor, fine motor, speech and social skills are estimated at 3–4 months. SARS2‐Deficient Patient P1SARS2 presented in the first year of life with clonus and lower limb spasticity, accompanied by developmental stagnation after 1 year of age. At the age of 2 years and 8 months, SARS2 genetic variants were identified, and serine treatment was initiated shortly afterwards. Before treatment, he had delayed speech development consisting of one‐word sentences and non‐verbal communication (Figure 3B). During treatment, he showed catch‐up language development and was able to formulate three‐ to‐six‐word sentences (development within p50–p90), allowing him to attend regular preschool, only 6 months after treatment initiation (Figure 3B). At preschool, he started to read three‐letter words and learned the English alphabet. Currently, he is attending kindergarten where he has learned to write three‐letter words. His lower limb weakness and spasticity have not changed since treatment. LARS2 Patient This patient was the second child of consanguineous Syrian parents born after an uncomplicated pregnancy. He was small for gestational age (2830 g) and had a poor start with Apgar scores 5–6–6. The postnatal course was complicated by congenital anemia (lowest Hb 6.9 g/dL) for which he received an erythrocyte concentrate, he had a spontaneous pneumothorax requiring drainage and respiratory adaptation problems requiring CPAP. A penoscrotal hypospadia was noted as well as an atrial septal defect, a small perimembranous ventricular septal defect and a small muscular ventricular septal defect without hemodynamic relevance. He failed neonatal hearing screening (OAE). BERA (Brainstem Evoked Response Audiometry) confirmed complete hearing loss. Recurrent elevations in lactate values were noted (5 mmol/L, Ref < 2). Upon physical examination, truncal hypotonia and mild dysmorphic features (low implanted ears, bell‐shaped thorax) were noted. Upon diagnosis at the age of 15 months, leucine supplementation was initiated (95 mg/kg/day). At treatment initiation, his development was already delayed (biological age 14 months: motor milestones at the 7/8th month of life, fine motor skills at the 12th month of life, contact and play behavior: 15–18th months). Follow‐up after 3 months of leucine treatment showed that development was still delayed but slightly improved (biological age 16 months: motor milestones at the 9/10th month of life, fine motor skills at the 12–18th month of life, contact and play behavior: 15–18th months). Parents and physiotherapy noticed that he was “stronger” and had more energy. At the last follow‐up at 2 years and 7 months (after 17 months of treatment) his weight and height had improved from the 50th to the 90th percentile. He is not reacting to sounds, does not speak and only produces certain sounds. One year after treatment initiation, he has learned to pull himself up and walk along furniture but his development plateaued since then (biological age 31 months: fine motor skills at the 15th month of life, contact and play behavior: 6–8 months). EARS2‐Deficient Patient P1EARS2 was born to non‐consanguineous parents. She was admitted to the ICU at 3 weeks of age, with a COVID‐19 infection, severe lethargy and lactic acidosis (lactate: 24 mmol/L). A ketogenic diet was initiated, but 1 week later, she died due to enterocolitis and sepsis. Genomic sequencing revealed pathogenic EARS2 variants. Her younger brother, P2EARS2, was diagnosed with the same EARS2 variants antenatally. Six hours after birth, he had his first seizure, after which anti‐epileptic medication, 40 mg/kg/day glutamate and a high protein diet (3 g/kg/day) were initiated. At 1 month of age, glutamate dosage was increased to 80 mg/kg/day. The following 2 months, his seizures persisted and he developed proximal hypotonia and brief clonus upon neurological examination. He showed some developmental progression, moving both arms and legs, smiling at his parents and visually tracking objects. Severe nausea and feeding difficulties limited his intake, and glutamate treatment was discontinued at 3 months of age by her parents, to see whether nausea would decrease, but it did not. At the age of 6 months, protein fortification was changed to Neocate LCP (a hypoallergenic protein formula lacking L‐glutamate). After switching to Neocate LCP and discontinuing glutamate supplementation, the patient developed sinus tachycardia, cyclic vomiting, and progressive lactic acidosis within 3 months (Figure S2). His spasticity worsened and he showed sustained clonus upon examination. He lost previously acquired developmental skills, as he could not smile at parents, move arms or legs, or track faces anymore. He developed uncontrolled epilepsy and somnolence, and palliative care was initiated. WARS2‐Deficient Patients P1WARS2 presented with spasticity and agitation at the age of 3 months. MRI showed a thin corpus callosum and frontal lobe atrophy, with a normal myelination pattern. Her development was significantly delayed. Additionally, she developed a complex movement disorder consisting of tremor and dystonia. She had severe sleeping difficulties. She was completely dependent on tube feeding and vomited frequently. After initiation of Tryptophan (60 mg/kg/day), the vomiting stopped. Parents noted her to have much more energy and less agitation. Before treatment, she could not join family dinners due to her agitation, but she is now able to. She sleeps better through the night; where she used to wake up several times without being able to sleep thereafter, she now sleeps much better. She only needs her chloralhydrate right before bedtime instead of several times per night. Where she used to have a fever several times a month, this now occurs less frequently. While in the 2 years before Tryptophan treatment initiation, her development did not improve (developmental age 7–8 months at age 20 months, developmental age 7–11 months (Bayley‐III) right before treatment at age 42 months), her developmental age improved significantly to 14 months (actual age: 52 months) with treatment. Her sitting and walking appear more stable, and she has improved on the social and language domains. She now engages in pretend play, can point at objects in a book and leaf through it. While her dystonia has not changed, she was able to wean off levodopa, which was used to control tremors before Tryptophan initiation. After weaning off levodopa, her tremor did not return. P2WARS2 presented with developmental delay and tremors at the age of 6 months. Brain MRI was normal. Tryptophan was initiated at the age of 2 years. Her constipation resolved within the first 2 weeks of treatment. She had more energy and was much more active during the day. In the first year of treatment, it was noted that she talked more, started producing three‐word sentences, and her tremors decreased. In addition, she learned how to walk within the first year of treatment. FARS2‐Deficient Patients P1FARS2 is the younger sibling of P2FARS2. At the age of 2, developmental delay and increased leg tone was noted by parents. At the age of 4 years, a tremor of the right arm was observed, which started after an infection accompanied by reduced food intake. Since then, the patient has only been able to move this arm to a limited extent. Three months after the start of phenylalanine at the age of 4, parents attest to improved stamina. The patient pulls himself up more often and tries to drink independently. He plays more role‐playing games. He had learned additional words and forms sentences. Emotional participation and interaction had also improved. There was an improvement of 20 s in the 9‐Hole‐Peg Test. Two years after the initiation of phenylalanine, he experienced a period of illness concomitant with severely decreased intake and a temporary pause in phenylalanine supplementation. A month after this episode began, he experienced a metabolic stroke, after which he could not move both of his arms and left leg, speak, swallow, chew, crawl or pull himself up anymore. Intake and phenylalanine were reinitiated through PEG feeding, and a few months later, protein fortification was added (total protein intake increased from 1.3 to 2 g/kg/day). After re‐initiation of intake and protein fortification, he regained some of his speech, energy levels and upper body tone. However, spasticity in his left arm remains. P2FARS2 showed an unremarkable development in early life. At 16 months of age, he experienced a metabolic crisis and lost all previously acquired developmental skills. Up until the age of 10, he did not reach any new developmental milestones and started developing focal epileptic seizures at the age of 9 years. After phenylalanine supplementation, his mother reports that he is much more alert and has improved stamina. The teachers independently report that he is less tired at school and laughs more. He shows an increased walking time on his walker (previously 15 min, now 30–45 min). He is more attentive and makes more contact. After taking phenylalanine, he no longer vomits during the day. Unfortunately, he still vomits during the night. He started taking protein fortification of 0.2 kg/day 2 years after the initiation of Phenylalanine. Discussion Here, we found significant mitochondrial dysfunction during cognate amino acid deprivation in fibroblasts of mitochondrial ARS deficiencies. To prevent depletion‐induced mitochondrial dysfunction, we supplemented VARS2, SARS2, LARS2, WARS2, FARS2 and EARS2‐deficient patients with cognate amino acids. Amino acid supplementation was well tolerated in all patients. Our clinical observations show that during amino acid supplementation, a subset of symptoms improved, including cardiomyopathy, development, sleeping problems, growth, agitation, and feeding difficulties, while certain neurological symptoms (epilepsy, spasticity, dystonia) did not change. Our preliminary results concur with recently published case reports showing beneficial effects of amino acids in children with pathogenic FARS2, YARS2 and RARS2 variants [12, 13, 14]. In adults with AARS2‐ and DARS2 deficiencies, there were no improvements after supplementation [15], suggesting that amino acid supplementation might be most beneficial for patients in an early stage. We hypothesize that cognate amino acid concentrations are directly correlated with ARS2 enzyme activity, similar to what has previously been demonstrated for ARS1 enzymes [8, 19]. Unfortunately, aminoacylation activity measurements for ARS2 enzymes are still under development, and therefore, these experiments could not be repeated for mitochondrial ARS. However, our fibroblast studies identified a relationship between amino acid availability and mitochondrial function, suggesting that amino acid supplementation may help prevent shortages during catabolic states and thereby support mitochondrial translation. Consistent with this concept, experimental models have demonstrated that impaired mitochondrial ARS2 activity disrupts mitochondrial protein synthesis. For example, FARS2 knockout or RNAi knockdown in Drosophila markedly reduces tRNA‐Phe charging, leading to decreased incorporation of phenylalanine into mitochondrial proteins and reduced abundance of mitochondrially encoded OXPHOS components and respiratory chain complexes I, III, IV, and V [20]. Similar reductions in mitochondrial respiratory complex abundance and/or activity have been reported in Drosophila with MARS2 mutations, in patients with MARS2 deficiency [21], and in DARS2 knockout mice [22]. Together, these results indicate that decreased aminoacylation activity affects mitochondrial translation, leading to decreased abundance of mitochondrial complexes and decreased mitochondrial respiration. There was considerable variability in the response to cognate amino‐acid depletion among ARS2‐deficient patient fibroblasts. One possible explanation may lie in the biochemical consequences of the individual ARS2 variants, as the location and type of mutations differ between patients. These variants may affect protein abundance, impair aminoacylation activity or editing capacity, directly affect amino acid binding and/or a combination of several of these mechanisms. Consequently, the residual enzymatic activity may vary substantially between cell lines. Although it is tempting to speculate that substrate therapy is more effective with higher residual ARS2‐enzyme activity, this is impossible to ascertain without reliable ARS2‐enzyme activity assays and will be complicated by the many different mutations affecting different protein domains and irreversible versus reversible symptoms. In addition, mitochondrial amino‐acid availability is strongly influenced by transport processes, metabolic production, and recycling pathways. For example, glutamate is quite abundant within mitochondria [23], meaning that depletion may require a longer period before functional consequences may become apparent. This could explain why oxygen consumption is already reduced—reflecting slowed mitochondrial metabolism—while the mitochondrial membrane potential remains preserved and has not yet collapsed. Whether differences in residual enzyme activity influence the cellular response to cognate amino‐acid depletion thus remains unclear, and will require reliable quantification of ARS2 activity, which is currently not possible. The small number of patients, differences in clinical presentation, and fluctuating disease course made it difficult at times to discriminate treatment effects from disease course fluctuations and ongoing development. Specifically, P2VARS2 already showed clinical improvement before valine supplementation was reintroduced. However, her clinical improvement concurred with better oral (valine) intake. A comparison to her untreated sibling does suggest efficacy of treatment, as she survived well beyond the age where the untreated sibling passed away, similar to the other patients that were compared to their older sibling (N = 2). In addition, the fact that P2EARS and P1FARS2 showed significant clinical deterioration upon treatment discontinuation suggests that treatment is indeed effective. However, double‐blind, placebo‐controlled crossover trials should be used to unambiguously assess amino acid treatment effects and would allow direct comparison of clinical improvement within the same patient. The insights gained from our observations can help shape these future studies. Since we noticed improvements in the majority of patients within 6 months, treatment should be given for 6 months or longer. In addition, treatment effects were largest in younger patients, where neurological damage might be less severe and still reversible. While standardized measures can enhance the interpretability and generalizability of future trials, the highly divergent phenotypes of ARS2 deficiencies should be taken into account when choosing outcome measures. However, the majority of patients showed improvement in terms of feeling more energetic, better development, improved growth (height) and improved cardiac function. Developmental scales, growth parameters, gross motor function, 6‐min walk test, cardiac markers and patient‐reported outcome measures might thus be useful for follow‐up trials assessing amino acid supplementation on a larger scale. In conclusion, our study shows that amino acid depletion leads to mitochondrial dysfunction in fibroblasts derived from patients with ARS2 deficiencies, suggesting that prevention of amino acid shortages by supplementation could represent an effective strategy to prevent deterioration in patients. Although anecdotal, our clinical observations show beneficial effects for the majority of patients and no unwanted side effects of treatment with cognate amino acids. Future studies are needed to objectively assess treatment effects, however, the clinical heterogeneity of ARS2 deficiencies and their fluctuating disease course should be taken into account when designing these trials. Author Contributions I.J.J.M., G.K., and S.A.F. conceptualized and wrote the original draft of the manuscript. G.K., I.J.J.M., Z.T., and K.S. did the laboratory investigation, validation, and analysis. I.J.J.M. performed the visualization of results (figures, tables). K.K., P.M.v.H., C.M.A.L., I.V.D.S., L.S.M., F.O.P., S.K.S., J.P.H., A.M.K.K., R.R., E.E.S.N., S.B.W., L.A., A.M.B., E.M.M.H.v.K., and S.A.F. were involved in clinical investigation. I.J.J.M., G.K., W.P.V., K.K., E.M.M.H.v.K., and S.A.F. reviewed and edited the final version of the manuscript. Funding This work was funded by the Tjallingh Roorda Foundation. Ethics Statement Written informed consent was obtained from all patients' parents to store, expand and experiment with patient‐derived fibroblasts (TCBio 19‐489/B and TCBio 22‐538, https://tcbio.umcutrecht.nl). The same biobank has been used previously [24]. All patient data was de‐identified before usage. All procedures in this study are in accordance with the ethical standards of the institutional and/or national research committee(s) and with the 2013 Helsinki Declaration. All medical studies were performed as clinical care following local institutional guidelines [24]. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: Seahorse and membrane potential assessment in healthy controls and ARS1/ARS2 deficient patients. (A) Seahorse analyses in healthy controls treated with normal (100%) valine, lysine, leucine, isoleucine, glutamate and tryptophan, or depletion medium (1%) for 2 days. The graphs represent one representative experiment with one healthy donor. The error bars represent the SEM of three or four technical replicates. (B) Seahorse analyses performed in fibroblasts of patients with LARS1 (N = 2) and IARS1 (N = 2) deficiency, treated with normal (100%) leucine or isoleucine (respectively) or depletion medium (1%) for 2 days. The graphs represent one representative experiment with one donor. The error bars represent the SEM of three or four technical replicates. (C) Seahorse analyses showing the non‐normalized OCR in P4VARS2 treated with normal (100%) valine or depletion medium (1%) for 2 days. The error bars represent the SEM of three technical replicates. (D) Seahorse analyses showing the normalized OCR in a patient with a clinical phenotype of MELAS (Mitochondrial Encephalopathy, Lactic Acidosis, and Stroke‐like episodes) with the 100% homoplastic MT‐TL1 variant (m.3243A>G) treated with normal (100%) valine or depletion medium (1%) for 2 days. The error bars represent the SEM of four technical replicates. (E) Bar graphs showing the average basal and maximal OCR of healthy controls and ARS1‐deficient patients. The OCR values of the 1% condition are normalized to the 100% condition and depicted as a percentage. The bars represent the mean percentage of the two individuals ±SD, the dots represent the two individuals. Statistics were calculated using two‐way repeated measures ANOVA.(F) Membrane potential measurements in patients with LARS1 (N = 2) or IARS1 (N = 2) deficiency. In the patients, the bars represent membrane potential normalized to healthy controls treated with the same condition ±SD. In the healthy control, the 1% condition is normalized to the 100% condition of the healthy control. Figure S2: Laboratory values of patients before and after amino acid treatment. Grey areas indicate the period where amino acid treatment was given. The dotted horizontal lines indicate reference values in the healthy pediatric population. The vertical dotted lines indicate the age of treatment initiation. ALP = Alkaline Phosphatase, ALT = Alanine Aminotransferase, AST = Aspartate Aminotransferase, BUN = Blood Urea Nitrogen, GGT = Gamma Glutamyl Transpeptidase, Hb = Hemoglobin, WBC = White Blood Cell Count. Figure S3: Growth parameters (in SDS) for patients before and after amino acid supplementation. Grey areas indicate the period where amino acid treatment was given. The vertical dotted lines indicate the age of treatment initiation. Note S1: Pathogenicity assessment of VARS2, SARS2, LARS2, FARS2, EARS2, WARS2 and KARS1 genetic variants. Figure S4: VARS2 western blot. Western blot of P2VARS2 isolated mitochondria from fibroblasts. The patient was compared to three different healthy control fibroblast lines. Membranes were probed with VARS2 antibody (ProteinTech). Table S1: Showing the complex activities of P2VARS2 compared to the reference values. Figure S5: Western blot of fibroblasts of P1AARS2 and P2AARS2 and two healthy controls. Membranes were probed with AARS2 antibody (HPA035636, Atlas) and β‐tubulin (CST 2128S, Cell Signaling). Table S2: Amino acid treatment Protocol.