Clinical, neuroradiological, and molecular characterization of mitochondrial threonyl-tRNA-synthetase (TARS2)-related disorder

Purpose Biallelic variants in TARS2, encoding the mitochondrial threonyl-tRNA-synthetase, have been reported in a small group of individuals displaying a neurodevelopmental phenotype but with limited neuroradiological data and insufficient evidence for causality of the variants. Methods Exome or genome sequencing was carried out in 15 families. Clinical and neuroradiological evaluation was performed for all affected individuals, including review of 10 previously reported individuals. The pathogenicity of TARS2 variants was evaluated using in vitro assays and a zebrafish model. Results We report 18 new individuals harboring biallelic TARS2 variants. Phenotypically, these individuals show developmental delay/intellectual disability, regression, cerebellar and cerebral atrophy, basal ganglia signal alterations, hypotonia, cerebellar signs, and increased blood lactate. In vitro studies showed that variants within the TARS2301-381 region had decreased binding to Rag GTPases, likely impairing mTORC1 activity. The zebrafish model recapitulated key features of the human phenotype and unraveled dysregulation of downstream targets of mTORC1 signaling. Functional testing of the variants confirmed the pathogenicity in a zebrafish model. Conclusion We define the clinico-radiological spectrum of TARS2-related mitochondrial disease, unveil the likely involvement of the mTORC1 signaling pathway as a distinct molecular mechanism, and establish a TARS2 zebrafish model as an important tool to study variant pathogenicity.


Cell culture, transient transfection, and co-immunoprecipitation
HEK293T cells were maintained in Dulbecco's Modified Eagle Medium (DMEM) (Wisent) supplemented with 10% fetal bovine serum (FBS) (Wisent) and 1% streptomycin-penicillin (Wisent) in a humidified atmosphere containing 5% CO2 at 37℃.Transient transfection of plasmids was performed using Lipofectamine 2000 (Invitrogen)   according to the manufacturer's instructions.For co-IP assay, 24 h after transfection, cells were washed by ice-cold PBS and collected by scraping in the pre-chilled lysis buffer (40 mM HEPES-KOH pH 7.5, 120 mM NaCl, 1 mM EDTA, 50 mM NaF, 0.3% CHAPS, supplemented with a complete EDTA-free protease inhibitor tablet and phosphatase inhibitor cocktail).Cells were lysed using a rotator for 20 min at 4℃, and the lysates were incubated with Protein G agarose beads (Millipore) for at least 30 min at 4℃. Pre-cleared lysates were incubated with anti-hemagglutinin (HA).11epitope tag antibody (Covance) for 1 hour at 4℃, followed by incubated with Protein G agarose beads overnight at 4℃. Beads were washed with wash buffer (lysis buffer supplemented with 50 mM HEPES-KOH pH 7.5, 150 mM NaCl) three times for 10 min each.Proteins bound to beads were eluted in 2x SDS sample buffer.TrueBlot anti-rabbit IgG HRP and anti-mouse IgG HRP (Rockland) were used as secondary antibodies for immunoblotting to eliminate signals from IgG heavy and light chains.

Preparation of mutant TARS2 plasmids, cell culture, transient transfection, and co-immunoprecipitation
Plasmids overexpressing the F323C, R327Q, E342D, and R346C TARS2 mutants were derived from pcDNA4 TARS2-His/Myc wild-type (WT) plasmid 13 using the QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent Inc.USA) according to the manufacturer's instructions.Cell culture, transient transfection, and coimmunoprecipitation (co-IP) assay were performed as described previously 13 with minor modifications as described in supplementary methods.

Zebrafish functional studies
We performed zebrafish experiments in WT strain NHGRI-1 according to the protocol approved by the Institutional Animal Care Committee (IACUC) of Oklahoma Medical Research Foundation (22-76).All animals were raised and maintained in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) accredited facility under standard conditions.

Whole-mount in situ hybridization (WISH)
The whole-mount in situ hybridization was performed using a method described earlier 2 .
Briefly, 1100-bp tars2 sequence was PCR amplified from zebrafish cDNA using primers pairs listed in Table S3.Amplified DNA fragment was used as a template to synthesize digoxigenin-UTP-labeled antisense riboprobe.24 hours post-fertilization (hpf) stage wild-type embryos were incubated with 0.003% 1-Phenyl-2-thiourea (PTU) (Millipore Sigma, MO, USA) to prevent pigments formation.Embryos were then fixed with 4% (V/V) paraformaldehyde (PFA)/ phosphate-buffered saline (PBS), followed by dehydration using increasing methanol series, then stored in 100% methanol at -20°C for at least overnight.Next day, samples were rehydrated with decreasing methanol series and permeabilized by 10 µg/mL proteinase K as described earlier 3 .The color development was performed using the BM purple alkaline phosphatase substrate (Millipore Sigma, MO, USA).
(Philomath, OR, USA).Approximately 8 nanogram of MO oligonucleotides were injected into 1-cell stage embryos as per guidelines for morpholino use in zebrafish 4 .

Cloning, site directed mutagenesis, synthesis of capped mRNA and microinjections
The full-length coding sequence of zebrafish tars2 was amplified from WT zebrafish cDNA.Full-length human TARS2 was PCR amplified from cDNA clone (# MHS6278-202827455, Horizon Discovery Ltd.) using Q5 high-fidelity DNA polymerase (New England Biolabs, Ipswich, MA, USA).The purified PCR products were cloned into the pCS2+ plasmid (a kind gift by Dr. David Turner lab, University of Michigan).Variants from affected individual were introduced in zebrafish tars2 cloned into pCS2+ at homologous positions using Quick Change XL site directed mutagenesis kit (Agilent Inc.), and primers listed in table S3.Wild-type tars2 and tars2 variant-specific plasmids were verified by sequencing, linearized, and then used as templates for synthesizing capped messenger RNA (mRNA) as described earlier 3 .One hundred fifty picogram of wild-type mRNA or variant-specific mRNA, together with tars2 MO was injected for the rescue experiments.The amount of mRNA in a mixture was adjusted to achieve the lowest number of dysmorphic embryos and highest efficiency.

Morphological phenotyping and imaging
Morphological phenotyping was carried out by randomly selecting animals from control MO, tars2 MO and tars2 MO rescue groups at 3 dpf and anesthetized in Tricaine/MS-222 (Sigma-Aldrich, MO, USA).Animals were oriented within an agarose cavity mold or 2% methylcellulose (Sigma Aldrich, MO, USA) under a stereomicroscope and imaged using either an Olympus SZX12 stereomicroscope fitted with an Olympus DP71 color digital camera (Olympus, Tokyo, Japan) or a high-definition camera Nikon DS-Fi2 mounted on a Nikon SMZ18 (Nikon, Japan) stereomicroscope.The head and eye sizes were measured from scale-calibrated images using ImageJ software (NIH).

Immunohistochemistry
Phalloidin staining was performed as described earlier 3 .Whole-mount embryo cell apoptosis was visualized using acridine orange (AO) as per methods described previously 5 .Briefly, 20 animals for each group at 72 hpf were incubated in 10 μg/mL AO (Invitrogen, Eugene, OR, USA) in E3 buffer for 1 hour, and then rinsed thoroughly in E3 buffer for three times.Next, animals were anaesthetized and mounted in 1.5% low melting point agarose and images were acquired with a Zeiss LSM-710 confocal microscope.

RNA extraction, reverse transcription and quantitative PCR (RT-qPCR)
Total RNA was extracted using the TRIzol Reagent (Thermo Fisher Scientific, CA, USA) and purified using RNA clean and concentrator-5 kit (Zymo Research, CA, USA) as per manufacturer's instructions.The cDNA was synthesized using iScript RT Supermix (Bio-Rad, USA), and used as a template for RT-qPCR with SYBR Green Supermix (Thermo Fisher Scientific, CA, USA) and the Light Cycler® 96 System (Roche, CA, USA).All RT-qPCR reactions were performed using biological (n = 10) and technical triplicates.The housekeeping gene 18S was used as a reference gene.The primer sequences for RT-qPCR are listed in table S3.The cycle threshold values (Ct) data were imported into Microsoft Excel for the relative gene expression analysis.Quantification was based on 2^(-ΔΔCT) method and using 1-hour post-fertilization (hpf) for tars2 temporal expression, skin for tars2 in tissue-specific expression, and the control MO-injected group for tars2 MO as well as mRNA-injected group for normalization.

Statistical analysis
Each experiment was performed three times independently.Data are presented as mean value ± standard deviation (SD).Statistical analysis was performed using GraphPad Prism version 9.3 (GraphPad Software, San Diego, CA, USA).In all analyses, the significance level was set to 0.05.The p value was determined by oneway analysis of variance (ANOVA) with Tukey's multiple comparisons test.

Analysis of amino-acid conservation of TARS2 residues impacted by diseaseassociated missense variations
The amino acid conservation of residues impacted by disease-associated missense variants is analyzed within a robust multiple sequence alignment of ThrRS protein sequences from archaea, bacteria and eukaryotes (See Figure S1).It shows that E440 is strictly conserved (100%), P155 and R684 are highly conserved except in archaea, and F323 is conserved in eukaryotes but variable in other species.In addition, R109 is mainly conserved (>80%) but only in eukaryotes; and the residue corresponding to L544 is preferentially non-polar (i.e.V/I/L/M/A/F).The other affected positions (R327, A337, E342, R346, R419, A714) are not conserved.
A similar variability of conservation of the affected positions is observed for the previously reported variants of TARS2, with residues strictly (D560, P613) or highly conserved (R452, except in rodents that actually display a W), conserved in only eukaryotes (P282), physico-chemically conserved (E or D for E425), or not conserved (T157, S258, E715).Noteworthy, the vast majority of the so-far reported pathologyrelated variants identified within mitochondrial aaRSs affect positions mainly not conserved (for updated statistics, see www.misynpat.org 6).Indeed, a strict conservation of residues is the hallmark of a strong selective pressure and would reflect a strong structural and/or functional role.

Analysis of the structural environment of the residues affected by missense variants
The human TARS2 is an homodimeric class II aaRS with the catalytic core made of three conserved motifs 7 .It possesses a catalytic and tRNA anticodon binding domain (Figure 3A, B), an editing domain to hydrolyze incorrectly charged amino acids 8 , and an N-terminal TGS domain (named after "ThrRS, GTPase and SpoT proteins") of unknown function.All disease-related TARS2 variants are positioned in a schematic representation of the modular organization of the enzyme (Figure 3C-G).The identification of 12 new missense variants (in red) greatly extends the pattern of reports with descriptions in all constitutive domains of the enzyme.
The investigation of the structural environment of the residues affected by the missense variants is made using known crystallographic structures (Staphylococcus aureus ThrRS, PDB: 1NYR 9 (Figure 3A) and one monomer of Escherichia coli ThrRS in complex with tRNA Thr , PDB: 1QF6 10 (Figure 3B) and a 3D model of the human TARS2 (downloaded from www.misynat.org).The correspondence of the residues from the three enzymes is given in Figure 3H.
As for the previously described D560H and R452W variants, E440K and R346C affect residues near the catalytic core (Figure 3C).R346 is located close to catalytic motif 1 and interacts with the backbone of a nearby helix.Its mutation may thus create local rearrangements.E440 belongs to motif 2 and interacts with the threonyl-adenylate, strongly suggesting a catalytic impact for the variant E440K.A functional impact is also anticipated for R684Q, since the corresponding conserved amino acid (R609) was shown in E. coli to directly interact with the tRNA anticodon nucleotide U36 (Figure 3E) 10 .As for P613, R327 is situated at the interface between the anticodon binding and the catalytic domains (Figure 3F).The replacement of a positively charged by a negatively charged residue (R327Q) might engender repulsive local changes and alter the inter-domain communication.Similarly, L544 is non-polar and surrounded by hydrophobic residues (Figure 3G).Its change into F (L544F) might engender a staking interaction with the aromatic ring of Y483, thus creating a local distortion.As for the editing domain, none of the reported variants impact residues identified to perform the editing function 8 (in orange in Figure 3D), P155 is however at close vicinity and its replacement by a leucine (P155L) might create local rearrangements.Finally, the other variants described in the present study impact residues that are located at the dimer interface (F323C, A337, E342D), in a loop that points to the solvent (R419W, E425G), or in the TGS domain (R109Q) (these are not shown).
Although all these observations must be experimentally confirmed, they serve the purpose of illustrating the diversity of the possible molecular impacts of the pathologyrelated missense variants of TARS2.

Zebrafish tars2 mRNA was strongly expressed in developing brain
Zebrafish tars2 gene has not been well-annotated in Ensembl (LO017852.1-201)   but predicted full length cDNA sequence can be found in NCBI database (Accession number: XM_021466791) as well as amino acid sequence (Accession number: XP_021322466.1).
To investigate the expression level of tars2 during zebrafish embryonic development, we performed RT-qPCR that showed tars2 transcript is expressed at about the same level starting from 1-hour post-fertilization (hpf) up to 120 hpf (Figure S3A).Analysis of tars2 mRNA expression in different adult tissues of zebrafish revealed a significant enrichment in brain, eye, heart, muscle, notochord and testis (Figure S3B).Zebrafish expression data is Our data is in agreement with the RNA expression data from the Genotype-Tissue Expression (GTEx) database, where TARS2 mRNA in humans is enriched in the cerebellum, testis and skeletal muscles.
We performed whole-mount in situ hybridization (WISH) to investigate the spatial and temporal expression of tars2 mRNA during zebrafish embryonic development.
WISH data show starting from 1 hpf onward maternally-supplied and ubiquitous expression of tars2 mRNA during early embryogenesis (Figure S3C, D).By 24 hpf, the expression is more prominent in the forebrain, the presumptive optic tectum (TeO), eye, and the trunk muscles (Figure S3E).At 72 hpf, the expression is abundant in the head and pectoral fin but decreased visibly in trunk muscles (Figure S3F, F').By 120 hpf, tars2 mRNA expression was further decreased throughout the embryo but showed brain and pectoral fin specificity with additional expression in the intestine (black arrowhead, Figure S3H, H').In conclusion, RT-qPCR, and WISH data show that tars2 mRNA was initially ubiquitously expressed but gradually became more prominent in the brain, eye, muscles, pectoral fin, and intestine.Our findings suggest distinct tissue requirements for Tars2 during development that strongly correlated with the pleiotropic symptoms observed in individuals carrying TARS2 variants.

Table S1
Characteristics of TARS2 variant.

Table S2
Genetic, clinical, and neuroradiological features of TARS2 deficiency individuals.

Table S3
List of oligos used in this study.

Figure S1. TARS multiple sequence alignment.
A multiple sequence alignment (MSA) of TARS sequences from 52 organisms, representative of bacterial, archaeal and eukaryotic phylogenetic diversity.Sequences were aligned using the MAFFT program.The MSA was generated inside the workbench ORDALIE (ORDered ALIgnment Information Explorer) and drawn in SnapGene software.The consensus sequence with a threshold of >70% is provided on the top of the alignment, and the sequence conservation is highlighted with gray bars.Amino acids are marked with color highlighting properties and conservation (following the ClustalX option embedded in SnapGene).The sequence nomenclature is: a T for the Threonine specificity of the aaRS, followed by four letters indicating the Bacterial (bact), Archaeal (arch), or Eukaryotic cellular location (cyto for cytosolic, and mito for mitochondria), then an underscore followed by two four-letter codes for the Genus and species separated by a dot (e.g.Tmito_Homo.sapi stands for human TARS2).Missense variants described in the present study are indicated in red on the top of the alignment, those previously reported (see Figure 1 and Table S2 for corresponding references) are indicated in black.

Figure S2. TARS2 c.387+6T>C leads to RNA transcripts lacking exons 3 and 4 containing a portion of the B-terminal editing domain. (A)
Lymphoblastoid cell lines isolated from the proband, father and mother were treated +/-puromycin (lanes 1 and 2), +/-cycloheximide (lanes 3 and 4), or nothing (lane 5) and tested in duplicate.PCR primers were designed to amplify exons encompassing those predicted to be spliced by in silico prediction tools.Cells from the proband and mother carrying the c.387+6T>C variant produced RNA transcripts that resulted in PCR products ~250bp smaller than the wild type 700bp PCR product regardless of treatment.This reduction in size is equal to the loss of TARS2 exons 3 and 4. The alternative splice product was not observed in the father nor in blood (bld), breast (brst), or colon (cln) samples, but was detected in brain (brn), suggesting it may be a tissue specific isoform.(B) CloneSeq on these same samples was performed as previously described.Relative quantification of CloneSeq results is shown as the percent spliced, confirming the presence of transcripts lacking TARS2 exons 3 and 4 in the proband and mother, but not the father.

Figure S4. RT-PCR analysis of tars2 splice-blocking morpholino (MO). (A)
Schematic representation of partial tars2 pre-mRNA.The MO was designed targeting exon 11/ intron11, as were primer binding sites used for RT-PCR.(B) mRNA from uninjected, control morphants and tars2 morphants was used as a template to amplify sequence.Both uninjected and ctrl MO samples showed a single band (size around 500 bp) as indicated by the black arrow.tars2 morphants showed bands of multiple different size (asterisk), indicating aberrant spliced mRNAs produced after MO injection.Quantification of the head size of ctrl MO, or tars2 morphants and morphants rescued with either human wild-type (WT) or different variants, as labeled.The size measurements were calculated as the percentage of the mean value of ctrl MO embryos.n = 30 embryos.Error bars = mean ± SD.The statistical significances were calculated by one-way ANOVA with Tukey's multiple comparisons test: ns, not significant p ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 were compared to tars2 MO group and displayed in blue text.ns, not significant p ≥ 0.05, # p < 0.05, ## p < 0.01, ### p < 0.001, and #### p < 0.0001 were compared to the uninjected control group and displayed in red text.

Figure S3 .
Figure S3.Detection of tars2 mRNA expression in zebrafish.(A) RT-qPCR of tars2 mRNA expression from 1 to 120 hpf in WT animals.The tars2 mRNA expression can be detected starting from 18 hpf and then increased throughout development.Data shown are mean ± SD and compared to 18 hpf.(B) RT-qPCR of tars2 mRNA expression in different adult tissues.Data shown are mean ± SD and compared to skin.(C-G') WISH was performed for detecting tars2 expression during zebrafish embryogenesis.Each embryonic stage is shown at the top right of the panels (hpf, hours post-fertilization). (C) tars2 expression measured by maternal mRNA.Dorsal view.(D) An 18 hpf embryo shown in lateral view, anterior to the left.(E) A 24 hpf embryo, flat mounted and shown in dorsal view (anterior to the left).fb: forebrain, mb: midbrain, hb: hindbrain, MHB midbrain and hindbrain boundary, TeO: optic tectum.A 72 hpf embryo shown in lateral view (F) and dorsal view (F') A 120 hpf embryo is shown in lateral view (G) and dorsal view (G').Arrowheads indicate the intestine.Scale bar = 250 µm.

Figure S6 .
Figure S6.Functional validation of human TARS2 variants in zebrafish.