The Association Between Mitochondrial tRNA^Glu Variants and Hearing Loss: A Case-Control Study

Introduction

Deafness is one of the most common human health problems, affecting one in 1000 newborns.¹ It is anticipated that the number of deaf people will be more than 28 million by the end of 2030 in the US.² To date, the etiology of hearing loss is not well understood, but increasing evidence has suggested that deafness can be caused by genetic and environmental factors.³ In fact, the genetic impact has been found more than 50% patients with hearing loss. To date, around 124 genes, as well as 1000 variants have been identified to be related to hearing loss (https://hereditaryhearingloss.org/).⁴ In addition to nuclear genes, mitochondrion is very important organelle whose primary role is to generate ATP via oxidative phosphorylation (OXPHOS). Moreover, mitochondria have their own genetic codes, named mtDNA, which is 16,569 bp in length.⁵ Variants in mtDNA are important causes of aminoglycoside-induced and non-syndromic hearing loss (AINSHL) in many families worldwide.^6–8 In addition to the well-known m.A1555G or m.C1494T mutation, mitochondrial tRNA (mt-tRNA) gene is another hot spot for pathogenic variants associated with deafness.⁹ Variants in mt-tRNA genes may alter tRNA structure and functions, including the processing of RNA precursors, modification of specific nucleotides, and maintenance of secondary and tertiary structures. Failure of mt-tRNA metabolism and protein synthesis caused by these variants may lead to mitochondrial dysfunction, which is involved in deafness.^10,11 More recently, several mt-tRNA variants had been reported to be related to hearing loss, including mt-tRNA^Ile A4317G;¹² mt-tRNA^Cys C5783T;¹³ mt-tRNA^Ser(UCN) T7505C,¹⁴ however, the pathophysiology of these variants remained poorly elucidated.

In the present study, we screened for the frequencies of mt-tRNA^Glu variants in a cohort of 180 children with NSHL and 100 controls using PCR and direct sequencing analyses. Phylogenetic conservation and mtDNA haplogroup analyses were performed to assess the pathogenicity of mt-tRNA variants. We also performed clinical, genetic, molecular, and biochemical assessments of a Chinese pedigree carrying the putative pathogenic mt-tRNA^Glu variant.

Materials and Methods

Patients

From January 2020 to January 2023, 180 deaf children including 100 males and 80 females, aged from 1 to 5 years, with an average age of 3.5 years were recruited from Hangzhou First People’s Hospital and Quzhou People’s Hospital. Furthermore, 100 healthy subjects including 60 males and 40 females, aged 3–8 years, with an average age of 6 years, were enrolled as controls. This study was reviewed by the ethics committees of Hangzhou First People’s Hospital (Approval No: 2020–285-01) and Quzhou People’s Hospital (Approval No: 2021–028). Informed consent was obtained by each individual (parents) enrolled in the study. All the procedures were performed with the principles of the Declaration of Helsinki.

Audiological Examinations

Age-appropriate audiological and neurotological examinations of hearing loss were performed as detailed previously, including pure-tone audiometry (PTA), auditory brainstem response, acoustic immittance measurement and distortion product otoacoustic emission.¹⁵ The PTA was performed in a sound-controlled room at frequencies ranging from 250 to 8000 Hz, as suggested in a recent study.¹⁶ The levels of hearing loss were divided into five grades according to a previous study: normal: <20 decibels (dB); mild: 20–40 dB; moderate: 41–70 dB; severe: 71–95 dB; and profound >95 dB.¹⁷

Mutational Analysis of Mt-tRNA^Glu Gene

To screen for mt-tRNA^Glu variants, we first performed PCR amplification of mt-tRNA^Glu in all participants enrolled in this study. The primer sequence for amplification of mt-tRNA^Glu was: forward-5’-GCA TAA TTA AAC TTT ACT TC-3’, reversed-5’-AGA ATA TTG AGG CGC CAT TG-3’. After amplification, the PCR products were purified and analyzed by direct Sanger sequencing. The sequence data were then compared with the revised Cambridge reference sequences (rCRS) to detect the mt-tRNA^Glu variants (GenBank accession number: NC_012920.1).¹⁸

Data Analysis

A total of 14 species were selected for conservation analysis, and the conservation index (CI) was calculated by comparing the human mtDNA with that of the other 13 vertebrates. A CI of ≥ 75% was considered functional significance.¹⁹

Molecular Characterization of One Chinese Family with Mt-tRNA^Glu A14692G Variant

We ascertained a Chinese pedigree (Figure 1) via Hangzhou First People’s Hospital. The entire mitochondrial genomes of matrilineal relatives (II-1, II-6, III-6, and IV-3) was PCR-amplified using 24 primers and sequenced. The mtDNA variants were detected by comparison with the rCRS (GenBank accession number: NC_012920.1).¹⁸ Furthermore, Phylotree (http://www.phylotree.org/) and East Asian phylogeny were used to determine the mtDNA haplogroup of this pedigree.^20,21

Figure 1 One Han Chinese family with NSHL, arrow indicated the proband.

Mutational Analysis of Common Deafness-Related Genes

To further explore the roles of nuclear genes variants to deafness expression, we conducted mutational screening of common deafness associated genes (GJB2, GJB3, GJB6, TRMU and SLC26A4) in matrilineal relatives (II-1, II-6, III-6 and IV-3), together with other healthy subjects from this family. The primers for GJB2 amplification were: forward-5’-TAT GAC ACT CCC CAG CAC AG-3’, and reverse-5’-GGG CAA TGC TTA AAC TGG C-3’.²² The primers for GJB3 were: forward-5’- GTC ACC TAT TCA TTC ATA CGA TGG-3’ and reverse-5’-TCA CTC AGC CCC TGT AGG AC-3’.²³ The primer sequences for amplification of the GJB6 were: forward-5’-CCT TAA AAT AAA GTT GGC TTC AG-3’, reverse-5’-GGA ACT TTC AGG TTG GTA TTG-3’.²⁴ The primer sequences for TRMU were: forward-5’- ACA GCG CAG AAG AAG AGC AGT-3’, reverse-5’- ACA ACG CCA CGA CGG ACG-3’.²⁵ The five primer sequences for SLC26A4 were as follows: forward-5’-CGT GTA GCA GCA GGA AGT AT-3’, and reverse-5’-TTA AAT AAA AAA GAC TGA CT-3’; forward-5’-TGG GGA AAA AGG ATG GTG GT-3’, and reverse-5’-CCA ACC CCT TCT TTA GCT GA-3’; forward-5’-GCA GGA TAG CTC AAG GAA TT-3’, and reverse-5’-TCA TCA GGG AAA GGA AAT AA-3’; forward-5’-TCT CCT TGA TGT CTT GCT TA-3’, and reverse-5’-CCC ATG TAT TTG CCC TGT TG-3’; and forward-5’-CTG GGC AAT AGA ATG AGA CT-3’, and reverse-5’-ATC TGT AGA AAG GTT GAA TA-3’.²⁶ After PCR amplification and direct Sanger sequencing, the data were compared with the wild-type sequences of GJB2, GJB3, GJB6, TRMU and SLC26A4 (GenBank accessible numbers: M86849, AF052692, NG_008323, AF448221 and NM_000441.1, respectively) to detect variants.

Qualification of mtDNA Copy Number

mtDNA copy number was assessed using real-time PCR in four patients with NSHL (II-1, II-6, III-6 and IV-3) carrying the m. A14692G and m.G7444A variants, as well as in four control subjects (II-2, II-5, III-1 and III-5). The primers used for amplification of mt-ND1 were as follows: forward: 5’-AAC ATA CCC ATG GCC AAC CT-3’, revised: 5’-AGC GAA GGG TTG TAG TAG CCC-3’. The primers for the amplification of β-globin were as follows: forward:5’-GAA GAG CCA AGG ACA GGT AC-3’; reversed: 5’-CAA CTT CAT CCA CGT TCA CC-3’. Real-time PCR was performed using a LightCycler^® 480II system (Roche Diagnostics GmbH, Mannheim Germany) and each measurement was repeated in triplicate.

Isolation of Polymononuclear Leukocytes (PMNs)

The PMNs from four subjects with NSHL (II-1, II-6, III-6 and IV-3) carrying the m. A14692G and m.G7444A variants, as well as in four control subjects (II-2, II-5, III-1 and III-5) without these variants were isolated according to our previous study.²⁷ PMNs were counted in a Neubauer chamber and cultured in Hanks’ buffered salt solution.

ATP Analysis

The cellular ATP levels in eight cells were measured by using the Cell Titer-Glo^® Luminescent Cell Viability Assay kit (Promega), according to the protocols provided by the manufacturer.²⁸

Analysis of ROS Production in Cells

The fluorometry was used to determine the ROS level. Approximately 2×10⁶ cells were first incubated with the fluorescent probe 2,7-dichlorodihydrofluorescein (DCFH) for about 30 min, subsequently, the cells were analyzed by fluorescence plate reader, as mentioned in our previous study.²⁹

Statistical Analysis

Statistical significance was evaluated by an independent Student’s t-test using SPSS software (version 22.0; SPSS, Inc., Chicago, IL, USA). p<0.05 was considered statistically significant.

Results

Screening for Deafness-Associated Mt-tRNA^Glu Variants

To determine the frequencies of the deafness-related mt-tRNA^Glu variants, we first performed PCR amplification of mt-tRNA^Glu gene in all participants (Figure 2), which indicated that the targeted PCR product was 1050-bp. Sanger-sequencing suggested that four variants were identified in the current study: m.T14709C, m.A14683G, m.A14692G, and m.A14693G (Table 1). Among these, the m.T14709C variant was found in two subjects with hearing loss (1.11%), the m.A14692G and m.A14683G variants were identified in one individual with deafness (0.56%), and the m.A14693G variant occurred in four patients with hearing loss (2.22%). These variants were not detected in the control groups.

Table 1 Molecular Characterization of Deafness-Associated Mt-tRNAGlu Mutations

Figure 2 PCR amplification of mt-tRNAGlu gene in subjects with NSHL, arrow indicated the PCR product, which was 1050-bp.

Molecular Features of Deafness-Associated Mt-tRNA^Glu Variants

Next, we performed phylogenetic conservation analysis of the variants identified in this study. For this purpose, a total of 14 species of mtDNA sequences were selected. As shown in Table 1 and Figure 3, we found that all variants were highly conserved (CI=100% for all).

Figure 3 Sequence alignment of mt-tRNAGlu gene from different species, arrows indicated the positions of 37, 54, 55 and 64, corresponding to the m.T14709C, m.A14693G, m.A14692G and m.A14683G variants.

Structurally, as shown in Figure 4, the m.T14709C variant was located at the anticodon stem of mt-tRNA^Glu gene (position 37), whereas the m.A14683G variant was localized at the TψC loop of mt-tRNA^Glu gene (position 64), which created a novel base-pairing (64A-50G) and may change the secondary structure of this mt-tRNA. Moreover, both m.A14692G and m.A14693G variants occurred in the TψC loop (positions 55 and 54). Nucleotide at positions 55 and 54 are often chemically modified and thus contribute to the structural and functional importance of mt-tRNA.^30,31 Therefore, it can be anticipated that the alteration of mt-tRNA structure by these variants may lead to the failure of mt-tRNA^Glu metabolism.

Figure 4 Cloverleaf structure of mt-tRNAGlu gene, arrows indicated the positions of m.T14709C, m.A14693G, m.A14692G and m.A14683G variants.

Clinical and Molecular Analysis of One Pedigree with NSHL

We ascertained a Chinese pedigree with maternally inherited NSHL at Hangzhou First People’s Hospital. Of the seven matrilineal relatives, four had suffered varying degrees of hearing loss. The proband (IV-3), aged 2 years, suffered from profound bilateral hearing loss (105 dB for the right ear and 116 dB for the left ear). As shown in Figure 5 and Table 2, subjects (II-1, II-6 and III-6) were also deaf. Intriguingly, none of the patients had a history of aminoglycoside use.

Table 2 Summary of Clinical and Molecular Data for Several Members of the Chinese Family

Figure 5 Audiological examination of matrilineal relatives of one pedigree with NSHL, (X) left ear; (O) right ear.

Screening for mtDNA Variants

Because of maternal transmission, which suggested that mtDNA dysfunction may be involved in the pathogenesis of NSHL. We conducted mutational screening of the entire mitochondrial genome of four matrilineal relatives (II-1, II-6, III-6 and IV-3). As shown in Table 3, PCR-Sanger sequencing revealed a set of genetic polymorphisms belonging to mitochondrial haplogroup B4c1.²¹ Of these, there were eight variants in the D-loop, two variants in 12S rRNA, three variants in 16S rRNA, and two variants in mt-tRNA genes (m.G7444A and m.A14692G), while the rest were mainly located in protein-coding genes. In addition, seven missense variants were identified: ND2 C5178A (Leu to Met), CO1/tRNA^Ser(UCN) G7444A (Ter to Lys), A6 A8701G (Thr to Ala) and A8860G (Thr to Ala), ND3 A10398G (Thr to Ala), Cytb C14766T (Thr to Ile) and A15326G (Thr to Ala). These missense variants and tRNA variants were further assessed by phylogenetic analysis, especially in bovine,³² mouse³³ and Xenopus laevis.³⁴ We found that except for the m.G7444A and m.A14692G variants (Figures 6 and 7), others were not well conserved, suggesting that these variants may be involved in deafness expression.

Table 3 mtDNA Variants in One Chinese Family with Hearing Impairment

Figure 6 Identification of m.A14692G and m.G7444A variants by direct sequencing analysis.

Figure 7 Location of deafness-associated variants in tRNASer(UCN) and adjacent CO1. Arrow indicated the m.G7444A variant in the precursor of this tRNA and adjacent sequence of CO1 from wild-type (WT) and mutant (MT).

Mutational Analysis of Nuclear Genes

Variants in GJB2, GJB3, GJB6, TRMU and SLC26A4 played active roles in NSHL.^35,36 To examine the contributions of these gene variants to NSHL, we analyzed the exons of these genes in matrilineal relatives of this pedigree (II-1, II-6, III-6, and IV-3), as well as in controls. However, we did not identify any variants of these genes, suggesting that nuclear modified genes may not play a role in hearing impairment.

Analysis of mtDNA Content

To test whether the m.A14692G and m.G7444A variants affected mitochondrial function, we examined the mtDNA content in four individuals with hearing loss and four control subjects using Real-time PCR. As can be seen in Figure 8A, deaf patients exhibited much lower mtDNA copy numbers than controls (p=0.0010).

Figure 8 Analysis of mtDNA copy number in matrilineal relatives of the pedigree with NSHL. (A) mtDNA copy number; (B) ATP content; (C) ROS qualifications.

ATP Decreased in Mutant Cells

We next measured the ATP levels in mutant and control cell lines, as shown in Figure 8B, cells with m.A14692G and m.G7444A variants showed markedly decreased in ATP levels as compared with controls (p<0.0001).

ROS Levels Increased in Cells with m.A14692G and m.G7444A Variants

We also examined the levels of ROS production in cell lines with and without m.A14692G and m.G7444A variants (Figure 8C), as compared with controls, we noticed that mutant cells exhibited much higher levels of ROS production (p=0.0001).

Discussion

In the present study, we examined the frequency of mt-tRNA^Glu variants in a cohort of 180 children with NSHL. Four potential pathogenic variants: m.T14709C, m.A14692G, m.A14693G and m.A14683G were identified using Sanger-Sequencing. These variants were further evaluated using the following criteria: (1) present in <1% of controls; (2) evolutionary conservation; and (3) potential structural and functional alterations. Of these, the m.T14709C variant was first reported in a patient with maternally inherited diabetes and deafness (MIDD).³⁷ Biochemical analysis revealed that the m.T14709C variant affected the Complexes I and IV activities,³⁸ furthermore, this variant decreased the steady-state level of tRNA^Glu and subsequently influenced the protein synthesis.³⁹ Thus, the m.T14709C was definitely pathogenic for NSHL.

Moreover, the m.A14693G variant occurred at the extremely conserved position of mt-tRNA^Glu (position 55), a position that was critical for tRNA chemical modification.³ Previously study indicated that the m.A14693G variant may modulate the clinical expression of deafness-associated m.A1555G mutation.⁴⁰ Besides, the m.A14683G variant created a new Watson-Crick base-paring (64A-50G).⁴¹ Intriguingly, m.T15965C variant which occurred at the same position as the TψC stem of mt-tRNA^Pro gene has been associated with Parkinson diseases.⁴² Therefore, the m.A14683G variant, which was similar to the m.T15965C variant, may also cause mitochondrial dysfunction that was responsible for NSHL.

Among the cases harboring the mt-tRNA^Glu variants, only one child with the m.A14692G variant had an obvious family history of NSHL. Interestingly, none of the members in this pedigree had a history of aminoglycoside use, notably, the age at onset of NSHL varied from 1 to 69 years, with an average of 38 years. Sequence analysis of the entire mitochondrial genome from matrilineal relatives suggested the presence of two interesting variants: mt-tRNA^Glu A14692G and CO1/tRNA^Ser(UCN) G7444A. At the molecular level, the m.A14692G variant affected a highly conserved uridine at position 55 in the TΨC loop of mt-tRNA^Glu. The uridine was modified to pseudouridine (Ψ55), which plays an important role in the structure and function of this mt-tRNA.^43,44 The destabilization of base-pairing (18A-Ψ55) caused by the m.A14692G variant perturbed the conformation and stability of mt-tRNA^Glu. An approximately ~65% decrease in the steady-state level of mt-tRNA^Glu was observed in the cells carrying this variant. Failure in mt-tRNA^Glu metabolism impaired mitochondrial translation, especially for polypeptides with a high proportion of glutamic acid codons such as MT-ND1, MT-ND6 and MT-CO2 in mutant cells.⁴⁵ Thus, the m.A14692G variant was definitely pathogenic as it caused mitochondrial dysfunction.

In addition, the m.G7444A variant resulted in a read-through of the stop codon AGA of the CO1 message, thereby adding three amino acids (Lys-Gln-Lys) to the C-terminal of the polypeptide.⁴⁶ Thus, the variant polypeptide may retain a partial function (Figure 7). Alternatively, the m.G7444A variant was adjacent to the site of 3’ end endonucleolytic processing of the L-strand RNA precursor, spanning mt-tRNA^Ser(UCN) and ND6 mRNA.⁴⁷ Furthermore, the homoplasmic m.A7445G variant was reported to reduce tRNA^Ser(UCN) levels by ~70% and to cause a 45% reduction in mitochondrial protein synthesis in cybrid cells containing this variant.⁴⁷ Therefore, the m.G7444A variant, which was similar to the m.A7445G variant, may have an impact on mitochondrial function. Our results indicated that patients with both the m.A14692G and m.G7444A variants had lower mtDNA copy numbers, ATP production than controls, whereas the levels of ROS increased significantly. In fact, mtDNA copy number was a relative measure of the cellular number or mass of mitochondria.⁴⁸ Recent experimental study suggested that alterations in mtDNA copy number played a fundamental role in the increase in ROS, maintenance of mtDNA copy number was essential for the preservation of mitochondrial function and cell growth.⁴⁹ Moreover, the increased ROS would affect the quality of gametes, lead to the cochlear cell death and apoptosis,^50,51 and in turn, impaired OXPHOS and decreased the ATP production. Therefore, the m.A14692G and m.G7444A variants led to mitochondrial dysfunctions that were involved in NSHL. However, the absent of any functional variants in GJB2, GJB3, GJB6, TRMU and SLC26A4 suggested that nuclear genes may not play important roles in deafness expression. Hence, the combination of m.A14692G and m.G7444A variants may account for high penetrance and expressivity of NSHL in this pedigree.

In conclusion, our study indicated that mt-tRNA^Glu variants were important risk factors for NSHL, m.T14709C, m.A14692G, m.A14693G and m.A14683G were associated with NSHL. Screening for mt-tRNA^Glu variants was recommended for early diagnosis and prevention of NSHL. However, the homoplasmic forms of mtDNA variants suggested that the mutation itself was not sufficient to produce the clinical phenotypes, hence, other modified factors such as environmental factors, epigenetic modifications and personal lifestyles contributed to deafness expression.

Limitations

The main limitations of the current study were the relatively small sample size, and further studies including more deaf patients and controls are needed to verify these conclusions.

Conclusions

In summary, our data indicated that variants in the mt-tRNA^Glu gene were important contributors to NSHL, and screening for variants in this gene is recommended for the early diagnosis and detection of NSHL.