Functional Characterization of a Novel Loss-of-Function Mutation of PRPS1 related to Early-Onset Progressive Nonsyndromic Hearing Loss in Koreans (DFNX1): Potential Implications on Future Therapeutic Intervention

Abstract

Background

The symptoms of phosphoribosyl pyrophosphate synthetase 1 (PRPS1) deficiency diseases have been reported to be alleviated by medication. Herein, we report biochemical data to favor PRPS1 deficiency-related hearing loss as a potential target for pharmaceutical treatment.

Methods

We recruited 42 probands from subjects under the age of 15 years with a moderate degree of nonsyndromic autosomal-recessive or sporadic sensorineural hearing loss (SNHL) in at least one side. Molecular genetic testing, including targeted exome sequencing (TES) of 129 deafness genes, and in silico prediction were performed.

Results

A strong candidate variant—p.A82P—of PRPS1 is co-segregated with SNHL in X-linked recessive inheritance from one Korean multiplex SNHL family. Subsequent measurement of in vitro enzymatic activities of PRPS1 from erythrocytes of affected and unaffected family members as well as unrelated normal controls has confirmed a pathogenic role of this variant. In detail, compared with normal hearing controls (0.23 – 0.26 nmol/ml/h), the proband, affected sibling, and their normal hearing mother demonstrated a significantly decreased PRPS1 enzymatic activity (0.07, 0.03, and 0.11 nmol/ml/h, respectively). This novel loss-of-function mutation of PRPS1—p.A82P—is the 9^th and 6^th most reported mutation in the world and in Asia, respectively.

Conclusions

DFNX1 turned out to account for about 2.4% (1/42) of moderate SNHL in a Korean pediatric population. Confirmation of PRPS1 activity deficiency and an audiologic phenotype that initially begins in a milder form of SNHL, as in our family, should indicate the necessity for rigorous genetic screening as early as possible.

Introduction

Mild-to-moderate sensorineural hearing loss (SNHL) is considerably attributed to genetic causes. In our previous study, genetic causes accounted for approximately 45.4% of all SNHL simplex cases [1]. Several genes, including OTOG, OTOGL, STRC, SERPINB6, GPR98, and PDZ7, are related to moderate SNHL [1,2,3,4]. In addition, some genes of severe to profound SNHL are also associated with moderate SNHL [5,6]. Several molecular genetic approaches, including phenotypic-driven genetic testing strategy [7], targeted exome sequencing (TES)[8], and whole exome sequencing (WES)[1], have been used to efficiently investigate rare but important causative mutations.

Because many moderate SNHL mutations show a delayed onset during childhood, it is highly probable to miss the early intervention period without suspicion. Identifying the molecular etiology of moderate SNHL is crucial for early intervention, allowing to minimize related complications and to alleviate its progression. Delayed intervention, on the other hand, can result in the neurological and intellectual developmental delays and can compromise cognitive functions [9].

The treatment of phosphoribosylpyrophosphate synthetase 1 (PRPS1) (MIM 311850)-related diseases using a metabolic compensation with S-adenosylmethionine (SAM) has been suggested recently [10]. The X-linked mutations of PRPS1 are associated with both syndromic and nonsyndromic (DFNX1) SNHL [11]. The audiologic phenotypes of PRPS1 mutations show a wide spectrum of moderate to profound, prelingual or postlingual, and progressive or non-progressive forms [12]. The SNHL can be manifested in all of the PRPS1-related disorders, including Arts syndrome, X-linked Charcot-Marie-Tooth disease-5 (CMTX5), DFNX1, and PRS1 superactivity with various degrees [13,14,15,16,17,18]. However, the association between audiologic phenotypes and PRS1 activities has not been well established thus far.

Recently, we identified X-linked missense mutation of PRPS1, using TES in a multiplex family. The frequency of DFNX1 in moderate nonsyndromic SNHL pediatric subjects with recessive or sporadic inheritances was estimated. Additionally, their audiologic phenotypes and PRS1 activities were analyzed and the association between the two has been discussed. Importantly, our data favors PRPS1 deficiency-related SNHL as the potential target of pharmaceutical treatment by showing a narrow margin between the enzymatic activity of the affected male siblings and the normal carrier mother. To the best of our knowledge, this is the first report that presents PRPS1-related nonsyndromic SNHL in a Korean family.

Material & method

Ethical considerations

This study was approved by the institutional review boards (IRBs) at Seoul National University Hospital (IRBY-H-0905-041-281) and Seoul National University Bundang Hospital (IRB-B-1007-105-402). Written informed consent was obtained from all participating subjects. For children, written informed consent was obtained from their parents or guardians.

Clinical evaluation

The 42 probands under the age of 15 years, with a moderate degree of nonsyndromic autosomal-recessive or sporadic SNHL in at least one side, and their families were examined for any history of hearing loss and other syndromic features at the otolaryngology clinics in the two institutions aforementioned between September 2010 and February 2016. Clinical characteristics included gender, date of birth, medical history, physical examination, pure-tone audiometry, auditory brainstem response (ABR), and speech evaluation. The presence of any phenotypic markers, indicating syndromic SNHL, was thoroughly investigated in the proband and related family members. Imaging evaluations were conducted using temporal bone computed tomography (TBCT) and internal auditory canal magnetic resonance imaging (IAC MRI) to identify inner ear anomalies, such as an enlarged vestibular aqueduct in subjects. All family members were investigated for the presence of any type of hearing loss. The pure-tone thresholds were recorded at 0.25, 0.5, 1, 2, 4, and 8 kHz. The hearing threshold was calculated by averaging the thresholds of 0.5, 1, 2, and 4 kHz, and was classified as subtle (16–25 dB), mild (26–40 dB), moderate (41–70 dB), severe (71–95 dB), or profound (> 95 dB). Based on the results of the retrieved clinical and audiologic evaluations, probands and their families without any phenotypic markers were screened for the presence of deafness mutation with phenotypic–driven genetic testing approaches [7].

DNA samples and phenotypic-driven genetic testing focused on the candidate genes

Genomic DNA was extracted from the peripheral blood using standard protocols (Gentra Puregene Blood Kit, Qiagen, Venlo, Limburg, Netherlands). Sanger sequencing of GJB2 was performed, as previously described [5,19]. Further genetic tests were conducted for Sanger sequencing of p.P240L of CDH23, which is the founder mutation of CDH23 in Koreans [20]. Next, the presence of causative mutations in other deafness genes was investigated by a targeted capture technique of exons and flanking sequences of 129 deafness genes, followed by massively parallel sequencing of DNA libraries, as previously described [21]. Variant detection was accomplished as previously described [21].

Investigation of pathogenicity of the PRPS1 variant

In silico predictions of the pathogenicity of PRPS1 variant were performed using PolyPhen-2 [22]. The segregation analysis was performed for the detected variations of PRPS1. The prevalence of PRPS1 variants among the 120 Korean normal-hearing control X chromosomes and 1244 Korean control alleles derived from the Korean Reference Genome Database (KRGDB) (http://152.99.75.168/KRGDB/menuPages/firstinfo.jsp) was estimated. The conservation of residues of the PRPS1 variants among different species was evaluated. The in vitro assay of phospho-ribosylpyrophosphate synthetase 1 (PRS1) enzymatic activity was performed in the proband (SH 178–396), sibling (SH 178–411), and their heterozygote carrier mother (SH 178–397) using patient-derived erythrocytes with the ELISA kit from the NOVOCIB Company (Lyon, France). Normal hearing unrelated controls of one male and a female were compared to the PRS1 enzyme activity. Each PRS1 activity assay was conducted three times to ensure reliability of the results.

Statistical Analysis

The difference of PRS1 activity between the groups of affected, unaffected, and carrier subjects were compared by using Mann-Whitney U test. P-values of less than 0.05 were considered to indicate significance. The results were statistically analyzed using SPSS ver. 21.0 (IBM, Armonk, NY, USA).

Results

Clinical characteristics

Proband from a boy (SH178–396) showed moderate-to-severe bilateral SNHL, which was first detected at age 3 years. At that time, his pure-tone threshold in the right and left ears was 80 and 65 dB HL, respectively (Fig. 1a). However, this was aggravated to 90 dB HL bilaterally over a 7-year period (Fig. 1a). Finally, he underwent right-sided cochlear implantation at age 12 years. He did not show any syndromic features, including peripheral or central neuropathy, mental retardation, and ataxia. He did not show uric acid abnormality or susceptibility to infections. The inner ear structure was intact in TBCT and IAC MRI.

Figure 1.

(a) Segregation study and pure tone audiograms of SH178, (b) PRS1 activities in the erythrocytes of affected subjects, carrier, and normal controls

His older brother (SH178–411) also suffered from moderate bilateral SNHL, which was first detected at age 7 years and gradually progressed (Fig. 1a). Syndromic features or inner ear anomaly was identified in this patient. All other family members did not have any history of hearing impairment or syndromic features.

Phenotypic –driven molecular genetic approaches

Because GJB2 mutation is the most prevalent causative deafness mutation in Koreans, we initially conducted GJB2 screening for autosomal recessive or sporadic cases, followed by CDH23 p.240L sequencing. In the aforementioned boy, SH178–396, there were no mutations in either deafness genes. Through TES, we identified a candidate variant of p.A82P (c.244G>C) of PRPS1, co-segregating with SNHL in an X-linked recessive inheritance manner (Figure 1a).

Validation of pathogenic potential of PRPS1 p.A82P variants

PolyPhen-2 analysis predicted that PRPS1 p.A82P was probably damaging with a score of 0.977 (sensitivity: 0.76; specificity; 0.96). Moreover, the p.A82 residue was well conserved from C.elegans to humans. There was no p.A82P variant in 120 Korean normal-hearing control X chromosomes and 1244 Korean control alleles derived from the KRGDB. While normal hearing controls showed PRS1 activities of 0.23–0.26, both the affected siblings showed a significantly lower PRS1 activity—0.07 in the younger brother (SH178–396) and 0.03 in the older brother (SH178–411) (Figure 1b) (P = 0.01, Mann-Whitney U test), confirming the pathogenic role of this variant in SNHL. Their carrier mother (SH178–397) showed a PRS1 activity of 0.11, which was only slightly higher than the two siblings (0.03 and 0.07) (P = 0.13, Mann-Whitney U test), but still markedly lower than the normal controls (0.23–0.26); it is worth noting that there was no statistical significance, probably due to a small number of samples (P = 0.07, Mann-Whitney U test).

Discussion

We identified a novel PRPS1 mutation through phenotypic-driven molecular genetic approaches. According to this study, DFNX1 turned out to account for about 2.4% (1/42) of moderate SNHL in a Korean pediatric population. Because the X-linked inheritance pattern could not be convinced from the beginning, prevalent autosomal recessive genes including GJB2 and CDH23 were initially investigated, followed by TES, identifying possible pathogenic variants of PRPS1. The pathogenicity of the selected variant of PRPS1 was examined by an in silico analysis of PolyPhen2 and the presence of variant in the normal control group. Moreover, the PRS1 activity assay provided solid evidence for the pathogenicity.

To date, 24 PRPS1 mutations have been identified [14]. It was presumed that the protein structural positions of PRPS1 mutations—related to the PRS1 activity—determine the severities of PRPS1-related disorders [14]. The active hexamer configuration of PRPS1 has a catalytic site that provides the binding sites for ATP and ribose 5-phosphate, and an allosteric site that regulates the enzyme activity [23,24]. PRPS1 mutations in the allosteric site hinder the feedback regulation of PRS1 enzyme activity, thereby causing PRS1 superactivity [25,26]. On the other hands, the destabilization of ATP-binding site result in PRPS1 deficiency-related disorders, encompassing the most severe forms of PRPS1-deficiency syndrome [15], Arts syndrome, intermediate form between CMTX5 and Arts syndrome [27,28], and CMTX5, in accordance with the residual PRS1 activities. DFNX1 mutations usually influence neither the allosteric site nor the ATP-binding site, but induce the local protein structural changes or effect the interface of trimers [27]. However, the structural loci of PRPS1 mutations is not always the determining factor of the types of PRPS1-deficiency disorders. Identical PRPS1 mutations were reported to cause nonsyndromic and syndromic PRPS1-deficiency disorders in a family [28,29].

The severity of PRPS1-related disorders from DFNX1 to recently reported most severe PRPS1-deficiency syndrome with prenatal growth restriction, retinal dystrophy, diabetes insipidus and white matter diseases was suggested to be related with PRS1 activities [14,15]. A few studies on DFNX1 demonstrated the significantly decreased PRS1 activity in the affected subjects compared with the unaffected or carrier subjects [11,28,29]. However, it is still elusive that audiologic phenotypes of PRPS1 mutations are also correlated with PRS1 activities. Our affected siblings—SH178–396 and SH178–411—showed severe and moderate SNHL, respectively. SH178–411 with moderate SNHL showed a lower PRS1 activity than SH178–396 with severe SNHL. Subtle discrepancy between the degree of SNHL and residual PRS1 activities in our affected siblings is not unprecedented [27]. A previous study also reported with similar results of intrafamilial phenotypic heterogeneities [29]. A previous study showed that even an identical PRPS1 mutation of p.Gly306Glu caused both nonsyndromic and syndromic features [29]. In females, this can be originated from the skewed inactivation of mutant X-chromosome [27,28]. However, in our affected male subjects, this discrepancy between the degree of SNHL and residual PRS1 activities cannot be explained by the skewed X-chromosome inactivation. Instead, intrafamilial variable PRPS1 expression can be mediated by tissue-specific modifier functions or by test errors [11]. For instance, microRNA-376 is known to regulate PRPS1 levels only in specific organs, including cerebral cortex, heart, and kidney [30]. The differences in the expressions of tissue-specific microRNA can lead to the variation of PRPS1 expression with an identical genetic configuration. In addition, compensatory metabolic pathways or paralogs of PRS, such as PRPS2 (NM_001039091.2), PRPS3 (NM_175886.2), PRPSAP1 (NM_002766.2), and PRPSAP2 (NM_002767.3), may compensate PRPS1 functions [10,29].

Our female heterozygote carrier (SH178–397) showed a normal hearing level and did not show any phenotypic markers. Her PRS1 activity was significantly decreased compared with that of normal controls, but it was not as low as that of the affected male siblings aforementioned (SH178–396 and SH178–411). Although most female carriers in families with PRPS1 mutations were noted to be asymptomatic [16], some studies reported that female carriers in families with Art syndrome were known to occasionally exhibit a delayed onset of hearing impairment combined with ataxia and neuropathy [13,27,28,31]. The tissue-specific PRS1 expressions due to the skewed X-chromosome inactivation and unrevealed modifiers could result in a relatively higher PRS1 activity of cochlear than that of erythrocytes in our carrier. It is likely that the residual PRS1 activity of our female carrier, albeit significantly reduced, is still enough to maintain normal hearing function. Because there have been only anecdotal reports on PRS1 activities of DFNX1 cases [11,28,29], the degree of deficiency of PRS1 activity leading to SNHL has not been documented to date. Previously, significant differences in PRS1 activities between syndromic patients and non-symptomatic carriers were reported [32,33]. However, the enzyme activity from non-syndromic SNHL, which is obligatorily related to a lesser degree of PRS1 deficiency, would provide a better clue to the minimal therapeutic level of PRS1. Indeed, our current study showed a much smaller difference of PRS1 activities between the affected nonsyndromic SNHL subjects and no symptomatic PRPS1 carriers and also provided an experimental value of PRS1 activity (0.11), at or above the normal level in which hearing is maintained. This level of PRS1 activity, which needs to prevent the symptoms of PRS1 deficiency including SNHL, can be useful in selecting the indication to compensate for PRS1 activity using SAM.

Our proband (SH178–396) demonstrated the progression of SNHL. Because his hearing impairments were associated with diminished PRS1 activity, which resulted in the depletion of nucleotides synthesis, we can presume that early SAM supplement prior to the progression of hearing impairments might attenuate SNHL. Therefore, early diagnosis of PRPS1-related hearing impairment is important to prevent and alleviate further progression of SNHL. Moreover, because DFNX1 is the mildest form of PRPS1-deficiency disorders and all of the PRPS1-deficiency disorders can accompany variable range of SNHL, hearing loss could function as an index symptom for PRPS1-deficiency syndromes, including Arts syndrome, CMTX5, and PRS1 superactivity. Identification of PRPS1 mutations in these syndromic SNHL subjects might be helpful for the early diagnosis and management of accompanying syndromic features.

Our proband (SH 178–396) underwent cochlear implantation and showed good hearing outcomes. It has been established that hearing loss in DFNX1 subjects is originated from the impairments in auditory hair cells [11]. To date, there have not been any reported cases of auditory neuropathy. Therefore, hearing rehabilitation with cochlear implantation may be favorable for DFNX1 subjects with intact auditory neural pathway.. However, the possible effects of PRPS1 on spiral ganglion or auditory neural systems should not be excluded because PRPS1 is also expressed in spiral ganglion cells [11]. Therefore, the pathogenic roles of PRPS1 in the inner ear needs further exploration in future studies with various types of PRPS1 mutations.

Because PRPS1 is a rare cause of genetic SNHL, studies pertaining to PRPS1 could be expensive and time-consuming. However, phenotypic-driven molecular genetic approaches efficiently ruled out the common deafness genes, and subsequent TES could identify a DFNX1 family with novel missense mutation of PRPS1 p.A82P. In this current study, the progression of SNHL was shown by a proband. This suggests the possibility for medical interventions in preventing and alleviating genetic SNHL. The differences of PRS1 activity among normal controls, normal hearing female carriers, and affected male subjects may suggest a correlation between PRS1 activity and audiologic phenotypes. Moreover, the small differences of PRS1 activity between the normal carrier mother and the affected siblings provide a good indication and target level for possible medical interventions to normalize PRS1 deficiency. Further studies on the expression and regulation of PRPS1 will allow genetic counseling and customized medical or surgical treatments possible for affected subjects.