Prevalence of the 35delg Mutation in the Gjb2 Gene in Two Samples of Non-syndromic Deaf Subjects from Chile

Hearing loss is the most common inherited sensorial defi ciency in humans; about 1 in 1000 children suff er from severe or profound hearing loss at birth. Mutations in the GJB2 gene are the most common cause of prelingual, non-syndromic autosomal recessive deafness in many populations; the c.35delG mutation is the most common in Caucasian populations. The frequency of the c.35delG mutation was estimated in two samples of deaf patients from Santiago, Chile. Unrelated non-syndromic sensorioneural deaf patients were examined: Group 1 consisted of 47 unrelated individuals with neurosensory deafness referred to the Chilean Cochlear Implant Program; Group 2 included 66 school children with prelingual deafness attending special education institutions for deaf people. Individuals with profound to moderate isolated neurosensory hearing loss with unknown etiology were included. The presence of the c.35delG mutation was evaluated by the allele-specifi c polymerase chain reaction method (PCR), and in some cases it was confi rmed by direct DNA sequencing of the coding region of the GJB2 gene. Deaf relatives were present in 20.3% of the cases. We found 19.5% (22/113) patients with the c.35delG mutation, 6 of them homozygous; these rates are similar to frequencies found in other Latin American countries.


INTRODUCTION
Hearing loss is the most common inherited sensory defi ciency in humans; approximately 1 in 1000 children suff er from severe or profound bilateral permanent hearing loss at birth or during early childhood (Morton, 1991).A congenital hypoacusis rate of 2.8 per 1000 newborns has been reported in Chile (Nazar et al., 2009).In developed countries, around 60% of the cases of childhood deafness have a genetic origin and 70% of them are non-syndromic forms (Petit et al., 2001;Marazita et al., 1993).The most common pattern of inheritance among hereditary non-syndromic deafness is autosomal recessive, about 80% (Zelante et al., 1997).To date, numerous loci (about 60) have been involved in this kind of deafness (http:// hereditaryhearingloss.org2013); however, the mutations in the GJB2 gene (encoding the gap-junction protein connexin-26 in the DFNB1 locus in 13q12) are the major cause of prelingual, non-syndromic autosomal recessive deafness in many populations (Denoyellle et al., 1999;Kenneson et al., 2002;Álvarez et al., 2005;Gravina et al., 2,010).About one hundred GJB2 mutations have been associated with autosomal recessive non-syndromic deafness (http://davinci.crg.es/deafness2013).Different mutations are found among populations from diff erent ethnic backgrounds.One of these, c.35delG, consists of a deletion of one guanine in a sequence of six guanines, leading to a frame shift and premature stop of the synthesis of connexin-26; this mutation has a high prevalence in Caucasian populations (Petit et al., 2001;Denoyelle et al., 1999), accounting for 70% of all pathogenic GJB2 alleles (Denoyelle et al., 1999: Cryns et al., 2004).Analyses of the GJB2 gene in patients with autosomal recessive inherited deafness, especially due to the c.35delG mutation, have shown that 10% to 50% of them presented only one mutant allele (Kenneson et al., 2002).Studies of these heterozygous patients for GJB2 mutations have found other mutations in the GJB6 gene, which encodes connexin-30 (also located at the DFNB1 locus, 35 kb from the GJB2 gene); in these cases there is a digenic origin of deafness (Del Castillo et al., 2005, Cordeiro-Silva et al., 2011).
Among the Italian and Spanish populations a carrier frequency of the c.35delG mutation of 3.2% has been found (Estivill et al., 1998); however, this mutation has not been found in Asian populations (Kudo et al., 2000), where the c.235delC is the most prevalent (Abe et al., 2000;Park et al., 1999).No individuals with the c.35delG mutation were found among 365 students with profound sensorineural hearing loss in Ghana, Africa (Brobby et al., 1998).
Some studies have been conducted in Latin American populations: e.g. two studies in Brazil reported c.35delG frequencies of 12.4% and 11.8%, respectively among nonsyndromic deaf patients (Batissoco et al., 2009;Cordeiro-Silva et al., 2011), while the frequency of c.35delG carriers in the Brazilian general population was 0.94% (Nivolini et al., 2010).In Venezuela, the c.35delG was found in 27.5% of deaf children and the carrier frequency was 4% (Utrera et al., 2007).Studies in deaf children from Argentina reported that c.35delG, the most frequent mutation, was present in 26.5% and 24.0 % of the cases (Gravina et al., 2010;Dalamón et al., 2005).
There are two local preliminary publications related to GJB2 mutations in the Chilean population: Arancibia et al. (2012) found the c.35delG mutation in 11.25% of deaf school children and López et al. (2012) found two patients heterozygous for the c.35delG mutation among 8 patients with congenital nonsyndromic sensorineural deafness.The current Chilean mixed population is of biracial origin, it was generated by the admixture between aboriginal populations (Amerindians of Asian origin) and Spanish conquerors of European origin that arrived in the country in the latter part of the sixteenth century (Valenzuela et al., 1987;Cifuentes et al., 2004).The aim of this study was to estimate the frequency of c.35delG in a sample of deaf patients from Santiago, Chile, and compare the fi ndings to other populations with similar ethnic background.

Patients
We examined two groups of unrelated non-syndromic sensorineural deaf patients from Chile: Group 1 consisted of 47 unrelated individuals with neurosensory deafness referred to the Chilean Cochlear Implant Program at the Otolaryngology Service of the Barros Luco Hospital in Santiago, arriving from all over the country with ages from 2 to 39 years.Group 2 included 66 school children with prelingual hearing loss attending three special education institutions for deaf people in Santiago, the capital city of Chile (Jorge Otte School, Anne Sullivan School, and San Francisco School); the children ages ranged from 6 to 22 years at the moment of the study (Arancibia et al., 2012).The inclusion criterion for the study was profound to moderate isolated neurosensory hearing loss with unknown etiology; thus individuals with syndromic deafness or secondary acquired deafness (caused by trauma, ototoxic drugs, prematurity or infectious diseases) were excluded.Clinical information was obtained from the individuals' medical and school records, as well as a questionnaire answered by their parents in case of children.We excluded those cases with incomplete clinical records.The protocol followed the Declaration of Helsinki guidelines and was approved by the Ethics Board of the Universidad de Chile School of Medicine; an appropriate informed consent was obtained from all participants or their parents in the case of minors.

Molecular genetic analysis
Genomic DNA was extracted by a salting out protocol (Miller et al., 1998) from peripheral blood leukocytes in Group 1 and from buccal cells collected on buccal swabs in Group 2.
After DNA isolation, all samples were evaluated for the presence of the c.35delG mutation using the allelespecific polymerase chain reaction (PCR).Briefly, two PCR amplifi cations of genomic DNA of each sample were performed, one to identify the normal allele and the other to identify the c.35delG mutation.Specifi c primers for the normal or the c.35delG mutation and the common primer were used (Scott et al., 1998).
All cases from group 2 harboring the c.35delG mutation and a random sample of 15 cases with wild genotype were further analyzed by direct sequencing of exon 2 of the GJB2 gene.Exon 2 of the connexin 26 gene was sequenced after DNA amplification with a set of external primers (5' TGCTTACCCAGACTCAGAGAA 3' -5' CGACTGAGCCTTGACAGCTGA 3') and a set of internal primers (5' AGTGGCCATGCACGTGGCCTA 3' -5' TGATCTCCTCGATGTCCTTAA 3') according to Wu et al., 2003.Chromatograms obtained after sequencing were analyzed according the information available in the NCBI Gene Bank (Access number: M86849)

Statistical analysis
The 35delG frequencies were compared by a Chi square test.The statistical signifi cance level was 0.05.

RESULTS
A total of 113 patients (90 sporadic and 23 familial cases) with idiopathic non-syndromic sensorineural hearing loss were genotyped for the c.35delG mutation at the GJB2 gene.Among these, 46 were males and 67 were females.Profound deafness was more frequent (84 patients) than severe (16 patients) and moderate cases (3 patients).Table I presents the results obtained in the c.35delG genotyping, according to familial records of deafness.We found 22 patients (19.5%) who carried the c.35delG mutation (95% CI: 12.2 -26.8%); six were homozygous for the mutation; accordingly, we could establish the etiology of deafness in these 6 cases.The c.35delG mutation was more frequent in familial cases of deafness (65%) than sporadic cases (7.8%), p = 0.00025.The frequency of the c.35delG pathogenic allele in the sample was 12.4% (95% CI: 10.2% -14.6%).In addition we found the non-pathogenic variant V27I in 7 deaf children.

DISCUSSION
As in many other countries, our study demonstrated that the c.35delG mutation is a frequent cause of hearing impairment in Chile.Also, as is expected in a hereditary disease, this mutation was more frequent in familial than sporadic cases of deafness.
We could explain the etiology of deafness in only 6 cases that were homozygotes for the 35del G mutation.The presence of a single pathogenic allele does not explain the cause of deafness, therefore other mutations in connexin 26 or in other related genes may be associated with the deafness in the 16 patients that carried only one 35delG allele.This situation is not uncommon; in fact, several studies in patients with autosomal recessive deafness have shown that about 10% to 50% of them presented only one mutated allele in the GJB2 gene (Kenneson et al., 2002).The fi nding that there is an important number of non-syndromic sensorineural deaf persons carrying a single GJB2 mutation led to a search for other mutations in or near GJB2.Del Castillo (2002Castillo ( , 2005) ) identifi ed two large deletions in the connexin 30 gene: GJB6-D13S1830 and GJB6-D13S1854; these deletions are usually found accompanying the GJB2 coding mutation, causing the deafness.
The deletion del(GJB6-D13S1830) was the accompanying mutation in about 50% of Spanish patients with only one GJB2 coding mutation (Del Castillo, 2005).Latin American studies have found the D13S1830 deletion in 11% to 37% of patients with non-syndromic sensorineural deafness that carried only one copy of the c.35delG mutation (Batissoco et al., 2009;Cordeiro-Silva et al., 2011;Dalamón et al., 2005;Gravina et al., 2007;Utrera et al., 2007).López et al. (2012) did not identify GJB6 deletions in the 8 patients they studied.The D13S1830 deletion accounted for the deafness in 15.9% of the patients with only one GJB2 mutation in a study performed by Pandya (2003) in the USA.Even if the whole coding region of the GJB2 gene and the known mutations at the GJB6 genes were studied, there still would be cases with a monoallelic pathogenic mutation that is not enough to explain the deafness.
Our patients with only one 35delG mutation could be compound heterozygous with another mutation in the GJB2 gene, either in the coding region or outside of it in the promoter or splicing recognition site.Moreover, some of these patients with only one 35delG mutation could have another mutation at the GJB6 gene (digenic inheritance).
The genetic heterogeneity of non-syndromic hearing loss makes its molecular diagnosis diffi cult, given the number of mutations that have been described in deafness-related genes.Additionally, we have to consider the diffi culties in distinguishing between genetic and non-genetic deafness in families presenting a single deaf subject (90 cases in our sample -Table I), notwithstanding the eff orts to fi nd possible environmental causes of deafness in their clinical records.
In this study, as in many others, we found no explanation for the deafness in several heterozygous patients.Some studies regarding mutations in the GJB2 gene have been performed in other mixed Latin-American countries with a Caucasian ethnic component (comparable to the Chilean population).These studies had the same inclusion and exclusion criteria (syndromic hearing loss, probable environmental cause of deafness, transmission deafness) to select the patients to be analyzed.Table II compares their results with our study: we see that the frequency of the c.35delG mutation (in biallelic or monoallelic situations) among non-syndromic deaf patients from Santiago, Chile is in line with the frequencies found in the majority of other Latin American countries.
Comparing the c.35delG frequency among these diff erent studies, we fi nd that three of them are diff erent from the rest (p = 0.043): a higher frequency in Argentina (Dalamón et al., 2010) and a lower frequency in Mexico (Arenas-Sordo et al., 2012) and Brazil (Cordeiro-Silva et al., 2011); the remaining studies, including our study in Chile, found similar results.The proportion of homozygotes for the c.35delG mutation among those carrying this allele was similar in all the studies cited in Table II (p = 0.23).
We can conclude that the c.35delG mutation plays an important role in the etiology of hearing loss in Chile, as in other countries with similar genetic backgrounds; however, there is need of further research to fi nd other pathogenic alleles involved in genetic hearing loss in our country.

AKNOWLEDGEMENTS
We thank the Sociedad Chilena de Otorrinolaringología, Medicina y Cirugía de Cabeza y Cuello of Chile for fi nancial support.We also thank all the persons who consented to participate in this study and Dr. Daisy Pezoa for the ascertainment of patients from the Chilean Cochlear Implant Program.

TABLE II
Studies performed to detect the c.35delG mutation in non-syndromic sensorineural deaf patients from different countries from Latin America