- Email: [email protected]
Minimally invasive genetic screen for GJB2 related deafness using dried blood spots
International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Minimally invasive genetic screen for GJB2 related deafness using dried blood spots§ Attila L. Nagy a,*, Ro´bert Csa´ki b, Jo´zsef Klem c, La´szlo´ Rovo´ a, Ferenc To´th a, Gyula Ta´losi d, Jo´zsef Jo´ri a, Korne´l Kova´cs c, Jo´zsef Ge´za Kiss a a University of Szeged, Albert Szent-Gyo¨rgyi Pharmaceutical and Medical Centre, Department of Oto-rhino-laryngology and Head and Neck Surgery, 6725, Tisza Lajos krt. 111, Szeged, Hungary b Alfa-Biosoft Ltd., 6720, Na´dor u. 10, Szeged, Hungary c University of Szeged, Faculty of Sciences, Institute of Biotechnology, 6726, Ko¨ze´p fasor 52, Szeged, Hungary d University of Szeged, Albert Szent-Gyo¨rgyi Pharmaceutical and Medical Centre, Department of Pediatrics, 6720, Kora´nyi fasor 14-15, Szeged, Hungary
A R T I C L E I N F O
A B S T R A C T
Article history: Received 28 May 2009 Received in revised form 18 October 2009 Accepted 21 October 2009 Available online 24 November 2009
Objective: Nonsyndromic hearing loss is one of the most abundant human sensory disorders, and can be found in 1 out of 1000 newborns. In 60–70% of the cases this disorder is hereditary. The phenotype varies from moderate hearing loss to almost complete deafness, often only revealed in late childhood. Early detection of hearing related genetic variations in the first few weeks of life would allow planning of the audiological and logopedical procedures to maintain the children’s normal audiological and speech development, and if required a cochlear implantation can be planned in time. We wanted to evaluate, whether the blood samples collected from neonates onto Guthrie cards (dried blood spots, or DBS), and blood collected from people of various ages into blood collecting tubes is equally usable for genetic testing. The quality of the samples on DBS’s for genetic tests after an extended period of storage was evaluated. The methods for sample preparation and analysis were also evaluated. Methods: Two DNA extraction methods were compared on the samples. We extracted DNA from whole blood with the Versagene Blood Kit from Gentra, and from DBS’s with boiling. Allele-specific PCRs (ASPCR) were carried out on each sample. Samples were analyzed with AS-PCR and sequencing, for the 35delG mutation in the GJB2 (Cx26) gene. Freshly drawn and dried blood spot samples stored for several years were used in the experiments. Results: An AS-PCR method for detecting 35delG mutation on DNA extracted from Guthrie cards was validated. Blood samples up to 10 years of storage were applicable in the screen. 84 patients were found with 35delG mutations, both heterozygous (with no detected hearing related phenotypical discrepancies), and homozygous (phenotipically with moderate to severe hearing loss) forms. Conclusions: The dried blood spots on Guthrie cards require only three drops of blood to be collected from children, which causes less stress than taking 3 ml of blood. The blood stored on Guthrie cards can be used to store DNA samples for at least 10 years. Even under suboptimal storage conditions the samples’ DNA remains intact for genetic testing. Compared to blood collection tubes Guthrie cards cost less, are easier to transport and store. ß 2009 Elsevier Ireland Ltd. All rights reserved.
Keywords: Dried blood spot Polymerase chain reaction Whole blood Nonsyndromic deafness 35delG Screening
1. Introduction According to current knowledge physiological processes related to hearing are regulated by at least 100–150 genes. To date 146
§ The work described here was supported by the grant GVOP-3.1.1-2004-050498/3.0. * Corresponding author. Tel.: +36 62 54 58 50; fax: +36 62 54 58 48. E-mail addresses: [email protected] (A.L. Nagy), [email protected] (R. Csa´ki), [email protected] (L. Rovo´), [email protected] (F. To´th), [email protected] (J. Jo´ri), [email protected] (J.G. Kiss).
0165-5876/$ – see front matter ß 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2009.10.021
alleles of 42 genes have been identified (57 dominant, 77 recessive, 8 X-linked, 1 Y-linked, 2 modifier, 1 auditory neuropathy) [1]. The large number of genes and loci complicate the genetic analysis of nonsydromic hearing losses. DFNB1 was the first identified locus; its autosomal recessive mutation causes nonsyndromic hearing loss [2]. Using co-segregation analysis, this locus was mapped to 13q12-13 [3]. In 1994, Chaib et al. described the first dominant nonsyndromic hearing loss of genetic origin to 13q12-13 [4]. Mutation in the coding region of Gap Junction protein Beta 2 (GJB2) – also called Connexin26 (Cx26) – was the first to be linked to nonsyndromic hearing loss of genetic origin [5]. Connexin 26 (Cx26) is a protein, which belongs to the family of connexins. In the
76
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
cochlea [6,7] connexins facilitate the flux, and the recycling of K+ ions from the intracellular space to the endolympha [8]. The length of the functional, protein coding exon of GJB2 is 681 bp, and it codes a 226 amino acid polypeptide. 35delG causes a frameshift mutation, the deletion of one guanine residue in a stretch of six guanines in the coding region of the gene at position 35, resulting in a nonsense mutation at the 13th codon. 35delG mutation of the GJB2 gene accounts for about 7–15% of nonsyndromic hearing losses of genetic origin in the European population [9–11], and regarding this mutation, about 1–5% of the European population is a carrier [12]. Since the 1950’s [13] and 1960’s [14] national neonatal screening programs begun to operate in the advanced countries; they mostly perform screening for metabolic diseases [15,16]. These programs generated an enormous amount of dried blood spots (DBS) on so called Guthrie cards. These cards have the advantage of easy transportation, and their storage conditions are much cheaper than any other way of storing blood samples. The question arises whether they can be used as a basis of genetic screening, and if yes, to what extent. As national hearing screening programs gain more and more attention, and financial funds, they become more and more ‘‘sample-hungry’’. If these cards could be utilized to give proper results, they could serve as a complementary tool to the existing auditory and genetic screening methods. These blood spots are available for most of the newborns in those countries, and each individual can be tested in their absence. The procedure is also less traumatic to children. 2. Materials and methods All participants involved in the trials were informed according to the University’s Ethical Committee’s Guidelines, and all have signed a written consent. Double blind tests were performed on the spot blood samples taken for the routine, population wide metabolic screening tests of neonates. The samples were taken on Guthrie cards at the 3rd-4th life day of newborns at any neonatal ward where the neonates were cared in the eastern part of Hungary and sent via conventional mail. The test cards were stored at room air at the Department of Paediatrics, University of Szeged. The Guthrie cards were selected as follows: 48 pieces from the years 1996, and 1997, and further 96 pieces from the years 1999, 2000, 2001, 2004, 2005, respectively. The total number of DBS’s was n = 576, as these were selected randomly, we could use them as a ‘‘generic population sample’’. The other population we examined consists of 318 of patients. These patients were Cochlear Implant (CI) users, their relatives, CI candidates, their relatives, and a few individual patients from the ENT Clinic. CI users and CI candidates were selected based on two criteria. They were selected if there was a family history of hearing losses, and there were no organic abnormalities (anatomical variations, or developmental problems) or other diseases that are known to cause hearing loss or deafness in the patient’s history. In case of the CI users (n = 20 + 32 = 52) and CI candidates (n = 56) the average hearing threshold level was bellow 70 dB and speech
recognition performance was under 25%. We excluded those patients and CI users whose patient history contained some form of disease that can cause deafness, or who suffered a head trauma, or head injury, that can account for their hearing problems. The CI users’ and CI candidates’ family members (n = 163) have various levels of hearing loss, from no hearing loss at all to severe to moderate levels. There were some individuals who wanted to participate in our study with various levels but with unknown origin of hearing loss (n = 47). 20 CI users (out of n = 52) were individuals with no screened relatives. Our group of control persons consisted of people with hearing threshold levels at 5 dB or better on both ears (n = 20) and no family history of hearing losses. 2.1. Preparation of DNA extracts from DBS for AS-PCR 4 mm diameter pieces from the bloodspot test cards were punched out and put into PCR tubes along with 200 ml 1 PCR buffer (Eppendorf HotMaster taq). The DNA was extracted from the DBS by 20 min of incubation at 96 8C in a thermalcycler. The samples were then centrifuged at 16 000 g for 2 min. The supernatant was then transferred into a sterile microcentrifuge tube, and stored at 4 8C until utilization [17,18]. 2.2. Preparation of DNA from venous blood for AS-PCR We collected 3 ml blood from patients having Cochlear Implant, their relatives, and CI candidates. Blood anti-coagulant was EDTA. Genomic DNA (gDNA) was purified from 400 ml of blood using Versagene Blood Kit (Gentra) according to the manufacturer instructions. The concentration of the DNA was measured with spectrophotometer and was calculated by the adsorption at 260 and 360 nm. 2.3. Polymerase chain reactions DNA integrity test reactions were carried out using HotMaster Taq DNA Polymerase (Eppendorf). 30 ml final volume of the reaction mix contained 3 ml (10) HotMasterTaq buffer (Eppendorf), 2.5 ml, dNTP (2.5 mM), 0.5–0.5 ml (15 pM) DF2F-DF2R primer pair (Table 1); 1 U HotMasterTaq (Eppendorf); 6 ml gDNS template and 16.5 ml water. PCR program was as follows: 96 8C 2 min, for 35 cycles (96 8C 30 s–61 8C 30 s–68 8C 35 s), and after these 35 cycles 96 8C for 5 min. Negative control experiment was also done with paper discs originating from the ‘‘blood-free’’ parts of the Guthrie papers. PCR fragments were analyzed by agarose gel electrophoresis on 1.5% agarose gel (AbGen) with 1X TBE buffer. The internal control used in all the AS-PCR experiments were the primer pair ICF and ICR. They amplificate a part of the serine proteinase inhibitor gene (Table 1). AS-PCR reactions: 30 ml final volume of the reaction mix contained: 3 ml (10) HotMasterTaq buffer (Eppendorf), 2.5 ml, dNTP (2.5 mM), 0.5–0.5 ml primers GJC-GJW pair for wild allele detection and GJC-GJM for 35delG mutant allele detection (Table 1). (15 pM), 0.4–0.4 ml (15 pM) internal control primer
Table 1 The primers used in this work and their sequences. Primer name
Sequence
Description
ICF ICR GJC GJW GJM DF2F DF2R
CCC ACC TTC CCC TCT CTC CAG GCA AAT GGG GGG CCT CAG TCC CAA CAT GGC TAA GAG GTG AGT GAT CGT AGC ACA CGT TCT TGC A GCA CGC TGC AGA CGA TCC TGG GGA G CAC GCT GCA GAC GAT CCT GGG GAT TCT CCC TGT TCT GTC CTA GC TTT CCC AAG GCC TCT TCC AC
Internal control (serine proteinase inhibitor gene) Internal control (serine proteinase inhibitor gene) Common reverse primer for GJB2 Primer for 35delG Wild allele detection Primer for 35delG mutant allele detection GJB2 exon and flanking region for sequencing GJB2 exon and flanking region for sequencing
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
pair (ICF-ICR); 1 U HotMasterTaq (Eppendorf); 6 ml gDNS template and 16.5 ml water. The PCR program was as follows: first denaturation step at 95 8C for 5 min, then 35 cycles at 96 8C 30 s, at 65 8C 35 s, and at 68 8C 38 s, and as the last step, 10 min incubation time at 68 8C. PCR fragments were analyzed by agarose gel electrophoresis on 1.5% agarose gel (AbGen) with 1X TBE buffer. 2.4. DNA Sequencing and sequence analysis PCR fragments were generated as described in section 2.3. PCR products were desalted on Microcon columns (Millipore). The purified PCR products were eluted in 30 ml of water. The DNA sequences were determined by automated sequencing at Macrogen Inc. (Korea) on both strands. The sequences were aligned to the wild-type reference sequence by the CLUSTALW program [19]. 3. Results We have tested randomly selected cards dated from 1996 to 2006. 576 (6 96) dried blood spot (DBS) samples, and 318 DNA samples, that were extracted from EDTA coagulated blood were analyzed with AS-PCR during our test period, totaling a number of 894 samples. 3.1. Sample preparation, PCR validation First we evaluated the influence of the DNA source on PCR based genetic screens. Two types of DNA sources were tested: EDTA anticoagulated blood samples and blood spots on Guthrie cards. 200 ml of DNA solution were obtained from one tube of EDTA
77
anticoagulated blood and 200 ml DNA solution from one 4 mm diameter piece of the bloodspot test cards. The amount and the integrity of the purified DNA samples was tested by PCR with the primers DF1F and DF2F which amplify parts of GJB2’s coding exon. The length of the PCR products was 420 and 324 bp. They covered the whole exon. The experiments were performed with 6 ml of genomic DNA (gDNA) template. The samples were analyzed by gel electrophoresis. 1, 5, 50 dilutions were made from blood spots and tested by PCR. The 200 ml gDNA solution when diluted 50 times is still acceptable for PCR testing, and that volume (10 000 ml) is sufficient theoretically for more than a thousand PCR reactions per purified sample, or 4 mm paper disc (Fig. 1a). As a negative control we used a bloodfree paper disc from a Guthrie card. As Fig. 1 shows, even the DNA extracted from a 12 years old blood spot gave good result. There were no experiments that produced no results, this means that we were able to screen for 35delG with AS-PCR on all 576 DBS samples. The image shows randomly selected samples from our tests. As we can see on the image there are no smaller sized fragments in detectable quantity bellow our AS-PCR product on the gel. The gDNA solution prepared from whole blood could be diluted by 100 and it still gave acceptable results (Fig. 1b). When diluted further (500, 1000, 5000), the PCR experiments did not give reliable results. The longest PCR product we have worked with is the 809 bp long product of the primers GJB2F and GJB2R (Table 1). We used this PCR product to validate the AS-PCR experiments by sequence analysis. As can be seen on (Fig. 1c) there were no aspecific PCR products in these PCR experiments. We were able to reliably detect the 35delG mutation on all gDNA samples from both DNA sources.
Fig. 1. (a) DNA dilution test. DNA was isolated from Dried Blood Spots from 1996 to 2005 and GJB2’s coding exon was amplified with the GJB2F and GJB2R primers. The PCR runs resulted in a 809 bp fragment. Every sample was diluted by 50 (1); 5 (2); and 1 (3). As a negative control we used a blood-free paper disc from a Guthrie card. The DNA size marker was labeled with M. Even the 50 diluted sample gave good results from a 12 years old blood spot. (b) DNA dilution test. DNA was isolated from peripheral EDTA anticoagulated blood and GJB2’s coding exon was amplified with the GJB2F and GJB2R primers. The gDNA solution was diluted by 1, 5, 10, 50, 100, 500, 1000, 5000. Up to 100 dilution the PCR reactions gave acceptable results, but at and above 500 the PCR product’s quantity noticeably drop. (c) The PCR product of the GJB2F and GJB2R primers on a gel electrophoresis picture. As can be seen on the image, no aspecific fragments were produced during the PCR run. These examples were amongst the samples we sent later for sequencing. (d) A few examples of other regions we examined. The numbers represent specific regions (data not shown) of the following genes: 1,2,36: GJB2; 3,4: 12SrRNA; 5,6,7: COCH; 8: GJA1; 9,10: GJB3; 31,32,33,34: MYO6; 37,38,39: GJB6; and 40: POU3F4. As these fragments were optimized for dHPLC experiments, their sizes are in the range 135–420 bp.
78
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
Other DNA regions were tested as well. On Fig. 1d, we show 8 genes’ (GJB2, 12SrRNA, COCH1, GJA1, GJB3, MYO6, GJB6, POU3F4) various regions (data not shown) that we could amplify with ASPCR from gDNA extracted from Guthrie cards. No other fragment is amplified in detectable amounts in these experiments. The fragment lengths were between 135 and 420 bp, as they were optimized for dHPLC experiments (data not shown). 3.2. Sequencing Homozygous 35delG, heterozygous, and homozygous wildtype samples were sequenced. As all AS-PCR experiments gave consequent and reproducible results when gDNA was purified from EDTA-anticoagulated blood and from dried blood spots, we compared the two methods representing all the three genotypes: wild-type, heterozygous 35delG and homozygous 35delG (Fig. 2). The GJB2 gene was sequenced on all samples, both from the DBS and from the venous blood (576 + 318 = 894). With the optimized reaction conditions no false positive or false negative results were obtained. As the PCR products of GJB2F and GJB2R were 809 bp long, the sequence analysis could cover the full length of the GJB2 gene’s coding exon on both strands (Fig. 3a). All sequencing chromatograms were clear and nice and readable up to 760 bps (Fig. 3b). 3.3. Patient data All the 576 DBS, and 318 samples from peripheral blood samples were sequenced and evaluated for 35delG. We treated these two groups separately, as the samples on Guthrie cards can be considered as a randomized population sample, whereas the patients who came to our department because of various hearing problems cannot be treated a randomized group. On Guthrie cards we have found 13 heterozygotes regarding the GJB2’s 35delG mutation, which means that the carrier frequency is around 2.3% in this population (Hungary) [12,20] (Table 2A). No homozygotes were found, and that can be attributed to the fact that the
Fig. 2. AS-PCR validation. AS-PCR primers were validated on several blood spot samples with a known GJB2 sequence. IC indicates the PCR product amplified by the internal control primer pair ICF-ICR. (w) indicates the PCR product from wild-type allele with DC-GJW primer pair. (m) shows the 35delG allele generated by DC-GJM primer pair. Sample 1, 2, 3, 4, 6, 7, all have wild-type genotype without 35delG mutation. Sample 5 shows a heterozygous genotype because both primer pairs produced the 809 bp product. Sample 8 is homozygous for 35delG mutation. No false positive or false negative signals were detected. No other aspecific products were detectable during gel electrophoresis.
incidence of the homozygous 35delG mutation is about 1/1000-2/ 1000 [20]. Three samples were found to belong to the same persons on blood spots and EDTA-anticoagulated blood samples. All three samples gave the same results when sequenced from whole blood and from DBSs as well, indicating that blood spots of several years of age may be a good source for GJB2 35delG AS-PCR tests. From the DNA extracted from peripheral blood we have found 24 subjects with homozygous recessive genotype, which takes around 7.6% of the screened population, and have found 51 heterozygous patients out of the 318 (16.0%). The total number of patients with homozygous wild genotype was 243 (76.4%). 11 cochlear implantees were homozygous for 35delG, and five CI users, or CI candidates were heterozygous (Table 2C). All of the
Fig. 3. (a) The figure shows a typical result of sequencing reactions of PCR products from the GJB2 gene. The quality of the sequences is the same in all samples, both from the DNA prepared from whole blood and from Dried blood spots. Wild-type homozygous (top), heterozygous (middle), and 35delG homozygous (bottom) alleles are shown. (b) Example of whole sequence of the 809 bp long PCR product from the coding exon of the GJB2 gene.
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
79
Fig. 3. (Continued ).
35delG homozygous recessives’ relatives (whom we could investigate) were heterozygous. 36 of the implantees, and 78 of all the implantees’ relatives were homozygous wild out of the total of 318 patients. Table 2 The abundance of the 35delG mutation in the randomly selected population in Hungary (A), in the selected population of our patients (B), and in the group of our CI users who had hearing loss of unknown origin (C). 35delG +/
35delG
/
No 35delG allele
A Total number of persons: 576 Percent: 100
13 2.3
0 0
563 97.7
Total number of persons: 318 Percent: 100
51 16.0
24 7.6
243 76.4
5 9.6
11 21.2
36 69.2
B
C Investigated CI users: 52 Percent: 100
4. Discussion Different blood storage methods were tested if they can produce appropriate DNA samples for AS-PCR tests and for sequencing the GJB2 gene. DNA was obtained from EDTAanticoagulated venous blood, and from dried blood spots. We have evaluated the possibility of the use of dried blood spots on Guthrie cards as a source of DNA for genetic testing after an extended period of storage under suboptimal conditions. According to our experiments both DNA sources gave satisfactory results. The usability of the PCR products from either template is equal when used in AS-PCR experiments, or in sequencing. Previous works have shown that extracting DNA or RNA is possible from DBSs [17,18], but none have evaluated the effect of storage conditions. AS-PCR primers were validated on samples that contain the 35delG mutation in the GJB2 gene’s coding exon. All the DBS samples with the 35delG allele were sequenced and all sequences certified the AS-PCR results. Whole GJB2 gene sequences of samples resulted in wild-type signals with AS-PCR proved that the
80
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
AS-PCR experiments did not give false negative results. Our AS-PCR test is suitable for large scale screening of dried blood spots as well as for simple and cost-effective detection of selected – and not only, or necessarily GJB2-related – point mutations on individuals or in family samples. Carrier rates of mutations that cause nonsyndromic deafness show strong variation according to the literature. The frequency of some of these mutations is not even known, as they are only analyzed on one or two families [21]. Some of these mutations are researched in more detail, and we do know their carrier rates [21] According to some researchers the 35delG mutation is the single most responsible mutation for nonsyndromic hearing losses in the European population [22]. Still, if not the single cause for most of the nonsydromic hearing losses, this is one of the leading causes for autosomal recessive nonsydromic hearing losses (ARNSLs) [21]. Our findings indicate that in the Hungarian population the carrier frequency of the 35delG mutation is around 2.3%, as we have found 13 heterozygotes on Guthrie cards. We got similar results as To´th et al. showed in their work in a population in Northern Hungary [20]. It seems that geographic (and hence minor ethnic) differences do not play an important role in the distribution of the 35delG alleles in Hungary, because our randomized samples came mostly from the Southern-, and South-Eastern parts of the country. The incidence of the homozygous 35delG mutation is roughly 1/1000 to 1/2000, the carrier rates are in the range of 1–3% [9,12,23] and, as expected, we found no 35delG homozygous patient in the randomized group. According to the literature far more 35delG alleles are found amongst CI users than in the normal population, but the genetic background of hearing loss does not seem to make any difference in the success of the later rehabilitation [24,25]. Speech development, however, can be normal, or close to normal, when the child is fitted with cochlear implant in the early ages. As a consequence, the procedure for the selection of a Cochlear implantation must be carried out in the very early years of life—ideally between 1.5–3 years, or even earlier if possible [26]. In the cochlear implanted population we found 8 patients with homozygous 35delG genotype. Five cochlear implantees were heterozygous for 35delG. All of the 11 homozygous recessives’ relatives (whom we could investigate) were heterozygous. 28 of the implantees, and 78 of all the implantees’ relatives were homozygous wild out of the total of 318 patients. We have found three samples on blood spots in the randomized group that belonged to our patients from whom we drew blood in EDTA-anticoagulated tubes. The samples were analyzed both with AS-PCR and sequencing, and these two methods showed the same results, not only on these three samples, but on all that we compared using the two approaches. However, all six experiments (AS-PCR and sequencing on all three corresponding DBS and whole-blood samples) showed the respective, matching results as well. We acknowledge that carrying out more experiments from the same patients’ whole blood and DBS might have been more convincing, and this is one limitation of the present study. Mainly financial reasons and partly organizational reasons played a critical role in this regard. About 1/4th of our country belongs to our department and that takes around 2 million people. According to statistics, there are around 800–1000 people per 10 million inhabitants per year who may need cochlear implantation, and the number of severe or profound hearing losses is higher, around 1–2% in the European population. That takes about 100 000–200 000 subjects who ideally should be screened for the background of their hearing losses, just based on the severity of their hearing loss, because this could help them to be a potential CI receiver. These are estimates based on our daily work, and our own experiences with patients. Others suggested somewhat lower numbers [27].
The costs of traveling this amount of people to hospitals, or university hospitals, just to draw blood are enormous. With DBS’s the costs can be cut down. Blood can be drawn by their physician, and blood transport do not need to take place in a controlled manner, the temperature for transportation and the time it takes to transport the anticoagulated blood to the screening centers is of no consideration anymore. On the other hand if the need arises to carry out further genetic testing, the DNA we can get from a DBS is enough to carry out hundreds or even more than a thousand PCR experiments. In fact very little amount of gDNA solution is needed to carry out a successful PCR. Whole blood cannot be stored long until it noticeably degrades, and the costs of storing the gDNA solution from peripheral blood (buying, or maintaining a refrigerator for example), and the laboratory room consumed by the needed equipment cannot be compared to the storage demands of the dried blood spots. With the methods presented dried blood spots can be used not only test for metabolic diseases but to carry out genetic experiments as well. As calculated by our lab we can get enough gDNA for a few hundreds of PCR runs from about 2/3 of the money if working with DBS, than needed to do the same number of experiments with EDTA-anticoagulated whole blood, and this is only the financial calculation regarding the acquisition and the maintenance of the equipment. We can add to this that the time consumed is even lower, because to purify 96 samples one needs about an hour with the previously described method from DBS, but almost a whole day to purify it from whole blood even with a very good and productive kit. Significantly more work can be done in a given time-frame when we use dried blood spots to extract gDNA. Purifying gDNA from DBS’s is more simple, requires less work and lab equipment, and the gDNA’s quality – based on the PCR experiments we made – is on par with the gDNA solution we got from the EDTAanticoagulated blood. Beyond all these advantages, AS-PCR is a cost effective, precise, and quick tool that can help us to screen newborns for specific alleles. Acknowledgements We would like to thank all our patients for accepting our participating in our work, and Dr. Krisztina Boda for helping us to randomize the selection of samples in the DBS experiments. We also would like to thank to both Institutes of Hard of Hearing in Szeged and Kaposva´r, for helping us to collect blood samples. We would also like to thank Barna Fodor PhD for carefully reading the manuscript and helping us with his advices. References [1] G. Van Camp, RJH Smith, Hereditary Hearing Loss Homepage. URL: http://webh01. ua.ac.be/hhh/January, 2009. [2] P. Guilford, S. Ben Arab, S. Blanchard, J. Levilliers, J. Weissenbach, A. Belkahia, et al., A non-syndrome form of neurosensory, recessive deafness maps to the pericentromeric region of chromosome 13q, Nat. Genet. 6 (1) (1994) 24–28. [3] K.A. Brown, A.H. Janjua, G. Karbani, G. Parry, A. Noble, G. Crockford, et al., Linkage studies of nonsyndromic recessive deafness (NSRD) in a family originating from the Mirpur region of Pakistan maps DFNB1 centromeric to D13S175, Hum. Mol. Genet. 5 (1) (1996) 169–173. [4] H. Chaib, G. Lina-Granade, P. Guilford, H. Plauchu, J. Levilliers, A. Morgon, et al., A gene responsible for a dominant form of neurosensory nonsyndromic deafness maps to the NSRD1 recessive deafness gene interval, Hum. Mol. Genet. 3 (12) (1994) 2219–2222. [5] D.P. Kelsell, J. Dunlop, H.P. Stevens, N.J. Lench, J.N. Liang, G. Parry, et al., Connexin 26 mutations in hereditary nonsyndromic sensorineural deafness, Nature 387 (6628) (1997) 80–83. [6] R. Rabionet, P. Gasparini, X. Estivill, Molecular genetics of hearing impairment due to mutations in gap junction genes encoding beta connexins, Hum. Mutat. 16 (September (3)) (2000) 190–202. [7] H.B. Zhao, T. Kikuchi, A. Ngezahayo, T.W. White, Gap junctions and cochlear homeostasis, J. Membr. Biol. 209 (February–March (2–3)) (2006) 177–186. [8] K. Willecke, S. Kirchhoff, A. Plum, A. Temme, E. Thonnissen, T. Ott, Biological functions of connexin genes revealed by human genetic defects, dominant
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
[9]
[10]
[11]
[12]
[13] [14] [15] [16] [17]
negative approaches and targeted deletions in the mouse, Novartis Found Symp. 219 (1999) 76–88 (discussion 88–96). A. Pampanos, J. Economides, V. Iliadou, P. Neou, P. Leotsakos, N. Voyiatzis, et al., Prevalence of GJB2 mutations in prelingual deafness in the Greek population, Int. J. Pediatr. Otorhinolaryngol. 65 (September (2)) (2002) 101–108. M. Tekin, T. Duman, G. Bog˘oc¸lu, A. Incesulu, E. Comak, I. Ilhan, et al., Spectrum of GJB2 mutations in Turkey comprises both Caucasian and Oriental variants: roles of parental consanguinity and assortative mating, Hum. Mutat. 21 (May (5)) (2003) 552–553. Y.C. Wang, C.Y. Kung, M.C. Su, C.C. Su, H.M. Hsu, C.C. Tsai, et al., Mutations of Cx26 gene (GJB2) for prelingual deafness in Taiwan, Eur. J. Hum. Genet. 10 (August (8)) (2002) 495–498. A. Bors, H. Andrikovics, L. Kalma´r, N. Erdei, S. Galambos, A. Losonczi, et al., Frequencies of two common mutations (c.35delG and c.167delT) of the connexin 26 gene in different populations of Hungary, Int. J. Mol. Med. 14 (December (6)) (2004) 1105–1108. M. Downing, R. Pollitt, Newborn bloodspot screening in the UK—past, present and future, Ann. Clin. Biochem. 45 (January (Pt 1)) (2008) 11–17. B.L. Therrell, J. Adams, Newborn screening in North America, J. Inherit. Metab. Dis. 30 (August (4)) (2007) 447–465. J.L. Dhondt, Neonatal screening: from the ‘Guthrie age’ to the ‘genetic age’, J. Inherit. Metab. Dis. 30 (August (4)) (2007) 418–422. G.D. Vladutiu, Third Robert Guthrie Memorial Lecture, Mol. Genet. Metab. 91 (August (4)) (2007) 303–304. C. Carducci, L. Ellul, I. Antonozzi, A. Pontecorvi, DNA elution and amplification by polymerase chain reaction from dried blood spots, Biotechniques 13 (November (5)) (1992) 735–737.
81
[18] Y. Matsubara, H. Ikeda, H. Endo, K. Narisawa, Dried blood spot on filter paper as a source of mRNA, Nucleic Acids Res. 20 (April (8)) (1992) 1998. [19] http://www.ebi.ac.uk/Tools/clustalw2/index.html. [20] T. To´th, S. Kupka, B. Haack, K. Riemann, S. Braun, F. Fazakas, et al., GJB2 mutations in patients with nonsyndromic hearing loss from Northeastern Hungary, Hum. Mutat. 23 (June (6)) (2004) 631–632. [21] J. Finsterer, J. Fellinger, Nuclear and mitochondrial genes mutated in nonsyndromic impaired hearing, Int. J. Pediatr. Otorhinolaryngol. 69 (May 5) (2005) 621– 647. [22] E.S. Cohn, P.M. Kelley, Clinical phenotype and mutations in connexin 26 (DFNB1/ GJB2), the most common cause of childhood hearing loss, Am. J. Med. Genet. 89 (September (3)) (1999) 130–136. [23] L.P. Gravina, M.E. Foncuberta, R.C. Estrada, C. Barreiro, L. Chertkoff, Carrier frequency of the 35delG and A1555G deafness mutations in the Argentinean population. Impact on the newborn hearing screening, Int. J. Pediatr. Otorhinolaryngol. 71 (April (4)) (2007) 639–643. [24] L.R. Lustig, D. Lin, H. Venick, J. Larky, J. Yeagle, J. Chinnici, et al., GJB2 gene mutations in cochlear implant recipients: prevalence and impact on outcome, Arch. Otolaryngol. Head Neck Surg. 130 (May (5)) (2004) 541–546. [25] T.C. Fischer, J. Samanich, B.E. Morrow, J. Chobot-Rodd, A. Shanske, S.R. Parikh, Genetic evaluation of American minority pediatric cochlear implant recipients, Int. J. Pediatr. Otorhinolaryngol. 73 (February (2)) (2009) 195–203. [26] L. Colletti, Long-term follow-up of infants (4-11 months) fitted with cochlear implants, Acta Otolaryngol. (January) (2009) 1–6. [27] A. Davis, H. Fortnum, G. O’Donoghue, Children who could benefit from a cochlear implant: a European estimate of projected numbers, cost and relevant characteristics, Int. J. Pediatr. Otorhinolaryngol. 31 (March (2–3)) (1995) 221–233.
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Minimally invasive genetic screen for GJB2 related deafness using dried blood spots§ Attila L. Nagy a,*, Ro´bert Csa´ki b, Jo´zsef Klem c, La´szlo´ Rovo´ a, Ferenc To´th a, Gyula Ta´losi d, Jo´zsef Jo´ri a, Korne´l Kova´cs c, Jo´zsef Ge´za Kiss a a University of Szeged, Albert Szent-Gyo¨rgyi Pharmaceutical and Medical Centre, Department of Oto-rhino-laryngology and Head and Neck Surgery, 6725, Tisza Lajos krt. 111, Szeged, Hungary b Alfa-Biosoft Ltd., 6720, Na´dor u. 10, Szeged, Hungary c University of Szeged, Faculty of Sciences, Institute of Biotechnology, 6726, Ko¨ze´p fasor 52, Szeged, Hungary d University of Szeged, Albert Szent-Gyo¨rgyi Pharmaceutical and Medical Centre, Department of Pediatrics, 6720, Kora´nyi fasor 14-15, Szeged, Hungary
A R T I C L E I N F O
A B S T R A C T
Article history: Received 28 May 2009 Received in revised form 18 October 2009 Accepted 21 October 2009 Available online 24 November 2009
Objective: Nonsyndromic hearing loss is one of the most abundant human sensory disorders, and can be found in 1 out of 1000 newborns. In 60–70% of the cases this disorder is hereditary. The phenotype varies from moderate hearing loss to almost complete deafness, often only revealed in late childhood. Early detection of hearing related genetic variations in the first few weeks of life would allow planning of the audiological and logopedical procedures to maintain the children’s normal audiological and speech development, and if required a cochlear implantation can be planned in time. We wanted to evaluate, whether the blood samples collected from neonates onto Guthrie cards (dried blood spots, or DBS), and blood collected from people of various ages into blood collecting tubes is equally usable for genetic testing. The quality of the samples on DBS’s for genetic tests after an extended period of storage was evaluated. The methods for sample preparation and analysis were also evaluated. Methods: Two DNA extraction methods were compared on the samples. We extracted DNA from whole blood with the Versagene Blood Kit from Gentra, and from DBS’s with boiling. Allele-specific PCRs (ASPCR) were carried out on each sample. Samples were analyzed with AS-PCR and sequencing, for the 35delG mutation in the GJB2 (Cx26) gene. Freshly drawn and dried blood spot samples stored for several years were used in the experiments. Results: An AS-PCR method for detecting 35delG mutation on DNA extracted from Guthrie cards was validated. Blood samples up to 10 years of storage were applicable in the screen. 84 patients were found with 35delG mutations, both heterozygous (with no detected hearing related phenotypical discrepancies), and homozygous (phenotipically with moderate to severe hearing loss) forms. Conclusions: The dried blood spots on Guthrie cards require only three drops of blood to be collected from children, which causes less stress than taking 3 ml of blood. The blood stored on Guthrie cards can be used to store DNA samples for at least 10 years. Even under suboptimal storage conditions the samples’ DNA remains intact for genetic testing. Compared to blood collection tubes Guthrie cards cost less, are easier to transport and store. ß 2009 Elsevier Ireland Ltd. All rights reserved.
Keywords: Dried blood spot Polymerase chain reaction Whole blood Nonsyndromic deafness 35delG Screening
1. Introduction According to current knowledge physiological processes related to hearing are regulated by at least 100–150 genes. To date 146
§ The work described here was supported by the grant GVOP-3.1.1-2004-050498/3.0. * Corresponding author. Tel.: +36 62 54 58 50; fax: +36 62 54 58 48. E-mail addresses: [email protected] (A.L. Nagy), [email protected] (R. Csa´ki), [email protected] (L. Rovo´), [email protected] (F. To´th), [email protected] (J. Jo´ri), [email protected] (J.G. Kiss).
0165-5876/$ – see front matter ß 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2009.10.021
alleles of 42 genes have been identified (57 dominant, 77 recessive, 8 X-linked, 1 Y-linked, 2 modifier, 1 auditory neuropathy) [1]. The large number of genes and loci complicate the genetic analysis of nonsydromic hearing losses. DFNB1 was the first identified locus; its autosomal recessive mutation causes nonsyndromic hearing loss [2]. Using co-segregation analysis, this locus was mapped to 13q12-13 [3]. In 1994, Chaib et al. described the first dominant nonsyndromic hearing loss of genetic origin to 13q12-13 [4]. Mutation in the coding region of Gap Junction protein Beta 2 (GJB2) – also called Connexin26 (Cx26) – was the first to be linked to nonsyndromic hearing loss of genetic origin [5]. Connexin 26 (Cx26) is a protein, which belongs to the family of connexins. In the
76
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
cochlea [6,7] connexins facilitate the flux, and the recycling of K+ ions from the intracellular space to the endolympha [8]. The length of the functional, protein coding exon of GJB2 is 681 bp, and it codes a 226 amino acid polypeptide. 35delG causes a frameshift mutation, the deletion of one guanine residue in a stretch of six guanines in the coding region of the gene at position 35, resulting in a nonsense mutation at the 13th codon. 35delG mutation of the GJB2 gene accounts for about 7–15% of nonsyndromic hearing losses of genetic origin in the European population [9–11], and regarding this mutation, about 1–5% of the European population is a carrier [12]. Since the 1950’s [13] and 1960’s [14] national neonatal screening programs begun to operate in the advanced countries; they mostly perform screening for metabolic diseases [15,16]. These programs generated an enormous amount of dried blood spots (DBS) on so called Guthrie cards. These cards have the advantage of easy transportation, and their storage conditions are much cheaper than any other way of storing blood samples. The question arises whether they can be used as a basis of genetic screening, and if yes, to what extent. As national hearing screening programs gain more and more attention, and financial funds, they become more and more ‘‘sample-hungry’’. If these cards could be utilized to give proper results, they could serve as a complementary tool to the existing auditory and genetic screening methods. These blood spots are available for most of the newborns in those countries, and each individual can be tested in their absence. The procedure is also less traumatic to children. 2. Materials and methods All participants involved in the trials were informed according to the University’s Ethical Committee’s Guidelines, and all have signed a written consent. Double blind tests were performed on the spot blood samples taken for the routine, population wide metabolic screening tests of neonates. The samples were taken on Guthrie cards at the 3rd-4th life day of newborns at any neonatal ward where the neonates were cared in the eastern part of Hungary and sent via conventional mail. The test cards were stored at room air at the Department of Paediatrics, University of Szeged. The Guthrie cards were selected as follows: 48 pieces from the years 1996, and 1997, and further 96 pieces from the years 1999, 2000, 2001, 2004, 2005, respectively. The total number of DBS’s was n = 576, as these were selected randomly, we could use them as a ‘‘generic population sample’’. The other population we examined consists of 318 of patients. These patients were Cochlear Implant (CI) users, their relatives, CI candidates, their relatives, and a few individual patients from the ENT Clinic. CI users and CI candidates were selected based on two criteria. They were selected if there was a family history of hearing losses, and there were no organic abnormalities (anatomical variations, or developmental problems) or other diseases that are known to cause hearing loss or deafness in the patient’s history. In case of the CI users (n = 20 + 32 = 52) and CI candidates (n = 56) the average hearing threshold level was bellow 70 dB and speech
recognition performance was under 25%. We excluded those patients and CI users whose patient history contained some form of disease that can cause deafness, or who suffered a head trauma, or head injury, that can account for their hearing problems. The CI users’ and CI candidates’ family members (n = 163) have various levels of hearing loss, from no hearing loss at all to severe to moderate levels. There were some individuals who wanted to participate in our study with various levels but with unknown origin of hearing loss (n = 47). 20 CI users (out of n = 52) were individuals with no screened relatives. Our group of control persons consisted of people with hearing threshold levels at 5 dB or better on both ears (n = 20) and no family history of hearing losses. 2.1. Preparation of DNA extracts from DBS for AS-PCR 4 mm diameter pieces from the bloodspot test cards were punched out and put into PCR tubes along with 200 ml 1 PCR buffer (Eppendorf HotMaster taq). The DNA was extracted from the DBS by 20 min of incubation at 96 8C in a thermalcycler. The samples were then centrifuged at 16 000 g for 2 min. The supernatant was then transferred into a sterile microcentrifuge tube, and stored at 4 8C until utilization [17,18]. 2.2. Preparation of DNA from venous blood for AS-PCR We collected 3 ml blood from patients having Cochlear Implant, their relatives, and CI candidates. Blood anti-coagulant was EDTA. Genomic DNA (gDNA) was purified from 400 ml of blood using Versagene Blood Kit (Gentra) according to the manufacturer instructions. The concentration of the DNA was measured with spectrophotometer and was calculated by the adsorption at 260 and 360 nm. 2.3. Polymerase chain reactions DNA integrity test reactions were carried out using HotMaster Taq DNA Polymerase (Eppendorf). 30 ml final volume of the reaction mix contained 3 ml (10) HotMasterTaq buffer (Eppendorf), 2.5 ml, dNTP (2.5 mM), 0.5–0.5 ml (15 pM) DF2F-DF2R primer pair (Table 1); 1 U HotMasterTaq (Eppendorf); 6 ml gDNS template and 16.5 ml water. PCR program was as follows: 96 8C 2 min, for 35 cycles (96 8C 30 s–61 8C 30 s–68 8C 35 s), and after these 35 cycles 96 8C for 5 min. Negative control experiment was also done with paper discs originating from the ‘‘blood-free’’ parts of the Guthrie papers. PCR fragments were analyzed by agarose gel electrophoresis on 1.5% agarose gel (AbGen) with 1X TBE buffer. The internal control used in all the AS-PCR experiments were the primer pair ICF and ICR. They amplificate a part of the serine proteinase inhibitor gene (Table 1). AS-PCR reactions: 30 ml final volume of the reaction mix contained: 3 ml (10) HotMasterTaq buffer (Eppendorf), 2.5 ml, dNTP (2.5 mM), 0.5–0.5 ml primers GJC-GJW pair for wild allele detection and GJC-GJM for 35delG mutant allele detection (Table 1). (15 pM), 0.4–0.4 ml (15 pM) internal control primer
Table 1 The primers used in this work and their sequences. Primer name
Sequence
Description
ICF ICR GJC GJW GJM DF2F DF2R
CCC ACC TTC CCC TCT CTC CAG GCA AAT GGG GGG CCT CAG TCC CAA CAT GGC TAA GAG GTG AGT GAT CGT AGC ACA CGT TCT TGC A GCA CGC TGC AGA CGA TCC TGG GGA G CAC GCT GCA GAC GAT CCT GGG GAT TCT CCC TGT TCT GTC CTA GC TTT CCC AAG GCC TCT TCC AC
Internal control (serine proteinase inhibitor gene) Internal control (serine proteinase inhibitor gene) Common reverse primer for GJB2 Primer for 35delG Wild allele detection Primer for 35delG mutant allele detection GJB2 exon and flanking region for sequencing GJB2 exon and flanking region for sequencing
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
pair (ICF-ICR); 1 U HotMasterTaq (Eppendorf); 6 ml gDNS template and 16.5 ml water. The PCR program was as follows: first denaturation step at 95 8C for 5 min, then 35 cycles at 96 8C 30 s, at 65 8C 35 s, and at 68 8C 38 s, and as the last step, 10 min incubation time at 68 8C. PCR fragments were analyzed by agarose gel electrophoresis on 1.5% agarose gel (AbGen) with 1X TBE buffer. 2.4. DNA Sequencing and sequence analysis PCR fragments were generated as described in section 2.3. PCR products were desalted on Microcon columns (Millipore). The purified PCR products were eluted in 30 ml of water. The DNA sequences were determined by automated sequencing at Macrogen Inc. (Korea) on both strands. The sequences were aligned to the wild-type reference sequence by the CLUSTALW program [19]. 3. Results We have tested randomly selected cards dated from 1996 to 2006. 576 (6 96) dried blood spot (DBS) samples, and 318 DNA samples, that were extracted from EDTA coagulated blood were analyzed with AS-PCR during our test period, totaling a number of 894 samples. 3.1. Sample preparation, PCR validation First we evaluated the influence of the DNA source on PCR based genetic screens. Two types of DNA sources were tested: EDTA anticoagulated blood samples and blood spots on Guthrie cards. 200 ml of DNA solution were obtained from one tube of EDTA
77
anticoagulated blood and 200 ml DNA solution from one 4 mm diameter piece of the bloodspot test cards. The amount and the integrity of the purified DNA samples was tested by PCR with the primers DF1F and DF2F which amplify parts of GJB2’s coding exon. The length of the PCR products was 420 and 324 bp. They covered the whole exon. The experiments were performed with 6 ml of genomic DNA (gDNA) template. The samples were analyzed by gel electrophoresis. 1, 5, 50 dilutions were made from blood spots and tested by PCR. The 200 ml gDNA solution when diluted 50 times is still acceptable for PCR testing, and that volume (10 000 ml) is sufficient theoretically for more than a thousand PCR reactions per purified sample, or 4 mm paper disc (Fig. 1a). As a negative control we used a bloodfree paper disc from a Guthrie card. As Fig. 1 shows, even the DNA extracted from a 12 years old blood spot gave good result. There were no experiments that produced no results, this means that we were able to screen for 35delG with AS-PCR on all 576 DBS samples. The image shows randomly selected samples from our tests. As we can see on the image there are no smaller sized fragments in detectable quantity bellow our AS-PCR product on the gel. The gDNA solution prepared from whole blood could be diluted by 100 and it still gave acceptable results (Fig. 1b). When diluted further (500, 1000, 5000), the PCR experiments did not give reliable results. The longest PCR product we have worked with is the 809 bp long product of the primers GJB2F and GJB2R (Table 1). We used this PCR product to validate the AS-PCR experiments by sequence analysis. As can be seen on (Fig. 1c) there were no aspecific PCR products in these PCR experiments. We were able to reliably detect the 35delG mutation on all gDNA samples from both DNA sources.
Fig. 1. (a) DNA dilution test. DNA was isolated from Dried Blood Spots from 1996 to 2005 and GJB2’s coding exon was amplified with the GJB2F and GJB2R primers. The PCR runs resulted in a 809 bp fragment. Every sample was diluted by 50 (1); 5 (2); and 1 (3). As a negative control we used a blood-free paper disc from a Guthrie card. The DNA size marker was labeled with M. Even the 50 diluted sample gave good results from a 12 years old blood spot. (b) DNA dilution test. DNA was isolated from peripheral EDTA anticoagulated blood and GJB2’s coding exon was amplified with the GJB2F and GJB2R primers. The gDNA solution was diluted by 1, 5, 10, 50, 100, 500, 1000, 5000. Up to 100 dilution the PCR reactions gave acceptable results, but at and above 500 the PCR product’s quantity noticeably drop. (c) The PCR product of the GJB2F and GJB2R primers on a gel electrophoresis picture. As can be seen on the image, no aspecific fragments were produced during the PCR run. These examples were amongst the samples we sent later for sequencing. (d) A few examples of other regions we examined. The numbers represent specific regions (data not shown) of the following genes: 1,2,36: GJB2; 3,4: 12SrRNA; 5,6,7: COCH; 8: GJA1; 9,10: GJB3; 31,32,33,34: MYO6; 37,38,39: GJB6; and 40: POU3F4. As these fragments were optimized for dHPLC experiments, their sizes are in the range 135–420 bp.
78
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
Other DNA regions were tested as well. On Fig. 1d, we show 8 genes’ (GJB2, 12SrRNA, COCH1, GJA1, GJB3, MYO6, GJB6, POU3F4) various regions (data not shown) that we could amplify with ASPCR from gDNA extracted from Guthrie cards. No other fragment is amplified in detectable amounts in these experiments. The fragment lengths were between 135 and 420 bp, as they were optimized for dHPLC experiments (data not shown). 3.2. Sequencing Homozygous 35delG, heterozygous, and homozygous wildtype samples were sequenced. As all AS-PCR experiments gave consequent and reproducible results when gDNA was purified from EDTA-anticoagulated blood and from dried blood spots, we compared the two methods representing all the three genotypes: wild-type, heterozygous 35delG and homozygous 35delG (Fig. 2). The GJB2 gene was sequenced on all samples, both from the DBS and from the venous blood (576 + 318 = 894). With the optimized reaction conditions no false positive or false negative results were obtained. As the PCR products of GJB2F and GJB2R were 809 bp long, the sequence analysis could cover the full length of the GJB2 gene’s coding exon on both strands (Fig. 3a). All sequencing chromatograms were clear and nice and readable up to 760 bps (Fig. 3b). 3.3. Patient data All the 576 DBS, and 318 samples from peripheral blood samples were sequenced and evaluated for 35delG. We treated these two groups separately, as the samples on Guthrie cards can be considered as a randomized population sample, whereas the patients who came to our department because of various hearing problems cannot be treated a randomized group. On Guthrie cards we have found 13 heterozygotes regarding the GJB2’s 35delG mutation, which means that the carrier frequency is around 2.3% in this population (Hungary) [12,20] (Table 2A). No homozygotes were found, and that can be attributed to the fact that the
Fig. 2. AS-PCR validation. AS-PCR primers were validated on several blood spot samples with a known GJB2 sequence. IC indicates the PCR product amplified by the internal control primer pair ICF-ICR. (w) indicates the PCR product from wild-type allele with DC-GJW primer pair. (m) shows the 35delG allele generated by DC-GJM primer pair. Sample 1, 2, 3, 4, 6, 7, all have wild-type genotype without 35delG mutation. Sample 5 shows a heterozygous genotype because both primer pairs produced the 809 bp product. Sample 8 is homozygous for 35delG mutation. No false positive or false negative signals were detected. No other aspecific products were detectable during gel electrophoresis.
incidence of the homozygous 35delG mutation is about 1/1000-2/ 1000 [20]. Three samples were found to belong to the same persons on blood spots and EDTA-anticoagulated blood samples. All three samples gave the same results when sequenced from whole blood and from DBSs as well, indicating that blood spots of several years of age may be a good source for GJB2 35delG AS-PCR tests. From the DNA extracted from peripheral blood we have found 24 subjects with homozygous recessive genotype, which takes around 7.6% of the screened population, and have found 51 heterozygous patients out of the 318 (16.0%). The total number of patients with homozygous wild genotype was 243 (76.4%). 11 cochlear implantees were homozygous for 35delG, and five CI users, or CI candidates were heterozygous (Table 2C). All of the
Fig. 3. (a) The figure shows a typical result of sequencing reactions of PCR products from the GJB2 gene. The quality of the sequences is the same in all samples, both from the DNA prepared from whole blood and from Dried blood spots. Wild-type homozygous (top), heterozygous (middle), and 35delG homozygous (bottom) alleles are shown. (b) Example of whole sequence of the 809 bp long PCR product from the coding exon of the GJB2 gene.
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
79
Fig. 3. (Continued ).
35delG homozygous recessives’ relatives (whom we could investigate) were heterozygous. 36 of the implantees, and 78 of all the implantees’ relatives were homozygous wild out of the total of 318 patients. Table 2 The abundance of the 35delG mutation in the randomly selected population in Hungary (A), in the selected population of our patients (B), and in the group of our CI users who had hearing loss of unknown origin (C). 35delG +/
35delG
/
No 35delG allele
A Total number of persons: 576 Percent: 100
13 2.3
0 0
563 97.7
Total number of persons: 318 Percent: 100
51 16.0
24 7.6
243 76.4
5 9.6
11 21.2
36 69.2
B
C Investigated CI users: 52 Percent: 100
4. Discussion Different blood storage methods were tested if they can produce appropriate DNA samples for AS-PCR tests and for sequencing the GJB2 gene. DNA was obtained from EDTAanticoagulated venous blood, and from dried blood spots. We have evaluated the possibility of the use of dried blood spots on Guthrie cards as a source of DNA for genetic testing after an extended period of storage under suboptimal conditions. According to our experiments both DNA sources gave satisfactory results. The usability of the PCR products from either template is equal when used in AS-PCR experiments, or in sequencing. Previous works have shown that extracting DNA or RNA is possible from DBSs [17,18], but none have evaluated the effect of storage conditions. AS-PCR primers were validated on samples that contain the 35delG mutation in the GJB2 gene’s coding exon. All the DBS samples with the 35delG allele were sequenced and all sequences certified the AS-PCR results. Whole GJB2 gene sequences of samples resulted in wild-type signals with AS-PCR proved that the
80
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
AS-PCR experiments did not give false negative results. Our AS-PCR test is suitable for large scale screening of dried blood spots as well as for simple and cost-effective detection of selected – and not only, or necessarily GJB2-related – point mutations on individuals or in family samples. Carrier rates of mutations that cause nonsyndromic deafness show strong variation according to the literature. The frequency of some of these mutations is not even known, as they are only analyzed on one or two families [21]. Some of these mutations are researched in more detail, and we do know their carrier rates [21] According to some researchers the 35delG mutation is the single most responsible mutation for nonsyndromic hearing losses in the European population [22]. Still, if not the single cause for most of the nonsydromic hearing losses, this is one of the leading causes for autosomal recessive nonsydromic hearing losses (ARNSLs) [21]. Our findings indicate that in the Hungarian population the carrier frequency of the 35delG mutation is around 2.3%, as we have found 13 heterozygotes on Guthrie cards. We got similar results as To´th et al. showed in their work in a population in Northern Hungary [20]. It seems that geographic (and hence minor ethnic) differences do not play an important role in the distribution of the 35delG alleles in Hungary, because our randomized samples came mostly from the Southern-, and South-Eastern parts of the country. The incidence of the homozygous 35delG mutation is roughly 1/1000 to 1/2000, the carrier rates are in the range of 1–3% [9,12,23] and, as expected, we found no 35delG homozygous patient in the randomized group. According to the literature far more 35delG alleles are found amongst CI users than in the normal population, but the genetic background of hearing loss does not seem to make any difference in the success of the later rehabilitation [24,25]. Speech development, however, can be normal, or close to normal, when the child is fitted with cochlear implant in the early ages. As a consequence, the procedure for the selection of a Cochlear implantation must be carried out in the very early years of life—ideally between 1.5–3 years, or even earlier if possible [26]. In the cochlear implanted population we found 8 patients with homozygous 35delG genotype. Five cochlear implantees were heterozygous for 35delG. All of the 11 homozygous recessives’ relatives (whom we could investigate) were heterozygous. 28 of the implantees, and 78 of all the implantees’ relatives were homozygous wild out of the total of 318 patients. We have found three samples on blood spots in the randomized group that belonged to our patients from whom we drew blood in EDTA-anticoagulated tubes. The samples were analyzed both with AS-PCR and sequencing, and these two methods showed the same results, not only on these three samples, but on all that we compared using the two approaches. However, all six experiments (AS-PCR and sequencing on all three corresponding DBS and whole-blood samples) showed the respective, matching results as well. We acknowledge that carrying out more experiments from the same patients’ whole blood and DBS might have been more convincing, and this is one limitation of the present study. Mainly financial reasons and partly organizational reasons played a critical role in this regard. About 1/4th of our country belongs to our department and that takes around 2 million people. According to statistics, there are around 800–1000 people per 10 million inhabitants per year who may need cochlear implantation, and the number of severe or profound hearing losses is higher, around 1–2% in the European population. That takes about 100 000–200 000 subjects who ideally should be screened for the background of their hearing losses, just based on the severity of their hearing loss, because this could help them to be a potential CI receiver. These are estimates based on our daily work, and our own experiences with patients. Others suggested somewhat lower numbers [27].
The costs of traveling this amount of people to hospitals, or university hospitals, just to draw blood are enormous. With DBS’s the costs can be cut down. Blood can be drawn by their physician, and blood transport do not need to take place in a controlled manner, the temperature for transportation and the time it takes to transport the anticoagulated blood to the screening centers is of no consideration anymore. On the other hand if the need arises to carry out further genetic testing, the DNA we can get from a DBS is enough to carry out hundreds or even more than a thousand PCR experiments. In fact very little amount of gDNA solution is needed to carry out a successful PCR. Whole blood cannot be stored long until it noticeably degrades, and the costs of storing the gDNA solution from peripheral blood (buying, or maintaining a refrigerator for example), and the laboratory room consumed by the needed equipment cannot be compared to the storage demands of the dried blood spots. With the methods presented dried blood spots can be used not only test for metabolic diseases but to carry out genetic experiments as well. As calculated by our lab we can get enough gDNA for a few hundreds of PCR runs from about 2/3 of the money if working with DBS, than needed to do the same number of experiments with EDTA-anticoagulated whole blood, and this is only the financial calculation regarding the acquisition and the maintenance of the equipment. We can add to this that the time consumed is even lower, because to purify 96 samples one needs about an hour with the previously described method from DBS, but almost a whole day to purify it from whole blood even with a very good and productive kit. Significantly more work can be done in a given time-frame when we use dried blood spots to extract gDNA. Purifying gDNA from DBS’s is more simple, requires less work and lab equipment, and the gDNA’s quality – based on the PCR experiments we made – is on par with the gDNA solution we got from the EDTAanticoagulated blood. Beyond all these advantages, AS-PCR is a cost effective, precise, and quick tool that can help us to screen newborns for specific alleles. Acknowledgements We would like to thank all our patients for accepting our participating in our work, and Dr. Krisztina Boda for helping us to randomize the selection of samples in the DBS experiments. We also would like to thank to both Institutes of Hard of Hearing in Szeged and Kaposva´r, for helping us to collect blood samples. We would also like to thank Barna Fodor PhD for carefully reading the manuscript and helping us with his advices. References [1] G. Van Camp, RJH Smith, Hereditary Hearing Loss Homepage. URL: http://webh01. ua.ac.be/hhh/January, 2009. [2] P. Guilford, S. Ben Arab, S. Blanchard, J. Levilliers, J. Weissenbach, A. Belkahia, et al., A non-syndrome form of neurosensory, recessive deafness maps to the pericentromeric region of chromosome 13q, Nat. Genet. 6 (1) (1994) 24–28. [3] K.A. Brown, A.H. Janjua, G. Karbani, G. Parry, A. Noble, G. Crockford, et al., Linkage studies of nonsyndromic recessive deafness (NSRD) in a family originating from the Mirpur region of Pakistan maps DFNB1 centromeric to D13S175, Hum. Mol. Genet. 5 (1) (1996) 169–173. [4] H. Chaib, G. Lina-Granade, P. Guilford, H. Plauchu, J. Levilliers, A. Morgon, et al., A gene responsible for a dominant form of neurosensory nonsyndromic deafness maps to the NSRD1 recessive deafness gene interval, Hum. Mol. Genet. 3 (12) (1994) 2219–2222. [5] D.P. Kelsell, J. Dunlop, H.P. Stevens, N.J. Lench, J.N. Liang, G. Parry, et al., Connexin 26 mutations in hereditary nonsyndromic sensorineural deafness, Nature 387 (6628) (1997) 80–83. [6] R. Rabionet, P. Gasparini, X. Estivill, Molecular genetics of hearing impairment due to mutations in gap junction genes encoding beta connexins, Hum. Mutat. 16 (September (3)) (2000) 190–202. [7] H.B. Zhao, T. Kikuchi, A. Ngezahayo, T.W. White, Gap junctions and cochlear homeostasis, J. Membr. Biol. 209 (February–March (2–3)) (2006) 177–186. [8] K. Willecke, S. Kirchhoff, A. Plum, A. Temme, E. Thonnissen, T. Ott, Biological functions of connexin genes revealed by human genetic defects, dominant
A.L. Nagy et al. / International Journal of Pediatric Otorhinolaryngology 74 (2010) 75–81
[9]
[10]
[11]
[12]
[13] [14] [15] [16] [17]
negative approaches and targeted deletions in the mouse, Novartis Found Symp. 219 (1999) 76–88 (discussion 88–96). A. Pampanos, J. Economides, V. Iliadou, P. Neou, P. Leotsakos, N. Voyiatzis, et al., Prevalence of GJB2 mutations in prelingual deafness in the Greek population, Int. J. Pediatr. Otorhinolaryngol. 65 (September (2)) (2002) 101–108. M. Tekin, T. Duman, G. Bog˘oc¸lu, A. Incesulu, E. Comak, I. Ilhan, et al., Spectrum of GJB2 mutations in Turkey comprises both Caucasian and Oriental variants: roles of parental consanguinity and assortative mating, Hum. Mutat. 21 (May (5)) (2003) 552–553. Y.C. Wang, C.Y. Kung, M.C. Su, C.C. Su, H.M. Hsu, C.C. Tsai, et al., Mutations of Cx26 gene (GJB2) for prelingual deafness in Taiwan, Eur. J. Hum. Genet. 10 (August (8)) (2002) 495–498. A. Bors, H. Andrikovics, L. Kalma´r, N. Erdei, S. Galambos, A. Losonczi, et al., Frequencies of two common mutations (c.35delG and c.167delT) of the connexin 26 gene in different populations of Hungary, Int. J. Mol. Med. 14 (December (6)) (2004) 1105–1108. M. Downing, R. Pollitt, Newborn bloodspot screening in the UK—past, present and future, Ann. Clin. Biochem. 45 (January (Pt 1)) (2008) 11–17. B.L. Therrell, J. Adams, Newborn screening in North America, J. Inherit. Metab. Dis. 30 (August (4)) (2007) 447–465. J.L. Dhondt, Neonatal screening: from the ‘Guthrie age’ to the ‘genetic age’, J. Inherit. Metab. Dis. 30 (August (4)) (2007) 418–422. G.D. Vladutiu, Third Robert Guthrie Memorial Lecture, Mol. Genet. Metab. 91 (August (4)) (2007) 303–304. C. Carducci, L. Ellul, I. Antonozzi, A. Pontecorvi, DNA elution and amplification by polymerase chain reaction from dried blood spots, Biotechniques 13 (November (5)) (1992) 735–737.
81
[18] Y. Matsubara, H. Ikeda, H. Endo, K. Narisawa, Dried blood spot on filter paper as a source of mRNA, Nucleic Acids Res. 20 (April (8)) (1992) 1998. [19] http://www.ebi.ac.uk/Tools/clustalw2/index.html. [20] T. To´th, S. Kupka, B. Haack, K. Riemann, S. Braun, F. Fazakas, et al., GJB2 mutations in patients with nonsyndromic hearing loss from Northeastern Hungary, Hum. Mutat. 23 (June (6)) (2004) 631–632. [21] J. Finsterer, J. Fellinger, Nuclear and mitochondrial genes mutated in nonsyndromic impaired hearing, Int. J. Pediatr. Otorhinolaryngol. 69 (May 5) (2005) 621– 647. [22] E.S. Cohn, P.M. Kelley, Clinical phenotype and mutations in connexin 26 (DFNB1/ GJB2), the most common cause of childhood hearing loss, Am. J. Med. Genet. 89 (September (3)) (1999) 130–136. [23] L.P. Gravina, M.E. Foncuberta, R.C. Estrada, C. Barreiro, L. Chertkoff, Carrier frequency of the 35delG and A1555G deafness mutations in the Argentinean population. Impact on the newborn hearing screening, Int. J. Pediatr. Otorhinolaryngol. 71 (April (4)) (2007) 639–643. [24] L.R. Lustig, D. Lin, H. Venick, J. Larky, J. Yeagle, J. Chinnici, et al., GJB2 gene mutations in cochlear implant recipients: prevalence and impact on outcome, Arch. Otolaryngol. Head Neck Surg. 130 (May (5)) (2004) 541–546. [25] T.C. Fischer, J. Samanich, B.E. Morrow, J. Chobot-Rodd, A. Shanske, S.R. Parikh, Genetic evaluation of American minority pediatric cochlear implant recipients, Int. J. Pediatr. Otorhinolaryngol. 73 (February (2)) (2009) 195–203. [26] L. Colletti, Long-term follow-up of infants (4-11 months) fitted with cochlear implants, Acta Otolaryngol. (January) (2009) 1–6. [27] A. Davis, H. Fortnum, G. O’Donoghue, Children who could benefit from a cochlear implant: a European estimate of projected numbers, cost and relevant characteristics, Int. J. Pediatr. Otorhinolaryngol. 31 (March (2–3)) (1995) 221–233.