Advertisement

Validation of a Commercially Available Screening Tool for the Rapid Identification of CGG Trinucleotide Repeat Expansions in FMR1

Open ArchivePublished:March 14, 2015DOI:https://doi.org/10.1016/j.jmoldx.2014.12.005
      Recently developed PCR-based methods for fragile X syndrome testing are often regarded as screening tools because of a reduced reliance on Southern blot analysis. However, existing PCR methods rely essentially on capillary electrophoresis for the analysis of amplicons. These methods not only require an expensive capillary electrophoresis instrument but also involve post-PCR processing steps. Here, we evaluated a commercially available PCR-based assay that uses melt curve analysis as a screening tool for the rapid detection of CGG repeat expansions in the fragile X mental retardation 1 (FMR1) gene. On the basis of testing with well-characterized DNA samples, the assay gave a detection limit of 10 ng per reaction and an analytic specificity beyond 150 ng per reaction. Furthermore, the melt temperatures critical for result interpretation were found to be closely linked to the CGG expansion lengths with great consistency (CV < 0.55%). The clinical performance of the assay was established with 528 blinded and previously analyzed clinical samples, yielding results of 100% sensitivity (95% CI, 91.0%–100%) and 99.6% specificity (95% CI, 98.5%–99.9%) in detecting expansions >55 CGG repeats in FMR1. This new approach eliminates post-PCR handling for all non-expanded samples, and exemplifies a truly efficient screening procedure.
      Fragile X syndrome (FXS) is the most common cause of inheritable intellectual disabilities and autism, affecting an estimated 1 in 4000 males and 1 in 5000 to 8000 females.
      • Hill M.K.
      • Archibald A.D.
      • Cohen J.
      • Metcalfe S.A.
      A systematic review of population screening for fragile X syndrome.
      • Tassone F.
      Newborn screening for fragile X syndrome.
      The medical condition is attributed to a trinucleotide CGG repeat expansion (>200 CGG repeats, full mutation) in the 5′ untranslated region of the fragile X mental retardation 1 (FMR1) gene, which results in aberrant hypermethylation of the FMR1 promoter.
      • Pieretti M.
      • Zhang F.
      • Fu Y.
      • Warren S.
      • Oostra B.
      • Caskey C.
      • Nelson D.
      Absence of expression of the FMR-1 gene in fragile X syndrome.
      • Verkerk A.J.
      • Pieretti M.
      • Sutcliffe J.S.
      • Fu Y.H.
      • Kuhl D.P.
      • Pizzuti A.
      • Reiner O.
      • Richards S.
      • Victoria M.F.
      • Zhang F.P.
      • et al.
      Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome.
      • Fu Y.H.
      • Kuhl D.P.
      • Pizzuti A.
      • Pieretti M.
      • Sutcliffe J.S.
      • Richards S.
      • Verkerk A.J.
      • Holden J.J.
      • Fenwick Jr., R.G.
      • Warren S.T.
      • et al.
      Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox.
      The CGG repeat expansion varies considerably in size among persons, giving rise to a spectrum of clinical outcomes through different pathogenic mechanisms.
      • Chen L.
      • Hadd A.
      • Sah S.
      • Filipovic-Sadic S.
      • Krosting J.
      • Sekinger E.
      • Pan R.
      • Hagerman P.J.
      • Stenzel T.T.
      • Tassone F.
      • Latham G.J.
      An information-rich CGG repeat primed PCR that detects the full range of fragile X expanded alleles and minimizes the need for southern blot analysis.
      • Hoffman G.E.
      • Le W.W.
      • Entezam A.
      • Otsuka N.
      • Tong Z.B.
      • Nelson L.
      • Flaws J.A.
      • McDonald J.H.
      • Jafar S.
      • Usdin K.
      Ovarian abnormalities in a mouse model of fragile X primary ovarian insufficiency.
      • Gallagher A.
      • Hallahan B.
      Fragile X-associated disorders: a clinical overview.
      • Hagerman P.
      Fragile X-associated tremor/ataxia syndrome (FXTAS): pathology and mechanisms.
      According to the American College of Medical Genetics, there are four FMR1 allelic forms: normal (NL), with 5 to 44 repeats; gray zone (GZ), with 45 to 54 repeats; premutation (PM), with 55 to 200 repeats; and full mutation (FM), with >200 repeats.
      • Monaghan K.G.
      • Lyon E.
      • Spector E.B.
      American College of Medical Genetics and Genomics
      ACMG Standards and Guidelines for fragile X testing: a revision to the disease-specific supplements to the Standards and Guidelines for Clinical Genetics Laboratories of the American College of Medical Genetics and Genomics.
      Persons with a FM allele of >200 CGG repeats usually have intellectual disabilities and present with symptoms of autism spectrum disorder.
      • Gallagher A.
      • Hallahan B.
      Fragile X-associated disorders: a clinical overview.
      • de Vries B.
      • Halley D.
      • Oostra B.
      • Niermeijer M.
      The fragile X syndrome.
      Other phenotypic presentations pertain to behavior, cognition, hyperactivity, language deficits, aggressions, and psychiatric problems such as social anxiety and withdrawal. Physical traits such as an elongated face, prominent ears, flat feet, and macroorchidism in puberty are also apparent.
      • Tassone F.
      Newborn screening for fragile X syndrome.
      • Gallagher A.
      • Hallahan B.
      Fragile X-associated disorders: a clinical overview.
      • Loesch D.
      • Hagerman R.
      Unstable mutations in the FMR1 gene and the phenotypes.
      Male and female persons with FMR1 PM alleles (55 to 200 CGG repeats) are carriers because they do not present symptoms of FXS. However, PM alleles are meiotically unstable and can expand to FM within one generation through females.
      • Cronister A.
      • Teicher J.
      • Rohlfs E.M.
      • Donnenfeld A.
      • Hallam S.
      Prevalence and instability of fragile X alleles: implications for offering fragile X prenatal diagnosis.
      • Nolin S.L.
      • Sah S.
      • Glicksman A.
      • Sherman S.L.
      • Allen E.
      • Berry-Kravis E.
      • Tassone F.
      • Yrigollen C.
      • Cronister A.
      • Jodah M.
      • Ersalesi N.
      • Dobkin C.
      • Brown W.T.
      • Shroff R.
      • Latham G.J.
      • Hadd A.G.
      Fragile X AGG analysis provides new risk predictions for 45-69 repeat alleles.
      Initial concerns for carriers, especially females, were focused on the intergenerational instability of expanded alleles rather than clinical presentations.
      • de Vries B.
      • Halley D.
      • Oostra B.
      • Niermeijer M.
      The fragile X syndrome.
      Accumulated evidence later found that PM carriers may develop fragile X-associated conditions such as fragile X–associated tremor/ataxia syndrome
      • Hagerman P.
      Fragile X-associated tremor/ataxia syndrome (FXTAS): pathology and mechanisms.
      and fragile X–associated primary ovarian insufficiency, a condition defined as menopause before 40 years of age.
      • Cronister A.
      • Schreiner R.
      • Wittenberger M.
      • Amiri K.
      • Harris K.
      • Hagerman R.J.
      Heterozygous fragile X female: historical, physical, cognitive, and cytogenetic features.
      Female PM carriers have an estimated 20% risk of fragile X–associated primary ovarian insufficiency,
      • Loesch D.
      • Hagerman R.
      Unstable mutations in the FMR1 gene and the phenotypes.
      whereas approximately 8% of female PM carriers and 40% of male PM carriers develop fragile X–associated tremor/ataxia syndrome.
      • Jacquemont S.
      • Hagerman R.J.
      • Leehey M.A.
      • Hall D.A.
      • Levine R.A.
      • Brunberg J.A.
      • Zhang L.
      • Jardini T.
      • Gane L.W.
      • Harris S.W.
      • Herman K.
      • Grigsby J.
      • Greco C.M.
      • Berry-Kravis E.
      • Tassone F.
      • Hagerman P.J.
      Penetrance of the fragile X-associated tremor/ataxia syndrome in a premutation carrier population.
      Some studies now suggest that PM carriers may have various degrees of learning disabilities and anxiety disorders.
      • Bourgeois J.A.
      • Seritan A.L.
      • Casillas E.M.
      • Hessl D.
      • Schneider A.
      • Yang Y.
      • Kaur I.
      • Cogswell J.B.
      • Nguyen D.V.
      • Hagerman R.J.
      Lifetime prevalence of mood and anxiety disorders in fragile X premutation carriers.
      • Grigsby J.
      • Brega A.G.
      • Engle K.
      • Leehey M.A.
      • Hagerman R.J.
      • Tassone F.
      • Hessl D.
      • Hagerman P.J.
      • Cogswell J.B.
      • Bennett R.E.
      • Cook K.
      • Hall D.A.
      • Bounds L.S.
      • Paulich M.J.
      • Reynolds A.
      Cognitive profile of fragile X premutation carriers with and without fragile X-associated tremor/ataxia syndrome.
      • Hagerman R.
      • Hagerman P.
      Advances in clinical and molecular understanding of the FMR1 premutation and fragile X-associated tremor/ataxia syndrome.
      The consequence of carrying the FMR1 GZ allele is less apparent than with PM and FM alleles.
      • Tassone F.
      Newborn screening for fragile X syndrome.
      Recent findings suggest that persons with FMR1 GZ alleles may have increased risks of developing adult-onset fragile X-associated conditions.
      • Loesch D.Z.
      • Tassone F.
      • Lo J.
      • Slater H.R.
      • Hills L.V.
      • Bui M.Q.
      • Silburn P.A.
      • Mellick G.D.
      New evidence for, and challenges in, linking small CGG repeat expansion FMR1 alleles with Parkinson's disease.
      However, more extensive studies are necessary to establish the effects of GZ alleles.
      Early diagnosis of FXS at a young age enables timely therapeutic interventions that can considerably improve quality of life.
      • Tassone F.
      Newborn screening for fragile X syndrome.
      • Murray J.
      • Cuckle H.
      • Taylor G.
      • Hewison J.
      Screening for fragile X syndrome.
      Because of the defined phenotypic presentations associated with various FMR1 allelic forms, molecular diagnoses of FXS and associated disorders are enabled by detecting CGG expansions in FMR1.
      • Murray J.
      • Cuckle H.
      • Taylor G.
      • Hewison J.
      Screening for fragile X syndrome.
      • Oostra B.A.
      • Jacky P.B.
      • Brown W.T.
      • Rousseau F.
      Guidelines for the diagnosis of fragile X syndrome. National Fragile X Foundation.
      Existing approaches for detecting such aberrant expansions include Southern blot analysis and various permutations of PCR tests.
      • Chen L.
      • Hadd A.
      • Sah S.
      • Filipovic-Sadic S.
      • Krosting J.
      • Sekinger E.
      • Pan R.
      • Hagerman P.J.
      • Stenzel T.T.
      • Tassone F.
      • Latham G.J.
      An information-rich CGG repeat primed PCR that detects the full range of fragile X expanded alleles and minimizes the need for southern blot analysis.
      • Monaghan K.G.
      • Lyon E.
      • Spector E.B.
      American College of Medical Genetics and Genomics
      ACMG Standards and Guidelines for fragile X testing: a revision to the disease-specific supplements to the Standards and Guidelines for Clinical Genetics Laboratories of the American College of Medical Genetics and Genomics.
      • Oostra B.A.
      • Jacky P.B.
      • Brown W.T.
      • Rousseau F.
      Guidelines for the diagnosis of fragile X syndrome. National Fragile X Foundation.
      • Lyon E.
      • Laver T.
      • Yu P.
      • Jama M.
      • Young K.
      • Zoccoli M.
      • Marlowe N.
      A simple, high-throughput assay for Fragile X expanded alleles using triple repeat primed PCR and capillary electrophoresis.
      Southern blot analysis is notoriously labor-intensive, time-consuming, and requires large amounts of DNA. The workflow is not optimized for high-throughput testing because limited numbers of samples can be processed simultaneously.
      • Chen L.
      • Hadd A.
      • Sah S.
      • Filipovic-Sadic S.
      • Krosting J.
      • Sekinger E.
      • Pan R.
      • Hagerman P.J.
      • Stenzel T.T.
      • Tassone F.
      • Latham G.J.
      An information-rich CGG repeat primed PCR that detects the full range of fragile X expanded alleles and minimizes the need for southern blot analysis.
      • Filipovic-Sadic S.
      • Sah S.
      • Chen L.
      • Krosting J.
      • Sekinger E.
      • Zhang W.
      • Hagerman P.J.
      • Stenzel T.T.
      • Hadd A.G.
      • Latham G.J.
      • Tassone F.
      A novel FMR1 PCR method for the routine detection of low abundance expanded alleles and full mutations in fragile X syndrome.
      As a result, various PCR-based methods were developed over the past decade as alternatives to Southern blot analysis.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      • Zhou Y.
      • Law H.Y.
      • Boehm C.D.
      • Yoon C.S.
      • Cutting G.R.
      • Ng I.S.
      • Chong S.S.
      Robust fragile X (CGG)n genotype classification using a methylation specific triple PCR assay.
      PCR approaches designed to probe for CGG repeat expansions in FMR1, regardless of permutation, are typically either conventional PCR or triplet repeat-primed PCR (TP-PCR). Conventional PCR uses primers that target the flanking sequence of the CGG repeat region.
      • Oostra B.A.
      • Jacky P.B.
      • Brown W.T.
      • Rousseau F.
      Guidelines for the diagnosis of fragile X syndrome. National Fragile X Foundation.
      • Filipovic-Sadic S.
      • Sah S.
      • Chen L.
      • Krosting J.
      • Sekinger E.
      • Zhang W.
      • Hagerman P.J.
      • Stenzel T.T.
      • Hadd A.G.
      • Latham G.J.
      • Tassone F.
      A novel FMR1 PCR method for the routine detection of low abundance expanded alleles and full mutations in fragile X syndrome.
      • Haddad L.A.
      • Mingroni-Netto R.C.
      • Vianna-Morgante A.M.
      • Pena S.D.
      A PCR-based test suitable for screening for fragile X syndrome among mentally retarded males.
      TP-PCR includes flanking and additional primers that target within the trinucleotide repeat region.
      • Chen L.
      • Hadd A.
      • Sah S.
      • Filipovic-Sadic S.
      • Krosting J.
      • Sekinger E.
      • Pan R.
      • Hagerman P.J.
      • Stenzel T.T.
      • Tassone F.
      • Latham G.J.
      An information-rich CGG repeat primed PCR that detects the full range of fragile X expanded alleles and minimizes the need for southern blot analysis.
      • Oostra B.A.
      • Jacky P.B.
      • Brown W.T.
      • Rousseau F.
      Guidelines for the diagnosis of fragile X syndrome. National Fragile X Foundation.
      • Warner J.P.
      • Barron L.H.
      • Goudie D.
      • Kelly K.
      • Dow D.
      • Fitzpatrick D.R.
      • Brock D.J.
      A general method for the detection of large CAG repeat expansions by fluorescent PCR.
      Such PCR approaches have advantages and inherent limitations compared with Southern blot analysis. For instance, conventional PCR is ineffective in amplifying large CGG repeats because of the repetitive nature of the region and the high GC-rich content. The reaction often fails beyond a certain size range, resulting in nonamplification of large PM or FM alleles.
      • Oostra B.A.
      • Jacky P.B.
      • Brown W.T.
      • Rousseau F.
      Guidelines for the diagnosis of fragile X syndrome. National Fragile X Foundation.
      • Filipovic-Sadic S.
      • Sah S.
      • Chen L.
      • Krosting J.
      • Sekinger E.
      • Zhang W.
      • Hagerman P.J.
      • Stenzel T.T.
      • Hadd A.G.
      • Latham G.J.
      • Tassone F.
      A novel FMR1 PCR method for the routine detection of low abundance expanded alleles and full mutations in fragile X syndrome.
      • Haddad L.A.
      • Mingroni-Netto R.C.
      • Vianna-Morgante A.M.
      • Pena S.D.
      A PCR-based test suitable for screening for fragile X syndrome among mentally retarded males.
      In fact, such a shortcoming known as the null-allele was proposed as a rapid negative identification of CGG expansions in male samples.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      Because of the nonamplification of larger alleles in conventional PCR, heterozygous female specimens with one NL allele and a second expanded allele that does not amplify (one normal band after PCR) cannot be distinguished from homozygous female samples with two NL alleles (one normal band).
      • Lyon E.
      • Laver T.
      • Yu P.
      • Jama M.
      • Young K.
      • Zoccoli M.
      • Marlowe N.
      A simple, high-throughput assay for Fragile X expanded alleles using triple repeat primed PCR and capillary electrophoresis.
      Additional testing is unavoidable in such situations; exclusion of the presence of an expanded allele is necessary because approximately 40% females are homozygous NL.
      • Lyon E.
      • Laver T.
      • Yu P.
      • Jama M.
      • Young K.
      • Zoccoli M.
      • Marlowe N.
      A simple, high-throughput assay for Fragile X expanded alleles using triple repeat primed PCR and capillary electrophoresis.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      TP-PCR overcomes the inadequacies of conventional PCR by priming from within the CGG repeat region and enabling amplification of large PM or FM alleles. Although full-length amplicons of large expansions may not be obtained, amplicons up to a certain size range will still be generated and subsequently analyzed.
      Regardless of approach, both conventional PCR and TP-PCR require coupling with either gel or capillary electrophoresis to resolve amplicons of varying sizes.
      • Lyon E.
      • Laver T.
      • Yu P.
      • Jama M.
      • Young K.
      • Zoccoli M.
      • Marlowe N.
      A simple, high-throughput assay for Fragile X expanded alleles using triple repeat primed PCR and capillary electrophoresis.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      In theory, existing PCR approaches are considered screening tools that reduce reliance on Southern blot analysis.
      • Chen L.
      • Hadd A.
      • Sah S.
      • Filipovic-Sadic S.
      • Krosting J.
      • Sekinger E.
      • Pan R.
      • Hagerman P.J.
      • Stenzel T.T.
      • Tassone F.
      • Latham G.J.
      An information-rich CGG repeat primed PCR that detects the full range of fragile X expanded alleles and minimizes the need for southern blot analysis.
      • Lyon E.
      • Laver T.
      • Yu P.
      • Jama M.
      • Young K.
      • Zoccoli M.
      • Marlowe N.
      A simple, high-throughput assay for Fragile X expanded alleles using triple repeat primed PCR and capillary electrophoresis.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      In practice, however, a true screening tool with a PCR-based workflow should ideally be streamlined and not require additional post-PCR processing or expensive equipment.
      Here, we validate the FastFraX FMR1 Identification kit (the FastFraX ID kit) as a screening tool for the rapid detection of CGG expansions in the FMR1 gene. The underlying technology was developed on the basis of a previous study by Teo et al.
      • Teo C.R.
      • Law H.Y.
      • Lee C.G.
      • Chong S.S.
      Screening for CGG repeat expansion in the FMR1 gene by melting curve analysis of combined 5' and 3' direct triplet-primed PCRs.
      The assay uses a direct TP-PCR, coupled with melting curve analysis (MCA) for detecting CGG expansions. The melting temperatures of the PCR amplicons are determined from the melt curve profile and depend on amplicon size. Through a simple MCA and without post-PCR processing, samples are rapidly classified as having non-expanded (<55 CGG repeats) or expanded (>55 CGG repeats) alleles. Our study aims to verify the accuracy and effectiveness of the TP-PCR with MCA approach as a first-line PCR-only screening tool in detecting CGG repeat expansions in clinical samples.

      Materials and Methods

       DNA Samples

      Thirty-eight individual cell line-derived DNA reference samples with various CGG repeat lengths were acquired from the Coriell Cell Repositories (Camden, NJ) (Table 1). Of these, three male DNA samples (GM06890, GM20244, and GM20230) with 30, 41, and 53 FMR1 CGG repeats were used for the initial assessment as recommended controls for the TP-PCR assay with MCA. The study included testing the analytic sensitivity, analytic specificity, and consistency of the assay when used on different PCR assay platforms. The genotypes of all Coriell DNA samples tested in this study are listed in Table 1. All 38 samples (Table 1) were used in an initial blinded (G.X.Y.L.) small-scale study to evaluate performance of the TP-PCR assay.
      Table 1Genotypes of DNA Samples from Coriell Cell Repositories Used in Blinded Study (n = 35)
      Coriell sample IDSexCGG repeats reported by Coriell, n
      Allele 1Allele 2
      Normal alleles (≤30 CGG repeats)
       GM06895M23
       GM07538F2929
       GM06904F2923
       GM07543F2920
       GM06890
      Samples are used as cutoff reference controls.
      M30
       GM06911F30NL
       GM06889F3023
      High normal alleles (30–44 CGG repeats)
       GM20244
      Samples are used as cutoff reference controls.
      M41
       GM20243F4129
      Gray zone alleles(45–54 CGG repeats)
       GM20232M46
       GM20234F4631
       GM20230
      Samples are used as cutoff reference controls.
      M53
       GM20236F5331
      Premutation/full mutation alleles (≥55 CGG repeats)
       CD00014M56
       GM06905F7023
       GM20242F7330
       GM06910F75–8930
       GM20231M76
       GM06894F7830
       GM20240F8030
       GM06892M80–85
       GM06907F8529
       GM06906M85–90
       GM20241F93–11029
       GM06903F9523
       GM06896F95–14023
       GM06968F10723
       GM20237M100–104
       GM06891M100–117
       GM20233M117
       GM20239F183–19320
       GM06852M>200
       GM07537F>20028–29
       GM06897M477
       GM07862M501–550
       GM04025M645
       GM05847F∼65021
       GM09237M931–940
      F, female; M, male; NL, normal.
      Samples are used as cutoff reference controls.
      The analytic specificity of the TP-PCR assay was evaluated with a DNA sample (GM23378 from Coriell Cell Repositories) of no direct relevance to FXS to test for potential interference. GM23378 (Coriell Cell Repositories) harbors a CTG repeat expansion in the dystrophia myotonica-protein kinase gene.
      To evaluate the sensitivity of the TP-PCR assay to detect mosaicism, a male NL sample (GM06890; 30 CGG repeats) was mixed with a male PM sample (GM06906; 85 to 90 CGG repeats) or a FM sample (GM06852; >200 CGG repeats) to create simulated mosaic content of different concentrations. These samples were mixed in various proportions, yet maintained at a total DNA input of 50 ng per reaction. The resulting simulated mosaic samples contained 1%, 2.5%, 5%, 7.5%, 10%, 20%, 50%, and 100% of the PM or FM sample. Melt curve profiles of the simulated mosaic samples were compared with that of 53 CGG repeat control (GM20230) and 30 CGG repeat control (GM06890).

       DNA Samples from Clinical Archive

      The performance of the TP-PCR assay with MCA was ultimately validated in a blinded (G.X.Y.L. and Y.L.L.) study of 528 archived clinical samples from an Indonesian population with intellectual disabilities. These genomic DNA samples were originally isolated from whole blood by using salting out methods as described,
      • Miller S.A.
      • Dykes D.D.
      • Polesky H.F.
      A simple salting out procedure for extracting DNA from human nucleated cells.
      with slight modifications. The samples were previously characterized for FMR1 CGG repeat length by using a combination of the following methods: a two-primer conventional (flanking) PCR,
      • Fu Y.H.
      • Kuhl D.P.
      • Pizzuti A.
      • Pieretti M.
      • Sutcliffe J.S.
      • Richards S.
      • Verkerk A.J.
      • Holden J.J.
      • Fenwick Jr., R.G.
      • Warren S.T.
      • et al.
      Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox.
      followed by fragment length analysis on an ABI Prism 3730 DNA Analyzer (Life Technologies, Carlsbad, CA) with the GeneMapper version 4.0 (Apache, Los Angeles, CA) or Southern blot analysis.
      • Mundhofir F.E.
      • Winarni T.I.
      • Nillesen W.M.
      • van Bon B.W.
      • Schepens M.
      • Ruiterkamp-Versteeg M.
      • Hamel B.C.
      • Yntema H.G.
      • Faradz S.M.
      Prevalence of fragile X syndrome in males and females in Indonesia.
      For samples with NL or PM alleles, exact CGG repeat lengths were determined with a combination of PCR and fragment analysis. For samples with results indicative of a FM, or for female samples that produced a single PCR product, Southern blot analysis was performed to confirm the FM or to differentiate between female heterozygous and homozygous samples.
      • Mundhofir F.E.
      • Winarni T.I.
      • Nillesen W.M.
      • van Bon B.W.
      • Schepens M.
      • Ruiterkamp-Versteeg M.
      • Hamel B.C.
      • Yntema H.G.
      • Faradz S.M.
      Prevalence of fragile X syndrome in males and females in Indonesia.
      In a few cases, the confirmation was based on TP-PCR as previously described.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      The frequency of CGG repeat lengths in the archived cohort was tabulated on the basis of the collective results generated by the aforementioned approaches (Figure 1). Genomic DNA concentrations were re-quantified with a NanoVue Plus spectrophotometer (GE Healthcare, Little Chalfont, UK). Usage of these archived clinical samples for the present study was approved by the Medical Ethical Committee of the Faculty of Medicine (Diponegoro University, Semarang, Indonesia). All samples were stored at −20°C until use.
      Figure thumbnail gr1
      Figure 1Frequency of CGG repeat lengths in the clinical archive cohort of 528 samples. In samples with two alleles such as female heterozygous specimens or male mosaic samples, only the larger allele is considered in the analysis. FMR1, fragile X mental retardation 1 gene.

       TP-PCR Assay with MCA

      The FastFraX ID kits (labeled For Research Use) were obtained from the Biofactory Pte Ltd (Singapore, Republic of Singapore). The kit was developed on the basis of a previous study.
      • Teo C.R.
      • Law H.Y.
      • Lee C.G.
      • Chong S.S.
      Screening for CGG repeat expansion in the FMR1 gene by melting curve analysis of combined 5' and 3' direct triplet-primed PCRs.
      It uses a direct TP-PCR, which includes a combination of primers that target a flanking region and from within the CGG repeats. The primer mix was designed to amplify from the 3′ end of the FMR1 CGG repeat region of non-modified genomic DNA.
      • Teo C.R.
      • Law H.Y.
      • Lee C.G.
      • Chong S.S.
      Screening for CGG repeat expansion in the FMR1 gene by melting curve analysis of combined 5' and 3' direct triplet-primed PCRs.
      The required PCR mixture and DNA polymerase were also included in the kit.
      TP-PCR assays were performed according to the manufacturer's instructions in 25-μL volumes with 50 ng of genomic DNA per test. The PCR assays that assessed analytic performance were conducted in triplicates. Unless otherwise specified, PCR assays and subsequent MCAs were performed with the Rotor-Gene Q HRM (Qiagen, Hilden, Germany) at the Center for Biomedical Research of Diponegoro University. The thermal cycling conditions comprised an initial denaturation step at 95°C for 15 minutes, followed by 40 cycles of 99°C for 2 minutes, 65°C for 2 minutes, 72°C for 3 minutes, and then a final extension step at 72°C for 10 minutes. To study the potential variations of different PCR systems, a few sets of assays were also performed under identical thermal cycling conditions on the LightCycler 480 (Roche Diagnostics, Mannheim, Germany) or ABI 7500 Fast Real-Time PCR system (Life Technologies).
      PCR amplicons were automatically melted after thermal cycling for MCA. The MCA parameters on the Rotor-Gene Q HRM comprised a denaturation step at 95°C for 1 minute and a temperature ramp from 75°C to 99°C at a rate of 0.5°C with a 5-second hold at each step. The MCA parameters on the LightCycler 480 comprised a denaturation step at 95°C for 1 minute and a temperature ramp from 60°C to 99°C at a rate of 0.01°C/second with 50 acquisitions per °C. The MCA parameters on the ABI 7500 Fast comprised a denaturation step at 95°C for 1 minute and a temperature ramp from 60°C to 99.9°C at a ramp percentage of 0.5%.

       Data Analysis and Result Interpretation

      On completion of the MCA, melt curve profiles were automatically generated with the negative first derivative of fluorescence plotted against temperature. Data were analyzed with software from the manufacturer of the real-time PCR platform, following the manufacturer's instructions. Briefly, the generated melt curve profiles were visually checked for aberrant results (eg, significantly lower signals than the controls). The temperatures at which baseline negative first derivative of fluorescence versus temperature resumed (resumed baseline −dF/dT temperature) were extracted from the melt curve profiles of cutoff controls. Samples were classified as nonexpanded, expanded, or indeterminate on the basis of their respective resumed baseline −dF/dT temperatures relative to that of the cutoff control used. The classification data for the clinical archived samples were sorted with Microsoft Excel 2010 (Microsoft Corp, Redman, WA). Indeterminate and expanded samples warrant further testing and are considered positive detections by the TP-PCR assay.

      Results

       Analytic Sensitivity

      Four Coriell DNA samples representing the male NL, GZ, PM, and FM genotypes (GM06890, GM20230, GM06906, and GM06852, respectively) were used to examine the analytic sensitivity of the TP-PCR with MCA assay. The amount of genomic DNA per test (reaction) recommended by the manufacturer is 50 ng. To determine the detection limit of the TP-PCR assay, we varied the amount of input DNA per reaction at 10, 25, 50, and 100 ng. We did not observe significant variations in the melt curve profiles with different amounts of DNA input (Figure 2A). Furthermore, the temperature at which baseline −dF/dT temperature levels resumed was consistent across the range of DNA input levels for all four Coriell male DNA samples. With the lowest input of genomic DNA tested at 10 ng, melt curve profiles of the PM and FM genotypes remained clearly distinguishable from the NL sample.
      Figure thumbnail gr2
      Figure 2Melt curve profiles that show analytic sensitivity of the triplet repeat-primed PCR assay with melting curve analysis on the ABI 7500 Fast real-time PCR platform. Melt curve profiles of male (A) and female (B) Coriell DNA samples that cover four FMR1 allelic forms tested in varying amounts, from 10 ng to 100 ng per reaction. Each reaction was conducted in triplicates; representative profiles are shown. The resumed baseline −dF/dT temperature cutoffs refer to the GZ samples for each sex: GM20230 (A) and GM20236 (B). FM, full mutation; FMR1, fragile X mental retardation 1 gene; GZ, gray zone; NL, normal; NTC, no-template control; PM, premutation; resumed baseline −dF/dT temperature, temperature at which baseline negative first derivative of fluorescence versus temperature resumed.
      Similarly, the analytic sensitivity of the assay was tested with four Coriell female DNA samples (GM07538, GM20236, GM06906, and GM07537, respectively) representing distinct genotypes (Figure 2B). Although the melt curve profile of GM07538 at 100 ng had a lower signal, the corresponding resumed baseline −dF/dT temperature was not altered significantly (Figure 2B). Regardless of DNA input amount, the resumed baseline −dF/dT temperatures of the PM and FM samples were distinct from the NL and GZ samples. The results indicated that the TP-PCR assay with MCA has a detection limit of 10 ng (or lower) per test.

       Analytic Specificity

      The analytic specificity of the TP-PCR assay was assessed with a non-relevant DNA sample (GM23378) to gauge if nucleic acids distinct from CGG repeat expansions in FMR1 will affect kit performance. A Coriell male DNA sample (GM23378) that harbored a CTG repeat expansion in the dystrophia myotonica-protein kinase gene (DMPK) was added in increasing amounts (0, 50, 100, and 150 ng) to 50 ng each of four Coriell DNA samples, representing the female NL, GZ, PM, and FM genotypes (GM07538, GM20236, GM06907, and GM07537, respectively). With increasing amounts of non-relevant DNA, the melt curve profiles of all four samples tested did not exhibit significant variation in terms of amplitude, shape, and resumed baseline −dF/dT temperatures (Figure 3). Importantly, the NL and GZ samples (GM07538 and GM20236) produced a melt curve profile distinct from the expanded samples (GM06907 and GM07537) (Figure 3) even with 150 ng of potentially cross-reactive DNA (GM23378). The results indicated that the TP-PCR assay with MCA approach can differentiate expanded FMR1 alleles from non-expanded alleles in the presence of other non-relevant DNA species.
      Figure thumbnail gr3
      Figure 3Melt curve profiles that show the analytic specificity of the triplet repeat-primed PCR assay with melting curve analysis on the ABI 7500 Fast real-time PCR platform. Melt curve profiles of four female DNA samples from Coriell spiked with increasing amounts of non-relevant DNA (GM23378) that harbor expanded CTG repeats in the DMPK gene. Zero, 50, 100, or 150 ng of GM23378 was added to 50 ng of each female DNA sample in each reaction. Each reaction was conducted in triplicates; representative profiles are shown. The resumed baseline −dF/dT temperature cutoff refers to the GZ sample, GM20236. DMPK, dystrophia myotonica-protein kinase gene; FM, full mutation; GZ, gray zone; NL, normal; NTC, no-template control; PM, premutation; resumed baseline −dF/dT temperature, temperature at which baseline negative first derivative of fluorescence versus temperature resumed.

       Consistency

      According to the manufacturer's recommendation, the TP-PCR assay required inclusion of an appropriate control sample to serve as a cutoff reference for the temperature at which the test sample resumes baseline −dF/dT temperature. The consistency of this temperature parameter was evaluated with three Coriell DNA samples that are recommended controls for the TP-PCR assay. The number of FMR1 CGG repeats in male DNA samples GM06890, GM20244, and GM20230 are 30, 41, and 53 repeats, respectively. The resumed baseline −dF/dT temperatures of these samples were obtained from a total of nine valid runs performed on the Rotor-Gene Q HRM system at two sites by two operators. The average resumed baseline −dF/dT temperatures were distinct among the three samples (Figure 4A); replicates produced consistent results with CV < 0.5% (Figure 4B).
      Figure thumbnail gr4
      Figure 4Melt curve profiles that show consistency of the resumed baseline −dF/dT temperature as a parameter across PCR platforms. A: Melt curve profiles of potential cutoff controls: GM06890 (30 CGG repeats), GM20244 (41 CGG repeats), and GM20230 (53 CGG repeats) on the Rotor-Gene Q HRM platform. B: Average resumed baseline −dF/dT temperatures of GM06890, GM20244, and GM20230 obtained on the Rotor-Gene Q HRM platform across nine valid runs operated independently at two sites. C: Intrarun average resumed baseline −dF/dT temperatures of GM06890, GM20244, and GM20230 on the ABI and the Rotor-Gene Q HRM platforms. Data are expressed as means ± SD. n = 9 (B); n = 96 (C, ABI); n = 12 (C, RGQ). ABI, ABI 7500 Fast Real-Time PCR; resumed baseline −dF/dT temperature, temperature at which baseline negative first derivative of fluorescence versus temperature resumed; RGQ, Rotor-Gene Q HRM platform.
      We further tested the intrarun consistency of the resumed baseline −dF/dT temperatures by using different recommended platforms, namely the ABI 7500 Fast Real-time PCR platform and the Qiagen Rotor-Gene Q HRM. We performed the assay by using the same three potential controls with 30, 41, and 53 CGG repeats. The average resumed baseline −dF/dT temperatures were obtained for 96 replicates of each sample on the ABI 7500 Fast Real-time PCR platform and for 12 replicates of each sample on the Rotor-Gene Q HRM (Figure 4C and Table 2). The average resumed baseline −dF/dT temperatures were consistent within the same run on each platform, giving a CV < 0.55% (Table 2).
      Table 2Resumed Baseline −dF/dT Temperatures of Controls Determined with Different PCR Platforms
      PlatformCGG repeats, nAverage resumed baseline temperature, °CCV, %
      ABI 7500 Fast (n = 96)3089.70.41
      4190.20.54
      5392.40.51
      Rotor-Gene Q HRM (n = 12)3090.20.11
      4191.10.45
      5392.80.18
      Resumed baseline −dF/dT temperature, temperature at which baseline negative first derivative of fluorescence versus temperature resumed.

       Detection of Mosaic Samples

      The performance of the TP-PCR assay with MCA in detecting mosaicism was initially evaluated with simulated samples and was subsequently verified with clinical samples (see Detection of Clinical Mosaic Samples). The simulated samples were generated by mixing a male normal sample (GM06890, 30 CGG repeats) with a male PM or FM sample (GM06906, 85 to 90 CGG repeats; GM06852, >200 CGG repeats). Samples were mixed in various proportions, yet maintained at a total DNA input of 50 ng per reaction. The simulated samples contained 1%, 2.5%, 5%, 7.5%, 10%, 20%, 50%, and 100% of the expanded sample. Melt curve profiles of the simulated samples were generated with the Rotor-Gene Q HRM and compared with that of a control sample harboring either 30 or 53 CGG repeats (Figure 5).
      Figure thumbnail gr5
      Figure 5Melt curve profiles of simulated mosaic samples relative to a 53 CGG control (GM20230). Melt curve profiles of a male normal sample (GM06890) mixed with a male PM sample (GM06906) (A) or a male FM sample (GM06852) (B) in varying proportions while maintaining total DNA input at 50 ng. Resumed baseline −dF/dT temperatures are compared with a 53 CGG control (GM20230) and a 30 CGG sample (GM06890). All reactions are performed in triplicates on the Rotor-Gene Q HRM; representative profiles are shown. A: 100% to 7.5% PM mosaic samples are in the top panel, whereas 5% to 1% PM mosaic samples are in the lower panel. B: 100% to 20% FM mosaic samples are in the top panel, whereas 10% to 1% FM mosaic samples are in the lower panel. FM, full mutation; PM, premutation; resumed baseline −dF/dT temperature, temperature at which baseline negative first derivative of fluorescence versus temperature resumed.
      We observed that the melt curve profiles of the 7.5% to 100% PM-simulated mosaic samples gave distinct and higher resumed baseline −dF/dT temperatures than the 53 CGG repeat control (Figure 5A). The 2.5% and 5% PM mosaic samples had resumed baseline −dF/dT temperatures that overlapped with the 53 CGG repeat control and yet were distinct from the 30 CGG repeat normal sample (GM06890) (Figure 5A). We also observed that the melt curve profiles of the 20% to 100% FM mosaic samples gave higher resumed baseline −dF/dT temperatures than the 53 CGG repeat control (Figure 5B). The 5% to 10% FM mosaic samples had resumed baseline −dF/dT temperatures overlapping with the 53 CGG repeat control and yet were distinct from the 30 CGG repeat normal samples (GM06890) (Figure 5B). The results indicated that the TP-PCR assay with MCA detected expanded FMR1 alleles in the simulated mosaic samples at levels as low as 7.5% but no less than 20% when using a 53 CGG repeat control.

       Kit Performance Evaluated with Reference Samples

      The performance of TP-PCR assay with MCA was evaluated in a blinded (G.X.Y.L.) study by using 35 Coriell reference DNA samples on the Rotor-Gene Q HRM platform. The genotypes of the DNA samples tested are listed in Table 1. Three male samples with 30, 41, and 53 CGG repeats (GM06890, GM20244, and GM20230, respectively) were designated as cutoff reference controls for result interpretation.
      Our test results were summarized and compared with CGG repeat data provided by Coriell Cell Repositories (Table 3). As expected, the performance of the TP-PCR assay in detecting CGG repeat expansions depended on the choice of cutoff control (GM06890, GM20244, or GM20230) used for result interpretation. Samples classified as expanded have alleles of >30, >41, and >53 CGG repeats, respectively. The agreements between our test results and data provided by Coriell were 89%, 97%, and 94% when GM06890 (30 CGG repeats), GM20244 (41 CGG repeats), and GM20230 (53 CGG repeats) were used, respectively, as the cutoff control (Table 3). Notably, most of the disagreements derived from false detections by the assay, especially when GM06890 (30 CGG repeats) and GM20244 (41 CGG repeats) were used as the control (Table 3). The false detections arose from samples with CGG repeat lengths close to (within one repeat) those of the respective controls. The only disagreement due to a false negative when GM20230 (53 CGG repeats) was used as the control derived from a male sample with 56 CGG repeats, which marginally met the criterion for PM.
      Table 3Classification of Reference DNA Samples Using TP-PCR Assay with MCA in Comparison with Characterized Genotypes from Coriell
      Classification using TP-PCR + MCAGenotype from Coriell cell repositories (n = 35)Agreement, %
      Percentage was determined on the basis of the fraction of samples undetected (nonexpanded) and detected (expanded and indeterminate) that was in agreement with the Coriell-reported genotypes with the different cutoff controls.
      Normal (≤30 CGG repeats)High normal (30–45 CGG repeats)Gray zone (45–55 CGG repeats)Pre-/Full mutation (≥55 CGG repeats)
      Cutoff control: GM20230 (53 CGG repeats)
       Nonexpanded (≤53 CGG repeats)6121
       Indeterminate
      A sample classified as indeterminate warrants further testing and is considered a positive detection.
      000094.2
       Expanded (>53 CGG repeats)00124
      Cutoff control: GM20244 (41 CGG repeats)
       Nonexpanded (≤41 CGG repeats)6000
       Indeterminate
      A sample classified as indeterminate warrants further testing and is considered a positive detection.
      000097.1
       Expanded (>41 CGG repeats)01325
      Cutoff control: GM06890 (30 CGG repeats)
       Nonexpanded (≤30 CGG repeats)2000
       Indeterminate
      A sample classified as indeterminate warrants further testing and is considered a positive detection.
      100088.6
       Expanded (>30 CGG repeats)31325
      Total samples per class, n61325
      MCA, melting curve analysis; TP-PCR, triplet repeat-primed PCR.
      Percentage was determined on the basis of the fraction of samples undetected (nonexpanded) and detected (expanded and indeterminate) that was in agreement with the Coriell-reported genotypes with the different cutoff controls.
      A sample classified as indeterminate warrants further testing and is considered a positive detection.

       Kit Performance Validated with Clinical Samples

      The performance of the TP-PCR assay with MCA was ultimately evaluated in a blinded (G.X.Y.L. and Y.L.L.) and retrospective study by using genomic DNA derived from 528 patient whole blood samples from a population with intellectual disabilities.
      • Mundhofir F.E.
      • Winarni T.I.
      • van Bon B.W.
      • Aminah S.
      • Nillesen W.M.
      • Merkx G.
      • Smeets D.
      • Hamel B.C.
      • Faradz S.M.
      • Yntema H.G.
      A cytogenetic study in a large population of intellectually disabled Indonesians.
      These clinical samples were previously characterized for their FMR1 CGG repeat size by using a combination of flanking PCR,
      • Fu Y.H.
      • Kuhl D.P.
      • Pizzuti A.
      • Pieretti M.
      • Sutcliffe J.S.
      • Richards S.
      • Verkerk A.J.
      • Holden J.J.
      • Fenwick Jr., R.G.
      • Warren S.T.
      • et al.
      Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox.
      TP-PCR, and/or Southern Blot analysis.
      • Mundhofir F.E.
      • Winarni T.I.
      • Nillesen W.M.
      • van Bon B.W.
      • Schepens M.
      • Ruiterkamp-Versteeg M.
      • Hamel B.C.
      • Yntema H.G.
      • Faradz S.M.
      Prevalence of fragile X syndrome in males and females in Indonesia.
      Again, three male samples from Coriell Cell Repositories with 30, 41, and 53 CGG repeats (GM06890, GM20244, and GM20230, respectively) were included in the blinded study and used as cutoff controls for result interpretation. The study was performed on the Rotor-Gene Q HRM platform.
      The acquired data from the study were interpreted on the basis of the three cutoff reference controls (Table 4). With the 53 CGG repeat cutoff control (GM20230), the assay detected all 39 PM/FM samples (true positives), of which 38 were classified as expanded (>53 CGG repeats) and one was indeterminate (equal to 53 CGG repeats). Consequently, the detection of all PM/FM samples gave a clinical sensitivity of 100.00% (95% CI, 91.0%–100%) (Table 5). The assay also detected two of three GZ samples (false positives) in addition to the 39 PM/FM samples, resulting in a positive predictive value of 95.12%. Conversely, the assay scored as nonexpanded for all 370 NL samples, all 116 high normal samples, and one of three GZ samples (Table 4), resulting in a clinical specificity of 99.59% (95% CI, 98.5%–99.9%) (Table 5). None of the PM/FM alleles were scored as nonexpanded (false negatives), thereby giving a negative predictive value of 100.00% (Table 5).
      Table 4Classification of Clinical Samples Using the TP-PCR Assay with MCA in Comparison with Previously Characterized Methods from Mundhofir et al
      • Mundhofir F.E.
      • Winarni T.I.
      • Nillesen W.M.
      • van Bon B.W.
      • Schepens M.
      • Ruiterkamp-Versteeg M.
      • Hamel B.C.
      • Yntema H.G.
      • Faradz S.M.
      Prevalence of fragile X syndrome in males and females in Indonesia.
      Classification using TP-PCR + MCAClassification from Mundhofir et al
      • Mundhofir F.E.
      • Winarni T.I.
      • Nillesen W.M.
      • van Bon B.W.
      • Schepens M.
      • Ruiterkamp-Versteeg M.
      • Hamel B.C.
      • Yntema H.G.
      • Faradz S.M.
      Prevalence of fragile X syndrome in males and females in Indonesia.
      Normal (≤30 CGG repeats)High normal (30–45 CGG repeats)Gray zone (45–55 CGG repeats)Premutation/full mutation (≥55 CGG repeats)
      Cutoff control: GM20230 (53 CGG repeats)
       Nonexpanded (≤53 CGG repeats)37011610
       Indeterminate
      A sample classified as indeterminate warrants further testing and is considered a positive detection.
      0021
       Expanded (>53 CGG repeats)00038
      Cutoff control: GM20244 (41 CGG repeats)
       Nonexpanded (≤41 CGG repeats)3709710
       Indeterminate
      A sample classified as indeterminate warrants further testing and is considered a positive detection.
      01000
       Expanded (>41 CGG repeats)09239
      Cutoff control: GM06890 (30 CGG repeats)
       Nonexpanded (≤30 CGG repeats)3011000
       Indeterminate
      A sample classified as indeterminate warrants further testing and is considered a positive detection.
      31400
       Expanded (>30 CGG repeats)38102339
      Total samples per class, n370116339
      MCA, melting curve analysis; TP-PCR, triplet repeat-primed PCR.
      A sample classified as indeterminate warrants further testing and is considered a positive detection.
      Table 5Clinical Performance of TP-PCR Assay with MCA
      Cutoff controlClinical performance, %
      Sensitivity (95% CI)Specificity (95% CI)PPVNPV
      53 CGG repeats100.00 (91.0–100)99.59 (98.5–99.9)95.12100.00
      41 CGG repeats97.62 (87.7–99.6)96.09 (93.8–97.5)68.3399.79
      30 CGG repeats93.67 (88.7–96.5)81.35 (77.1–85.0)68.2096.78
      MCA, melting curve analysis; NPV, negative predictive value; PPV, positive predictive value; TP-PCR, triplet repeat-primed PCR.
      With the 41 CGG repeat cutoff control, expanded samples would comprise CGG repeats >41, and will include GZ alleles. Consequently, any GZ, PM, and FM sample that was undetected (scored as neither indeterminate nor expanded) would be considered as a false negative (Table 4). The TP-PCR assay failed to detect one of three GZ samples, yet falsely detected 19 high normal samples, resulting in an overall sensitivity of 97.62% (95% CI, 87.7%–99.6%) and an overall specificity of 96.09% (95% CI, 93.8%–97.5%) (Table 5).
      When the cutoff reference was extended to 30 CGG repeats, detection included high normal samples in addition to GZ, PM, and FM samples. Consequently, any sample classified as high normal, GZ, PM, and FM that was undetected (scored neither as indeterminate nor expanded) would be considered as a false negative (Table 4). On the basis of this criterion, the assay failed to detect 10 of 116 high normal samples, yet falsely detected 67 of the 370 NL samples. The performance of the assay in this case resulted in 93.67% sensitivity (95% CI, 88.7%–96.5%) and 81.35% specificity (95% CI, 77.1%–85.0%) (Table 5).

       Detection of Clinical Mosaic Samples

      Of the 528 clinical samples tested, three PM mosaic male samples harbored FMR1 alleles with CGG repeats of 86 of 166, 80 of 103, and 86 of 103 determined with other methods.
      • Fu Y.H.
      • Kuhl D.P.
      • Pizzuti A.
      • Pieretti M.
      • Sutcliffe J.S.
      • Richards S.
      • Verkerk A.J.
      • Holden J.J.
      • Fenwick Jr., R.G.
      • Warren S.T.
      • et al.
      Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox.
      • Mundhofir F.E.
      • Winarni T.I.
      • Nillesen W.M.
      • van Bon B.W.
      • Schepens M.
      • Ruiterkamp-Versteeg M.
      • Hamel B.C.
      • Yntema H.G.
      • Faradz S.M.
      Prevalence of fragile X syndrome in males and females in Indonesia.
      These three samples were identified retrospectively on testing completion; respective melt curve profiles generated were specifically examined. The resumed baseline −dF/dT temperatures of all three mosaic samples were clearly higher than that of the 53 CGG cutoff control (GM20230) (Figure 6), indicating detection of expanded alleles in clinical mosaic samples by using the TP-PCR assay.
      Figure thumbnail gr6
      Figure 6Melt curve profiles of male mosaic clinical samples. Male mosaic samples from the clinical archive were detected through melt curve analysis on the Rotor-Gene Q HRM platform when a 53 CGG control (GM20230) was used. Resumed baseline −dF/dT temperature cutoff refers to that of GM20230. NTC, no template control; resumed baseline −dF/dT temperature, temperature at which baseline negative first derivative of fluorescence versus temperature resumed.

       Detection of Female Heterozygotes

      The study cohort included 139 female homozygous samples that carried two NL FMR1 alleles and 67 female heterozygotes that carried either two NL alleles of different sizes or one NL and one expanded allele. These samples were identified retrospectively on testing completion, and their data were specifically examined. The assay detected all female homozygous samples (n = 139) with NL alleles and heterozygous samples that carried NL alleles of different sizes (n = 44) as nonexpanded when the 53 CGG cutoff control (GM20230) was used (Table 6). Further, all such samples (n = 139 + 44) generated a resumed baseline −dF/dT temperature that was much lower than that of the 53 CGG control (Figure 7). Conversely, all 23 female heterozygotes with PM and FM alleles were detected as expanded; their resumed baseline −dF/dT temperatures were higher than that of the selected male 53 CGG control (GM20230). The temperature profiles of female heterozygotes with expanded alleles clearly differ from homozygous and heterozygous samples that contained only NL alleles (Figure 7).
      Table 6Genotypes of Female Clinical Samples Detected by Using TP-PCR with MCA
      Genotype (females)nDetection by TP-PCR and MCA
      Homozygous (n = 139)
       Normal139Nonexpanded
      Heterozygous (n = 67)
       Normal44Nonexpanded
       PM or FM23Expanded
      Includes one indeterminate sample, which warrants further testing and is considered as expanded.
      FM, full maturation; MCA, melting curve analysis; PM, premutation; TP-PCR, triplet repeat-primed PCR.
      Includes one indeterminate sample, which warrants further testing and is considered as expanded.
      Figure thumbnail gr7
      Figure 7Results of female heterozygous clinical samples with expanded alleles. Representative melt curve profiles of female homozygous NL samples and heterozygous expanded samples relative to a male 53 CGG cutoff control (GM20230) on the Rotor-Gene Q HRM platform. FM, full mutation; NL, normal; PM, premutation; resumed baseline −dF/dT temperature, temperature at which baseline negative first derivative of fluorescence versus temperature resumed.

      Discussion

      The lack of a screening tool that is both rapid and cost-effective for detecting FMR1 expansions in large populations is well recognized.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      To tackle the problem, Tassone et al
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      adopted an approach that combined two PCRs: first a standard flanking PCR followed by a second chimeric CGG repeat-targeted PCR. The first PCR could rapidly identify the normal male and the obvious heterozygous normal female, but a subsequent TP-PCR was required to resolve the ambiguity between a homozygous normal female and a female heterozygous for a non-amplified large expansion. Collectively, the two PCRs will reduce the need for the Southern blot testing.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      Other groups incorporated capillary-based methods into their TP-PCRs to increase throughput and to minimize reliance on Southern blot analysis.
      • Chen L.
      • Hadd A.
      • Sah S.
      • Filipovic-Sadic S.
      • Krosting J.
      • Sekinger E.
      • Pan R.
      • Hagerman P.J.
      • Stenzel T.T.
      • Tassone F.
      • Latham G.J.
      An information-rich CGG repeat primed PCR that detects the full range of fragile X expanded alleles and minimizes the need for southern blot analysis.
      • Lyon E.
      • Laver T.
      • Yu P.
      • Jama M.
      • Young K.
      • Zoccoli M.
      • Marlowe N.
      A simple, high-throughput assay for Fragile X expanded alleles using triple repeat primed PCR and capillary electrophoresis.
      A third approach developed by Teo et al
      • Teo C.R.
      • Law H.Y.
      • Lee C.G.
      • Chong S.S.
      Screening for CGG repeat expansion in the FMR1 gene by melting curve analysis of combined 5' and 3' direct triplet-primed PCRs.
      was unique, using MCA of TP-PCR amplicons to detect FMR1 CGG expansions.
      The combinatorial approach by Teo et al
      • Teo C.R.
      • Law H.Y.
      • Lee C.G.
      • Chong S.S.
      Screening for CGG repeat expansion in the FMR1 gene by melting curve analysis of combined 5' and 3' direct triplet-primed PCRs.
      rapidly generates positive identification of PM and FM FMR1 alleles even in the presence of non-expanded alleles regardless of sex or zygosity. Such an approach contrasts with the method by Tassone et al
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      which relies on a null-allele result for identifying expansions in males. Additional testing is also required for resolving ambiguity in females. Consequently, the TP-PCR with MCA approach by Teo et al
      • Teo C.R.
      • Law H.Y.
      • Lee C.G.
      • Chong S.S.
      Screening for CGG repeat expansion in the FMR1 gene by melting curve analysis of combined 5' and 3' direct triplet-primed PCRs.
      reduces reliance on subsequent Southern blot analysis and also removes the requirement for post-PCR analysis via gel or capillary electrophoresis. In terms of equipment needs and reagent costs, the TP-PCR with MCA approach requires minimally a real-time PCR platform, rather than an additional more expensive capillary electrophoresis instrument. Indeed, the TP-PCR with MCA approach is much more cost-effective and efficient and truly exemplifies a first-line PCR-only screening assay that requires no post-PCR processing.
      The FastFraX ID kit is a commercially available screening test developed using the TP-PCR approach with MCA. In our present work, we assessed the analytic performance of the assay by using reference DNA samples from the Coriell Cell Repositories. Importantly, we also evaluated its clinical performance in a blinded retrospective study with 528 clinical DNA samples from persons with intellectual disabilities; these DNA samples span a wide range of CGG repeat sizes.
      Our study verified that melt curve profiles generated by the TP-PCR with MCA approach distinguished expanded PM and FM samples from NL samples when the recommended 50 ng of DNA input was used (Figure 2). The robustness of the assay was demonstrated by the degree of clear separation between the melt curve profiles of expanded and the non-expanded samples, even when DNA inputs were as low as 10 ng per test (Figure 2). Melt curve profiles of the expanded and non-expanded profiles were maintained in the presence of ≥150 ng of non-relevant DNA (Figure 3). Further, the resumed baseline −dF/dT temperatures of the melt curve profiles were consistent when the assay was validated with the ABI 7500 Fast and Rotor-Gene Q HRM. The former generated a CV < 0.55% and the latter <0.5% (Figure 4C and Table 2). The variation across platforms, however small, suggested that the resumed baseline −dF/dT temperatures are not absolute. Rather, the resumed baseline −dF/dT temperatures are relative to CGG repeat sizes within each run (Figure 4). As a result, a temperature established with a chosen control from one validation run should not be used as a standing cutoff parameter for future assay runs. A normal cutoff control sample is essential and should be included with every assay run to establish a cutoff temperature.
      On the basis of our findings, we selected three Coriell samples (GM06890, GM20244, and GM20230 with 30, 41, and 53 CGG repeats, respectively) as potential cutoff controls in our small-scale evaluation of 35 well-characterized reference DNA samples. The performance of the TP-PCR assay depended on the choice of control used. Results generated with the assay had a high degree of agreement with the genotypes provided by Coriell (Table 3). Notably, most of the disagreements arose from false detections of samples with at least one FMR1 allele having CGG repeat length close to that of the control.
      The clinical performance was established by extending the study to test archived clinical samples. In particular, all PM and FM samples were detected as expanded when the male 53 CGG sample (GM20230) was used as the cutoff control, yielding 100% clinical sensitivity (95% CI, 91.0%–100%) and 99.6% clinical specificity (95% CI, 98.5%–99.9%). Conversely, the clinical performance of the assay was less than perfect when the 41 and 30 CGG samples (GM20244 and GM06890, respectively) were used as cutoff controls. In fact, the CGG repeats of the cutoff controls used in our study are close, but not identical, to the standard boundaries of high normal, GZ, and PM/FM classifications. For instance, the 41 CGG sample used as the control differs from the standardized GZ cutoff of 44 CGG repeats. Thus, unless a precise cutoff is available and used, the high normal samples with 41 to 44 CGG repeats will inevitably be detected as expanded. In addition, the varied performance can be explained with the size distribution of FMR1 alleles within the tested population (Figure 1). Essentially, a high frequency of samples had CGG repeat lengths close to or identical to the 30 and 41 CGG cutoff controls, more so than the 53 CGG cutoff control. Melt curve profiles of most NL FMR1 alleles that harbor 29 ± 1 CGG repeats will have resumed baseline −dF/dT temperatures overlapping with that of the 30 CGG control yet distinct from the 53 CGG control, thus resulting in performance differences.
      Although the differences in performance were apparent among the three cutoff controls used, we emphasize that the choice of the control affects the scope of the detected expansion and largely depends on the purpose of the study. Expanded samples detected with the 53 CGG cutoff control comprise only PM and FM, whereas expanded samples detected with the 41 CGG cutoff control will include GZ, PM, and FM. Thus, the TP-PCR assay with MCA is easily adaptable to a wide range of applications by simply adjusting the choice of controls. For example, although the 53 CGG cutoff control is useful for detecting carriers and probing cases of mental disability, the 41 CGG cutoff control is a valuable alternative for more conservative screening to include detection of GZ samples. Furthermore, should a research laboratory be interested in studying high normal FMR1 alleles, the 30 CGG cutoff control will be more appropriate. A 100% detection of all PM and FM samples is guaranteed with the 41 and 30 CGG cutoff controls.
      The clinical sensitivity of the TP-PCR assay exhibited no deficiencies when the 53 CGG cutoff control was used in the study. However, a Coriell sample with 56 CGG repeats was undetected with the 53 CGG cutoff control in the preliminary study (Table 3). The favorable results demonstrated by the clinical study are likely because of the cohort containing few (although naturally representative in proportion) samples with FMR1 CGG repeats between 45 and 60 (Figure 1). Such excellent performance provides further evidence that the 53 CGG repeat sample is perfectly adequate as a control when identifying carriers (PM) or persons with FXS (FM) in a population with intellectual disabilities. A lower CGG cutoff may be more appropriate when testing a cohort with fragile X–associated tremor/ataxia syndrome; this population likely has a higher frequency of GZ and PM alleles that is readily detectable with a 41 CGG cutoff control.
      • Liu Y.
      • Winarni T.I.
      • Zhang L.
      • Tassone F.
      • Hagerman R.J.
      Fragile X-associated tremor/ataxia syndrome (FXTAS) in grey zone carriers.
      With the 53 CGG cutoff control, all 39 PM and FM samples from the 528 clinical samples were detected as expanded with one exception classified as indeterminate (Table 4). Closer inspection revealed a female PM sample with previously characterized FMR1 alleles of 30 and 78 CGG repeats. The melt curve profile of this PM sample exhibited a resumed baseline −dF/dT temperature that overlaps with the 53 CGG cutoff control, resulting in the sample being scored as indeterminate and warranting further testing. Interestingly, the overlapping resumed baseline −dF/dT temperature of the sample in question with that of the 53 CGG cutoff control may suggest similarities in PCR amplicon size distributions. Nevertheless, most samples with normal FMR1 alleles of 29 or 30 CGG repeats had melt curve profiles with much lower and distinctly different resumed baseline −dF/dT temperatures than that of the 53 CGG repeat cutoff control (Figure 2).
      A key concern with standard PCR is the ambiguity associated with a single PCR product in 40% to 50% of female samples.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      Although the single band frequently indicates homozygosity, there are also heterozygous expanded samples with a non-amplifying FM or a large PM allele that are undetected by conventional PCR.
      • Tassone F.
      • Pan R.
      • Amiri K.
      • Taylor A.K.
      • Hagerman P.J.
      A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
      These samples produce only a single product when analyzed with standard PCR. In our clinical study, the assay effectively identified all 23 heterozygous cases with one expanded FMR1 allele >55 CGG repeats (eg, 30, 98, and 30, >200 CGG repeats) when used with a 53 CGG cutoff control. This finding indicated that the TP-PCR assay with MCA approach is unambiguous in differentiating heterozygous expanded samples with one NL and one expanded allele from homozygous NL samples (eg, 30, 30 CGG repeats, or two alleles of the same size or alleles differing by one to two repeats) (n = 139) and heterozygous NL samples (n = 44) (Table 6). Therefore, this test enables rapid screening for identification of all samples with at least one expanded FMR1 CGG repeat allele, without capillary electrophoresis or Southern blot analysis. Undoubtedly, this assay takes into consideration CGG repeat size but not methylation status, and so it is unable to provide information about skewed X-chromosome inactivation in females.
      TP-PCR is used to detect mosaic samples when coupled with capillary electrophoresis, by relying on a ladder motif generated.
      • Juusola J.S.
      • Anderson P.
      • Sabato F.
      • Wilkinson D.S.
      • Pandya A.
      • Ferreira-Gonzalez A.
      Performance evaluation of two methods using commercially available reagents for PCR-based detection of FMR1 mutation.
      As such, detection of simulated mosaic samples with the combinatorial TP-PCR and MCA approach is unsurprising. A simulated mosaic sample with a PM allele present at 7.5% of total DNA input was detected when a 53 CGG cutoff control was used (Figure 5). The 53 cutoff control is stringent in detecting simulated samples; the 30 CGG control has a much lower resumed baseline −dF/dT temperature and would have enabled the detection of low-level mosaics. Nevertheless, all three male mosaic samples in the clinical archive were detected with the stringent 53 CGG cutoff control (Figure 6). Although these clinical mosaic samples were easily identified because of the presence of two PM alleles, detection served as a confirmation for the utility of the assay.
      The TP-PCR assay with MCA approach appears to have its limitation in distinguishing PM from FM alleles because of the overlap of resumed baseline −dF/dT temperatures between the PM and FM samples. In theory, the resumed baseline −dF/dT temperatures of the PM and FM amplicons should be distinguishable because the latter would contain longer amplicons due to larger alleles. In practice however, the assay is consistent with other TP-PCR approaches, producing only limited amounts of CGG amplicons beyond a certain size. As a result, the products of larger alleles are present in low proportions in the mixture of CGG amplicons. Thus, differences between the mixtures of amplicons that contain inconsequential amounts of a large and an even larger allele might not be detectable by MCA used in our approach. Nevertheless, our approach meets the intended utility for detecting samples with defined CGG expansions, markedly reducing the number of samples that require further characterization. Regardless, this screening assay will have to be complemented with additional approaches to precisely determine the CGG repeat size and FMR1 methylation status of the samples being interrogated. This MCA identification will be applicable also for other trinucleotide repeat disorder diseases.

      References

        • Hill M.K.
        • Archibald A.D.
        • Cohen J.
        • Metcalfe S.A.
        A systematic review of population screening for fragile X syndrome.
        Genet Med. 2010; 12: 396-410
        • Tassone F.
        Newborn screening for fragile X syndrome.
        JAMA Neurol. 2014; 71: 355-359
        • Pieretti M.
        • Zhang F.
        • Fu Y.
        • Warren S.
        • Oostra B.
        • Caskey C.
        • Nelson D.
        Absence of expression of the FMR-1 gene in fragile X syndrome.
        Cell. 1991; 66: 817-822
        • Verkerk A.J.
        • Pieretti M.
        • Sutcliffe J.S.
        • Fu Y.H.
        • Kuhl D.P.
        • Pizzuti A.
        • Reiner O.
        • Richards S.
        • Victoria M.F.
        • Zhang F.P.
        • et al.
        Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome.
        Cell. 1991; 65: 905-914
        • Fu Y.H.
        • Kuhl D.P.
        • Pizzuti A.
        • Pieretti M.
        • Sutcliffe J.S.
        • Richards S.
        • Verkerk A.J.
        • Holden J.J.
        • Fenwick Jr., R.G.
        • Warren S.T.
        • et al.
        Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox.
        Cell. 1991; 67: 1047-1058
        • Chen L.
        • Hadd A.
        • Sah S.
        • Filipovic-Sadic S.
        • Krosting J.
        • Sekinger E.
        • Pan R.
        • Hagerman P.J.
        • Stenzel T.T.
        • Tassone F.
        • Latham G.J.
        An information-rich CGG repeat primed PCR that detects the full range of fragile X expanded alleles and minimizes the need for southern blot analysis.
        J Mol Diagn. 2010; 12: 589-600
        • Hoffman G.E.
        • Le W.W.
        • Entezam A.
        • Otsuka N.
        • Tong Z.B.
        • Nelson L.
        • Flaws J.A.
        • McDonald J.H.
        • Jafar S.
        • Usdin K.
        Ovarian abnormalities in a mouse model of fragile X primary ovarian insufficiency.
        J Histochem Cytochem. 2012; 60: 439-456
        • Gallagher A.
        • Hallahan B.
        Fragile X-associated disorders: a clinical overview.
        J Neurol. 2012; 259: 401-413
        • Hagerman P.
        Fragile X-associated tremor/ataxia syndrome (FXTAS): pathology and mechanisms.
        Acta Neuropathol. 2013; 126: 1-19
        • Monaghan K.G.
        • Lyon E.
        • Spector E.B.
        • American College of Medical Genetics and Genomics
        ACMG Standards and Guidelines for fragile X testing: a revision to the disease-specific supplements to the Standards and Guidelines for Clinical Genetics Laboratories of the American College of Medical Genetics and Genomics.
        Genet Med. 2013; 15: 575-586
        • de Vries B.
        • Halley D.
        • Oostra B.
        • Niermeijer M.
        The fragile X syndrome.
        J Med Genet. 1998; 35: 579-589
        • Loesch D.
        • Hagerman R.
        Unstable mutations in the FMR1 gene and the phenotypes.
        Adv Exp Med Biol. 2012; 769: 78-114
        • Cronister A.
        • Teicher J.
        • Rohlfs E.M.
        • Donnenfeld A.
        • Hallam S.
        Prevalence and instability of fragile X alleles: implications for offering fragile X prenatal diagnosis.
        Obstet Gynecol. 2008; 111: 596-601
        • Nolin S.L.
        • Sah S.
        • Glicksman A.
        • Sherman S.L.
        • Allen E.
        • Berry-Kravis E.
        • Tassone F.
        • Yrigollen C.
        • Cronister A.
        • Jodah M.
        • Ersalesi N.
        • Dobkin C.
        • Brown W.T.
        • Shroff R.
        • Latham G.J.
        • Hadd A.G.
        Fragile X AGG analysis provides new risk predictions for 45-69 repeat alleles.
        Am J Med Genet A. 2013; 161A: 771-778
        • Cronister A.
        • Schreiner R.
        • Wittenberger M.
        • Amiri K.
        • Harris K.
        • Hagerman R.J.
        Heterozygous fragile X female: historical, physical, cognitive, and cytogenetic features.
        Am J Med Genet. 1991; 38: 269-274
        • Jacquemont S.
        • Hagerman R.J.
        • Leehey M.A.
        • Hall D.A.
        • Levine R.A.
        • Brunberg J.A.
        • Zhang L.
        • Jardini T.
        • Gane L.W.
        • Harris S.W.
        • Herman K.
        • Grigsby J.
        • Greco C.M.
        • Berry-Kravis E.
        • Tassone F.
        • Hagerman P.J.
        Penetrance of the fragile X-associated tremor/ataxia syndrome in a premutation carrier population.
        JAMA. 2004; 291: 460-469
        • Bourgeois J.A.
        • Seritan A.L.
        • Casillas E.M.
        • Hessl D.
        • Schneider A.
        • Yang Y.
        • Kaur I.
        • Cogswell J.B.
        • Nguyen D.V.
        • Hagerman R.J.
        Lifetime prevalence of mood and anxiety disorders in fragile X premutation carriers.
        J Clin Psychiatry. 2011; 72: 175-182
        • Grigsby J.
        • Brega A.G.
        • Engle K.
        • Leehey M.A.
        • Hagerman R.J.
        • Tassone F.
        • Hessl D.
        • Hagerman P.J.
        • Cogswell J.B.
        • Bennett R.E.
        • Cook K.
        • Hall D.A.
        • Bounds L.S.
        • Paulich M.J.
        • Reynolds A.
        Cognitive profile of fragile X premutation carriers with and without fragile X-associated tremor/ataxia syndrome.
        Neuropsychology. 2008; 22: 48-60
        • Hagerman R.
        • Hagerman P.
        Advances in clinical and molecular understanding of the FMR1 premutation and fragile X-associated tremor/ataxia syndrome.
        Lancet Neurol. 2013; 12: 786-798
        • Loesch D.Z.
        • Tassone F.
        • Lo J.
        • Slater H.R.
        • Hills L.V.
        • Bui M.Q.
        • Silburn P.A.
        • Mellick G.D.
        New evidence for, and challenges in, linking small CGG repeat expansion FMR1 alleles with Parkinson's disease.
        Clin Genet. 2013; 84: 382-385
        • Murray J.
        • Cuckle H.
        • Taylor G.
        • Hewison J.
        Screening for fragile X syndrome.
        Health Technol Assess. 1997; 1 (1-71): i-iv
        • Oostra B.A.
        • Jacky P.B.
        • Brown W.T.
        • Rousseau F.
        Guidelines for the diagnosis of fragile X syndrome. National Fragile X Foundation.
        J Med Genet. 1993; 30: 410-413
        • Lyon E.
        • Laver T.
        • Yu P.
        • Jama M.
        • Young K.
        • Zoccoli M.
        • Marlowe N.
        A simple, high-throughput assay for Fragile X expanded alleles using triple repeat primed PCR and capillary electrophoresis.
        J Mol Diagn. 2010; 12: 505-511
        • Filipovic-Sadic S.
        • Sah S.
        • Chen L.
        • Krosting J.
        • Sekinger E.
        • Zhang W.
        • Hagerman P.J.
        • Stenzel T.T.
        • Hadd A.G.
        • Latham G.J.
        • Tassone F.
        A novel FMR1 PCR method for the routine detection of low abundance expanded alleles and full mutations in fragile X syndrome.
        Clin Chem. 2010; 56: 399-408
        • Tassone F.
        • Pan R.
        • Amiri K.
        • Taylor A.K.
        • Hagerman P.J.
        A rapid polymerase chain reaction-based screening method for identification of all expanded alleles of the fragile X (FMR1) gene in newborn and high-risk populations.
        J Mol Diagn. 2008; 10: 43-49
        • Zhou Y.
        • Law H.Y.
        • Boehm C.D.
        • Yoon C.S.
        • Cutting G.R.
        • Ng I.S.
        • Chong S.S.
        Robust fragile X (CGG)n genotype classification using a methylation specific triple PCR assay.
        J Med Genet. 2004; 41: e45
        • Haddad L.A.
        • Mingroni-Netto R.C.
        • Vianna-Morgante A.M.
        • Pena S.D.
        A PCR-based test suitable for screening for fragile X syndrome among mentally retarded males.
        Hum Genet. 1996; 97: 808-812
        • Warner J.P.
        • Barron L.H.
        • Goudie D.
        • Kelly K.
        • Dow D.
        • Fitzpatrick D.R.
        • Brock D.J.
        A general method for the detection of large CAG repeat expansions by fluorescent PCR.
        J Med Genet. 1996; 33: 1022-1026
        • Teo C.R.
        • Law H.Y.
        • Lee C.G.
        • Chong S.S.
        Screening for CGG repeat expansion in the FMR1 gene by melting curve analysis of combined 5' and 3' direct triplet-primed PCRs.
        Clin Chem. 2012; 58: 568-579
        • Miller S.A.
        • Dykes D.D.
        • Polesky H.F.
        A simple salting out procedure for extracting DNA from human nucleated cells.
        Nucleic Acids Res. 1988; 16: 1215
        • Mundhofir F.E.
        • Winarni T.I.
        • Nillesen W.M.
        • van Bon B.W.
        • Schepens M.
        • Ruiterkamp-Versteeg M.
        • Hamel B.C.
        • Yntema H.G.
        • Faradz S.M.
        Prevalence of fragile X syndrome in males and females in Indonesia.
        World J Med Genet. 2012; 2: 15-22
        • Mundhofir F.E.
        • Winarni T.I.
        • van Bon B.W.
        • Aminah S.
        • Nillesen W.M.
        • Merkx G.
        • Smeets D.
        • Hamel B.C.
        • Faradz S.M.
        • Yntema H.G.
        A cytogenetic study in a large population of intellectually disabled Indonesians.
        Genet Test Mol Biomarkers. 2012; 16: 412-417
        • Liu Y.
        • Winarni T.I.
        • Zhang L.
        • Tassone F.
        • Hagerman R.J.
        Fragile X-associated tremor/ataxia syndrome (FXTAS) in grey zone carriers.
        Clin Genet. 2013; 84: 74-77
        • Juusola J.S.
        • Anderson P.
        • Sabato F.
        • Wilkinson D.S.
        • Pandya A.
        • Ferreira-Gonzalez A.
        Performance evaluation of two methods using commercially available reagents for PCR-based detection of FMR1 mutation.
        J Mol Diagn. 2012; 14: 476-486