- Email: [email protected]
Real time PCR assays to detect common mutations in the biotinidase gene and application of mutational analysis to newborn screening for biotinidase deficiency
Molecular Genetics and Metabolism 78 (2003) 100–107 www.elsevier.com/locate/ymgme
Real time PCR assays to detect common mutations in the biotinidase gene and application of mutational analysis to newborn screening for biotinidase deficiency Steven F. Dobrowolski, Janine Angeletti, Richard A. Banas, and Edwin W. Naylor* Neo Gen Screening, 90 Emerson Lane, Suite 1403, Abele Business Park, P.O. Box 219, Bridgeville, PA 15017, USA Received 27 November 2002; accepted 29 November 2002
Abstract Biotinidase deficiency is an autosomal recessive disorder of biotin metabolism caused by defects in the biotinidase gene. Symptoms of biotinidase deficiency are resolved or prevented with oral biotin supplementation and as such newborn screening is performed to prospectively identify affected individuals prior to the onset of symptoms. Biotinidase deficiency is detected by determining the activity of the biotinidase enzyme utilizing the newborn dried blood spot and colorimetric end point analysis. While newborn screening by enzyme analysis is effective, external factors may compromise results of the enzyme analysis and difficulty is encountered in distinguishing between complete and partial enzyme deficiencies. In the United States, the four mutations most commonly associated with complete biotinidase deficiency are c98:d7i3, Q456H, R538C, and the double mutation D444H:A171T. Partial biotinidase deficiency is almost universally attributed to the D444H mutation. To more effectively distinguish between profound and partial biotinidase deficiency, a panel of assays utilizing real time PCR and melting curve analysis using Light Cycler technology was developed. Employing DNA extracted from the original dried blood specimens from newborns identified through prospective newborn screening as presumptive positive for biotinidase deficiency, the specimens were analyzed for the presence of the five common mutations. Using this approach it was possible to separate newborns with partial and complete deficiency from each other as well as from many of those with false positive results. In most cases it was also possible to correlate the genotype with the degree of residual enzyme activity present. In newborn screening for biotinidase deficiency, we have shown that the analysis of common mutations is useful in distinguishing between partial and complete enzyme deficiency as well as improving specificity. Combining biotinidase enzyme analysis with genotypic data also increases the sensitivity of screening for biotinidase deficiency and provides information useful to clinicians earlier than would otherwise be possible. Ó 2003 Elsevier Science (USA). All rights reserved. Keywords: Newborn screening; Mutations; Genetic testing; Biotinidase deficiency
Introduction Biotinidase deficiency is an autosomal recessive metabolic disorder occurring in 1:80,000 live births [1]. Those affected by biotinidase deficiency exhibit irreversible neurological problems, seizures, developmental delay, hypotonia, ataxia, cutaneous, and other symptoms. Symptoms are preventable and in some cases reversible through oral biotin supplementation. Prospective newborn screening for biotinidase deficiency is, therefore, performed in much of the United States and in numerous * Corresponding author. Fax: 1-412-220-0784. E-mail address: [email protected] (E.W. Naylor).
other countries [2]. Biotinidase deficiency results from mutations in the biotinidase gene and depending upon the nature of the mutation(s), the enzyme deficiency may be either complete or partial. Mean biotinidase activity is 7.1 nmol/min/ml serum in normal newborns [3]. Those affected with complete biotinidase deficiency have enzymes that produce <10% of mean normal activity, while those affected with partial deficiency have enzymes that produce 10–30% of normal activity. In newborn screening laboratories, assaying for biotinidase deficiency is performed using an extract of whole blood derived from the universally collected newborn dried blood spot (DBS). Whole blood extract from a DBS specimen is an effective sample from which to assay for
1096-7192/03/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1096-7192(02)00231-7
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
biotinidase activity but not as precise as results obtained with serum. Activity of the biotinidase enzyme may be adversely affected if the DBS specimen is mishandled. DBS specimens that are incompletely dried, exposed to moisture after drying, or exposed to heat may not exhibit biotinidase activity [4]. Inconclusive or ambiguous results in screening for biotinidase deficiency are, therefore, often attributable to errors in sample collection and processing prior to their arrival in the screening laboratory. Another difficulty experienced in newborn screening for biotinidase deficiency involves differentiating between a complete and a partial enzyme deficiency by the screening laboratory using the initial DBS specimen. To aid in distinguishing between complete and partial biotinidase deficiency and subsequently increasing the sensitivity and specificity of screening for biotinidase deficiency mutational analysis has been employed. In the United States, the following five mutations are the most frequently observed in patients with biotinidase deficiency: c98:d7i3, Q456H, R538C, D444H, and the double mutant D444H:A171T. c98:d7i3, Q456H, and R538C are associated with complete biotinidase deficiency [5–7,9]. The D444H mutation has a carrier rate of 3.9% in the general population and when paired with a severe mutation causes partial enzyme deficiency [8]. This high frequency in the general population combined with its causing a partial enzyme deficiency makes the D444H mutation similar to the Duarte D2 N314D variant in galactosemia [9]. Interestingly, when D444H is in cis with the A171T mutation, the combined deleterious effects of both mutations result in a severe allele which when combined a second severe allele will cause complete enzyme deficiency. The D444H:A171T double mutation has been commonly observed in biotinidase deficient children that are ascertained by newborn screening [10]. Using Light Cycler technology and paired hybridization probes, assays were designed to detect these five mutations frequently observed in patients with biotinidase deficiency. Reported here are the assay procedures to detect these five commonly observed biotinidase mutations together with examples of complete and partial biotinidase deficiency derived from our prospective newborn screening program. Examples of how second tier molecular screening can reduce false positives are also presented. This use of mutational analysis supplements the biochemical enzymatic screening results and increases both the specificity and sensitivity of newborn screening for biotinidase deficiency.
Methods Specimens and DNA preparation Biotinidase deficient specimens, ascertained through either routine prospective newborn screening or high-
101
risk screening, were retrieved from the Neo Gen Screening specimen archive. Anonymous DBS specimens were utilized in assay development and specimens from known patients were retrieved from storage and used in the retrospective study. DNA was isolated from DBS specimens as previously described and 80–130 ng were utilized as template in each reaction [11]. Specimens, characterized as homozygous for Q456H, were provided by Adolf Muhl (Department of Pediatrics, University Hospital Vienna, Vienna, Austria). Prospective newborn screening for biotinidase deficiency DBS specimens are routinely submitted by hospitals, physicians, and parents that subscribe to Neo Gen ScreeningÕs Supplemental Newborn Screening program and screening for biotinidase deficiency has been part of that service since 1985. Analysis of biotinidase activity is routinely performed using the Astoria Pacific Continuous Flow Analyzer and the Astoria Pacific SPOTCHECK biotinidase enzyme assay reagents. This method, to assay biotinidase activity in DBS specimens, is based upon a modification of the methods described by Wolf and co-workers [12,13]. Samples demonstrating biotinidase activity below the critical cut-off level of 16.0 eru (enzyme response units, a proprietary unit of Astoria Pacific) were selected for genotype analysis. Polymerase chain reaction and hybridization probes Sequences of the human biotinidase gene (GenBank Accession Nos. NM00060, AF018630, and AF18631) were the basis from which primers and probes were designed. Primer Premier 5 (Premier Biosoft, Palo Alto, CA) and Tm Utility 1.5 IT (Idaho Technology, Salt Lake City, UT) software was utilized to design primers for polymerase chain reaction (PCR) and hybridization probes to detect mutations. Primers and fluorescent labeled probes were obtained from either Operon Technology (Alameda, CA) or Idaho Technology (Salt Lake City, UT). PCR buffers were obtained from Idaho Technology (Salt Lake City, UT). Amplification reactions utilize 1 PCR buffer, 2 mM MgCl2 , 200 lM dNTPs (Roche, Mannheim, Germany), and 0.6 U Klen taq (AB Peptides, St. Louis, MO) in a complex with TaqStart antibody (ClonTech, Palo Alto, CA). Preparing a complex between the polymerase and TaqStart antibody is performed according to manufacturerÕs instructions. The sequence of individual primers, the sequence of fluorescent hybridization probes, and the concentration at which each is used are found in Table 1. Ten microliter PCRs were performed in capillary tubes using a Roche Light Cycler (Mannheim, Germany). Temperature cycling conditions for PCR utilizes a modified two-step thermal cycling scheme. Specimens are ramped to 94 °C at 20°/s and held there for 0 s to
0.13
DNA sequences are shown 50 to 30 , Concentrations are in micromolar units, Anchor probes are utilized at 0.2 lM, detection probes are utilized at 0.1 lM, P, phosphate; fitc, fluoroscene isothiocynate; LCred640, Light Cycler red 640.
0.25
LCred640-GCTTGCTCTT TTCCTCTGCG-P TGGTGACCAATCTTGG GACA-fitc 0.5 1.0
ACAGGTGTCGAAGCC AAGAC TCCATTATTATTGCTGA ACACGAC A171T
R538C
GGGGAAAGGAAGGCT ATCTC CTCCAGCGCCTGAGTT GTAT
0.5
0.5 Q456H
D444H
1.0
1.0
Detection probe
LCred640-GCTTGCTCTTT TCCTCTGCG-P LCred640-ACTACATCCA CGTGTGTGCCC-P CTCTATGGGCGCTTGT ATGA-fitc
Concentration
Anchor probe
CTCATACACGGCAGCC ACAT GGTGTCGAAGCCAAG ACCC CTTGTAGCCTGTGGA AGTGC
Reverse primer
GCCCCATTACATTCCA GATTTG GCCCACCTTATCCAAA GAGC GCTTGGCTGGGAGAAT GACC 7d3i
0.5
Forward primer Assay
Concentration
1.0
TGGTCTGCATTATGTCT GGAGCCAGAAGTA-fitc TTTGATGGGCTTCACA CAGTACATGGCACT-fitc LCred640-AGGGACTAG GAAAAGTGTGTGGTCT GTGG-P TGGTCTGCATTATGTC TGGAGCCAGAAGTA-fitc LCred640-AGGAGCCTTG TCATAGCAGTGACCCA AGGT-P
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
Table 1 Primers for polymerase chain reaction, hybridization probes for mutation analysis, and concentrations at which they are utilized
102
denature the DNA strands. Temperature then ramps at 20°/s to 58 °C and holds at this temperature for 15 s at which time primers anneal and polymerization of new DNA begins. Polymerization is completed while ramping from 58 to 72 °C at 1.0°/s. The slow ramp speed allows polymerization to precede, thus negating the necessity of a hold time at 72 °C. Thermal cycling is repeated for 45 cycles. All amplifications are performed in an asymmetric manner. Asymmetric amplifications for c98:d7i3, Q456H, R538C, and A171T assays enrich the antisense strand of the amplicon while the asymmetric amplification in the D444H assay enriches the sense strand of the amplicon. Asymmetry produces an excess of the DNA strand to which the hybridization probes will bind in the analysis phase of the assay. Hybridization probes and genotyping analysis Genotyping is performed using paired hybridization probes, where each assay has a detection probe and an anchor probe [14–23]. Probes for the c98:d7i3, Q456H, R538C, and A171T assays hybridize to the antisense strand of the amplicon while the probes for the D444H assay hybridize to the sense strand of the amplicon. The detection probe hybridizes with a region of the amplicon that includes the mutation, while the anchor probe hybridizes with a region adjacent to the detection probe. When both probes are hybridized, there is a 1 base gap between the anchor and detection probes. For each set of hybridization probes, one is conjugated on the 30 end with fitc while the second is conjugated on the 50 end with LCred640 (see Table 1). The probe which is 50 conjugated with LCred640 is also 30 phosphorylated to prevent extension by Taq DNA polymerase. When both probes are hybridized with the amplicon, the fluorescent moieties are brought into close proximity, and this proximity allows fluorescence resonance energy transfer to occur between the donor fluorophore (fitc) and the acceptor fluorophore (LCred640). Anchor probes have a Tm that is at least 15% higher than the corresponding detection probe, which allows the anchor probe to remain hybridized during the melting transition of the detection probe that occurs during the analysis phase of the assay. After completing the thermal cycling program, the Light Cycler proceeds seamlessly to the analysis program. The analysis program ramps to 94 °C at 20°/s and after reaching 94 °C, immediately begins to ramp at l°/s to 35 °C. Upon reaching 35 °C the temperature ramps upward at 0.1°/s to 76 °C. During the entire analysis program, the excitation wavelength of fitc is provided and the fluorescence of LCred640 is continuously acquired. Melting curves are generated by plotting the fluorescence of LCred640 against temperature during the 35–76 °C upward temperature ramp. Melting peaks are generated computationally by calculating the dF =dT of the melting curve which is then plotted against temperature.
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
Results Detecting mutations in the biotinidase gene Figs. 1A–E display analysis of individual biotinidase mutations using melting peaks generated with the Light Cycler. Fig. 1A displays the assay results for the R538C
Fig. 1. Wild type specimens: ———, heterozygous specimens:
103
mutation and specimens that are homozygous wild type, heterozygous, and a no DNA control. No specimen that is homozygous for R538C has yet been identified by Neo Gen Screening and we were unable to obtain one from outside sources. The remainder of the assays, shown in Figs. 1B–E display specimens that are homozygous wild type, heterozygous, homozygous for the mutation, and a
, homozygous mutant specimens:
, no DNA control: .......
104
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
Fig. 1. (continued)
no DNA control. In the cases of the D444H, c98:d7i3, and R538C assays, the detection probe matches the wild type sequence. Therefore, the high temperature melting peak represents the wild type form of the gene while the lower temperature melting peak represents the mutant form of the gene. In the A171T and Q456H assays, the detection probe matches the mutant form of the gene and has a 1 bp mismatch with the wild type allele. In these assays, the high-temperature melting peak represents the mutant form of the gene while the low-temperature melting peak represents the wild type form of the gene. Analysis of specimens identified through newborn screening as presumptive positive for biotinidase deficiency Through prospective newborn screening of over 1.4 million newborns, Neo Gen Screening has identified numerous newborns with complete and partial biotinidase deficiency. These cases have been been referred to metabolic treatment centers where confirmatory quantitative serum biotinidase studies are performed. In addition to
those cases that are confirmed, a number of specimens have been identified as presumptively positive for biotinidase deficiency and many of these cases have turned out to be false positives due to environmental factors or mildly reduced enzyme activity resulting from homozygosity for the D444H mutation, or from heterozygosity for a severe mutation. With improvement in the sensitivity of the newborn screening methodologies the number of presumptive positives has increased. It is therefore important to utilize a second tier molecular assay to identify which of these increased presumptive positives are true positives and which are false positives. Table 2 lists the molecular results from nine confirmed complete biotinidase deficient patients. Six of these were screened between 1 and 3 days of age while the other three were high risk patients screened between 1 and 6 months of age. Six of these patients were homozygotes or compound heterozygotes with two copies of severe mutations, two had only one copy of one of the severe mutations that we screened for, and one had no copies. These last three cases, however, were confirmed as having complete deficiency as a result of very low
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107 Table 2 Genotypes of newborns and high-risk patients with complete biotinidase deficiency Patient Nos.
Sample date
Allele 1
Allele 2
1.
2 days
R538C
2.
2 days
D444H A171T D444H A171T c98:d7i3 c98:d7i3 c98:d7i3 Q456H Q456H Q456H Q456H ?
3. 4. 5. 6. 7. 8. 9. 10.
3 months 5.5 months 1 day 2 days 2 days 2 days 2 days 1 month
D444H A171T c98:d7i3 c98:d7i3 c98:d7i3 Q456H ? ? ? ?
Table 3 Genotypes of newborns and high-risk patients with partial biotinidase deficiency Patient Nos.
Sample date
Allele 1
Allele 2
1.
2 days
D444H
2.
2 days
3.
2 days
4. 5. 6. 7. 8. 9. 10. 11. 12.
2 days 2 days 1 day 2 days 3 days 2 days <1 day 14 days 2 days
D444H A171T D444H A171T D444H A171T c98:d7i3 c98:d7i3 Q456H Q456H Q456H Q456H Q456H R538C R538C Variant
D444H D444H D444H D444H D444H D444H D444H D444H D444H D444H D444H
screening and confirmatory enzyme activity. Table 3 lists 12 newborns screened between the ages of 1 and 14 days with confirmed partial biotinidase deficiency together with their respective genotypes. All have one severe mutation in combination with one copy of the D444H mild mutation. Among our false positives we have encountered a number of newborns homozygous for the D444H common mutation, heterozygous for a severe mutation, and some heterozygous for the D444H mild mutation. These false positives have mildly reduced enzyme activity on initial screening and normal or lownormal activity on repeat screening. In these cases, second tier molecular testing has been useful in reducing the repeat request rate which previously would have been based solely on the initial enzymatic screen.
Discussion Newborn screening with its deep roots in biochemical screening methods has been slow to embrace molecular
105
genetic analysis. With the genes responsible for many genetic diseases having been characterized and common causative mutations identified, incorporating second tier molecular testing into newborn screening is a natural progression. Recently, many newborn screening programs have expanded the list of disorders for which they screen. This expansion is due in large to broader application of tandem mass spectrometry and the publicÕs demand for more comprehensive analysis [24,25]. With this increased number of comprehensive newborn screening programs, the need for second tier molecular genetic testing will also increase. The first broad application of molecular genetic methods in newborn screening was as a second tier assay in screening for cystic fibrosis. The primary screen for cystic fibrosis is quantification of immunoreactive trypsinogen (IRT), however analysis of IRT alone is plagued with a high false positive rate. When second tier testing for the presence of the common cystic fibrosis transmembrane conductance regulator mutation (DF508) was performed on specimens having elevated IRT, the false positive rate fell significantly [26,27]. This was the first demonstration of a clear place for second tier molecular genetic analysis in newborn screening. Any assay that involves measuring enzyme activity is highly dependent upon proper sample handling and newborn screening assays are not exempted from this. Despite the fact that birthing centers have collected newborn screening samples for over 35 years, sample mishandling is still a persistent problem. Newborn screening for galactosemia, glucose-6-phosphate dehydrogenase (G-6-PD) deficiency, and biotinidase deficiency involve measuring enzyme activity. Improper sample collection, processing, and transportation have been demonstrated to be deleterious to the stability of these enzymes in the assays employed by newborn screening laboratories [4,29,30]. While it is the experience of this laboratory that the biotinidase enzyme is less sensitive to external factors than either galactose-1-phosphate uridyl transferase (GALT) or G-6-PD, environmental factors can still have an affect upon biotinidase activity and cannot be ignored. The physical appearance of the DBS specimen or a long delay between the date of specimen collection and its being received by the newborn screening laboratory are signs of possible mishandling. If multiple enzyme activities are deficient and mutational analysis of biotinidase and other genes (GALT, G-6-PD) shows no DNA sequence aberrations, it provides some evidence that the specimen was affected by environmental factors such as heat or humidity, and a replacement specimen should be requested. In newborn screening for biotinidase deficiency, it is frequently difficult to discern if a partial or complete enzyme deficiency has been encountered based solely on the initial enzymatic screening level. In the vast majority
106
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
of partial deficiencies, the D444H mutation is involved [8]. D444H has a carrier frequency of 3.9% in the general United States population and reduces the activity of the biotinidase enzyme by 48–52% [8,10,31]. It is noteworthy, that these percent reductions were determined with serum quantitative enzyme analysis, thus the percent enzyme reduction in a DBS derived whole blood specimen could be different. Partial deficiencies generally result from compound heterozygosity with one D444H mutation and a second severe mutation. In Table 3, there are 12 specimens with genotypes identifying them as partial deficiencies. Five specimens are compound heterozygous between D444H and Q456H; three with the D444H:A171T double mutation; two with c98:d7i3; and two with R538C. In the enzyme assay used in newborn screening, compound heterozygotes containing D444H and a mutation causing a complete deficiency (c98:d7i3, Q456H, etc.) may initially generate biochemical data mimicking complete biotinidase deficiency. A similar situation is frequently observed in the Beutler assay that is used to measures GALT activity when screening for galactosemia. Compound heterozygotes between the Duarte D2 N314D variant and a classical galactosemia mutation such as the common Q188R may initially generate biochemical data suggesting classical galactosemia. Identifying the N314D GALT mutation provides definitive proof that these newborns do not have classical galactosemia [28]. In a similar fashion, the D444H mutation is responsible for the vast majority of partial biotinidase deficiencies and therefore identifying this mutation provides strong evidence that a complete enzyme deficiency is not present. A complication surrounding the D444H mutation is the presence of double mutants. Three double mutations, involving D444H and a second mutation on the same gene, have been described. The commonly observed double mutant, D444H:A171T, that accounts for 17.3% of the mutations in complete biotinidase deficiencies ascertained by newborn screening, is part of this panel [10]. Two other double mutants, D444H:F403V and D444H:R157H, have been described, however, both are extremely rare [31]. The D444H mutation is very useful in helping to identify partial deficiencies, but the possibility of a double mutant resulting in a complete deficiency cannot be dismissed. After newborn screening results are reported and repeat specimens are requested and screened, the first clinical visit of a potentially biotinidase deficient newborn will involve determining quantitative serum biotinidase activity and possibly additional confirmatory molecular testing. Quantitative biotinidase analysis is the ultimate confirmation test to establish complete or partial biotinidase deficiency. Second tier mutation analysis is used to improve the sensitivity and specificity of the newborn screening program and to assist in clinical evaluation. However, in certain situations, the genotype data unambiguously
identified these specimens as having a complete enzyme deficiency. Such informative results can expedite patient care by bringing the newborn to immediate attention. The data shown in Tables 2 and 3, and discussed above, provide evidence to the utility of second tier mutation analysis in newborn screening for biotinidase deficiency. In a high throughput newborn screening laboratory, the most important issue is validity of results, but following closely behind is turn around time. Minimizing turn around time requires that assay platforms be fast, reliable, and easily interpreted within the context of a routine service laboratory. The Light Cycler platform is ideal for the high throughput newborn screening laboratory because all of those criteria are met. Air driven thermal cycling is fast, genotyping with fluorescent hybridization probes is simple because it involves no post-PCR manipulation, and melting peak data are easily interpreted. From isolation of DNA to data interpretation, the 5-mutation panel described here is completed in less than 2 h at a relatively low cost. At Neo Gen Screening second tier molecular testing is provided at no additional cost for newborns receiving routine newborn screening. Such rapid analysis assures that supplemental molecular data is reported along with the primary biochemical screening data. An additional benefit is that the closed tube format simplifies sample tracking and is favorable for avoiding amplicon contamination in the laboratory. Data files from the Light Cycler are easily stored and may be backed up in an offsite archive rendering them safe from loss. This is an ideal situation for the newborn screening laboratory where large quantities of sensitive clinical data are generated. Currently, 55 mutations that cause biotinidase deficiency have been identified [31]. Among specimens of domestic origin identified at Neo Gen Screening, the panel of five mutations proved useful in a very high percentage of presumptive positive newborns. Enzymatic analysis will remain the primary means by which biotinidase deficiency is detected in both newborn screening and high risk clinical diagnostics. Second tier mutation analysis provides valuable support to the enzymatic analysis and should be considered as a supplement to the biochemical data by those performing newborn screening for biotinidase deficiency.
References [1] B. Wolf, Worldwide survey of neonatal screening for biotinidase deficiency, J. Inher. Met. Dis. 14 (1991) 923–927. [2] National Newborn Screening and Genetics Resource Center http://genes-r-us.uthhscsa.edu. [3] B. Wolf, R.E. Grier, J. Secor-McVoy, G.S. Heard, Biotinidase deficiency: a novel vitamin recycling defect, J. Inher. Metab. Dis. 8 (Suppl. 1) (1985) 53–58.
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107 [4] J.R. Secor McCoy, H.C. Levy, M. Lawler, M.A. Schmidt, D.D. Ebers, P.S. Hart, M.G. Pettit, B. Wolf, Partial biotinidase deficiency: clinical and biochemical features, J. Pediatr. 115 (1) (1990) 78–83. [5] K.J. Norrgard, R.J. Pomponio, K.L. Swango, J. Hymes, T.R. Reynolds, G.A. Buck, B. Wolf, Mutation (Q456H) is the most common cause of profound biotinidase deficiency in children ascertained by newborn screening in the United States, Biochem. Mol. Med. 61 (1997) 22–27. [6] R.J. Pomponio, K.J. Norrgard, J. Hymes, T.R. Reynolds, G.A. Buck, R. Baumgartner, T. Suormala, B. Wolf, Arg538 to Cys mutation in a CpG dinucleotide of the human biotinidase gene is the second most common cause of profound biotinidase deficiency in symptomatic children, Hum. Genet. 99 (1997) 506–512. [7] R.J. Pomponio, T.R. Reynolds, H. Cole, G.A. Buck, B. Wolf, Mutation hotspot in the human biotinidase gene causes profound biotinidase deficiency, Nature Genet. 11 (1995) 96–98. [8] K.L. Swango, M. Demirkol, G. Huner, E. Pronicka, J. SykutCegielska, A. Schulze, B. Wolf, Partial biotinidase deficiency is usually due to the D444H mutation in the biotinidase gene, Hum. Genet. 102 (1998) 571–575. [9] H.C. Lin, L.T. Kirby, W.G. Ng, J.K.V. Reichardt, On the nature of the Duarte variant of galactose-1-phosphate uridyl transferase (GALT), Hum. Genet. 93 (1994) 167–169. [10] K.J. Norrgard, R.J. Pomponio, K.L. Swango, J. Hymes, T.R. Reynolds, G.A. Buck, B. Wolf, Double mutation (A171T and D444H) is a common cause of profound biotinidase deficiency in children ascertained by newborn screening in the United States, Hum. Mutat. 11 (1998) 410–414. [11] E. Heath, D. OÕBrien, R. Banas, S.F. Dobrowolski, Optimization of an automated protocol to isolate DNA from Guthrie cards, Arch. Pathol. Lab. Med. 123 (1999) 1154–1160. [12] G. Herd, J.S. McVoy, B. Wolf, A screening method for biotinidase deficiency in newborns, Clin. Chem. 30 (1) (1984) 125–127. [13] A. Denise, D. Pettit, P.D. Amadorand, B. Wolf, The quantitation of biotinidase activity in dried blood spots using microtiter transfer plates: identification of biotinidase deficient and heterozygous individuals, Anat. Biochem. 170 (1989) 371–374. [14] M.J. Lay, C.T. Wittwer, Real-time fluorescence genotyping of factor V Leiden during rapid-cycle PCR, Clin. Chem. 43 (1997) 2262–2267. [15] P.S. Bernard, M.J. Lay, C.T. Wittwer, Integrated amplification and detection of the C677T point mutation in the methylenetetrahydrofolate reductase gene by fluorescence resonance energy transfer and probe melting curves, Anal. Biochem. 255 (1998) 101–107. [16] P.S. Bernard, R.S. Ajioka, J.P. Kushner, C.T. Wittwer, Homogeneous multiplex genotyping of hemochromatosis mutations with fluorescence probes, Am. J. Pathol. 153 (1998) 1055–1061. [17] E. Lyon, A. Millson, T. Phan, C.T. Wittwer, Detection and identification of base alterations within the region of factor V
[18]
[19]
[20]
[21]
[22]
[23]
[24] [25] [26]
[27]
[28]
[29] [30]
[31]
107
Leiden by fluorescence melting curves, Mol. Diag. 3 (1998) 203– 210. C.N. Gundry, P.S. Bernard, M.G. Herrmann, G.H. Reed, C.T. Wittwer, Rapid F508 del and F508C assay using fluorescent hybridization probes, Genet. Test 3 (1999) 365–370. N. von Ahsen, E. Schultz, V.W. Armstrong, M. Oellerich, Rapid detection of prothrombic mutations of prothrombin (G20210A), factor V (G1691 A), and methylenetetrahydrofolate reductase (C677T) by real-time fluorescence melting curves, Mol. Diag. 45 (1999) 694–696. K. Mangassner-Stephan, C. Tag, A. Reiser, A.M. Gressner, Rapid genotyping of hemochromatosis gene mutations on the Light Cycler with fluorescent hybridization probes, Clin. Chem. 45 (1999) 1875–1878. C. Aslanidis, G. Schmitz, High-speed apolipoprotein E genotyping and apolipoprotein B3500 mutation detection using real-time fluorescence PCR and melting curves, Clin. Chem. 45 (1999) 1094–1097. S.F. Dobrowolski, C.T. Wittwer, C. Gundry, E.W. Naylor, Detection of CFTR DF508 using rapid cycler PCR and analysis of FRET probes in Neonatal Screening for cystic fibrosis, 115– 121. M.G. Herrmann, S.F. Dobrowolski, C.T. Wittwer, Rapid bglobin genotyping by multiplex probe melting temperature and color, Clin. Chem. 46 (2000) 425–428. D.H. Chace, Mass spectrometry in the clinical laboratory, Chem. Rev. 101 (2000) 445–477. E. Marshall, Fast technology drives new world of newborn screening, Science 294 (2001) 2272–2274. W.C. Spence, J. Paulus-Thomas, D.M. Orenstein, E.W. Naylor, Neonatal screening for cystic fibrosis: addition of molecular diagnosis to increase specificity, Biochem. Med. Metab. Biol. 49 (1993) 200–211. R.G. Gregg, A. Simantel, P. Farrell, et al., Newborn screening for cystic fibrosis in Wisconsin: comparison of biochemical and molecular methods, Pediatrics 99 (6) (1997) 819–824. S.F. Dobrowolski, R.A. Banas, J.G. Suzow, M. Berkley, E.W. Naylor, Analysis of common mutations in the galactose-1phosphate uridyl transferase gene: new assays to increase the sensitivity and specificity of newborn screening for galactosemia. J. Mol. Diag. (2002) (in press). P. Polana, S.F. Dobrowolski, E.W. Naylor, unpublished observation, 2002. V.E. Shih, H.L. Levy, V. Karolkewicz, S. Houghton, M.L. Efron, K.J. Isselbacher, E. Buetler, R.A. MacCready, Galactosemia screening in Massachusetts, N. Eng. J. Med. 284 (1971) 753–760. K.J. Norrgard, R.J. Pomponio, J. Hymes, B. Wolf, Mutations causing profound biotinidase deficiency in children ascertained by newborn screening in the United States occur at different frequencies than in symptomatic children, Pediatr. Res. 46 (1) (1999) 20–27.
Real time PCR assays to detect common mutations in the biotinidase gene and application of mutational analysis to newborn screening for biotinidase deficiency Steven F. Dobrowolski, Janine Angeletti, Richard A. Banas, and Edwin W. Naylor* Neo Gen Screening, 90 Emerson Lane, Suite 1403, Abele Business Park, P.O. Box 219, Bridgeville, PA 15017, USA Received 27 November 2002; accepted 29 November 2002
Abstract Biotinidase deficiency is an autosomal recessive disorder of biotin metabolism caused by defects in the biotinidase gene. Symptoms of biotinidase deficiency are resolved or prevented with oral biotin supplementation and as such newborn screening is performed to prospectively identify affected individuals prior to the onset of symptoms. Biotinidase deficiency is detected by determining the activity of the biotinidase enzyme utilizing the newborn dried blood spot and colorimetric end point analysis. While newborn screening by enzyme analysis is effective, external factors may compromise results of the enzyme analysis and difficulty is encountered in distinguishing between complete and partial enzyme deficiencies. In the United States, the four mutations most commonly associated with complete biotinidase deficiency are c98:d7i3, Q456H, R538C, and the double mutation D444H:A171T. Partial biotinidase deficiency is almost universally attributed to the D444H mutation. To more effectively distinguish between profound and partial biotinidase deficiency, a panel of assays utilizing real time PCR and melting curve analysis using Light Cycler technology was developed. Employing DNA extracted from the original dried blood specimens from newborns identified through prospective newborn screening as presumptive positive for biotinidase deficiency, the specimens were analyzed for the presence of the five common mutations. Using this approach it was possible to separate newborns with partial and complete deficiency from each other as well as from many of those with false positive results. In most cases it was also possible to correlate the genotype with the degree of residual enzyme activity present. In newborn screening for biotinidase deficiency, we have shown that the analysis of common mutations is useful in distinguishing between partial and complete enzyme deficiency as well as improving specificity. Combining biotinidase enzyme analysis with genotypic data also increases the sensitivity of screening for biotinidase deficiency and provides information useful to clinicians earlier than would otherwise be possible. Ó 2003 Elsevier Science (USA). All rights reserved. Keywords: Newborn screening; Mutations; Genetic testing; Biotinidase deficiency
Introduction Biotinidase deficiency is an autosomal recessive metabolic disorder occurring in 1:80,000 live births [1]. Those affected by biotinidase deficiency exhibit irreversible neurological problems, seizures, developmental delay, hypotonia, ataxia, cutaneous, and other symptoms. Symptoms are preventable and in some cases reversible through oral biotin supplementation. Prospective newborn screening for biotinidase deficiency is, therefore, performed in much of the United States and in numerous * Corresponding author. Fax: 1-412-220-0784. E-mail address: [email protected] (E.W. Naylor).
other countries [2]. Biotinidase deficiency results from mutations in the biotinidase gene and depending upon the nature of the mutation(s), the enzyme deficiency may be either complete or partial. Mean biotinidase activity is 7.1 nmol/min/ml serum in normal newborns [3]. Those affected with complete biotinidase deficiency have enzymes that produce <10% of mean normal activity, while those affected with partial deficiency have enzymes that produce 10–30% of normal activity. In newborn screening laboratories, assaying for biotinidase deficiency is performed using an extract of whole blood derived from the universally collected newborn dried blood spot (DBS). Whole blood extract from a DBS specimen is an effective sample from which to assay for
1096-7192/03/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1096-7192(02)00231-7
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
biotinidase activity but not as precise as results obtained with serum. Activity of the biotinidase enzyme may be adversely affected if the DBS specimen is mishandled. DBS specimens that are incompletely dried, exposed to moisture after drying, or exposed to heat may not exhibit biotinidase activity [4]. Inconclusive or ambiguous results in screening for biotinidase deficiency are, therefore, often attributable to errors in sample collection and processing prior to their arrival in the screening laboratory. Another difficulty experienced in newborn screening for biotinidase deficiency involves differentiating between a complete and a partial enzyme deficiency by the screening laboratory using the initial DBS specimen. To aid in distinguishing between complete and partial biotinidase deficiency and subsequently increasing the sensitivity and specificity of screening for biotinidase deficiency mutational analysis has been employed. In the United States, the following five mutations are the most frequently observed in patients with biotinidase deficiency: c98:d7i3, Q456H, R538C, D444H, and the double mutant D444H:A171T. c98:d7i3, Q456H, and R538C are associated with complete biotinidase deficiency [5–7,9]. The D444H mutation has a carrier rate of 3.9% in the general population and when paired with a severe mutation causes partial enzyme deficiency [8]. This high frequency in the general population combined with its causing a partial enzyme deficiency makes the D444H mutation similar to the Duarte D2 N314D variant in galactosemia [9]. Interestingly, when D444H is in cis with the A171T mutation, the combined deleterious effects of both mutations result in a severe allele which when combined a second severe allele will cause complete enzyme deficiency. The D444H:A171T double mutation has been commonly observed in biotinidase deficient children that are ascertained by newborn screening [10]. Using Light Cycler technology and paired hybridization probes, assays were designed to detect these five mutations frequently observed in patients with biotinidase deficiency. Reported here are the assay procedures to detect these five commonly observed biotinidase mutations together with examples of complete and partial biotinidase deficiency derived from our prospective newborn screening program. Examples of how second tier molecular screening can reduce false positives are also presented. This use of mutational analysis supplements the biochemical enzymatic screening results and increases both the specificity and sensitivity of newborn screening for biotinidase deficiency.
Methods Specimens and DNA preparation Biotinidase deficient specimens, ascertained through either routine prospective newborn screening or high-
101
risk screening, were retrieved from the Neo Gen Screening specimen archive. Anonymous DBS specimens were utilized in assay development and specimens from known patients were retrieved from storage and used in the retrospective study. DNA was isolated from DBS specimens as previously described and 80–130 ng were utilized as template in each reaction [11]. Specimens, characterized as homozygous for Q456H, were provided by Adolf Muhl (Department of Pediatrics, University Hospital Vienna, Vienna, Austria). Prospective newborn screening for biotinidase deficiency DBS specimens are routinely submitted by hospitals, physicians, and parents that subscribe to Neo Gen ScreeningÕs Supplemental Newborn Screening program and screening for biotinidase deficiency has been part of that service since 1985. Analysis of biotinidase activity is routinely performed using the Astoria Pacific Continuous Flow Analyzer and the Astoria Pacific SPOTCHECK biotinidase enzyme assay reagents. This method, to assay biotinidase activity in DBS specimens, is based upon a modification of the methods described by Wolf and co-workers [12,13]. Samples demonstrating biotinidase activity below the critical cut-off level of 16.0 eru (enzyme response units, a proprietary unit of Astoria Pacific) were selected for genotype analysis. Polymerase chain reaction and hybridization probes Sequences of the human biotinidase gene (GenBank Accession Nos. NM00060, AF018630, and AF18631) were the basis from which primers and probes were designed. Primer Premier 5 (Premier Biosoft, Palo Alto, CA) and Tm Utility 1.5 IT (Idaho Technology, Salt Lake City, UT) software was utilized to design primers for polymerase chain reaction (PCR) and hybridization probes to detect mutations. Primers and fluorescent labeled probes were obtained from either Operon Technology (Alameda, CA) or Idaho Technology (Salt Lake City, UT). PCR buffers were obtained from Idaho Technology (Salt Lake City, UT). Amplification reactions utilize 1 PCR buffer, 2 mM MgCl2 , 200 lM dNTPs (Roche, Mannheim, Germany), and 0.6 U Klen taq (AB Peptides, St. Louis, MO) in a complex with TaqStart antibody (ClonTech, Palo Alto, CA). Preparing a complex between the polymerase and TaqStart antibody is performed according to manufacturerÕs instructions. The sequence of individual primers, the sequence of fluorescent hybridization probes, and the concentration at which each is used are found in Table 1. Ten microliter PCRs were performed in capillary tubes using a Roche Light Cycler (Mannheim, Germany). Temperature cycling conditions for PCR utilizes a modified two-step thermal cycling scheme. Specimens are ramped to 94 °C at 20°/s and held there for 0 s to
0.13
DNA sequences are shown 50 to 30 , Concentrations are in micromolar units, Anchor probes are utilized at 0.2 lM, detection probes are utilized at 0.1 lM, P, phosphate; fitc, fluoroscene isothiocynate; LCred640, Light Cycler red 640.
0.25
LCred640-GCTTGCTCTT TTCCTCTGCG-P TGGTGACCAATCTTGG GACA-fitc 0.5 1.0
ACAGGTGTCGAAGCC AAGAC TCCATTATTATTGCTGA ACACGAC A171T
R538C
GGGGAAAGGAAGGCT ATCTC CTCCAGCGCCTGAGTT GTAT
0.5
0.5 Q456H
D444H
1.0
1.0
Detection probe
LCred640-GCTTGCTCTTT TCCTCTGCG-P LCred640-ACTACATCCA CGTGTGTGCCC-P CTCTATGGGCGCTTGT ATGA-fitc
Concentration
Anchor probe
CTCATACACGGCAGCC ACAT GGTGTCGAAGCCAAG ACCC CTTGTAGCCTGTGGA AGTGC
Reverse primer
GCCCCATTACATTCCA GATTTG GCCCACCTTATCCAAA GAGC GCTTGGCTGGGAGAAT GACC 7d3i
0.5
Forward primer Assay
Concentration
1.0
TGGTCTGCATTATGTCT GGAGCCAGAAGTA-fitc TTTGATGGGCTTCACA CAGTACATGGCACT-fitc LCred640-AGGGACTAG GAAAAGTGTGTGGTCT GTGG-P TGGTCTGCATTATGTC TGGAGCCAGAAGTA-fitc LCred640-AGGAGCCTTG TCATAGCAGTGACCCA AGGT-P
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
Table 1 Primers for polymerase chain reaction, hybridization probes for mutation analysis, and concentrations at which they are utilized
102
denature the DNA strands. Temperature then ramps at 20°/s to 58 °C and holds at this temperature for 15 s at which time primers anneal and polymerization of new DNA begins. Polymerization is completed while ramping from 58 to 72 °C at 1.0°/s. The slow ramp speed allows polymerization to precede, thus negating the necessity of a hold time at 72 °C. Thermal cycling is repeated for 45 cycles. All amplifications are performed in an asymmetric manner. Asymmetric amplifications for c98:d7i3, Q456H, R538C, and A171T assays enrich the antisense strand of the amplicon while the asymmetric amplification in the D444H assay enriches the sense strand of the amplicon. Asymmetry produces an excess of the DNA strand to which the hybridization probes will bind in the analysis phase of the assay. Hybridization probes and genotyping analysis Genotyping is performed using paired hybridization probes, where each assay has a detection probe and an anchor probe [14–23]. Probes for the c98:d7i3, Q456H, R538C, and A171T assays hybridize to the antisense strand of the amplicon while the probes for the D444H assay hybridize to the sense strand of the amplicon. The detection probe hybridizes with a region of the amplicon that includes the mutation, while the anchor probe hybridizes with a region adjacent to the detection probe. When both probes are hybridized, there is a 1 base gap between the anchor and detection probes. For each set of hybridization probes, one is conjugated on the 30 end with fitc while the second is conjugated on the 50 end with LCred640 (see Table 1). The probe which is 50 conjugated with LCred640 is also 30 phosphorylated to prevent extension by Taq DNA polymerase. When both probes are hybridized with the amplicon, the fluorescent moieties are brought into close proximity, and this proximity allows fluorescence resonance energy transfer to occur between the donor fluorophore (fitc) and the acceptor fluorophore (LCred640). Anchor probes have a Tm that is at least 15% higher than the corresponding detection probe, which allows the anchor probe to remain hybridized during the melting transition of the detection probe that occurs during the analysis phase of the assay. After completing the thermal cycling program, the Light Cycler proceeds seamlessly to the analysis program. The analysis program ramps to 94 °C at 20°/s and after reaching 94 °C, immediately begins to ramp at l°/s to 35 °C. Upon reaching 35 °C the temperature ramps upward at 0.1°/s to 76 °C. During the entire analysis program, the excitation wavelength of fitc is provided and the fluorescence of LCred640 is continuously acquired. Melting curves are generated by plotting the fluorescence of LCred640 against temperature during the 35–76 °C upward temperature ramp. Melting peaks are generated computationally by calculating the dF =dT of the melting curve which is then plotted against temperature.
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
Results Detecting mutations in the biotinidase gene Figs. 1A–E display analysis of individual biotinidase mutations using melting peaks generated with the Light Cycler. Fig. 1A displays the assay results for the R538C
Fig. 1. Wild type specimens: ———, heterozygous specimens:
103
mutation and specimens that are homozygous wild type, heterozygous, and a no DNA control. No specimen that is homozygous for R538C has yet been identified by Neo Gen Screening and we were unable to obtain one from outside sources. The remainder of the assays, shown in Figs. 1B–E display specimens that are homozygous wild type, heterozygous, homozygous for the mutation, and a
, homozygous mutant specimens:
, no DNA control: .......
104
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
Fig. 1. (continued)
no DNA control. In the cases of the D444H, c98:d7i3, and R538C assays, the detection probe matches the wild type sequence. Therefore, the high temperature melting peak represents the wild type form of the gene while the lower temperature melting peak represents the mutant form of the gene. In the A171T and Q456H assays, the detection probe matches the mutant form of the gene and has a 1 bp mismatch with the wild type allele. In these assays, the high-temperature melting peak represents the mutant form of the gene while the low-temperature melting peak represents the wild type form of the gene. Analysis of specimens identified through newborn screening as presumptive positive for biotinidase deficiency Through prospective newborn screening of over 1.4 million newborns, Neo Gen Screening has identified numerous newborns with complete and partial biotinidase deficiency. These cases have been been referred to metabolic treatment centers where confirmatory quantitative serum biotinidase studies are performed. In addition to
those cases that are confirmed, a number of specimens have been identified as presumptively positive for biotinidase deficiency and many of these cases have turned out to be false positives due to environmental factors or mildly reduced enzyme activity resulting from homozygosity for the D444H mutation, or from heterozygosity for a severe mutation. With improvement in the sensitivity of the newborn screening methodologies the number of presumptive positives has increased. It is therefore important to utilize a second tier molecular assay to identify which of these increased presumptive positives are true positives and which are false positives. Table 2 lists the molecular results from nine confirmed complete biotinidase deficient patients. Six of these were screened between 1 and 3 days of age while the other three were high risk patients screened between 1 and 6 months of age. Six of these patients were homozygotes or compound heterozygotes with two copies of severe mutations, two had only one copy of one of the severe mutations that we screened for, and one had no copies. These last three cases, however, were confirmed as having complete deficiency as a result of very low
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107 Table 2 Genotypes of newborns and high-risk patients with complete biotinidase deficiency Patient Nos.
Sample date
Allele 1
Allele 2
1.
2 days
R538C
2.
2 days
D444H A171T D444H A171T c98:d7i3 c98:d7i3 c98:d7i3 Q456H Q456H Q456H Q456H ?
3. 4. 5. 6. 7. 8. 9. 10.
3 months 5.5 months 1 day 2 days 2 days 2 days 2 days 1 month
D444H A171T c98:d7i3 c98:d7i3 c98:d7i3 Q456H ? ? ? ?
Table 3 Genotypes of newborns and high-risk patients with partial biotinidase deficiency Patient Nos.
Sample date
Allele 1
Allele 2
1.
2 days
D444H
2.
2 days
3.
2 days
4. 5. 6. 7. 8. 9. 10. 11. 12.
2 days 2 days 1 day 2 days 3 days 2 days <1 day 14 days 2 days
D444H A171T D444H A171T D444H A171T c98:d7i3 c98:d7i3 Q456H Q456H Q456H Q456H Q456H R538C R538C Variant
D444H D444H D444H D444H D444H D444H D444H D444H D444H D444H D444H
screening and confirmatory enzyme activity. Table 3 lists 12 newborns screened between the ages of 1 and 14 days with confirmed partial biotinidase deficiency together with their respective genotypes. All have one severe mutation in combination with one copy of the D444H mild mutation. Among our false positives we have encountered a number of newborns homozygous for the D444H common mutation, heterozygous for a severe mutation, and some heterozygous for the D444H mild mutation. These false positives have mildly reduced enzyme activity on initial screening and normal or lownormal activity on repeat screening. In these cases, second tier molecular testing has been useful in reducing the repeat request rate which previously would have been based solely on the initial enzymatic screen.
Discussion Newborn screening with its deep roots in biochemical screening methods has been slow to embrace molecular
105
genetic analysis. With the genes responsible for many genetic diseases having been characterized and common causative mutations identified, incorporating second tier molecular testing into newborn screening is a natural progression. Recently, many newborn screening programs have expanded the list of disorders for which they screen. This expansion is due in large to broader application of tandem mass spectrometry and the publicÕs demand for more comprehensive analysis [24,25]. With this increased number of comprehensive newborn screening programs, the need for second tier molecular genetic testing will also increase. The first broad application of molecular genetic methods in newborn screening was as a second tier assay in screening for cystic fibrosis. The primary screen for cystic fibrosis is quantification of immunoreactive trypsinogen (IRT), however analysis of IRT alone is plagued with a high false positive rate. When second tier testing for the presence of the common cystic fibrosis transmembrane conductance regulator mutation (DF508) was performed on specimens having elevated IRT, the false positive rate fell significantly [26,27]. This was the first demonstration of a clear place for second tier molecular genetic analysis in newborn screening. Any assay that involves measuring enzyme activity is highly dependent upon proper sample handling and newborn screening assays are not exempted from this. Despite the fact that birthing centers have collected newborn screening samples for over 35 years, sample mishandling is still a persistent problem. Newborn screening for galactosemia, glucose-6-phosphate dehydrogenase (G-6-PD) deficiency, and biotinidase deficiency involve measuring enzyme activity. Improper sample collection, processing, and transportation have been demonstrated to be deleterious to the stability of these enzymes in the assays employed by newborn screening laboratories [4,29,30]. While it is the experience of this laboratory that the biotinidase enzyme is less sensitive to external factors than either galactose-1-phosphate uridyl transferase (GALT) or G-6-PD, environmental factors can still have an affect upon biotinidase activity and cannot be ignored. The physical appearance of the DBS specimen or a long delay between the date of specimen collection and its being received by the newborn screening laboratory are signs of possible mishandling. If multiple enzyme activities are deficient and mutational analysis of biotinidase and other genes (GALT, G-6-PD) shows no DNA sequence aberrations, it provides some evidence that the specimen was affected by environmental factors such as heat or humidity, and a replacement specimen should be requested. In newborn screening for biotinidase deficiency, it is frequently difficult to discern if a partial or complete enzyme deficiency has been encountered based solely on the initial enzymatic screening level. In the vast majority
106
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107
of partial deficiencies, the D444H mutation is involved [8]. D444H has a carrier frequency of 3.9% in the general United States population and reduces the activity of the biotinidase enzyme by 48–52% [8,10,31]. It is noteworthy, that these percent reductions were determined with serum quantitative enzyme analysis, thus the percent enzyme reduction in a DBS derived whole blood specimen could be different. Partial deficiencies generally result from compound heterozygosity with one D444H mutation and a second severe mutation. In Table 3, there are 12 specimens with genotypes identifying them as partial deficiencies. Five specimens are compound heterozygous between D444H and Q456H; three with the D444H:A171T double mutation; two with c98:d7i3; and two with R538C. In the enzyme assay used in newborn screening, compound heterozygotes containing D444H and a mutation causing a complete deficiency (c98:d7i3, Q456H, etc.) may initially generate biochemical data mimicking complete biotinidase deficiency. A similar situation is frequently observed in the Beutler assay that is used to measures GALT activity when screening for galactosemia. Compound heterozygotes between the Duarte D2 N314D variant and a classical galactosemia mutation such as the common Q188R may initially generate biochemical data suggesting classical galactosemia. Identifying the N314D GALT mutation provides definitive proof that these newborns do not have classical galactosemia [28]. In a similar fashion, the D444H mutation is responsible for the vast majority of partial biotinidase deficiencies and therefore identifying this mutation provides strong evidence that a complete enzyme deficiency is not present. A complication surrounding the D444H mutation is the presence of double mutants. Three double mutations, involving D444H and a second mutation on the same gene, have been described. The commonly observed double mutant, D444H:A171T, that accounts for 17.3% of the mutations in complete biotinidase deficiencies ascertained by newborn screening, is part of this panel [10]. Two other double mutants, D444H:F403V and D444H:R157H, have been described, however, both are extremely rare [31]. The D444H mutation is very useful in helping to identify partial deficiencies, but the possibility of a double mutant resulting in a complete deficiency cannot be dismissed. After newborn screening results are reported and repeat specimens are requested and screened, the first clinical visit of a potentially biotinidase deficient newborn will involve determining quantitative serum biotinidase activity and possibly additional confirmatory molecular testing. Quantitative biotinidase analysis is the ultimate confirmation test to establish complete or partial biotinidase deficiency. Second tier mutation analysis is used to improve the sensitivity and specificity of the newborn screening program and to assist in clinical evaluation. However, in certain situations, the genotype data unambiguously
identified these specimens as having a complete enzyme deficiency. Such informative results can expedite patient care by bringing the newborn to immediate attention. The data shown in Tables 2 and 3, and discussed above, provide evidence to the utility of second tier mutation analysis in newborn screening for biotinidase deficiency. In a high throughput newborn screening laboratory, the most important issue is validity of results, but following closely behind is turn around time. Minimizing turn around time requires that assay platforms be fast, reliable, and easily interpreted within the context of a routine service laboratory. The Light Cycler platform is ideal for the high throughput newborn screening laboratory because all of those criteria are met. Air driven thermal cycling is fast, genotyping with fluorescent hybridization probes is simple because it involves no post-PCR manipulation, and melting peak data are easily interpreted. From isolation of DNA to data interpretation, the 5-mutation panel described here is completed in less than 2 h at a relatively low cost. At Neo Gen Screening second tier molecular testing is provided at no additional cost for newborns receiving routine newborn screening. Such rapid analysis assures that supplemental molecular data is reported along with the primary biochemical screening data. An additional benefit is that the closed tube format simplifies sample tracking and is favorable for avoiding amplicon contamination in the laboratory. Data files from the Light Cycler are easily stored and may be backed up in an offsite archive rendering them safe from loss. This is an ideal situation for the newborn screening laboratory where large quantities of sensitive clinical data are generated. Currently, 55 mutations that cause biotinidase deficiency have been identified [31]. Among specimens of domestic origin identified at Neo Gen Screening, the panel of five mutations proved useful in a very high percentage of presumptive positive newborns. Enzymatic analysis will remain the primary means by which biotinidase deficiency is detected in both newborn screening and high risk clinical diagnostics. Second tier mutation analysis provides valuable support to the enzymatic analysis and should be considered as a supplement to the biochemical data by those performing newborn screening for biotinidase deficiency.
References [1] B. Wolf, Worldwide survey of neonatal screening for biotinidase deficiency, J. Inher. Met. Dis. 14 (1991) 923–927. [2] National Newborn Screening and Genetics Resource Center http://genes-r-us.uthhscsa.edu. [3] B. Wolf, R.E. Grier, J. Secor-McVoy, G.S. Heard, Biotinidase deficiency: a novel vitamin recycling defect, J. Inher. Metab. Dis. 8 (Suppl. 1) (1985) 53–58.
S.F. Dobrowolski et al. / Molecular Genetics and Metabolism 78 (2003) 100–107 [4] J.R. Secor McCoy, H.C. Levy, M. Lawler, M.A. Schmidt, D.D. Ebers, P.S. Hart, M.G. Pettit, B. Wolf, Partial biotinidase deficiency: clinical and biochemical features, J. Pediatr. 115 (1) (1990) 78–83. [5] K.J. Norrgard, R.J. Pomponio, K.L. Swango, J. Hymes, T.R. Reynolds, G.A. Buck, B. Wolf, Mutation (Q456H) is the most common cause of profound biotinidase deficiency in children ascertained by newborn screening in the United States, Biochem. Mol. Med. 61 (1997) 22–27. [6] R.J. Pomponio, K.J. Norrgard, J. Hymes, T.R. Reynolds, G.A. Buck, R. Baumgartner, T. Suormala, B. Wolf, Arg538 to Cys mutation in a CpG dinucleotide of the human biotinidase gene is the second most common cause of profound biotinidase deficiency in symptomatic children, Hum. Genet. 99 (1997) 506–512. [7] R.J. Pomponio, T.R. Reynolds, H. Cole, G.A. Buck, B. Wolf, Mutation hotspot in the human biotinidase gene causes profound biotinidase deficiency, Nature Genet. 11 (1995) 96–98. [8] K.L. Swango, M. Demirkol, G. Huner, E. Pronicka, J. SykutCegielska, A. Schulze, B. Wolf, Partial biotinidase deficiency is usually due to the D444H mutation in the biotinidase gene, Hum. Genet. 102 (1998) 571–575. [9] H.C. Lin, L.T. Kirby, W.G. Ng, J.K.V. Reichardt, On the nature of the Duarte variant of galactose-1-phosphate uridyl transferase (GALT), Hum. Genet. 93 (1994) 167–169. [10] K.J. Norrgard, R.J. Pomponio, K.L. Swango, J. Hymes, T.R. Reynolds, G.A. Buck, B. Wolf, Double mutation (A171T and D444H) is a common cause of profound biotinidase deficiency in children ascertained by newborn screening in the United States, Hum. Mutat. 11 (1998) 410–414. [11] E. Heath, D. OÕBrien, R. Banas, S.F. Dobrowolski, Optimization of an automated protocol to isolate DNA from Guthrie cards, Arch. Pathol. Lab. Med. 123 (1999) 1154–1160. [12] G. Herd, J.S. McVoy, B. Wolf, A screening method for biotinidase deficiency in newborns, Clin. Chem. 30 (1) (1984) 125–127. [13] A. Denise, D. Pettit, P.D. Amadorand, B. Wolf, The quantitation of biotinidase activity in dried blood spots using microtiter transfer plates: identification of biotinidase deficient and heterozygous individuals, Anat. Biochem. 170 (1989) 371–374. [14] M.J. Lay, C.T. Wittwer, Real-time fluorescence genotyping of factor V Leiden during rapid-cycle PCR, Clin. Chem. 43 (1997) 2262–2267. [15] P.S. Bernard, M.J. Lay, C.T. Wittwer, Integrated amplification and detection of the C677T point mutation in the methylenetetrahydrofolate reductase gene by fluorescence resonance energy transfer and probe melting curves, Anal. Biochem. 255 (1998) 101–107. [16] P.S. Bernard, R.S. Ajioka, J.P. Kushner, C.T. Wittwer, Homogeneous multiplex genotyping of hemochromatosis mutations with fluorescence probes, Am. J. Pathol. 153 (1998) 1055–1061. [17] E. Lyon, A. Millson, T. Phan, C.T. Wittwer, Detection and identification of base alterations within the region of factor V
[18]
[19]
[20]
[21]
[22]
[23]
[24] [25] [26]
[27]
[28]
[29] [30]
[31]
107
Leiden by fluorescence melting curves, Mol. Diag. 3 (1998) 203– 210. C.N. Gundry, P.S. Bernard, M.G. Herrmann, G.H. Reed, C.T. Wittwer, Rapid F508 del and F508C assay using fluorescent hybridization probes, Genet. Test 3 (1999) 365–370. N. von Ahsen, E. Schultz, V.W. Armstrong, M. Oellerich, Rapid detection of prothrombic mutations of prothrombin (G20210A), factor V (G1691 A), and methylenetetrahydrofolate reductase (C677T) by real-time fluorescence melting curves, Mol. Diag. 45 (1999) 694–696. K. Mangassner-Stephan, C. Tag, A. Reiser, A.M. Gressner, Rapid genotyping of hemochromatosis gene mutations on the Light Cycler with fluorescent hybridization probes, Clin. Chem. 45 (1999) 1875–1878. C. Aslanidis, G. Schmitz, High-speed apolipoprotein E genotyping and apolipoprotein B3500 mutation detection using real-time fluorescence PCR and melting curves, Clin. Chem. 45 (1999) 1094–1097. S.F. Dobrowolski, C.T. Wittwer, C. Gundry, E.W. Naylor, Detection of CFTR DF508 using rapid cycler PCR and analysis of FRET probes in Neonatal Screening for cystic fibrosis, 115– 121. M.G. Herrmann, S.F. Dobrowolski, C.T. Wittwer, Rapid bglobin genotyping by multiplex probe melting temperature and color, Clin. Chem. 46 (2000) 425–428. D.H. Chace, Mass spectrometry in the clinical laboratory, Chem. Rev. 101 (2000) 445–477. E. Marshall, Fast technology drives new world of newborn screening, Science 294 (2001) 2272–2274. W.C. Spence, J. Paulus-Thomas, D.M. Orenstein, E.W. Naylor, Neonatal screening for cystic fibrosis: addition of molecular diagnosis to increase specificity, Biochem. Med. Metab. Biol. 49 (1993) 200–211. R.G. Gregg, A. Simantel, P. Farrell, et al., Newborn screening for cystic fibrosis in Wisconsin: comparison of biochemical and molecular methods, Pediatrics 99 (6) (1997) 819–824. S.F. Dobrowolski, R.A. Banas, J.G. Suzow, M. Berkley, E.W. Naylor, Analysis of common mutations in the galactose-1phosphate uridyl transferase gene: new assays to increase the sensitivity and specificity of newborn screening for galactosemia. J. Mol. Diag. (2002) (in press). P. Polana, S.F. Dobrowolski, E.W. Naylor, unpublished observation, 2002. V.E. Shih, H.L. Levy, V. Karolkewicz, S. Houghton, M.L. Efron, K.J. Isselbacher, E. Buetler, R.A. MacCready, Galactosemia screening in Massachusetts, N. Eng. J. Med. 284 (1971) 753–760. K.J. Norrgard, R.J. Pomponio, J. Hymes, B. Wolf, Mutations causing profound biotinidase deficiency in children ascertained by newborn screening in the United States occur at different frequencies than in symptomatic children, Pediatr. Res. 46 (1) (1999) 20–27.