Contents lists available at SciVerse ScienceDirect
Thrombosis Research
j ourna l homepage: www.e lsev ie r .com/ locate / thromres
Regular Article
Identification of an ancient haemophilia A splice site mutation
Sylvia Reitter-Pfoertner a,⁎,1, Arndt von Haeseler b, Birgit Horvath c, Raute Sunder-Plassmann c, Vera Tiedje a,Ingrid Pabinger a, Christine Mannhalter c
a Division of Haematology and Haemostaseology, Department of Medicine I, Medical University of Vienna, Austriab Center for Integrative Bioinformatics Vienna, Max F. Perutz Laboratories, University of Vienna, Medical University of Vienna, University of Veterinary Medicine Vienna, Austriac Department of Laboratory Medicine, Medical University of Vienna, Austria
⁎ Corresponding author at: Division of Haematologyment of Medicine I, Medical University of Vienna, WaehVienna. Tel.: +43 1 40400 2757; fax: +43 1 40400 4030
E-mail address: sylvia-elisabeth.reitter@meduniwien1 Dr. Reitter-Pfoertner is a recipient of the Bayer
Received 29 December 2011Received in revised form 9 February 2012Accepted 10 February 2012Available online 6 March 2012
Keywords:Splie site mutationHaemophiliaFounder
Introduction: To date, numerous mutations resulting in haemophilia A are known and recorded at HAMSTeRS.We identified a new splice site mutation in intron 6 of the F8 gene (T to G transition at position −14;c.788-14T>G) in seven not knowingly related patients, who all suffer from mild haemophilia A. RNA analysisof blood cells indicated that this mutation leads to the preferred generation of a transcript lacking the completeexon 7 (without frameshift).Methods: To determinewhether themutation represented a foundermutationwe analyzed intragenic (intronic)and extragenic short tandem repeat (STR) regions and constructed haplotypes in the 7 patients and 128 appar-ently healthy male control individuals.Results: In the 128 healthy control individuals, 109 different haplotypes were found. Surprisingly, also the 7
patients carried 3 different haplotypes. However, by genealogy reconstruction using BATWINGwe could identifyan ancestral haplotype on which the mutation apparently occurred. This haplotype - DXS9897:12-DXS1073:21-HA472:64-DXS1108:26 - was frequent and was found in three patients, but was also present infour control individuals who did not carry the splice site mutation.Conclusion:Our data indicate that the splice sitemutation occurred in an individual carrying a relatively commonhaplotype. While the mutation was passed on through generations, the haplotypes identified in the sevenpatients derived from this founder haplotype but were changed by later mutations in the STR regions.
Haemophilia A, a recessive X-linked bleeding disorder characterizedby factor VIII (FVIII) deficiency, is generally caused by mutations in theF8 gene. The gene is located at the guanine-cytosine-rich terminal re-verse band on the long arm of the X-chromosome at Xq28. The wholegene spans a length of 186 kb, consists of 26 exons, has a coding regionof 9 kb and encodes amature protein of 2,332 amino acids [1]. Overall, alarge number (>1,200) of different mutations in the F8 gene have beenfound to cause haemophilia. These are intron inversions, stop and mis-sense mutations, deletions, insertions and splice site mutations. Causa-tive mutations are recorded at HAMSTeRS [2], which serves as areference data base for haemophilia A diagnostic laboratories all overthe world. The most common mutation causing approximately40 – 50%of severe haemophilia A is the intron 22 inversion, an inversioncreated by homologous recombination of intron 22 and related
and Haemostaseology, Depart-ringer Guertel 18–20, A-1090..ac.at (S. Reitter-Pfoertner).Haemophilia Clinical Training
rights reserved.
sequences outside the F8 gene [3]. The second most common mutationis the inversion in intron 1with a prevalence of about 5% [4]. Other mu-tations, such asmissensemutations, nonsensemutations, deletions andinsertions are reported in all regions of the F8 gene. A number of indi-vidual mutations have been located in introns of the F8 gene. Recently,80 new, intragenic and extragenic STR loci on Xq28 were published [5].Furthermore, several genetic variants have been identified in the F8gene – some single nucleotide polymorphisms and some microsatel-lites. These polymorphic STR loci located in intragenic and extragenicregions, both upstream and downstream of the F8 gene, have beenreported and validated for indirect carrier-tracking in haemophilia A.They are a useful tool for monitoring inheritance of genetic alterationsand are still used for FVIII linkage analysis which, until recently, wasthe most commonly used technique for carrier detection. Now, thisapproach has been superseded by direct mutation analysis [6].However, the highly polymorphic STR markers are still helpful todiscriminate whether a mutation represents a mutation hotspot or afounder mutation, and they can also be used to trace patients' ancestry.
A study among 240 Austrian patients on the spectrum of haemophi-lia A mutations identified over 50 thus far not reported mutations [7].One of these was a novel splice site mutation in intron 6 of the F8gene (c.788-14T>G), which appeared in seven not knowingly relatedpatients, all suffering from mild haemophilia A.
446 S. Reitter-Pfoertner et al. / Thrombosis Research 130 (2012) 445–450
In the present study, we investigated whether this new mutationrepresents a thus far unrecognized mutation hotspot or appeared asa result of common ancestor genes (founder mutation).
Patients and Methods
Patients and Controls
Our study included seven patients with mild haemophilia A (FVIII:C0.05-0.25 IU/ml),whowere referred formutation analysis and inwhoma novel splice site mutation in intron 6 was identified as the causativemutation. All patients had a mild phenotype with no relevant bleedingevents; in particular, one patientwas diagnosed at the age of >50whensurgery became necessary. In the patients' family history, no closerelatives with bleeding tendency have been reported.
According to patient interviews, the patients were not knowinglyrelated.
We have also tested 128 male control individuals without knownbleeding problems.
All patients and control individuals gave their written informedconsent to DNA analysis. The studywas approved by the ethics commit-tee of the Medical University of Vienna.
Methods
DNA was extracted from venous blood anticoagulated with EDTAusing the MagNA Pure DNA-isolation system (Roche Diagnostics,Mannheim, Germany). Total RNA was isolated from leucocytes withthe RNeasy® Mini Kit (Qiagen, Hilden, Germany) on a QIAcube instru-ment (Qiagen, Hilden, Germany). In both cases, the manufacturer's in-structions were followed. RNA was reversely transcribed into cDNAusing the High Capacity cDNA Reverse Transcription Kit (ABI, FosterCity, CA, USA). For cDNA transcription, 2 μg of RNAwere used. The reac-tionwas performedwith the following temperature protocol: first, RNAwas incubated at 22 °C, then a temperature of 37 °C wasmaintained for45 minutes, followed by 5 minutes at 85 °C. Nested RT-PCR was per-formed essentially as described in El-Maarri et al. [8]. in an EppendorfCycler (Eppendorf, Hamburg, Germany) using 5 μl cDNA for the firstPCR and 2 μl of the resulting PCR product for the second PCR. For thefirst PCR, the PCR program was set to initial denaturation at 95 °C for10 min., followed by 30 cycles of amplificationwith 30 seconds of dena-turation at 95 °C, 30 seconds annealing at 52 °C, 1 minute elongation at72 °C and 7 minutes final extension at 72 °C. For the second PCR, thesame PCR protocol was used. The annealing temperature for the secondPCR, however, was 58 °C and there were 35 cycles of amplification. Theamplicons were evaluated on a 5% Criterion gel (Bio-Rad LaboratoriesGmbH, Munich, Germany).
Sequencing was performed as described below.All primers used for PCR amplifications were synthesized and PAGE-
purified by VBC Biotech (Vienna, Austria). The primer sequences andthe PCR protocols used in our laboratory can be provided upon request.
Table 1Intragenic (intronic) and extragenic markers used for linkage analysis with the respective
Locus name Concensus pattern Forward sequence (5′ ➔ 3′)1
Int1 AC CTG CCC TTG GAC ATA AGC ATInt6 TG TTC TCC TGC TTC AGC CTC TCInt 9.1 AG AGA TTC GAG CGA TTC TCC TGInt 22 GT AAG ACC CTT AGC TGT TTC ATA AGCInt 25.3 TG TCC AAG ATC AAG GGG TAG GCDXS 9897 CTAT TTC TGC TGT GCA ATA CAT CTG ADXS 1073 TG AAG AAT GCC CTC TCC GAG TTHA 472 CTT GCT CCT TTG ATT GGA TAA TTT CADXS 1108 CA GGG AGA TAG GAA TGA TGG AGT G
1 Machado FB,Medina-Acosta E. High resolution combined linkage physical map of short tanHaemophilia (2009), 15, 297-308.
Regarding the primers used for sequencing of exon 7 (genomic DNA) aswell as for sequencing of the cDNA, see Table 1.
All coding regions as well as the exon/intron boundaries and partof the promoter region were analyzed.
PCR products generated from genomic DNA and cDNA weresequenced following treatment with ExoSAP-it (USB, Cleveland, OH,USA). The sequencing reaction was performed with the Big Dye Termi-nator cycle sequencing kit (ABI, Foster City, CA, USA) using 5 pmol ofeach primer by applying an ABI 3130xl Genetic Analyzer. The bestpractice of DNA sequencing [9] was followed.
Analysis of Short Tandem Repeats (STRs)
Patients and control individuals were tested for the following previ-ously published extragenic STRs (in 3′➔ 5′ order): DXS 9897, DXS 1073,HA 472 and DXS 1108. Regarding the specific localisation of these STRs,see Table 1. All STRs were repeatedly tested in order to confirm therepeat number.
For each PCR, 1.7 pmol forward and reverse primer (Eurofins MWGOperon, Ebersberg, Germany) were used, respectively. The PCR pro-gram was set to initial denaturation at 95 °C for 10 min., followed by28 cycles of amplification with 45 seconds of denaturation at 95 °C,45 seconds annealing at 57 °C, 45 seconds elongation at 72 °C and 7 -minutes final extension at 72 °C. The ampliconswere tested on a Sprea-dex EL400 Wide Mini gel (Elchrom Scientific AG, Cham, Switzerland).
A 1.5 μl aliquot of the PCR product was added to 10 μl of Hi-Di Form-amide (ABI, Foster City, California, USA). Then, 0.3 μl size standardGeneScan 400HD (ABI, Foster City, California, USA) was added and themixture was heated at 95 °C for 2 minutes to denature the amplicon.Fragments were analysed at 60 °C on an ABI Prism 3100 GeneAnalyzerusing a 50-cm capillary array and POP 6 polymer (all from ABI, FosterCity, California, USA). The Genotyper software version 3.7 (ABI, FosterCity, California, USA) was used for data analysis.
Statistical Analysis
The analysis is based on four STRs. We followed the approach out-lined by Zivelin et al. [10,11]. However, the chi-square (χ2)-test wasadapted to multiple alleles. To check the reliability of the χ2-approxi-mation, we drew 10,000 times random samples with replacement ofseven (corresponding to the number of patients) from the allele fre-quency distribution obtained in the 128 control individuals to approxi-mate the chi-square distribution. The p-value of the observed χ2-valuewas then estimated from the simulated distribution. Finally, the datawas subject to a genealogy reconstruction using BATWING [12]. Fromthe resulting 5,000 trees a consensus tree was computed. Using a max-imumparsimony approach [13], thenumber ofmutations leading to thedifferent repeat numbers of the four STRs found in the patients and con-trol individuals could be evaluated. A symmetric step-wise mutationmodel was used, where each STR gains or loses one repeat unit.
primers.
Reverse sequence (5′ ➔ 3′)1 Physical position
CCA TAT GAT CCA GCA ACT CG within the F8 geneAGC ATA TCC ACC CTC ACC AC within the F8 geneCAG TCA TTG CTG TGG GTT TG within the F8 geneTTC ATA CAG TGG GAT CAT TCA TT within the F8 geneGCC TGG ACT ACA GAG GGA GA within the F8 geneCAG CAG ACA TTA TTG AGG GAG A approx. 1596 kb 3′ of the F8 geneATT GGT GGC CTT TGA AAC AC approx. 235 kb 3′ of the F8 geneTGC CTC AAC ATC AGA ATA GAC C approx. 163 kb 3′ of the F8 geneTAT TTT CTG GGC CAT CTT GG approx. 610 kb 5′ of the F8 gene
dem repeat loci on human chromosome Xq28 for indirect haemophilia A carrier detection.
447S. Reitter-Pfoertner et al. / Thrombosis Research 130 (2012) 445–450
Results
Mutation Analysis
By sequencing of the F8 gene we identified a missense mutation(T➔G) at position −14 in intron 6 of the F8 gene in proximity to theacceptor splice site (c.788-14T>G) in seven patients. This mutation isnot listed at HAMSTeRS and to the best of our knowledge, has notbeen published elsewhere. We did not find any other mutation or anypolymorphism in the coding region of the F8 gene.
In 5 patients sequence analysis of mRNA/cDNA from leucocytescould be performed. The data indicated that the T ➔ G mutation at po-sition−14 in intron 6 (c.788-14T>G) leads to preferred use of anothersplice site and results in the loss of the complete exon 7 (without frame-shift). Analysis of the leucocyte-derived cDNA in these 5 patientsshowed no expression of full-length mRNA (ie. mRNA containingexon 7).
Analysis of STRs
Regarding the analyzed intragenic (intronic) STRs (for details seeTable 1), all seven patients carried the same genotype. Following theadvice of our statistician, thesemarkerswere not included in the genea-logic analyses.
Regarding details of the four analysed extragenic STRs (concensuspattern, sequence and physical position) see Table 1.
Fig. 1 displays the allele frequencies of these STRs among patientsand controls.
For further analysis, we calculated the expected numbers for thedifferent alleles of the four STRs and compared these to the observednumbers in our patient cohort. Interestingly, two theoretically possi-ble alleles of the locus DXS1108 were not observed and could
Fig. 1. a. Allele frequencies of DXS9897. b. Allele frequencies of DXS1073. c. Allele frequencieof patients and control individuals, respectively.
therefore not be included in the further analysis. As pointed out byCochran [14] the approximation of the χ2 distribution may not beadequate, because some expected STR counts are less than one.Thus, we carried out a bootstrap procedure to obtain a simulated χ2
distribution for each STR. Regarding the detailed results of thisanalysis, see Table 2.
From the simulated χ2-square distribution (as outlined in theMethods section) the following p-values have been obtained: forDXS9897 0.114, for DXS1073b1∙10−5, for HA472 0.029, for DXS1108b1∙10−5, respectively.
Table 3 compares the results for the χ2-value, the p-value and thesimulated p-value for the four different STRs. Three of the four STRallele frequencies found in the patients clearly differ from theexpected distributions derived from the findings in the control indi-viduals. This is reflected by simulated p-values below 0.05, which isregarded as statistically significant. Merely, the locus DXS9897(the most remote STR on the 3′-side of the F8 gene) does not showa significant deviation from the expected distribution.
Haplotypes in Patients and Controls
Based on the four extragenic STRs, haplotypes for patients as wellas for control individuals were generated.
In the patients, three different haplotypes characterized by thefollowing repeat numbers could be identified; namely, 12-22-63-26,12-22-64-26 and 12-21-64-26 (see Table 2). At the locus DXS1073,two different alleles were found in the patients – the 21-repeat allelein three patients and the 22-repeat allele in four patients. Remarkably,the haplotypes with 22 repeats at locus DXS1073 were not identicalbut belonged to two different clades.
Also, for the locus HA472, two different alleles were found in the 7patients: 63 repeats in three patients and 64 repeats in four patients.
s of HA472. d. Allele frequencies of DXS1108. x-axis: number of repeats; y-axis: number
Table 2Observed number of alleles in the patients and expected number of alleles inferred from allele frequencies from the healthy individuals.
DXS9897 DXS1073 HA472 DXS1108
Allele / productlength
Observed Expected Allele /productlength
Observed Expected Allele /productlength
Observed Expected Allele /productlength
Observed Expected
10-repeat / 224 bp 0 0.164 19-repeat /130 bp 0 0.2734375 57-repeat /241 bp 0 0.0546875 18-repeat /112 bp 0 0.2187511-repeat 228 bp 0 1.695 20-repeat /132 bp 0 2.078125 58-repeat /244 bp 0 0.1640625 19-repeat /114 bp 0 1.64062512-repeat 232 bp 7 3.50 21-repeat / 134 bp 3 3.4453125 59-repeat / 247 bp 0 0.546875 20-repeat /116 bp 0 013-repeat 236 bp 0 1.422 22-repeat /136 bp 4 0.0546875 60-repeat /250 bp 0 0.875 21-repeat /118 bp 0 014-repeat 240 bp 0 0.219 23-repeat /138 bp 0 0.21875 61-repeat /253 bp 0 0.546875 22-repeat /120 bp 0 0.109375
24-repeat /140 bp 0 0.0546875 62-repeat 256 bp 0 0.65625 23-repeat / 122 bp 0 0.437525-repeat /142 bp 0 0.0546875 63-repeat /259 bp 3 0.3828125 24-repeat /124 bp 0 1.585937526-repeat /144 bp 0 0.7109375 64-repeat /262 bp 4 0.765625 25-repeat /126 bp 7 2.242187527-repeat /146 bp 0 0.109375 65-repeat /265 bp 0 0.8203125 26-repeat /128 bp 0 0.7109375
66-repeat /268 bp 0 0.765625 27-repeat /130 bp 0 0.054687567-repeat /271 bp 0 0.32812568-repeat / 274 bp 0 0.601562569-repeat / 277 bp 0 0.10937570-repeat / 280 bp71-repeat / 283 bp72-repeat / 286 bp73-repeat / 289 bp74-repeat / 292 bp
448 S. Reitter-Pfoertner et al. / Thrombosis Research 130 (2012) 445–450
In the 128 male control individuals, 109 different haplotypes weredetected.
Haplotype 12-21-64-26 is the most frequent in the whole sampleand was observed in three patients and four control individuals.Surprisingly, the haplotypes 12-22-63-26 and 12-22-64-26, whichwere found in four of seven patients, were not detected in the controlcohort.
Reconstruction of the Ancestral Haplotype
To illustrate the genealogical relationships and ancestry of the pa-tients, we inferred a genealogy of the samples and established a genea-logical tree. The patient samples comprised three haplotypes, whichform two distinct clusters in the genealogy. A consensus tree indicatedthat the patients form one common haplotype group. By computationwe determined the minimum number of mutations necessary toexplain the haplotype diversity in the sample. Fig. 2 shows the resultinghaplotype genealogy for the patient samples – the circle size is propor-tional to the number of haplotypes found. The haemophilia A causingmutation occurred on the most frequent haplotype D1 (12-21-64-26).The mutation 21 ➔ 22 at the DXS1073 locus led to haplotype D2(12-22-64-26), which was found in one patient, and subsequently themutation 64 ➔ 63 on locus HA472 generated the third patient haplo-type D3 (12-22-63-26), which was found in three patients. HaplotypeD1 is the one linked to the remaining control haplotypes. From thegenealogy we conclude that the haemophilia A causing mutationoriginated on the genetic background of haplotype D1, and mutationsof STRs led to the different haplotypes D2 and D3 present in haemophi-lia patients with the same mutation.
Discussion
The splice site mutation present in seven unrelated haemophilia Apatients is a not yet described T➔Gmissensemutation at position−14
Table 3Results for theχ2-value, the p-value and the simulated p-value for the four different STRs.
in intron 6 (c.788-14T>G). To our knowledge, sequencing of the F8gene usually does not cover this part of intron 6 and thus this mutationmay have escaped detection in patients with mild haemophilia andminor clinical symptoms. It is known thatmutations causinghaemophi-lia A can occur in introns of the F8 gene. It may be necessary to analyzegene regions beyond the coding region and the exon/intron boundariesat least in selected patient groups.
Analysis of cDNA to evaluate the effect of this splice site mutationconfirmed that this mutation leads to the preferred generation ofmRNA lacking exon 7. The predominant presence of RNA lacking exon7 seems to be responsible for the reduced FVIII activity and the mildhaemophilia A observed in the seven patients. Previously, the deletionof exon 7 from genomic DNA has been reported to cause a severe hae-mophilia A phenotype [15]. We speculate that the mild phenotype ob-served in our patients with a splice site mutation may be due to thepresence of a small amount of regularly spliced FVIII mRNA in livercells which was below the detection limit in mRNA obtained from leu-cocytes. Efforts to detect minute amounts of correctly spliced mRNA inblood cells were unsuccessful, and we did not have access to liver sam-ples, as none of the patients needed a liver biopsy for medical reasons.
To distinguish whether the mutation represented a mutation hotspot or a foundermutationwe analyzed intragenic (intronic) and extra-genic short tandem repeat polymorphisms (STRs). Statistical analysis ofthe allele frequency distribution for the analyzed STRs in the patientsand 128 individuals without any bleeding diathesis showed that theSTR-allele distribution of the patients shows less variation comparedto the control sample. The haplotype distribution found in patients dif-fered from the one expected on the basis of the haplotypes identified inthe controls. DXS9897 andDXS1108weremonomorphic among the pa-tients. In this regard, it should be pointed out that the genetic length be-tween DXS9897 and DXS1108 is 3.22 cM [16], an interval, which covers81 genes and 13 pseudogenes [5].
Genetic disequilibrium can be eroded by recombination events,which break down the ancestral haplotypes. This leads to a linkageequilibrium over time [17]. DXS9897 lies approximately 1596 kb 3′ ofthe F8 gene. Recombination events likely occur in DNA stretchesbetween STRs; more frequently if these stretches are longer. Probably,the STR locus DXS9897 is more frequently affected by recombinationevents due to the large distance to the FVIII gene.
Overall, four healthy control individuals belong to the same haplo-type cluster as our patient cohort, which means that they have thesame ancestral haplotype. We could analyze one healthy male withhaplotype D1. He had a normal FVIII:C level (155%) and never
Fig. 2. Genealogy of the patient cohort (relevant part of the consensus tree in concise illus-tration). Each node represents a different haplotype. Circle areas are proportional tohaplotype frequency. “C” stands for control individuals and “D” for patients. Filled circles represent haplo types not found in the samples. The dotted lines and circles indicate thelink to the remainingnetwork of control individuals (not shown in this figure). The haplotype groups C1 and D1 are patients and control individuals sharing the same haplotypeshown in parenthe sis. The labels at the branches connecting the circles specify the mutations at the respective loci leading to the different haplotypes. The repeat number onthe left side of the equivalence sign repre- sents the repeat number of the left haplotype or the haplotype above. This information suf- fices to reconstruct the other haplotypes.
449S. Reitter-Pfoertner et al. / Thrombosis Research 130 (2012) 445–450
experienced bleeding events. Sequencing proved that this person didnot carry the intron 6 splice site mutation found in the patients. This in-dicates that the missense mutation occurred once in a healthy individ-ual with haplotype D1 and stayed in the population. Further supportfor a single mutational event is the absence of the D2 and D3 haplotypein the control population and the low frequency of the D1 haplotype inthe control population. However,we note thatwe identified a very largenumber of haplotypes – 109 in 125 individuals. This might indicate thatthe STRs are unstable and/or that that there is a high rate ofrecombination in the population. It also indicates that the controlpopulation used was probably not large enough to enable an accurateestimation of the frequency of individual haplotypes.
Due to the unknown mutation rates of the STRs our data does notprovide accurate information on the age of the mutation. However,we can conclude that the splice site mutation occurred in one founder.From the number of STR mutations on the patient-D1 haplotypewe can assume that the origin of the mutation dates back at least500 years (25 generations), if we assume a mutation rate as high as0.075 per marker per generation [18].
Nowadays, life expectancy of patients withmild haemophilia equalsthe life expectancy of the normal population [19]. Also in earlier days,
when no effective treatment was available, patients with mild haemo-philia usually did not die at a very young age but reached adulthood[20]. This explains why the missense mutation found in our patientcohort, although based on a founder effect, was passed on over severalgenerations and many years.
Conflict of Interest Statement
None of the authors has a conflict of interest to declare.
Acknowledgements
The authors would like to thank Renate Freitag for her contributionconcerning the mutation analysis of the haemophilia A patients.Further, we would like to acknowledge the financial support renderedthrough the Anniversary Fund of the Oesterreichische Nationalbank(OeNB, project number AP12208ONB). Arndt von Haeseler also thanksthe Vienna Science and Technology Fund (WWTF) for financial support.In addition, Sylvia Reitter-Pfoertner thanks Bayer Healthcare for receiptof the Bayer Haemophilia Clinical Training Award, which enabled her to
450 S. Reitter-Pfoertner et al. / Thrombosis Research 130 (2012) 445–450
work on this project at the Department of Laboratory Medicine at theMedical University of Vienna.
References
[1] Bicocchi MP, Pasino M, Lanza T, Bottini F, Boeri E, Mori PG, et al. Analysis of 18novel mutations in the factor VIII gene. Br J Haematol 2003;122(5):810–7.
[2] The Haemophilia A Mutation, Structure, Test and Resource Site. 12–3–2010.RefType: Internet Communication.
[3] Lakich D, Kazazian Jr HH, Antonarakis SE, Gitschier J. Inversions disrupting the factorVIII gene are a common cause of severe haemophilia A.Nat Genet 1993;5(3):236–41.
[4] Bagnall RD,WaseemN, Green PM, Giannelli F. Recurrent inversion breaking intron 1of the factor VIII gene is a frequent cause of severe hemophilia A. Blood 2002;99(1):168–74.
[5] Machado FB,Medina-Acosta E. High-resolution combined linkage physicalmap of shorttandem repeat loci on human chromosome band Xq28 for indirect haemophilia Acarrier detection. Haemophilia 2009;15(1):297–308.
[6] Keeney S, Mitchell M, Goodeve A. Themolecular analysis of haemophilia A: a guidelinefrom the UK haemophilia centre doctors' organization haemophilia genetics laboratorynetwork. Haemophilia 2005;11(4):387–97.
[7] Reitter S, Sturn R, Horvath B, Freitag R, Male C, MunteanW, et al. Spectrum of causativemutations in patients with haemophilia A in Austria. Thromb Haemost 2010;104(1):78–85.
[8] El-Maarri O, Herbiniaux U, Graw J, Schroder J, Terzic A, Watzka M, et al. Analysis ofmRNA in hemophilia A patients with undetectable mutations reveals normalsplicing in the factor VIII gene. J Thromb Haemost 2005;3(2):332–9.
[9] CMGS website. Sequencing Best Practice Guidelines. http://www.cmgs.org/BPGs/best_practice_guidelines.htm20-9-2010 Ref Type: Internet Communication.
[10] Zivelin A, Griffin JH, Xu X, Pabinger I, SamamaM, Conard J, et al. A single genetic originfor a common Caucasian risk factor for venous thrombosis. Blood 1997;89(2):397–402.
[11] Zivelin A, RosenbergN, Faier S, KornbrotN, PeretzH,Mannhalter C, et al. A single genet-ic origin for the common prothrombotic G20210A polymorphism in the prothrombingene. Blood 1998;92(4):1119–24.
[12] Wilson IJ, Weale ME, Balding DJ. Inferences from DNA data: population histories,evolutionary processes and forensic match probabilities. J R Statist Soc A2003;166(Part 2):155–201.
[13] Fitch WM. Rate of change of concomitantly variable codons. J Mol Evol 1971;1(1):84–96.
[14] Cochran W. Some methods for strengthening the common Chi2 test. Biometrics1954;10:417–51.
[15] CitronM,GodmilowL, Ganguly T, Ganguly A. High throughputmutation screening ofthe factor VIII gene (F8C) in hemophilia A: 37 novel mutations andgenotype-phenotype correlation. Hum Mutat 2002;20(4):267–74.
[16] Matise TC, Chen F, Chen W, De LV, De La Vega FM, Hansen M, He C, et al. Asecond-generation combined linkage physical map of the human genome. GenomeRes 2007;17(12):1783–6.
[17] Sabatti C, Risch N. Homozygosity and linkage disequilibrium. Genetics 2002;160(4):1707–19.
[18] Ballantyne K. GMFRSOLOWA. Mutability of Y-Chromosomal Microsatellites:Rates, Characteristics, Molecular Bases, and Forensic Implications. Am J HumGenet 2010;87:341–53.
[19] Reitter S, Waldhoer T, Vutuc C, Lechner K, Pabinger I. Survival in a cohort of patientswith haemophilia at the haemophilia care center in Vienna, Austria, from 1983 to2006. Haemophilia 2009;15(4):888–93.
[20] Larsson SA. Life expectancy of Swedish haemophiliacs, 1831–1980. Br J Haematol1985;59(4):593–602.
