Recurrent Inversion with Concomitant Deletion and Insertion Events in The

HUMAN MUTATION Mutation in Brief #977 (2007) Online

MUTATION IN BRIEF

Recurrent Inversion with Concomitant Deletion and Insertion Events in the Coagulation Factor VIII Gene Suggests a New Mechanism for X-Chromosomal Rearrangements Causing Hemophilia A Christiane Mühle1,2, Martin Zenker3, Nadia Chuzhanova4, and Holm Schneider1,2*

1Experimental Neonatology, Department of Pediatrics, Medical University of Innsbruck, Austria; 2Children’s Hospital, University of Erlangen-Nuernberg, Erlangen, Germany; 3Institute of Human Genetics, University of Erlangen-Nuernberg, Germany; 4Department of Biological Sciences, University of Central Lancashire, Preston, United Kingdom

*Correspondence to Dr. Holm Schneider, Experimental Neonatology, Department of Pediatrics, Medical University of Innsbruck, Innrain 66, 6020 Innsbruck, Austria; Tel.: +43 512 50425732; Fax: +43 512 50424680; E-mail: [email protected]

Grant sponsor: The study was supported by a grant from Wyeth Pharma GmbH (Germany) to CM and HS.

Communicated by Stylianos E. Antonarakis

Recurrent int22h-related inversions in the coagulation factor VIII gene (F8) are the most common cause of severe hemophilia A. Such inversions have repeatedly been hypothesized to be associated with concomitant deletions that are responsible for an increased risk of immune responses against therapeutic exogenous factor VIII. However, exact DNA breakpoints have not yet been reported. In a patient with persistent factor VIII-inactivating antibodies, molecular analysis of F8 including Southern Blot, long-range PCR and primer walking techniques revealed a combination of an int22h2-related inversion, deletion of exons 16-22 and insertion of a duplicated part of the X-chromosomal MPP1 gene. This novel genomic rearrangement was also detectable in the patient’s mother, but absent in both maternal grandparents. The genetic defect most likely originated from a complex X- chromosomal recombination event during spermatogenesis due to the formation of a DNA loop stabilized by Alu and LINE repeat elements. Elucidation of such combined mutations may allow early identification of patients at high risk of developing factor VIII-neutralizing antibodies and will help to understand the mechanisms behind gross chromosomal rearrangements causing hemophilia A and other diseases. © 2007 Wiley-Liss, Inc.

KEY WORDS: hemophilia A; F8; X-chromosome; breakpoints; inversion; deletion; insertion

INTRODUCTION Hemophilia A (MIM# 306700), the most common severe coagulation disorder with an incidence of 1 in 5000 males, is caused by absence or impaired activity of clotting factor VIII (FVIII) resulting from various mutations of the FVIII gene (F8). This large gene, which has been mapped to the most distal band (Xq28) of the long arm of the X chromosome, comprises 26 exons spread over 186 kb. Approximately half of the hemophiliacs with less than 1 % FVIII activity carry a genomic inversion originating from a hot spot of intrachromosomal recombination

Received 9 March 2007; accepted revised manuscript 17 May 2007.

2 Mühle et al. between a 9.5 kb region within F8 intron 22 (int22h1) and one of its two extragenic telomeric copies on the X- chromosome, int22h2 or int22h3 (Lakich et al., 1993). Disruption of F8 may also occur by recombination between a 1 kb portion of F8 intron 1 and its duplicated extragenic version (Bagnall et al., 2002). In the remaining cases, hemophilia A has been attributed to a broad spectrum of mostly private mutations scattered over the entire gene (Kemball-Cook et al., 1998). Although a genomic inversion normally does not result in gain or loss of DNA, unusual patterns in Southern blots or long-range PCR for the detection of int22h-related inversions have led to the hypothesis of concomitant deletions (Schroder et al., 1996; Andrikovics et al., 2003) which would be associated with an increased risk of developing FVIII-inactivating antibodies (inhibitors). Such antibodies represent a major obstacle to FVIII replacement therapy. They are found in approximately 41 % of hemophiliacs carrying a large deletion in F8, whereas only 21 % of patients with a recurrent int22h-related inversion develop FVIII-neutralizing antibodies (Oldenburg et al., 2004). However, DNA breakpoints of few large deletions in F8 have been determined exactly. Although some deletions have been associated with non-homologous recombination (Woods-Samuels et al., 1991), more recent reports demonstrated unequal homologous recombination between Alu-derived sequences to be responsible for deletions in the Alu-rich F8 (Vidal et al., 2002; Rossetti et al., 2004). None of any fully characterized deletion in this gene was combined with an int22h-related inversion. Here, we report the first case of a hemophiliac with a complex genomic rearrangement including a recurrent int22h-related inversion, a deletion of F8 exons 16-22 and an insertion of an MPP1 gene fragment. We propose a mechanism for a multi-step recombination event which is likely to have occurred within one generation during spermatogenesis and which led to severe hemophilia A with high-titer inhibitors resistant to repeated attempts of immune tolerance induction.

MATERIAL AND METHODS

Case report The index patient is the second child of Caucasian parents with no family history of bleeding disorders. Severe hemophilia A was diagnosed at the age of three months subsequent to intramuscular vaccination. FVIII activity was found to be below 1 % with VWF:Ag of 90 % and VWF:RCo of 56 % and no abnormal changes in multimer analysis. Following treatment with recombinant FVIII, the child developed inhibitors within 40 exposure days, which were quantified using the Nijmegen modification of the Bethesda assay (Verbruggen et al., 1995). Two attempts of immune tolerance induction under different regimes and clotting factor concentrates failed. Inhibitor levels rose temporarily above 2000 BU/ml before a third trial with simultaneous immunosuppression finally lowered the antibody concentration to <2 BU/ml. The polyclonal FVIII-specific antibodies were found to bind to both the heavy and the light chain of FVIII with main epitopes in the domains A1, A2 and C1 (Muhle et al., 2004). The patient’s sister showed a FVIII activity of 10 % with VWF:Ag of 69 % and VWF:RCo of 28 % suggesting a carrier status, whereas the FVIII levels of the patient’s mother were within the normal range. The family gave informed consent to participate in the study in accordance with the Declaration of Helsinki.

DNA analysis Genomic DNA was extracted from peripheral blood lymphocytes and investigated for int22h-related inversions by Southern Blot (Lakich et al., 1993) and long-range PCR (Liu et al., 1998) employing two additional PCR primers with 5’-G/C-extensions, F8int22_A2 (5’-GCCTGCATTTCCCATCAAAATGCTAACATTGTTTTTCA- 3’) and F8int22_A3 (5’-GTTACGGGCCTTGCTGCTTTGTCCAGTTTCAGG-3’), to distinguish between recombinations with int22h3 and int22h2 (type 1 and type 2 inversion, respectively). This primer set differs from the recently published novel set for a discriminating long-range PCR (Bagnall et al., 2006). Routine PCR analysis (Oldenburg et al., 2001) was carried out to demonstrate the presence or absence of F8 exons. Further PCR amplifications of intronic regions for breakpoint localization, across the breakpoint and for haplotyping were performed according to standard PCR protocols based on the F8 sequence from the UCSC Human Genome Browser Project (genome.ucsc.edu, NCBI Build 36.1; primer sequences and conditions obtainable from the authors). The genomic DNA insert in the patient’s F8 was identified using the DNA Walking SpeedUp Premix Kit (Seegene, BioCat, Germany). DNA sequencing was performed with the BigDye Terminator sequencing mix (Applied BiosystemsSpecific) on an automated capillary ABI PRISM 310 sequencer. Paternity of the maternal grandfather was verified using the PowerPlex 16 system (Promega, Mannheim, Germany). Combined Inversion, Deletion and Insertion in F8 3

Karyotyping of blood lymphocytes was done by routine procedures. Interphase nuclei and metaphase chromosome spreads were investigated by fluorescence in situ-hybridization (FISH) with a F8-specific probe from the BAC clone RP11-671F22.

RESULTS AND DISCUSSION In the index patient, routine Southern blot analysis to detect int22h-related inversions in F8 resulted in an unusual pattern. While bands of 16.0 kb and 15.5 kb representing the int22h3 and the hybrid int22h2/1 fragment, respectively, were easily detected, the 20.0 kb band typical for the inversion (Lakich et al., 1993) was missing. Long-range PCR confirmed an int22h2-related inversion by yielding the expected A2Q product for int22h2/1, however, lack of an amplification product with primers P and B (Fig. 1A) indicated a deletion within the int22h1/2 counterpart. The newly designed primers A2 and A3 allowed here a PCR-based distinction between the frequent type 1 and the less common type 2 inversion involving int22h3 and int22h2, respectively, which is routinely done by more laborious and material-consuming Southern blot analysis (Lakich et al., 1993). Attempts to amplify F8 exons 16 to 22 by PCR failed consistently, although other exons of this gene were readily detected. PCR analysis of regions within introns 15 and 22 was performed to determine the extent of the deletion (Fig. 1B), but the presence of large sequence homology regions flanking both int22h2 and int22h3 as a palindrome (Fig. 2) precluded the exact localization of the distal breakpoint. Because of the inversion, the large deletion is not limited to exons 16-22 encompassing 15 kb but may extend up to the 3’-end of the 40 kb palindromic arm directed towards the telomere, which could be amplified from the patient’s DNA. However, a gross genomic rearrangement of chromosomes was excluded by demonstration of a normal 46;XY karyotype in the patient’s lymphocytes. In addition, FISH analysis using a F8-specific probe hybridizing with 83 kb and 89 kb on each side of the partial F8 deletion, respectively, yielded a single fluorescence signal with an insignificantly increased signal split ratio indicating that both regions flanking the deletion were located within a distance of 1 Mbp. As PCR amplification spanning the deletion site proved impossible, primer walking approaches were applied to the breakpoint of intron 15, which led to the detection of an inserted part of the X-chromosomal gene MPP1 replacing the deleted F8 exons. This finding was confirmed by PCR across the F8/MPP1 breakpoint junction (Fig. 1C) and DNA sequencing of the hybrid product (Fig. 1D). Further reactions with a series of MPP1 primers in combination with a F8 primer allowed an estimation of the size of the MPP1 insert to a range from 201 bp to 486 bp. The presence of an intact MPP1 region in the patient's X-chromosome as indicated by a PCR product of the expected size (Fig. 1C) amplified with primers, one of which binds to the part of MPP1 identified as insertion in F8 while the other one binds to the MPP1 gene outside the insertion sequence, provides evidence of a duplication rather than a translocation of the inserted MPPI fragment. Contrary to the primary assumption that the indel mutation had occurred in an ancestor carrying an X- chromosome with a pre-existing int22h-related inversion, the inversion (Fig. 1A) and the insertion (Fig. 1C) were detectable only in the patient’s mother and sister but were absent in both maternal grandparents. The possibility of a deletion present heterozygously in the female line as the first mutation step was judged improbable in view of several reports of concomitant inversions and deletions (Andrikovics et al., 2003). Because non-paternity was ruled out by multiplex analysis of short tandem repeats at 15 loci, all three mutational events most likely took place during a single gametogenesis. Although somatic and/or germline mosaicism cannot be formally excluded, the recently identified cases of somatic F8 mosaicism have been found to derive from point mutations of F8 that occurred in female ancestors (Leuer et al., 2001). However, the distribution of eight intragenic and two extragenic polymorphisms demonstrated that the patient’s as well as his mother’s and sister’s defective F8 was inherited from the maternal grandfather (Fig. 1E), who is expected to carry already the recently proposed polymorphic, non- pathogenic inversion (step 1 in Fig. 2) as a prerequisite for the int22h2-related inversion (Bagnall et al., 2005).

4 Mühle et al.

A D

AATTTTATAAATGATGGTTT TT GAAATAAAACAGGTTTTAATTT F8 GGTCCTGGGGAAAGTGTTGC TT CAAACAGGCCCGAAGAGAAAAG E

Figure 1. Analysis of the F8 mutation in the patient and his relatives. A: Detection of the int22h-related inversion: Long-range PCR was performed on genomic DNA from the patient (P), his sister (S), mother (M), maternal grandmother (MGM) and grandfather (MGF) using published primers together with the novel primers A2 and A3 to distinguish between type 1 and type 2 inversions. Sizes of the PCR products and primer binding sites in a normal X-chromosome are indicated. MW, DNA size marker. B: Localization of the DNA breakpoints of the additional deletion in F8 by PCR amplification of flanking fragments in intron 14, intron 15, exon 22 and intron 22 of patient’s and control genomic DNA. MW, 1 kb ladder. C: Investigation of the patient’s relatives for the additional MPP1 insertion: PCR with genomic DNA as template was performed using primer pairs flanking the breakpoint sequence either in F8 or in MPP1, or with a combination of a forward primer binding to F8 and a reverse primer binding to the MPP1 insert, which leads to hybrid PCR products (ins) spanning the breakpoint. MW, 100 bp ladder. D: Sequencing electropherogram of a hybrid PCR product spanning the junction between F8 and the MPP1 insert, which confirms replacement of the deleted F8 region by the MPP1 fragment. E: Determination of the origin of the defective X- chromosome by polymorphism analysis: Genomic DNA of the index patient and his relatives was investigated by DNA sequencing at ten loci, data of which were retrieved from the International HapMap Project (www.hapmap.org). The location of eight single nucleotide polymorphisms and two perfect microsatellites (variable number of tandem repeats, VNTR) is depicted in the scheme (custom track from the March 2006 assembly (NCBI Build 36.1) of the UCSC Human Genome Browser Project, genome.ucsc.edu) displaying the F8 region on the X-chromosome with predicted heterozygosity rates in brackets. The identified alleles and lengths of repeats are grouped by haplotype blocks with corresponding frequency data obtained from the International HapMap Project. Combined Inversion, Deletion and Insertion in F8 5

To elucidate the etiopathogenic mechanism of this complex mutation, the sequence at the telomeric end of Xq28 was analysed using the RepeatMasker Open-3.0 program (www.repeatmasker.org). Several pairs of AluSx and other Alu repeat elements as well as two LINE1 elements with 90 % identity within 4.3 kb were detected suggesting the formation of a loop structure, the stem of which was stabilized by these inverted repeat elements and allowed templating of the MPP1 fragment insertion in F8 intron 15 (step 2 in Fig. 2). The mechanism by which repeat sequences mediate mutational events has been described for microinsertions (Ball et al., 2005). If repeat elements were the key players, the insertion presumably occurred prior to the pathogenic inversion and the concomitant loss of genomic material during truncation and joining of DNA ends resulting in the deletion of F8 exons 16-22 and possibly part of the int22h2 flanking region (steps 3 and 4 in Fig. 2). Such complex rearrangements are probably facilitated by the higher flexibility of the telomeric end of the X chromosome in the absence of a second pairing X-chromosome during male meiosis (Rossiter et al., 1994).

Figure 2. Scheme of the proposed etiopathogenic mechanism. Data are based on the March 2006 assembly (NCBI Build 36.1) of the UCSC Human Genome Browser Project (genome.ucsc.edu). The figure is not drawn to scale. Values indicate the size of regions in kb, while the letters P, Q, A, A2, A3 and B show the binding site of primers used for long-range PCR to detect int22h-related inversions. Black boxes represent exons 1-22 and 23-26 of F8. Grey boxes indicate the homologous regions flanking the extragenic copies int22h2 and int22h3 that facilitate intragenic homologous recombination resulting in a polymorphic, non-pathogenic inversion (step 1) with int22h2 consequently located distal and in opposite orientation to int22h1 (Bagnall et al., 2005). The proposed formation and stabilization of a loop structure by inverted repeat elements (Alu and LINE, represented by triangles indicating the orientation) allowed templating of the MPP1 fragment insertion in F8 intron 15 (step 2). The recurrent pathogenic recombination between int22h1 and int22h2 (step 3) and the deletion of F8 exons 16-22 possibly extending to the 3’-end of the 40 kb palindromic arm directed towards the telomere (step 4) presumably occurred simultaneously. Steps 2, 3 and 4 of this complex mutation are assumed to have taken place in one generation during spermatogenesis. 6 Mühle et al.

In conclusion, the patient carries the recurrent X-chromosomal F8 int22h1/2-related inversion in combination with a deletion of at least 15 kb but not exceeding 65 kb and an insertion of between 201 and 486 bp which was duplicated from the first intron of the X-chromosomal MPP1 gene. Following established nomenclature guidelines (http://www.hgvs.org/mutnomen/) and numbering the A of the ATG translation initiation codon with +1, the indel mutation identified here was named F8(NM_00132.2):c.5373+615_6429+5746+?del15475…65200 ins oMPP1(NM_002436):c.102+2731_102+2531…2246, "?" representing an unknown deleted part of the int22h1/2 region and the flanking palindromic arm (Fig. 2) and “…” indicating the size ranges for the deletion and the insert.

This case is the first detailed report on a combination of an int22h-related inversion and a partial deletion of F8. It demonstrates that the chromosomal instability in Xq28 associated with the intron 22 homology regions may not only involve deletions but even additional insertions. Several independent reports of unusual patterns in diagnostic Southern blots or long-range PCR [summarized in (Andrikovics et al., 2003)] have led to the assumption that a number of patients may carry both an inversion and a deletion in F8 (“rare inversion type”). In a pilot study of DNA samples from 32 patients with severe hemophilia A carrying large deletions, we identified three cases with similar patterns (C. Muehle, unpublished results) suggesting that this combination may actually be more frequent than expected. However, screening for additional mutations is rarely performed in routine testing after an inversion has been detected, although a concurrent large deletion in F8 significantly increases the risk for inhibitor development during FVIII replacement therapy in comparison with the inversion alone. Unusual genetic findings, therefore, require an elaborate DNA analysis to provide sufficient data for the selection of the best therapeutic approach, rapid and reliable carrier detection and prenatal diagnosis. Furthermore, detailed analysis of such cases may help to disclose the molecular mechanisms behind gross genomic rearrangements causing hemophilia A and other diseases.

ACKNOWLEDGMENTS The authors thank Dr. Grischa Lischetzki (Children’s Hospital, University of Erlangen-Nuernberg), for providing blood samples and clinical data of the patient and establishing contact to family members, Dr. Anita Rauch (Institute of Human Genetics, University of Erlangen-Nuernberg) for performing the FISH analysis and Prof. Elisabeth Steichen (Department of Pediatrics, Medical University of Innsbruck) for her valuable advice.

REFERENCES Andrikovics H, Klein I, Bors A, Nemes L, Marosi A, Varadi A, Tordai A. 2003. Analysis of large structural changes of the factor VIII gene, involving intron 1 and 22, in severe hemophilia A. Haematologica 88(7):778-84.

Bagnall RD, Giannelli F, Green PM. 2005. Polymorphism and hemophilia A causing inversions in distal Xq28: a complex picture. J Thromb Haemost 3(11):2598-9.

Bagnall RD, Giannelli F, Green PM. 2006. Int22h-related inversions causing hemophilia A: a novel insight into their origin and a new more discriminant PCR test for their detection. J Thromb Haemost 4(3):591-8.

Bagnall RD, Waseem N, Green PM, Giannelli F. 2002. Recurrent inversion breaking intron 1 of the factor VIII gene is a frequent cause of severe hemophilia A. Blood 99(1):168-74.

Ball EV, Stenson PD, Abeysinghe SS, Krawczak M, Cooper DN, Chuzhanova NA. 2005. Microdeletions and microinsertions causing human genetic disease: common mechanisms of mutagenesis and the role of local DNA sequence complexity. Hum Mutat 26(3):205-13.

Kemball-Cook G, Tuddenham EG, Wacey AI. 1998. The factor VIII Structure and Mutation Resource Site: HAMSTeRS version 4. Nucleic Acids Res 26(1):216-9.

Lakich D, Kazazian HH, Jr., Antonarakis SE, Gitschier J. 1993. Inversions disrupting the factor VIII gene are a common cause of severe haemophilia A. Nat Genet 5(3):236-41.

Leuer M, Oldenburg J, Lavergne JM, Ludwig M, Fregin A, Eigel A, Ljung R, Goodeve A, Peake I, Olek K. 2001. Somatic mosaicism in hemophilia A: a fairly common event. Am J Hum Genet 69(1):75-87. Combined Inversion, Deletion and Insertion in F8 7

Liu Q, Nozari G, Sommer SS. 1998. Single-tube polymerase chain reaction for rapid diagnosis of the inversion hotspot of mutation in hemophilia A. Blood 92(4):1458-9.

Muhle C, Schulz-Drost S, Khrenov AV, Saenko EL, Klinge J, Schneider H. 2004. Epitope mapping of polyclonal clotting factor VIII-inhibitory antibodies using phage display. Thromb Haemost 91(3):619-25.

Oldenburg J, Ivaskevicius V, Rost S, Fregin A, White K, Holinski-Feder E, Muller CR, Weber BH. 2001. Evaluation of DHPLC in the analysis of hemophilia A. J Biochem Biophys Methods 47(1-2):39-51.

Oldenburg J, Schroder J, Brackmann HH, Muller-Reible C, Schwaab R, Tuddenham E. 2004. Environmental and genetic factors influencing inhibitor development. Semin Hematol 41(1 Suppl 1):82-8.

Rossetti LC, Goodeve A, Larripa IB, De Brasi CD. 2004. Homeologous recombination between AluSx-sequences as a cause of hemophilia. Hum Mutat 24(5):440.

Rossiter JP, Young M, Kimberland ML, Hutter P, Ketterling RP, Gitschier J, Horst J, Morris MA, Schaid DJ, de Moerloose P, Sommer SS, Kazazian HH, Jr., Antonarakis SE. 1994. Factor VIII gene inversions causing severe hemophilia A originate almost exclusively in male germ cells. Hum Mol Genet 3(7):1035-9.

Schroder W, Wehnert M, Herrmann FH. 1996. Intron 22 of factor VIII gene--a hot spot for structural aberrations causing severe hemophilia A. Blood 87(7):3067-8.

Verbruggen B, Novakova I, Wessels H, Boezeman J, van den Berg M, Mauser-Bunschoten E. 1995. The Nijmegen modification of the Bethesda assay for factor VIII:C inhibitors: improved specificity and reliability. Thromb Haemost 73(2):247-51.

Vidal F, Farssac E, Tusell J, Puig L, Gallardo D. 2002. First molecular characterization of an unequal homologous alu-mediated recombination event responsible for hemophilia. Thromb Haemost 88(1):12-6.

Woods-Samuels P, Kazazian HH, Jr., Antonarakis SE. 1991. Nonhomologous recombination in the human genome: deletions in the human factor VIII gene. Genomics 10(1):94-101.