Intron 1 and intron 22 inversions (inv1 and inv22) represent the most prevalent mutations in the F8 gene (F8) (1–5% and 40–45%, respectively) causative of severe haemophilia A (HA) . These alterations result from intrachromosomal recombination between intronic regions (int1h-1 and int22h-1, respectively) and their homologous extragenic copies in the telomeric position (int1h-2 and int22h-2/3, respectively) . Inv1 causes an altered F8 structure with the translocation and inversion of the 5′ region including exon 1 in the extragenic site, giving rise to two chimeric mRNAs: one containing the promoter and first exon of the F8 followed by some exons of the VBP1 gene; the other one formed by the promoter and coding sequences of the C6.1A gene joined to a part of the intron 1 and exons 2–26 of the F8 . The clinical effect of this rearrangement is a severe phenotype with the absence of functional factor (F)VIII. According to literature data, the prevalence of inv1 ranges from 0% to 5% in severe HA patients from different countries [3,4]. In our cohort of 135 unrelated severe HA patients from Southern Italy, inv1 was found in four patients (2.9%) . Here, we report the case of a severe HA patient, who showed an unusual pattern for inv1 associated with a complex gene rearrangement not described to date.
The propositus is a Caucasian young male diagnosed with severe HA at the age of 1 year because of a positive family history. The causative F8 mutation in his affected uncle or the carrier status in his mother had not been previously searched for. The patient was treated with recombinant FVIII concentrates with no personal or family history of inhibitor development. We first analyzed the patient for both F8 inversions. This analysis excluded inv22 but showed a novel abnormal pattern for inv1, compared with those previously reported in literature, with two bands: one corresponding to the normal int1h-1 region (1908 bp) and the other corresponding to the band with the inversion of the int1h-1 sequence (1323 bp), as a heterozygous female, suggesting the presence of more than two copies of the int1h region (Fig. 1A). No bands of the int1h-2 region were detected, indicating a deletion within the homologous extragenic copy or a new rearrangement of sequences flanking the int1h-2 repeat not detectable by the PCR method used (Fig. 1A). We also investigated the presence of inv1 in his relatives: both mother and sister were carriers of the same aberrant pattern plus the 1191-bp fragment specific for the normal homologous extragenic int1h-2 sequence. The affected uncle was not available for molecular analysis. To confirm the direct diagnosis, linkage analysis was performed in this family, using intragenic (intron 18 BclI and intron 19 HindIII RFLP) and extragenic (DXS52 VNTR) polymorphisms, matching up with the same genetic profile. Furthermore, in the patient sequencing analysis of the F8, including all 26 exons, flanking intron/exon boundaries, promoter, 5′ and 3′ regions, did not reveal other mutations. In this family, a prenatal diagnosis was also performed during the first pregnancy of the patient’s carrier sister. The fetus was diagnosed as a wild-type male without the aberrant inv1 pattern (Fig. 1A). Aberrant diagnostic patterns for inv1 have been occasionally described in severe HA, suggesting the possible existence of extra copies of the int1h region, as shown in int22h-related inversions [6–8]. In order to detect possible quantitative genomic changes, as copy number variations (CNVs), responsible of the unusual results, we applied different technical approaches, such as multiplex ligation-dependent probe amplification (MLPA), array comparative genomic hybridization (aCGH) and quantitative-PCR (qPCR) (details available upon request). MLPA analysis highlighted an unexpected novel large duplication of exons 2–6 of the F8 in the proband, also observed in his mother and sister (Fig. 1B). This duplication had not reported in the HAMSTeRS hemophilia A mutation database so far. Subsequently, aCGH screening was crucial, revealing a more complex genetic rearrangement not yet reported, that is, a 19.32-kb duplication in the F8 involving a part of intron 1 with the int1h-1 region and extending through intron 6 (154 209 461–154 228 780), combined with an extragenic deletion of 41.87 kb in the telomeric position, including an unknown region with no reference genes available (154 383 812–154 425 684) (Fig. 1C). This DNA rearrangement was also found in both patient’s mother and sister, as first shown by inv1 PCR products, compatible with the carrier state of the aberrant pattern. Quantitative changes of the specific F8 region have been further confirmed for introns 3 and 5 by qPCR, yielding concordant results. As aCGH enables the detection of gains and/or losses within genomic regions, but provides neither genome position nor orientation information , we could not establish if the duplication was arranged in tandem or translocated to a different genomic position outside the target gene. In our index case, the characterization of duplication breakpoints is still in progress, with no sequencing information available. In the attempt to elucidate the possible mechanisms behind this rearrangement, we performed a bioinformatic analysis, as repetitive elements, such as short interspersed nuclear elements (SINEs), are known to be frequently involved in duplication mechanisms . Both introns 1 and 6 harbor a SINE/AluY repeat with 87% homology at nucleotides 154 241 443–154 241 175 and 154 210 740–154 210 476, respectively (RefSeq NC_000023, on negative strand), according to RepeatMasker analysis. The latter might have mediated the specific non-recurrent CNVs by pairing ectopic AluY repeats through one of these different mechanisms, such as fork stalling and template switching (FoSTeS), microhomology-mediated break-induced replication (MMBIR), non-homologous end-joining (NHEJ) and, to a lesser extent, non-allelic homologous recombination (NAHR) . Therefore, we hypothesized that multiple operating mechanisms occurred as independent events; presumably, the duplication was the first event, followed by the inv1, as suggested by the amplification pattern of patient’s DNA, through an intrachromatid or intrachromosomal NAHR. The 41-kb extragenic deletion revealed by the aCGH analysis is localized to a chromosomal region rich in repetitive elements with no reference genes, according to the bioinformatic analysis, and does not include the int1h-2 sequence recombined. Therefore, a new rearrangement of sequences flanking the recombined int1h-2 repeat, not detectable by the PCR method used, presumably occurred.
Standard screening methods for molecular HA diagnostics do not envisage the analysis of complex genomic rearrangements, like that described in our case. Indeed, aCGH screening highlighted an unexpected extragenic deletion, otherwise unrevealed, allowing a more detailed and accurate genotyping of the patient and his family members. Nevertheless, in the limited perspective of HA molecular diagnosis, standard detection of abnormal inv1, combined with MLPA, led to a correct carrier identification in the patient’s relatives, confirming as an appropriate, reliable and fast technical strategy for carrier and prenatal diagnosis.