High-Throughput Genotyping of Haemophilia A and B Using Next-Generation Sequencing Technology in Lille University Hospital

Introduction

The identification of molecular defects in haemophilia is essential for the optimization of patient treatment and the formal characterization of female carriers. The Sanger method is the gold standard for sequencing F8 and F9 genes but is time-consuming and expensive. We aimed to develop a high-throughput method to genotype haemophilia A (HA) and B (HB) patients using the Next-Generation Sequencing (NGS) technology for an exhaustive and less expensive analysis of F8 and F9 genes.

Material & Methods

We developed a small panel containing F8 and F9 for exons and introns/exons junctions sequencing. We used two different methods for library preparation (AmpliSeq™, Life Technologies™ and HaloPlex™, Agilent™), performed in the same PCR emulsion system (Ion One Touch 2™, Life Technologies™) and sequenced with a Ion 316™ chip in a PGM™ Ion Torrent sequencer, or a Ion PI™ chip in a Proton™ sequencer (Life Technologies™) respectively. The promoter and 3' regions of F8 and F9 were always studied by Sanger. NGS analysis was first performed in 62 samples (HA: n= 42; HB: n=13; carriers: n=7) previously characterized for F8 or F9 mutations by Sanger method or Multiplex Ligation-Probe Amplification (MLPA). All types of mutations were studied (nonsense, missense, splice, small insertion/deletion and exons deletion/duplication) and were distributed in all exons of F8 and F9. NGS analysis was further performed in 42 haemophilia patients (HA: n=31, HB: n=11; of which 36 presented a mild phenotype), with unknown mutation status. All patients were included by the local Comprehensive Care Haemophilia Center of Lille University Hospital after written informed consent. Data were analyzed with SeqNext™ software (JCI Medical System™). A Normalized Reads Depth (NRD) ratio was used to detect exons deletion/duplication.

Results

All exons were well covered by AmpliSeq™ (average number of reads, ANR = 300) and overall by HaloPlex™ (ANR = 15000, except small parts of exons 14 and 19 in F8). The average quality value for mutation detection was 60 (risk of false result <0.0001%).

In previously-genotyped patients, 92% (57/62) of F8 and F9 mutations were detected by AmpliSeq™ and 85% (53/62) by HaloPlex™. The detection rate of small insertion/deletion in homopolymers of exon 14 in F8 was only 20% (1/5) with both methods and 71% (5/7) in other exons of F8. Four deletions and one duplication of exons accounting for a severe haemophilia phenotype were identified (3 HA, 1 HB and 1 HA-carrier). In 5 uncharacterized patients by Sanger method, neither AmpliSeq™ nor HaloPlex™ were able to find a mutation suggesting that the molecular defect is located in introns of F8 or F9.

In never-genotyped patients, a mutation was detected in 90% (38/42) of cases (including 34 missense, 2 nonsense and 2 splice mutations in both F8 and F9). A duplication of exons 10 to 14 was also detected in a severe HA patient and was confirmed by MLPA. In NGS negative patients, no mutation was found in promoter or 3' regions. Of the ten candidate mutations identified in our cohort, seven were predicted to be deleterious by in silico analysis and/or co-segregation studies. No mutation was found in 10% (4/42) of never-genotyped patients with mild haemophilia A, in consistence with the available data for the mild phenotype. The technical development and laboratory protocol was easier and less expensive ($530 vs $602 including reagents and technical/medical staff) with AmpliSeq™ than HaloPlex™.

Conclusion

We confirmed that NGS is able to detect the main types of mutations in F8 and F9 genes, albeit with a lower mutation detection rate with HaloPlex™ compared to AmpliSeq™. These detections were associated with an important depth of reads and high Quality Values, except for exons 14 and 19 in F8 with HaloPlex™. AmpliSeq™ seems also an interesting screening method for the detection of exons deletion/duplication using the NRD ratio. However, both strategies fail to detect small insertion/deletion located in homopolymers of exon 14 in F8, whom identification will still rely on Sanger sequencing. AmpliSeq™ protocol performed in the PGM™ sequencer appears as a new interesting tool in genotyping of HA and HB patients of the Lille University Hospital.