Abstract

Objectives:

Hemophilia A (HA) is an X-linked recessive bleeding disorder caused by mutations in the F8 gene, which exhibits complex molecular mechanisms and high genetic heterogeneity. This study aimed to perform carrier screening and genetic analysis in five phenotypically normal females to identify F8 variants and assess their implications for genetic counseling and prenatal diagnosis.

Methods:

Genomic DNA was extracted from peripheral blood samples. High-throughput sequencing was used to screen for mutations in 455 genes associated with genetic diseases. Long-range PCR (LR-PCR) was employed to detect the inversion of introns 1 and 22 of the F8 gene. Sanger sequencing validated small deletions. To clarify the complex variations of Subject 4, qPCR, CNV-seq and SV-seq analyses were conducted on her affected uncle.

Results:

Subject 1 carried an intron 22 inversion (Inv22) of the F8 gene. Subjects 2 and 4 carried intron 1 inversions (Inv1). Subject 3 had a small deletion (c.3168_3187del), and Subject 5 had a frameshift deletion (c.4379delA). Notably, Subject 4 was found to carry a rare complex structural variant involving Inv1. QPCR suggested potential duplications in the corresponding genomic region of Subject 4’s uncle. SV-seq identified two duplications on the X chromosome in her uncle, which may disrupt F8 gene function.

Conclusion:

All five subjects carried pathogenic F8 gene variants, and one affected individual had a pathogenic variant of the gene. Although carriers are asymptomatic, they can transmit the mutant allele to their offspring. Molecular genetics in carriers is crucial for improving genetic screening, prenatal diagnosis, and the development of targeted therapies for HA. This study further enriches the mutation spectrum of the F8 gene and underscores the importance of comprehensively detecting inversions, copy number variations, and structural rearrangements in the molecular diagnosis of hemophilia A.

1 Introduction

Hemophilia A (HA) is the most common inherited bleeding disorder, caused by variants in the factor VIII gene (F8) that lead to abnormal production or function of factor VIII (FVIII) protein (1). The causative gene, located on Xq28, follows an X-linked recessive pattern. In this model, female heterozygotes are typically asymptomatic carriers who can transmit the mutant allele to their offspring, while hemizygous males express the phenotype, with an incidence of approximately 1 in 5,000 males (2). The clinical severity of HA is classified into three categories based on plasma factor VIII clotting activity (FVIII: C): severe (FVIII: C < 1%), moderate (FVIII: C 1–5%), and mild (FVIII: C > 5–40%) (3).

Intron 22 inversion (Inv22), a major cause of severe HA, results from homologous recombination between a sequence within intron 22 (*Int22h-1*) and one of two extragenic homologous sequences (*Int22h-2* or *Int22h-3*). This inversion disrupts the F8 gene structure (4). The intron 1 inversion (Inv1) is another common cause of severe HA (5), arising from recombination between the *int1h-1* sequence in intron 1 and its telomeric homolog *int1h-2*, located approximately 125 kb away, thereby severing the F8 gene (6). The recombination severs the F8 gene, thereby preventing the synthesis of functional coagulation factor VIII and ultimately causing severe hemophilia A (6). Inv22 and Inv1 account for approximately 45% and 1–5% of all severe HA cases, respectively (7), with Inv1 identified in 2.94% of the Chinese severe HA population (8). Large F8 gene duplications are relatively rare, representing about 0.07% of all variants, in contrast to smaller genetic alterations (9).

Although carriers are asymptomatic, they risk transmitting the mutant allele. Molecular genetic analysis in carriers is therefore essential for enhancing genetic screening, prenatal diagnosis, and targeted therapy development for HA. This study further enriches the mutation spectrum of the F8 gene and highlights the importance of carrier screening and comprehensive molecular diagnosis of hemophilia A.

2 Materials and methods2.1 Sample collection

Five phenotypically normal subjects without a personal history of hemophilia or related bleeding disorders were enrolled. Subjects 1–5 participated in carrier screening as part of a preventive genetic counseling program in the Women and Children’s Hospital, Xiamen University, China (Xiamen, China). Subject 4 has a family history of HA, and her paternal uncle (Subject 6) and father were diagnosed with hemophilia A in an external hospital. Both her uncle and father suffered severe bleeding from minor injuries multiple times, and her father died from cerebral hemorrhage. The clinical characteristics and testing items of all subjects are listed in Table 1. Following pre-test genetic counseling, 2 mL of peripheral venous blood was collected from each subject into EDTA-K2 vacuum tubes, gently inverted eight times, and stored at 4 °C.

CharacteristicsSubject 1Subject 2Subject 3Subject 4Subject 5Subject 6SexFemaleFemaleFemaleFemaleFemaleMaleAge at assessment(years)363327292652Clinical featuresNANANANANAHAHistory of treatmentNANANANANAYesTesting itemsINR (0.88–1.16)0.961.071.001.081.06/PT (9.4–12.5 s)10.7011.9011.1012.0011.80/APTT (25.1–36.5 s)32.632.334.540.5 ↑36.5/TT (10.3–16.6 s)12.314.113.015.014.6/FIB-C (2–4 g/L)3.543.662.922.652.75/DD (0–500 ng/mL)225.0090.0047.00/185.08/FDP (0–5 ug/ml)/0.8////AT-IIIA (83–128%)//95.00///PC (70.0–140.0%)//100.810785.1/PS (63.5–149%)//89.023 ↓74.0/FII: C (70.0–120.0)///120.1 ↑//FX: C (70.0–120.0)////73.1/FXII: C (70.0–150.0)//103.448.8 ↓30.0 ↓/FV: C (70.0–120.0)///55.7 ↓63.4 ↓/FVII: C (70–120)////81.6/

Clinical features of subjects with mutations in F8.

INR, international normalized ratio; PT, prothrombin time; APTT, activated partial thromboplastin time; TT, thrombin time; FIB-C, fibrinogen concentration; FDP, fibrinogen degradation products; AT-IIIA, antithrombin III activity; PC, protein C activity; PS, protein S activity; FII: C, factor II coagulant activity; FX: C, factor X coagulation activity; FXII: C, factor XII coagulant activity; FV: C, factor V coagulation activity; FVII: C, factor VII coagulation activity; The values in brackets are the normal reference ranges of each index. INR, PT, APTT, TT, FIB-C, PS, FII: C, FX: C, FXII: C, FV: C and FVII: C were measured by the clotting assay. DD and FDP were measured by the immunoturbidimetric assay. PC was measured by the chromogenic substrate assay.

2.2 Genomic DNA extraction

Genomic DNA was extracted from the peripheral blood of participants using the QIAamp Blood DNA Kit (QIAGEN, Hilden, Germany). DNA purity was assessed with a Nanodrop spectrophotometer (A260/A280 ratios between 1.8 and 2.0), concentration was determined using a Qubit fluorometer (>70 ng/μL), and integrity was verified by 1% agarose gel electrophoresis.

2.3 High-throughput sequencing

Libraries were constructed to capture the entire exonic regions and flanking 20 bp sequences of 455 genes associated with 457 genetic diseases at Becon Medical Laboratory. High-throughput sequencing was subsequently performed. After raw data quality control, reads were aligned to the UCSC hg19 reference genome using BWA, and PCR duplicates were removed. Base quality score recalibration and joint genotyping of single-nucleotide variants (SNVs) and insertions/deletions (INDELs) were performed using the Genome Analysis Toolkit (GATK).

2.4 Sanger sequencing validation

Primers were designed using Oligo 6 software1 to amplify the genomic regions encompassing the F8 (NM_000132.4) c.3168_3187del and c.4379delA variants. PCR products were sequenced on an ABI 3130xl genetic analyzer (Applied Biosystems; Thermo Fisher Scientific) for validation.

2.5 Combined long-range PCR and multiplex PCR for Inv1 and Inv22

The intron 1 and intron 22 inversions of the F8 gene were detected using a combined LR-PCR and multiplex PCR strategy (Figure 1). The assay was performed using a modified four-pool multiplex PCR. Specific primers (sequences in Supplementary Table 1) targeting the wild-type and rearranged junctions of intron 1 (EF1, IF, DER1, IR1; 10 μM each) and intron 22 (NP, NQ, P, Q, A, B; 20 μM each) were combined into four primer pools at the ratios specified in Table 2. Each 25 μL PCR reaction was assembled with the components listed in Table 3, containing 50 ng of genomic DNA. After gentle vortexing and centrifugation, amplification was performed on a Bio-Rad T100 thermal cycler under these conditions: initial denaturation at 94 °C for 2 min; 10 cycles of 98 °C for 10 s and 70 °C for 6 min; 20 cycles of 98 °C for 10 s and 68 °C for 6 min; final hold at 4 °C. Subsequently, 4 μL of each PCR product was mixed with 1 μL of 10 × loading buffer and analyzed by electrophoresis on a 0.5% agarose gel in 1 × TAE buffer at 200 V for 30 min, using the DL15000 DNA ladder (Takara) as a size standard. Gels were stained with a nucleic acid dye, and amplicons were visualized under UV light using a GelDoc-It TS imaging system (UVP).

Schematic diagram of the F8 Inv22 and Inv1 inversion detection. (A) The normal structure of the F8 gene. Red and blue boxes represent intron 22 homologous regions (int22h-distal, int22h-2; proximal, and int22h-1; intragenic), yellow and green boxes indicate intron 1 homologous regions (int1h-2; distal and int1h-1; intragenic), and gray boxes represent exon regions. The upper labels indicate primer binding sites (A, B, IR1, IF, EF1, DER1, P, Q). Xqtel: X-chromosome q arm telomere, Xqcen: X-chromosome q arm centromere. (B) Intron 22 type 1 inversion (Inv22 I): homologous recombination occurs between int22h-1 and int22h-3, resulting in fusion of int22h-3 and int22h-1 at both recombination junctions. Exons 1 to 22 are displaced toward the telomere and oriented opposite to their normal orientation. (C) Intron 22 type 2 inversion (Inv22 II): homologous recombination occurs between int22h-1 and int22h-2, resulting in fusion of int22h-2 and int22h-1 at both recombination junctions. The FVIII promoter region is displaced toward the centromere. (D) Intron 1 inversion (Inv1): homologous recombination occurs between int1h-1 and int1h-2, resulting in fusion of int1h-2 and int1h-1 at both recombination junctions. Exon 1 and the FVIII promoter region are displaced toward the telomere. The illustration is a simplified model of gene structure, with the scale adjusted to not represent the actual sequence length. The actual size shall be based on the number of bases indicated in the figure.

Primer poolMultiplex primer panelPooling volume ratioPooling1EF1 + DER1 + NP + NQ1:1:2:2Pooling2IF+IR1 + A + B1:1:2:2Pooling3EF1 + IR1 + P + B1:1:2:2Pooling4IF+DER1 + A + Q1:1:2:2

Multiplex PCR primer-pool composition and pooling ratios.

ReagentVolumeKOD Fx neo0.4 μLdNTPs Mixture(2 mM)5.0 μL2 × PCR buffer for KOD Fx neo12.5 μLPooling1/ Pooling2/ Pooling3/ Pooling41.5 μLgDNA(50 ng)X μLNuclease-Free Waterbring up to 25 μL

Multiplex PCR reaction mixture.

2.6 Quantitative PCR (qPCR)

To assess potential copy number variation in the F8 intron 1 region for Subject 6, qPCR was performed using two primer pairs (Supplementary Table 1) targeting ChrX:154182158–154,233,526. Reactions (20 μL) contained 10 μL SYBR Premix Ex Taq II (Takara), 0.2 μM of each forward and reverse primer, 2 μL of 1:5 diluted cDNA, and nuclease-free water. Amplification was conducted on a Bio-Rad CFX96 system: pre-denaturation at 95 °C for 30 s; 40 cycles of 95 °C for 5 s and 61 °C for 30 s; followed by a melt curve analysis from 65 °C to 95 °C with increments of 0.5 °C for 5 s to verify amplification specificity. The 2^(-ΔΔCt) method was used for relative quantification, normalizing to GAPDH and β-actin. Statistical significance between groups was determined by Student’s t-test or one-way ANOVA, with a p-value < 0.05 considered statistically significant.

2.7 CNV-seq

To further validate the duplication in the F8 intron 1 in Subject 6, CNV-seq was performed on Subject 6’s DNA by BGI Clinical Laboratories (ShenZhen) Co., Ltd. Following previously established protocols (10). Briefly, genomic DNA was fragmented by acoustic shearing, and sequencing libraries were constructed through end repair, A-tailing, adapter ligation, and PCR amplification. Library quality was assessed by Qubit fluorometry and AgilentBioanalyzer 2,100 (concentration ≥2 ng/μL, fragment size ~150–300 bp). Qualified libraries then underwent a series of procedural steps, including single-strand separation, circularization, and rolling circle replication, to generate DNA nanoballs. Subsequently, sequencing was conducted using the combinatorial probe-anchor synthesis (cPAS) method on the MGISEQ-2000 sequencer (BGI, Shenzhen, China). Raw data underwent preprocessing steps, such as sequence alignment, deduplication, and GC correction. CNV analysis was then performed using statistical algorithms to derive the results.

2.8 SV-seq

To clarify whether there is a cryptic duplication in the F8 intron 1 in Subject 6, SV-seq was conducted by Shanghai Jingyin Biotechnology Co., Ltd. SV-seq was performed following previously established method (11). Briefly, genomic DNA is extracted from the sample. After DNA fragmentation, sequencing libraries are constructed through end repair, A-tailing, linker ligation, and PCR amplification. Subsequently, sequencing is performed using the Illumina platform (0.1 × −5 × coverage, 36–100 bp read length). The obtained reads are aligned to the human reference genome (GRCh37/hg19 or GRCh38) using standard bioinformatics algorithms. Structural variations (SVs), including deletions, duplications, insertions, inversions, and translocations, were detected through comprehensive analysis of discordant read pairs, split reads, and read depth signatures. SV calling was performed using integrated algorithms, and candidate variants were filtered based on quality scores, read support, and population frequency.

3 Results3.1 High-throughput sequencing and sanger validation

High-quality sequencing data (Table 4) were obtained via sequencing. No F8 point mutations were detected in Subjects 1, 2, and 4. In Subject 3, the c.3168_3187del variant was identified and confirmed by Sanger sequencing (Figure 2A). This variant is not listed in the Human Gene Mutation Database (HGMD) or ClinVar but had been documented in the Leiden Open Variation Database (LOVD) with one literature report (12). Subject 5 carried the c.4379delA variant, which was also confirmed by Sanger sequencing (Figure 2B). The LOVD database2 has recorded 41 independent occurrences and recognized this site as a pathogenic variant.

Subject12345Total data (GB)17.6616.0115.0318.0913.82Coverage (%)99.9299.9199.6299.6299.60Specificity (%)83.1381.3181.8482.8783.76Uniformity (%)98.0797.5997.8898.0997.47Proportion of > 30X (%)98.9898.1398.1198.6997.62Average depth (X)191.33173.23161.74201.77155.12

Quality of high-throughput sequencing data in the target area.

Sanger sequencing results of subject 3 and subject 5. (A) Heterozygous F8 c.3168_3187del variant identified by Sanger sequencing in Subject 3. Arrows denote the mutation sites. (B) Heterozygous F8 c.4379delA variant identified by Sanger sequencing in Subject 5. Arrows denote the mutation sites.

3.2 LR-PCR analysis

Subject 1 was identified as a heterozygous carrier of Inv 22. Subject 2 was identified as a heterozygous carrier of Inv1. Subject 4 showed an atypical three-band electrophoretic pattern for Inv1, differing from the classic carrier profile (Figure 3). Subject 6 (the affected uncle) showed a normal Inv22 pattern but the same atypical three-band pattern for Inv1 as Subject 4. To clarify the underlying genetic mechanism, we performed additional genetic analyses for Subject 6 (Subject 4’s uncle), including qPCR, CNV-seq, and SV-seq.

LR-PCR analysis of Subject 4 and Subject 6. Long-range PCR showed no inv22 in Subject 4 (S4) or Subject 6 (S6). An aberrant three-band profile was observed for Inv1 in Subject 4 and Subject 6, inconsistent with wild-type or the heterozygous carrier state. Line 1, 5, 9, and 13 are amplification products of primers EF1, DER1, NP, and NQ. Lanes 2, 6, 10, and 14 are amplification products of primers IF, IR1, A, and B. Lanes 3, 7, 11, and 15 are amplification products of primers EF1, IR1, P, and B. Lanes 4, 8, 12, and 16 are amplification products of primers IF, DER1, A, and Q. Green arrow: intron 22 primer sets; Red arrow: intron 1 primer sets. NC, normal control (wild-type); S4: subject 4; S6: subject 6; Inv1 carrier: F8 intron 1 inversion carrier; Inv22 carrier: F8 intron 22 inversion carrier; NTC, no-template control. M: DL15000 DNA ladder.

3.3 Quantitative PCR analysis

QPCR was subsequently performed to validate the duplication at locus ChrX:154182158–154,233,526 in Subject 6. The results demonstrated a significant increase in copy number in the sample from Subject 6 compared to the normal male control, confirming a potential duplication in this genomic region (Figure 4).

qPCR result for Subject 6. (A) Bar graph showing the elevated copy number in Subject 6 relative to female and male controls. (B) Relative quantification (RQ) values obtained with two independent primer pairs for the female control, male control, and Subject 6. The data indicate a heterozygous duplication at ChrX:154182158–154,233,526.

3.4 CNV-seq validation

CNV-seq identified two CNVs in Subject 6 (Table 5), but no copy number abnormalities were detected on the X chromosome.

Variant designationClassificationFragment sizeOrigin of the variantsseq[GRCh37]8p22p21.3(18,102,071_19263288)x3
chr8:g.18102071 19263288dupVariant of uncertain significance1.16 MbUnknownsseq[GRCh37]8p22p22(16,826,647_16977490)x1
chr8:g.16826647 16977490delVariant of uncertain significance150.84KbUnknown

CNV-seq findings for Subject 6.

3.5 SV-seq validation

SV-seq analysis of Subject 6 identified multiple CNVs, including two duplications on the X chromosome: chrX:g.154168455_154231120dup (~63 kb, involving F8 intron 1 to intron 13) and chrX:g.154321974_154373215dup (Table 6). Based on combined LR-PCR analysis and SV-seq results, two potential structural rearrangement models for the F8 gene are proposed (Figure 5).

CNV findingsVariant descriptionFragment sizeACMG pathogenicity classification1 duplicationseq[hg19] dup (8) (p22q21.3) chr8:g.18115980_19235079dup1.12 MbVariant of uncertain significance2 duplicationseq[hg19] dup(X) (q28) chrX:g.154168455_154231120dup63 KbVariant of uncertain significance3 duplicationseq[hg19] dup(X) (q28) chrX:g.154321974_154373215dup51 KbVariant of uncertain significance4 deletionseq[hg19] del(Y) (q11.223q11.23) chrY:g. 24825000_28025000del(mos 50%)3.2 MbVariant of uncertain significance

SV-seq findings for Subject 6.

Schematic overview of the SV-seq findings. A1/A2: the upstream and downstream breakpoint ends at chX:154168455. B1/ B2: the upstream and downstream breakpoint ends at chX:154231120. C1/ C2: the upstream and downstream breakpoint ends at chX:154321974. D1/ D2: the upstream and downstream breakpoint ends at chX:154373215.

4 Discussion

The F8 gene spans approximately 3,186 kb, making it one of the largest genes in its genomic region. It comprises 26 exons and 25 introns, encoding a 2,332 amino acids polypeptide organized in the domain structure A1-A2-B-A3-C1-C2. The A domains are critical for FVIII protein synthesis and activation, while the B domain modulates its secretion. Small deletions or insertions within the B domain can induce frameshift mutations that disrupt normal splicing or transcription (13). The 9.1 kb coding region of F8 contains 70 CpG dinucleotides, corresponding to 140 potential base-pair change sites. Its high GC content contributes to hypermutability, with approximately 30% of mutations arising de novo (14). The extensive allelic heterogeneity and broad spectrum of copy number variations further drive the genetic heterogeneity characteristic of F8 gene mutations (15). To date, the Human Gene Mutation Database (HGMD) has cataloged 4,412 distinct variants in the F8 gene,3 including inversions, missense and nonsense mutations, deletions, duplications, and insertions/deletions. Among these, inversions represent a pivotal molecular mechanism, accounting for approximately half of all severe HA cases (16). In the present study, all five subjects undergoing carrier screening were found to carry F8 gene mutations. The detected mutational spectrum included intronic inversions and small nucleotide deletions.

Inv22 in F8 was detected in Subject 1, while Inv1 was detected in Subject 2 and 4. The Inv22 rearrangement arises from intrachromosomal homologous recombination between the *int22h-1* sequence within F8 intron 22 and its extragenic homologs located telomeric to the gene (17). This event splits the F8 gene into two inversely oriented segments: exons 1–22 are translocated to a distal site on Xq in reverse orientation, whereas exons 23–26 remain at the original locus. Consequently, the reading frame is disrupted, preventing the transcription of full-length F8 mRNA and leading to a severe deficiency of FVIII (18). Inv22 accounts for approximately 45% of all severe HA cases (4). Inv1 occurs at a lower frequency, but it accounts for about 2.94% of HA cases in the Chinese population (8). In patients with Inv1, the inversion results from homologous recombination between the *Int1h-1* sequence in F8 intron 1 and its telomeric homolog *Int1h-2* (18). This structural rearrangement disrupts the F8 gene and abrogates the production of full-length mRNA, ultimately leading to the severe HA phenotype due to a complete lack of functional FVIII (19). The detection of these inversions in Subjects 1, 2, and 4 underscores their significant role in the pathogenesis of hemophilia A.

The deletions c.3168_3187del and c.4379delA were identified in the F8 (NM_000132.4) genes of Subject 3 and Subject 5, respectively. Both variants are located within exon 14, a region characterized by long poly-adenine (poly-A) tracts that confer high susceptibility to replication slippage, making it a well-recognized mutational hotspot for frameshift mutations (20). The c.3168_3187del deletion leads to the substitution of glutamic acid with phenylalanine at amino acid position 1,057 and introduces a premature termination codon four residues downstream, likely resulting in a truncated protein. Although this 20-bp deletion is located within the B-domain, it is close to the A2-domain boundary and may therefore affect A2-A3 domain interactions (1). This variant was first reported in a prenatal diagnosis case in 2009. The male neonate carrying this mutation exhibited typical clinical manifestations of severe HA. Laboratory tests revealed markedly reduced factor VIII clotting activity (FVIII: C) at 0.60% of normal, consistent with the diagnostic criterion for severe HA (FVIII: C < 1%) (12). However, the pathogenic mechanism of this mutation has not yet been experimentally elucidated. The deletions c.4379delA is located within the B-domain of exon 14 in the F8 gene. Although the B-domain is dispensable for FVIII coagulant activity, it facilitates protein secretion. This deletion mutation results in the substitution of asparagine with isoleucine at amino acid position 1,460 and introduces a premature termination codon five amino acids downstream, likely producing a truncated protein. It has been identified in a Colombian cohort with severe HA, where the carrier exhibited a truncated FVIII protein and severely impaired coagulation activity (21). In a study of the hemophilia population in India, a patient carrying c.4379delA showed reduced FVIII: C (2.6%) and factor VIII antigen levels (FVIII: Ag) below 1%, consistent with a severe HA phenotype (14). The LOVD database lists 41 recorded cases of this mutation, which is classified as pathogenic. The frameshift leads to a premature termination codon (TAG), causing aberrant termination of FVIII synthesis and protein truncation. This truncation prematurely terminates the B-domain of FVIII, impairing its normal interaction with activated factor IX (FIXa) and the tenase complex, and disrupting the coagulation cascade (21). Previous studies have proposed that during F8 gene replicati