A Strategy for the Molecular Diagnosis in Hemophilia A in Chinese Population

ORIGINAL PAPER

A Strategy for the Molecular Diagnosis in HemophiliaA in Chinese Population

Zhihui He • Juan Chen • Shiyan Xu •

Shufen Chen • Xiao Xiao • Hongyi Li •

Yibin Guo • Weiying Jiang

Published online: 31 October 2012

� Springer Science+Business Media New York 2012

Abstract Hemophilia A is an x-linked recessive inherited

bleeding disorder. So far, more than 1,885 disease-causing

mutations of factor VIII gene have been identified. Clinic

confers a great challenge for the molecular diagnosis. We aim

to make a better strategy for the molecular diagnosis in

Hemophilia A. First, factor VIII intron 22 inversion and intron

1 inversion mutations were detected using Inversion-PCR and

double-tube multiple PCRs. And then, non-inversion muta-

tions were analyzed by denaturing high performance liquid

chromatography and/or direct sequencing. Novel mutations

were further analyzed the conservation and 3D structures by a

B domain deleted crystallographic model and bioinformatics.

Finally, we can indirectly confirm the diagnosis by linkage

analysis for the patients with the confusing diagnosis by the

techniques mentioned above. Eleven patients with the factor

VIII Inv 22 were found, and the remaining 16 patients were

found with 11 different mutations, of which 3 was novel

mutations affecting A1, B domains and splicing site. More-

over, the prenatal diagnosis was performed on 14 fetuses. Ten

fetuses were successfully confirmed to be normal, 1 fetus to be

a heterozygote with factor VIII c.3275–3276 ins A and 3

fetuses to be hemizygotes with factor VIII Inv 22 mutation.

Keywords DHPLC � Direct sequencing � Genetic

diagnosis � Hemophilia A � Prenatal diagnosis

Introduction

Hemophilia A (HA) is an X-linked recessive inherited

bleeding disorder affecting approximately 1/5,000 males [1],

seldom seen in females. Genetic mutation of anti-hemophilic

globulin A causes the disease, as a result of deficiency of

factor VIII for blood coagulation. Patients suffer from fre-

quent spontaneous or traumatic bleeding in several organs.

Hemorrhage into joints is the most common feature, recur-

rence could cause arthropathy with severe disability. If

intracranial hemorrhage occurs, it always leads to death.

Although the factor VIII replacement therapy is the most

common approach used in clinic, the emergence of alloan-

tibodies against factor VIII (which inhibits the pro-coagulant

activity) is beyond doubt. Lacking in effective method for

curing this disease, it is important to assist the HA families to

make informed decisions through carrier and prenatal test-

ing. We aim to make a strategy for the molecular diagnosis in

HA in Chinese population. We also focused on the HA

pathogenesis mechanisms mediated by the different novel

mutations. The affected gene is located on Xq28, 186 kb in

length. The disease presents allelic heterogeneity with more

than 1,885 affected alleles, including point mutations, small

insertions and deletions. However, large deletions and

insertions are rare. A listing of mutation types can be found at

‘‘The Hemophilia A Mutation Structure, Test and Resource

Site (HAMSTeRS)’’ (http://hadb.org.uk/) and Human Gene

Mutation Database (HGMD, http://www.hgmd.cf.ac.uk).

The factor VIII inversion 22 (Inv 22) collectively account for

about one-half of severe HA cases [2, 3]. The inversion 1 (Inv

1) accounts for about 5 % of severe cases of HA [4]. Factor

VIII comprises 26 exons, encoding a polypeptide chain of

2,351 amino acids, which includes a signal peptide of 19

amino acids and a mature protein of 2,332 amino acids. The

factor VIII protein is a large multi-domain glycoprotein

Zhihui He and Juan Chen (co-fist author) have contributed equally to

this study.

Z. He � J. Chen � S. Xu � S. Chen � X. Xiao � H. Li � Y. Guo �W. Jiang (&)

Department of Medical Genetics, Medical School and Key

Laboratory of Tropic Disease Control, Ministry Education,

Sun Yat-sen University, 74#, Zhongshan Road 2,

Guangzhou 510080, China

e-mail: [email protected]

123

Cell Biochem Biophys (2013) 65:463–472

DOI 10.1007/s12013-012-9450-2

http://hadb.org.uk/

http://www.hgmd.cf.ac.uk

composed of a heavy chain (domains A1–A2–B) and a light

chain (domains A3–C1–C2) [5].

Subjects and Methods

Subjects

Sixty-six subjects (including 27 patients and 39 carrier from

32 unrelated families, Table 1) and 50 healthy controls

from 5 different provinces in China, have been studied.

Fourteen pregnant women from these families asked for

prenatal diagnosis. Amniotic fluid was aspirated by

amniocentesis from 16 to 20 weeks of gestation and fetal

blood samplings were carried out by cordocentesis from 22

to 26 weeks of gestation. Phenotypic diagnosis of fetus was

established by standard coagulation assays to determine

plasma factor VIII activity. Patients and their parents or

legal representatives have been informed and consented to

participate in the study. The investigation was approved by

the Ethics Committee of Sun Yat-sen University.

Methods

Clinical Classification

According to the residual plasma FVIII coagulant activity

(FVIII: C), patients with HA can be classified as severe

(\1 %), moderate (1–5 %) and mild ([5–40 %) [6].

DNA Sample Prepare

Five microliter of peripheral blood and 0.5–2 ml of

umbilical cord blood were collected in tubes with ethylene

diamine tetraacetic acid. Fifteen microliter of amniotic

fluid sample was collected in 20 ml centrifuge tube.

Genomic DNA was extracted from peripheral blood by

using phenol–chloroform. Kits were used for gaining fetal

DNA from umbilical cord blood and amniotic fluid

simultaneously because of a few volumes. The purity,

quality and concentration of DNA were assessed by

ultraviolet spectrophotometry (260 and 280 nm) and aga-

rose gel electrophoresis.

Inversion-PCR (I-PCR) for the Inv 22

I-PCR was carried out based on the protocol described

previously by Rossetti et al. [7]. The following modifica-

tion has been made: (1) After digestion and ligation,

extracting DNA fragments by phenol–chloroform is not

used, instead, we precipitate the sample in sodium glacial

acetic acid and ethanol. (2) We use sodium glacial acetic

acid to replace sodium chloride for separating DNA, as

sodium glacial acetic acid provides more suitable ionic

strength for DNA precipitation. (3) We run 2 % agarose gel

to analyze I-PCR products, which can improve the reso-

lution. (4) Parameters of reaction have also been adjusted.

Time for Bcl I digestion is shortened to 2 h and total

volume of self-ligation is reduced from 600 to 150 ul. The

modification saves time, labor and money with success. It

is also safer for no toxic chemicals being used.

Double-Tube Multiple PCRs for the Inv 1

Double-tube multiple PCRs were performed as previously

described by Liang et al. [8].

Factor VIII Mutation Analysis

For the factor VIII Inv 22 and Inv 1 negative subjects, we

screened the mutation by DHPLC and/or direct sequencing.

The causative nature of the novel mutations was estab-

lished by the absence of these changes in a set of 50

anonymous DNA samples from healthy female individuals.

Mutation Screening by DHPLC

The PCR amplifications were carried out using some

primers described by Johannes et al. [9], Steve Keeney

(HAMSTeRS) and University of California Santa Cruz

(UCSC). Some new primers are self-designed (Table 2).

Total of 39 sets of primers were specifically designed to

amplify coding exons, exon–intron boundaries, promoter,

50 and 30-untranslated regions (UTR) of the factor VIII

gene (Table 2). PCR was performed with 100 ng DNA

template in a volume of 30 ul, containing 0.5 U of Taq

DNA polymerase (Fermentas), 0.3 uM each primer,

0.2 mM deoxynucleotide triphosphates (dNTP), 2.5 mM

MgCl2, and 3 ul 109 buffer. The annealing temperature for

the primers ranged between 53 and 64 �C. PCR reactions

were performed on GeneAmp PCR system BD-044 (Don-

gshenglong) using the following PCR condition: The initial

denaturation step was 94 �C for 3 min, followed by 35

thermocycles of 94 �C for 30 s, annealing 53–64 �C for

45 s and 72 �C for 50 s. The final extension was 72 �C for

10 min. The PCR products were detected by electrophorese

on 1.5 % agarose gel stained by ethidium bromide.

Equal volumes of a 7 ul PCR reaction product from the

patient and a wild type male control were mixed, heated for

5 min at 96 �C, cooled slowly at room temperature to allow

for heteroduplex formation. Female heterozygote’s needn’t

mix with wild type. Analysis of the heteroduplex and

homoduplex mixture was performed on a WAVE DNA

Fragment Analysis System E 2100B (Transgenomic, Omaha,

NE, USA).

464 Cell Biochem Biophys (2013) 65:463–472

123

The PCR products were sequenced in both orientations

by Invitrogen Company when we observed heteroduplex

peak. Mutations were always validated on a second inde-

pendent PCR product.

Direct Sequencing

Five patients and one carrier without the Inv 22 and Inv 1

asked for direct sequencing. The PCR primers and

Table 1 Mutations of F8 gene identified in Chinese patients, carriers and affected fetuses with HA in this study

Family No. of

patients

No. of

carrier

Affected

fetus

Unaffected

fetus

Nucleotide

change

Amino acid change Mutation Severitya Note

1 0 2 0 0 c.1063C[T p.R355X (Arg336Term) Non-sense Unknown b

2 1 1 0 1 c.670?1G[C Splice site Moderate Novel

3 0 1 0 0 Inv 22 Unknown b

4 1 1 0 1 Inv 22 Unknown b

5 1 1 0 0 c.6046C[G p.Arg2016Gly (Arg1997Gly) Missense Mild

6 1 1 0 0 Moderate c

7 2 2 2 0 Inv 22 Moderate


9 1 1 0 1 Unknown b,d

11 1 1 0 0 c.88G[A p.Glu30Lys (Glu11Lys) Missense Mild

12 1 1 0 0 Inv 22 Severe

13 0 1 0 0 Moderate e

14 0 1 0 1 Unknown b

15 1 1 0 0 Severe d

16 1 1 1 0 Inv 22 Severe

17 1 1 0 1 c.98G[A p.W33X(Trp14Term) Non-sense Moderate


19 0 1 0 1 c.2571delG p.Arg857SerfsX20

(Arg838SerfsX20)

Frame

shift

Unknown Novelb

20 1 2 0 1 c.4379delA p.N1460fsX5 (N1441fsX5) Frame

shift

Moderate

21 1 1 0 0 c.6682C[T p.R2228X (Arg2209Term) Non-sense Moderate

22 1 1 0 0 Inv 22 Mild

23 1 1 0 1 Inv 22 Severe

24 2 1 0 0 Moderate d

25 1 1 0 0 c.219C[A p.Phe73Leu (Phe54Leu) Missense Moderate Novel

26 1 1 0 0 Moderate c

27 1 2 0 0 Mild d

28 0 1 0 0 Mild e

29 1 1 0 0 Inv 22 Severe

30 1 1 1 0 c.3275–3276 Ins

A

p.N1092fsX25 (N1073fsX25) Frame

shift

Moderate

31 1 1 0 0 c.388G[A p.Gly130Arg (Gly111Arg) Missense Mild


33 0 2 0 0 Moderate e

Total 27 39 4 10

a According to the residual FVIII activity, HA is classified as severe (\1 %), moderate (1–5 %) and mild (6–30 %) [6]b Six patients in six families were unknown for the severity, because 3 patients passed away, the other 2 patients could not provide the data of

factor VIII activity, although their DNA-samples were obtained, and the last patient could not offer his blood samplec We could not detect the disease-causing mutations in F8 gene. The further study will be under going for detecting the mutations in VWF gened The diagnosis was confirmed by linkage analysise Although we gained the data of factor VIII activity from the individuals, we could not gained the DNA samples from them

Cell Biochem Biophys (2013) 65:463–472 465

123

Table 2 Primers and Tm of factor VIII gene and conditions of DHPLC screening

Name Primer sequence PCR size (bp) T (�C) Oven temperature (�C) B buffer (%)

P-1-F 50-GAGCTCACCATGGCTACATTC-30 560 59 56.5 59

P-1-R 50-AATTTAAAACTATAAAGCGAGTCCTG-30 58 57

P-2-F 50-GGACCTAGGCCATGGTAAAGA-30 601 60 56 58

P-2-R 50-TGCAGAGCATTTTAAGGAACTTT-30 58.2 58

E-1-F 50-TAGCAGCCTCCCTTTTGCTA-30 480 60 56 58

E-1-R 50-CTAACCCGATGTCTGCACCT-30 59 54

E-2-F 50-CATTACTTCCAGCTGCTTTTTG-30 290 62 58.2 56

E-2-R 50-TTTGGCAGCTGCACTTTTTA-30

E-3-F 50-GTACTATCCCCAAGTAACCTT-30 205 54 59.3 47

E-3-R 50-CATAGAATGACAGGACAATAGG-30

E-4-F 50-TACAGTGGATATAGAAAGGAC-30 296 54 58.3 51

E-4-R 50-TGCTTATTTCATCTCAATCCTACGCTT-30

E-5-F 50-CCTCCTAGTGACAATTTCCTA-30 188 56 55.9 47

E-5-R 50-AGCAGAGGATTTCTTTCAGGAATCCAA-30

E-6-F 50-CAGGGAAGGAGAAAGGGG-30 252 58 56.3 54

E-6-R 50-GAACTCTGGTGCTGAATTTGG-30 55.7 53

E-7-F 50-CAGATTCTCTACTTCATAGCCATAG-30 324 54 57.5 52

E-7-R 50-ATTAAAAGTAGGACTGGATA-30

E-8-F 50-ATATAGCAAGACACTCTGACA-30 336 56 58.1 55

E-8-R 50-AGAGAGTACCAATAGTCAAA-30

E-9-F 50-AGAGTTGGATTTGAGCCTACC-30 284 56 54.6 57

E-9-R 50-CAGACTTTTTCTTCTTACCTGACCTT-30

E-10-F 50-GGATTTGATCTTAGATCTCGC-30 205 53 55 47

E-10-R 50-ATTTTAGTTGTTATTGATGA-30

E-11-F 50-TTGAGCTATTTATGGTTTTG-30 294 53 58.5 50

E-11-R 50-GACATACACTGAGAATGAA-30

E-12-F 50-GCATTTCTTTACCCCTTTCA-30 230 54 59.2 50

E-12-R 50-CTTTATTCACCACCCACTG-30

E-13-F 50-GATGTGTCTAAATCTCTTTTC-30 261 56 58.2 51

E-13-R 50-ATATAATAACTAACCTGGGTTTTCCATC-30

E14-1F 50-ATCTGTGTTATGAGTAACCA-30 430 58 57.8 56

E14-1R 50-TCATATTTGGCTTCTTGGAG-30

E14-2F 50-CATGGGCTATCCTTATCTGA-30 479 60 56.4 57

E14-2R 50-CATGGGCTATCCTTATCTGA-30

E14-3F 50-TCAAAGTTGTTAGAATCAGG-30 441 62 55.6 57

E14-3R 50-ATTTTGTGCATCTGGTGGAA-30

E14-4F 50-GTCCAACAGAAAAAAGAGGG-30 481 60 56.5 55

E14-4R 50-CTACATTTTGCCTAGTGCTC-30 54.3 57

E14-5F 50-CTGGCACTAAGAATTTCATG-30 429 63 57.4 54

E14-5R 50-CCTTCTCATTGTAGTCTATC-30 56.7 55

56 57

E14-6F 50-GAAACATTTGACCCCGAGCA-30 431 60 57.3 55

E14-6R 50-TTTTGGGCAAGTCTGGTTTC-30

E14-7F 50-CACATACAAGAAAGTTGAGA-30 436 54 58.5 55

E14-7R 50-CTCATTTATTGCTGCTATTG-30 57.8 55

56.8 56

E14-8F 50-GATACCATTTTGTCCCTGAA-30 418 58 57.4 57

E14-8R 50-GTCACAAGAGCAGAGCAAAG-30 56.5 57


123

condition were the same as above-mentioned. The PCR

products were sequenced in both orientations by Invitrogen

Company. Mutations were always validated on a second

independent PCR product.

The mutations detected were compared with the

HAMSTeRS, HGMD and new documents to establish

novelty.

Linkage Analysis

According with linkage analysis conditions, Hind III, Bcl I

and St14 (DXS52) linkage analysis were performed for 4

families (family 9, 15, 24 and 27) because we could not

directly confirm the diagnosis by molecular analysis with

DHPLC and/or Direct sequencing for them.

Description of the Mutant Protein

The description of the mutant protein was based on protein

sequence NP_000123.1, and the translation initiator

methionine is numbered as ?1. As the codon numbering

taken from the literature and reference mutation databases

(according to which the 19 amino acids containing signal

peptide are numbered in reverse, the initial methionine is

numbered as -19, and the first alanine of the mature

protein is numbered as ?1) differs from the journal

Table 2 continued

Name Primer sequence PCR size (bp) T (�C) Oven temperature (�C) B buffer (%)

E-15-F 50-CACCTAGGAAAATGAGGATGT-30 300 53 55.5 53

E-15-R 50-ATAGTCAGCAAGAAAATAAA-30

E-16-F 50-AAGATCCTAGAAGATTATTC-30 330 58 56.3 54

E-16-R 50-TTAGTACACAAAGACCATTT-30

E-17-F 50-TGATGAGAAATCCACTCTGG-30 349 58 57.8 55

E-17-R 50-GTGCAATCTGCATTTCACAG-30

E-18-F 50-TCCTTCTCCAGCAATCAAT-30 271 56 57.2 53

E-18-R 50-TCCCAGTGCCTAGACCAT-30

E-19-F 50-GCAAGCACTTTGCATTTGAG-30 342 60 56 50

E-19-R 50-AGCAACCATTCCAGAAAGGA-30 56.3 56

E-20-F 50-CCATTTTCATTGACTTACATTTGAG-30 195 59 58 48

E-20-R 50-AGATATAATCAGCCCAGGTTC-30

E-21-F 50-TTTATTCTCAAGTGTCTAGGACTAACC-30 278 58 55.2 53

E-21-R 50-CAAATCATTAAGGCATTCTGTTC-30

E-22-F 50-AAATAGGTTAAAATAAAGTGTTAT-30 206 53 58 50

E-22-R 50-GACTAATTACATACCATTAAG-30

E-23-F 50-CTCTGTATTCACTTTCCATG-30 250 54 56.8 53

E-23-R 50-ACAGTTAGTCACCCTACCCA-30

E-24-F 50-GCTCAGTATAACTGAGGCTG-30 249 56 58.3 52

E-24-R 50-CTCTGAGTCAGTTAAACAGT-30

E-25-F 50-CACCTAGGAAAATGAGGATGT-30 372 56 58.5 53

E-25-R 50-ATAGTCAGCAAGAAAATAAA-30 55 56

E26-1F 50-CTGTGCTTTGCAGTGACCAT-30 557 62 57.2 61

E26-1R 50-TTCTACAACAGAGGAAGTGGTGA-30 56 59

E26-2F 50-GGAGAAACCTGCATGAAAGC-30 596 60 54.5 60

E26-2R 50-TTGGCCATCACAAATTTCAA-30 54.8 59

E26-3F 50-TGCAAATGTGCATTTTTCTGA-30 580 60 55 59

E26-3R 50-CCTCCAGCCCCCTTTACTAT-30 54.3 60

E26-4F 50-CCACCCCCATAAGATTGTGA-30 580 64 55 58

E26-4R 50-CTGAAGAAACCAGCAGGAAAA-30 55.9 58

E26-5F 50-CCCCAAAGGTGATATGGTTTT-30 230 60 54.6 52

E26-5R 50-TCAGTGTTCACATTTTTATTTCCA-30 53 55

Sources of primers P-1, P-2, E-1, E-2, E26-(1-5): Steve Keeney (HAMSTeRS); E-6, E-21: UCSC; E-18: self-designed; the others: Ref. [8]


123

approved nomenclature, the traditional numbering is also

shown in the text and tables.

Conservation Analysis

Amino acids that were subjected to replace through novel

mutation were examined for their conservation in porcine,

murine and canine factor VIII using the publicly available

multiple-sequence alignment lineup on the factor VIII

mutation database (http://hadb.org.uk/WebPages/Database/

Protein/lineups.html).

Protein of Factor VIII Prediction Programs

A major concern in human molecular genetics is to deter-

mine whether a certain mutation is functionally neutral or

can alter the protein function and ultimately cause disease.

Thus, the pathogenic characteristic of each novel missense

mutation was analyzed based on the protein prediction of

three-dimensional structures to evaluate the possible impact

of an amino acid substitution on the structure and function of

the mutant protein using the CBS server (http://www.cbs.

dtu.dk/services/CPHmodels). Protein sequence segments

with the corresponding mutations were submitted for com-

parative homology modeling.

B domain deleted crystallographic model (2r7e) of factor

VIII was taken for domain structure prediction from the

Protein Data Bank (PDB, http://www.rcsb.org/pdb/home/

home.do/) [10]. Referent model and mutated model were

superimposed using Accelrys DS Visualizer 1.7 and rayed

using PyMOL 0.99. Hydrogen bonds were observed using

Accelrys DS Visualizer 1.7.

Aim to evaluate the pathogenicity of mutations located at

the boundaries between introns and exons, the consensus

splicing sequences was analyzed using the splice site predic-

tion program (http://www.fruitfly.org/seq_tools/splice.html)

Prenatal Diagnosis

Eleven fetuses were under prenatal diagnosis by a combi-

nation of I-PCR, double-tube multiple PCRs, DHPLC and/

or direct sequencing, linkage analysis and factor VIII

activity measurement of umbilical cord blood. One fetus

had not been analyzed by the methods mentioned above,

but the measurement of VIII activity. The other 2 fetuses’

amniotic fluid was gained for the prenatal diagnosis.

Results

In this study, we examined 27 patients clinically affected

with HA from different geographical areas of China. The

initial screening for factor VIII Inv 22 revealed that 11 out

of 27 cases were positive for this abnormality. None had

Inv 1. The remaining 16 patients were then detected for

mutations in factor VIII 26 exons, exon–intron boundaries,

promoter, 50 and 30-UTR. By DHPLC screening, we

observed heteroduplex peaks on exon 1, 5, 8 and 19. After

sequencing, we found 5 mutations, including 2 mutations

on exon 1. Using direct sequencing, we found 6 mutations

on exon 2, 3, 14-2, 14-3, 14-6 and 24, respectively. Two

patients were diagnosed by linkage analysis (Table 3). The

other 2 families could not provide polymorphism infor-

mation (Table 4).

Except factor VIII Inv 22, 11 different mutations (4

missense, 3 non-sense, 2 deletions, 1 insertion and 1 splice

site mutation) were also identified. Of these mutations, 1

were novel missense, 1 was novel deletion and 1 was novel

splice site mutation (neither deposited to date in the

HAMSTeRS database and HGMD, nor reported in recently

published articles). Figure 1 showed the sequencing results

of all novel mutations and recurrent mutations. All the

mutations were found in the families with a positive family

history. Any presence of the novel mutation was excluded

in a control sample of 50 health female DNAs. Factor VIII

c.219C[A, p.Phe73Leu (Phe54Leu) caused by novel mis-

sense mutation was highly conserved in porcine and canine

and had similar residues in murine. It suggested an

important role for protein function (Fig. 2). The newly

detected amino acid substitution was also analyzed for

conformational changes and influence on molecular sta-

bility for factor VIII domain with available structures,

using homology modeling. Factor VIII c.219C[A,

p.Phe73Leu (Phe54Leu) caused the formation of 6 helixes,

changed the location of hydrogen bond and reduced three

hydrogen bonds (Fig. 3).

The novel deletion mutation (c.2571delG) located in

exon 14 causes a frame shift mutation leading to intro-

duction of 19 new amino acids before a premature stop

codon.

One novel splicing mutation was located at c.670?1G[C

affecting the specific donor site. Splice site prediction for

normal c.670?1G with donor score cutoff 0.40, acceptor

score cutoff 0.40 (exon/intron boundary shown in larger

font) showed the wild splicing site: gatgaag/gttagtgagtct.

However, Splice site prediction for abnormal c.670?1C with

donor score cutoff 0.40, acceptor score cutoff 0.40 (exon/

intron boundary shown in larger font) showed the abnormal

variation in the splicing site: gatgaagctta/gtgagtct. One more

residue of Alamine was introduced.

. Eight fetuses hadn’t inherited the mutations from their

mothers. Combining with the normal factor VIII activity in

umbilical cord blood, we diagnosed them as normal fetuses

and all of them born as normal baby boys. One fetus hadn’t

the mutation of her mother’s by analyzing DNA, we

diagnosed her as normal fetus and she born as normal baby


123

http://hadb.org.uk/WebPages/Database/Protein/lineups.html

http://hadb.org.uk/WebPages/Database/Protein/lineups.html

http://www.cbs.dtu.dk/services/CPHmodels

http://www.cbs.dtu.dk/services/CPHmodels

http://www.rcsb.org/pdb/home/home.do/

http://www.rcsb.org/pdb/home/home.do/

http://www.fruitfly.org/seq_tools/splice.html

girl. One female fetus was diagnosed as a heterozygote

with factor VIII c. 3275_3276 ins A by analyzing DNA

from amniotic fluid and she has born as the heterozygote.

Three male fetuses were diagnosed as hemizygote with

factor VIII Inv 22 mutation. One fetus had not been ana-

lyzed by the methods mentioned above, but the measure-

ment of VIII activity. The factor VIII activity and factor

VIII mutation of cord blood samples showed Table 5. The

normal range of fetal factor VIII activity is 25–53 % [11].

Discussion

Factor VIII genotyping is now systematically assessed

using direct sequencing, but screening methods such as

DHPLC, conformational-sensitive gel electrophoresis, and

high resolution melting analysis are also used. Direct

sequencing is more expensive than DHPLC, but DHPLC is

unavailable for detecting the mutation of F8 Inv 22. Lab-

oratories may make choice according to their condition.

We worked out a better strategy for HA mutations in

Chinese population (Fig. 4). By the first step, the Inv 22

will be detected by I-PCR and the Inv 1 will be detected by

double-tube multiple PCRs. By the second step, factor VIII

mutations will be monitored by DHPLC and/or direct

sequencing. By the last step, linkage analysis will be done

for the patients with the confusing diagnosis by the tech-

niques mentioned above. We believe it is very helpful in

HA patients for clinical diagnosis and genetic counseling.

The prevalence of HA in China is not well-known. The

molecular genotype of HA was carried out to develop

mutational analysis and to evaluate genotype–phenotype

correlation that can help in carrier detection and prenatal

diagnosis. The putative role of the detected novel muta-

tions in causing disease is not always obvious. We thus also

focused on the HA pathogenesis mechanisms mediated by

the different novel mutations, integrating molecular, fam-

ily, and conservative parameters.

In this study, 27 HA patients were subjected to factor

VIII molecular analysis following clinical examination. In

addition, our analysis ruled out the possibility that the

novel mutations were related to a polymorphism, as we did

not find any of these among the 50 normal female controls.

The study confirms that the factor VIII Inv 22 is very

common genetic abnormality of HA in China. We do not

find Inv 1. The prevalence of Inv 1 is estimated to be 1.5 %

in Iranian patients and 1.8 % in patients of British origin

[12, 13]. It is difficult for us to find factor VIII Inv 1

mutation due to the relatively small sample size.

Missense mutations are usually of particular interest

because they pinpoint functionally important amino acids.

One novel missense mutation factor VIII c.219C[A,

p.Phe73Leu (Phe54Leu) in our study was conserved in

porcine, murine and canine. Amino acid substitutions at

interspecies-conserved positions appear to be of importance

in the development of HA. For 3D structure, factor VIII

c.219C[A, p.Phe73Leu (Phe54Leu) causes the formation

Table 3 Results of 32 unrelated families

Family Patient I-PCR Double-tube

multiple PCRs

DHPLC DS Linkage

analysis

1 0 - - - / /

2 1 - - ? ? /

3 0 ? / - / /

4 1 ? / / / /

5 1 - - ? ? /

6 1 - - - / -

7 2 ? / / / /

8 1 ? / / / /

9 1 - - - / ?

11 1 - - ? ? /

12 1 ? / / / /

13 0 - - - / /

14 0 - - - / /

15 1 - - - / -

16 1 ? / / / /

17 1 - - ? ? /

18 1 ? / / / /

19 0 - - ? ? /

20 1 - - ? ? /

21 1 - - ? ? /

22 1 ? / / / /

23 1 ? - / / /

24 2 - - - - -

25 1 - - ? - /

26 1 - - - / /

27 1 - - - / ?

28 0 - - - / /

29 1 ? / / / /

30 1 - - ? ? /

31 1 - - ? ? /

32 1 ? / / / /

33 0 - - - / /

Table 4 Results of linkage analysis

Family Severity FVIII level % Bcl I Hind III St14 (bp)

9 Unknowna Unknowna ± ; 700/1,300

15 Severe 0.3 ?/? -/- 700/700

24 Moderate 1.7, 2b ?/? -/- 700/1,570c

27 Mild 20 ± ; 700

a This patient could not provide the data of factor VIII activityb There were 2 patients in 1 familyc Parents were the same


123

of 6 helixes, changed the location of hydrogen bond and

reduced three hydrogen bonds (Fig. 3). This patient was

moderate and his FVIII activity was 1.3 %. Phe 54 is buried

in A1 domain and located on an irregular loop composed of

Val52 to Pro67 [14]. The pathogenic mechanism may be

involved in that the mutant protein destroyed the cupre-

doxin-like sub-domains that form an extensive interface.

From our data, we can conclude that the missense mutations

change the factor VIII domain topology and influence the

molecular stability of the corresponding chain segments or

protein regions. Compared to HAMSTeRS mutation data-

base and HGMD novel mutations, factor VIII c.219C[A,

p.Phe73Leu (Phe54Leu) was located at residues previously

related to HA, but we found different nucleotide substitu-

tion. This suggested the allelic heterogeneity and the

importance of the residue for maintaining normal structure

and function of factor VIII. Unlike genetic or ‘locus (non-

allelic) heterogeneity, in which mutations are in different

genes may explain one variant phenotype, allelic hetero-

geneity implies that different alleles in the same gene can

cause a similar variant phenotype. It is a different diagnosis

challenge for us to identify the pathogenic mutations. In this

study, the procedure, including analysis of conservation,

prediction of 3D structures of the mutant protein and

exclusion of polymorphisms, is available to confirm the

disease-causing mutations of F8 gene.

We found a novel deletion mutation involving frame

shift in codon Arg838 (p. Arg857). The novel deletion is

predicted to cause a frame shift mutation leading to

introduction of 19 new amino acids before a premature stop

Fig. 1 The sequencing results

of all novel mutations. a1Sequence analysis of normal

exon 2 of factor VIII gene. a2Sequence analysis of affected

exon 2 of factor VIII gene

showing a point mutation

(c.219C[A). The arrowsindicate the 219C and HA

219A, respectively. b1Sequence analysis of normal

exon 14-2 of factor VIII gene.

b2 Sequence analysis of

affected exon 14-2 of factor

VIII gene showing a deletion

mutation (c.2571delG). The

arrows indicate the normal G

and deleted G, respectively. c1Sequence analysis of normal

intron 5 of factor VIII gene. c2Sequence analysis of affected

intron 5 of factor VIII gene

showing a point mutation

(c.670?1G[C). The arrowsindicate the 670?1G and HA

670?1C, respectively

Fig. 2 Amino acid sequence alignment of human factor VIII and

other homologous proteins. The figure represents partial alignment of

A1 domain sequences of human, porcine, murine, and canine (HUM

FVIII, PIG FVIII, MUR FVIII, and CAN FVIII). The missense

mutation is denoted by an arrow pointing at the position above the

aligned sequences

Fig. 3 Model image of factor VIII A1 domain and mutation.

a Normal amino acid Phe54, p.Phe73 is presented in magenta.

b Mutated amino acid Phe54Leu, p.Phe73Leu is presented in red(Color figure online)


123

codon on exon 14. It leads to premature stop and causes

HA.

According to the splice site prediction program, the

novel mutation c.670?1G[C affecting the specific con-

sensus donor site (GT) causes a frame shift mutation

leading to introduction of a residue of alamine before a

premature stop codon. As a result of the aberration splic-

ing, we believe the mutation, c.670?1G[C, may contrib-

uted to the disease and lead to a moderate HA phenotype.

However, no patient RNA sample was available from liver,

spleen, lymph nodes, and kidney cells, in which F8-gene

expresses. It is limitations for us to further verify the

abnormal splice.

The clinical severity of HA dose not correlate with

genotypes and appears to vary among different patients

(Table 1). We found 8 mutations (c.88G[A, c.98G[A,

c.388G[A, c.1063C[T, c.3275–3276 ins A, c.4379del A,

c.6046C[G, and c.6682C[T), which had been reported.

Our patient (c.6046C[G) and that reported by Ma et al.

[15] were mild. Two of our patients (c.88G[A, c.388G[A)

were mild. But c.88G[A reported by Ghafar et al. [16] is

moderate and c.388G[A reported by Markoff et al. [17] is

moderate/severe. Five of them (c.98G[A, c.1063C[T,

c.3275–3276 ins A, c.4379del A, and c.6682C[T) lead to

severe or unknown phenotype of HA. But in our study, 4

patients are moderate. Another one is a carrier and the

proband of her family has passed away. One possible

reason for this phenomenon is the presence of an additional

molecular change on some of the mutated or wild alleles,

which regulates the gene expression pathologically. The

presence of a second mutation in the coding region of

factor VIII was excluded by the sequencing analysis or

DHPLC screening.

A better strategy for HA mutations is a combination of

running I-PCR for the Inv 22, double-tube multiple PCRs

for the Inv 1, and detecting non-inversion factor VIII

mutations by DHPLC and/or direct sequencing. It is rec-

ommended to undergo both DHPLC and/or direct

sequencing to detect the factor VIII mutation and bio-

chemical assay to measure the factor VIII activity of

umbilical cord blood in prenatal diagnosis.

Our cohort is a small group and does not cover the

whole Chinese population. Therefore, a larger study with

more Chinese patients is needed to establish a solid con-

clusion about the prevalence of various mutations in Chi-

nese patients with HA. It will be helpful to understand the

mechanism of HA and helpful to detect carriers and

affected fetus.

Acknowledgments We thank Qun Fang, Yanmin Luo of Fetal

Center of the First Affiliated Hospital of Sun Yat-sen University for

helping us to acquire umbilical cord blood and amniotic fluid.

Table 5 Factor VIII activity of

umbilical cord blood and

mutation results

? Positive, - negative, / had

not donea Two affected fetuses in this

familyb Diagnosed by linkage analysisc Amniotic fluid

Family FVIII activity and I-PCR

result of umbilical cord blood (%)

I-PCR result

of proband

Other

mutation

Diagnosis

2 45.1 / - - Normal

4 50.3 - ? - Normal

7a 0.7 ? ? / Abnormal

8 90.4 - ? / Normal

9 36 - - Unknownb Normal

14 88.6 - Passed away - Normal

16 / ? ? / Abnormal

17 41.8 / - - Normal

18 55.4 - ? / Normal

19 42.9 / Passed away - Normal

20 36 / - - Normal

30 /c / - ? Heterozygote

23 65 - ? / Normal

Fig. 4 Strategy for HA mutations in Chinese population. ? Positive/

Abnormal, - Negative/Normal, DT-PCR/Double-tube multiple PCR


123

W. Jiang is supported by Chinese National Natural Scientific Grant

(No. 31171214)

References

1. Bolton-Maggs, P. H., & Pasi, K. J. (2003). Haemophilias A and

B. Lancet, 361, 1801–1809.

2. Lakich, D., Kazazian, H. H., Jr., & Antonarakis, S. E. (1993).

Inversions disrupting the factor VIII gene are a common cause of

severe hemophilia A. Nature Genetics, 5, 236–241.

3. Naylor, J., Brinke, A., Hassock, S., Green, P. M., & Giannelli, F.

(1993). Characteristic mRNA abnormality found in half the

patients with severe haemophilia A is due to large DNA inver-

sions. Human Molecular Genetics, 2, 1773–1778.

4. Bagnall, R. D., Waseem, N., Green, P. M., & Giannelli, F. (2002).

Recurrent inversion breaking intron 1 of the factor VIII gene is a

frequent cause of severe hemophilia A. Blood, 99, 168–174.

5. Vehar, G. A., Keyt, B., Eaton, D., Rodriguet, H., O’Brien, D. P.,

Rotblat, F., et al. (1984). Structure of human factor VIII. Nature,312, 337–342.

6. White, G. C., I. I., Rosendaal, F., Aledort, L. M., Lusher, J. M.,

Rothschild, C., & Ingerslev, J. (2001). Definitions in hemophilia:

Recommendation of the scientific subcommittee on factor VIII

and factor IX of the scientific and standardization committee of

the International Society on Thrombosis and Haemostasis.

Thrombosis and Haemostasis, 85, 560.

7. Rossetti, L. C., Radic, C. P., Larripa, I. B., & Brasi, C. D. (2005).

Genotyping the hemophilia inversion hotspot by use of inverse

PCR. Clinical Chemistry, 51, 1154–1158.

8. Liang, Y., Yan, M., Xiao, B., & Liu, J. (2008). Analysis of intron

1 factor VIII gene inversion for direct diagnosis in patients with

severe haemophilia A in China. Prenatal Diagnosis, 28, 160–161.

9. Johannes, O., Vytautas, I., Simone, R., Andreas, F., Karen, W.,

Elke, H. F., et al. (2001). Evaluation of DHPLC in the analysis of

hemophilia A. Journal of Biochemical and Biophysical Methods,47, 39–51.

10. Shen, B. W., Spiegel, P. C., Chang, C. H., Huh, J. W., Lee, J. S.,

Kim, J., et al. (2008). The tertiary structure and domain organi-

zation of coagulation factor VIII. Blood, 111, 1240–1247.

11. Panigrahi, I., Ahmed, R. P., Kannan, M., Kabra, M., Deka, D., &

Saxena, R. (2005). Blood analysis for prenatal diagnosis of thal-

assemia major and hemophilia A. Indian Pediatrics, 42, 577–581.

12. Rastegar, L. G., Enayat, M. S., Arjang, Z., Lavergne, J. M., &

Ala, F. (2004). Identification of intron 1 and 22 inversion muta-

tions in the factor VIII gene of 124 Iranian families with severe

haemophilia A. Haemophilia, 10, 410–411.

13. Cumming, A. M., & UK Haemophilia Centre Doctors’ Organi-

zation Haemophilia Genetics Laboratory Network. (2004). The

factor VIII gene intron 1 inversion mutation: prevalence in severe

hemophilia A patients in the UK. Thrombosis and Haemostasis,2, 205–206.

14. Xue, F., Zhang, L., Sui, T., Ge, J., Gu, D., Du, W., et al. (2010).

Factor VIII gene mutations profile in 148 Chinese hemophilia A

subjects. European Journal of Haematology, 85, 264–272.

15. Ma, G. C., Chang, S. P., Chen, M., Kuo, S. J., Chang, C. S., & Shen,

C. (2008). The spectrum of the factor 8 (F8) defects in Taiwanese

patients with haemophilia A. Haemophilia, 14, 787–795.

16. Ghafar, A., Bogdanova, N., Lim, L. C., Zhao, Y., Markoff, A., &

Tien, S. L. (2010). Ten novel factor VIII (F8C) mutations in

eighteen haemophilia A families detected in Singapore. Haemo-philia, 16, 545–566.

17. Markoff, A., Gerke, V., & Bogdanova, N. (2009). Combined

homology modelling and evolutionary significance evaluation of

missense mutations in blood clotting factor VIII to highlight

aspects of structure and function. Haemophilia, 15, 932–941.


123

Documents

A Strategy for the Molecular Diagnosis in Hemophilia A in Chinese Population