From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
The Molecular Genetic Analysis of Hemophilia A: A Directed Search Strategy for
the Detection of Point Mutations in the Human Factor VI11 Gene
By J.K. Pattinson, D.S. Millar, J.H. McVey, C.B. Grundy, K. Wieland, R.S.Mibashan, U. Martinowitz, K. Tan-Un,
M. Vidaud, M. Goossens, M. Sampietro, P.M. Mannucci, M. Krawczak, J. Reiss, 6. 2011, D. Whitmore, S.Bowcock,
R. Wensley, A. Ajani, V. Mitchell, C. Rizza, R. Maia, P. Winter, E.E. Mayne, M. Schwartz, P.J. Green, V.V. Kakkar,
E.G.D. Tuddenham, and D.N. Cooper
-
A directed-search strategy for point mutations in the
factor Vlll gene causing hemophilia A was used to screen
eight potentially hypermutable CpG dinucleotides occurring at sites deemed to be of functional importance.
Polymerase chain reaction-amplified DNA samples from
793 unrelated individuals with hemophiliaA were screened
by discriminant oligonucleotide hybridization. Point mutations were identified in 16 patients that were consistent
with a model of 5-methylcytosine (5mC) deamination. Four
new examples of recurrent mutation were demonstrated
at the following codons: 336 (CGA
TGA), 372
(CGC
TGC), 372 (CGC
CAC). and 1689 (CGC
TGC).
These are functionally important cleavage sites for either
T
activated protein C or thrombin. Further novel C
transitionswere identified in the remaining arginine codons
screened -5, 427. 583, 795, and 1696). resulting in the
creation of TGA termination codons. Differences in mutation frequency were found both within and between the
CpG sites and between ethnic groups. These differences
are assumed to be due to differences in the level of
cytosine methylation at these sites. although direct evidence for this inference is lacking.
0 1990 by The American Society of Hematology.
H
DNA samples exhibited variant Taq I bands on a Southern
blot, two of these subsequently proving to be nonsense codons
(CGA TGA). This finding was confirmed by Youssoufian
et al, who also provided the first evidence for recurrent
mutation due to CG TG transitions at identical sites in the
factor VI11 gene.4 Further examples of recurrent mutation
have since been found in five different Taq I restriction sites
within exons 18,22,23,24, and 26.’
That CpG is a “hotspot” for mutation is evidenced by (a)
its under-representation or “suppression” in genomic DNA:
(b) the high frequency of polymorphism detected by restriction enzymes containing CpG in their recognition sequences:
(c) the high rate of CpG substitution observed in evolutionary studies on vertebrate genes’ and (d) the high frequency
(-33%) of CG
TG and CG
CA transitions among
point mutations causing human genetic disease.’.’’ The CpG
dinucleotide is the preferred site for cytosine methylation in
higher eukaryotes” and 5-methylcytosine (5mC) is prone to
mutate to thymidine by deamination.” C
T and G
A
(resulting from 5mC deamination on the antisense strand)
transitions are thus predicted to occur frequently in the
heavily methylated human genome.
The characterization of point mutations in the factor VI11
gene is important, not only as a means toward understanding
the basis of CpG hypermutability with all its diagnostic
implications, but also as a way of relating the structure of the
factor VI11 protein to its function.” Such studies should
eventually enable us to provide explanations at the molecular
level for the variety of different phenotypes reported to date.
There are 70 CpG sites in the factor VI11 gene coding
sequence, of which 7 occur in Taq I sites. Five of these are
T transition would
CGA (arginine) codons, which a C
convert to a TGA termination codon (Fig 1). Because these
lesions are certain to come to clinical attention, it is not
surprising that a high frequency of nonsense mutations have
already been found in the factor VI11 gene by Taq I site
screening. However, the majority of CpG dinucleotides do
not fall within available restriction enzyme recognition
sequences; these nevertheless may be studied using a combi-
-
-
- -
EMOPHILIA A is an X-linked bleeding disorder
affecting approximately 1 in 5,000 males, and is
caused by deficiency of factor VIII, a cofactor in the
activation of factor X by factor IXa. Characterization of
point mutations in the human factor VI11 gene has been
hampered by its large size and its complex structure; the 9-kb
factor VI11 mRNA is encoded by a gene comprising 26 exons
spanning 186 kb of chromosomal DNA.’ Hemophilia A is a
good model disease for the study of mutation because of its
relatively high prevalence, its variable phenotype (implying
allelic heterogeneity), and the high proportion of new mutations due to the lowered probability of survival of disease
Gitschier et a1 found that the restriction enzyme Taq I
(recognition sequence TCGA) detected point mutations in
the factor VI11 gene of hemophilia A patients.’ Four of 92
From the Haemostasis Research Group, MRC Clinical Research
Centre, Harrow, and the Molecular Genetics Section, Thrombosis
Research Institute, Chelsea, London, England (in collaboration
with King’s College Hospital, London;National Hemophilia Center, Tel-Hashomer, Israel; H6pital Henri Mondor, Crktkil. France;
A . Bianchi Bonomi Hemophilia and Thrombosis Center, Milan,
Italy; Institut fur Humangenetik. Gottingen, West Germany; Lewisham Hospital, London; Hammersmith Hospital, London; Royal
Infirmary, Manchester, England; Fatimid Foundation, Karachi,
Pakistan; Royal Infirmary. Leicester. England; Haemophilia Centre. Oxford, England; Centro Hospitalar. Coimbra,Portugal; Royal
Victoria Hospital. Belfast, Ireland; Rigshospitalet, Copenhagen,
Denmark; and Saint Mary’s Hospital, Portsmouth, England).
Submitted January 8.1990; accepted July 23.1990.
Supported by The Medical Research Council, the Thrombosis
Research Trust, and the Factor VIII Research Trust.
Address reprint requests to David N . Cooper. PhD. Molecular
Genetics Section. Thrombosis Research Institute, Manresa Rd,
Chelsea, London S W3 6LR. England.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement”in accordance with 18 U.S.C.section 1734 solely to
indicate this fact.
0 1990 by The American Society of Hematology.
OOO6-4971/90/7611-0102$3.OO/0
2242
-
-
-
-
-
-
-
Blood, Vol76, No 1 1 (December 1). 1990: pp 2242-2248
From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
MOLECULAR GENETIC ANALYSIS OF HEMOPHILIA A
2243
1941 2116 2147 2209 2307
>Y\ 1
a1
A1
CGA
Codons:
I
-5
B
A2
tTT
T
427 583
336
A3
T
795
I
CGC
Codons:
1689
Fig 1. Location of dmed CpQ dinucleotides in factor Vlll cDNA. Top line: Taq I sites titat contain CGA in the reading frame (cadon
numbering according to Vehar er ap). Second line: Diagram of the factor Vlll cDNA coding sequence with predicted domeins of the
protein.= Third line: All seven CGA codons not in Teq I sites. Bottom line: Two CGC codons that encode arginine in critical thrombin
cleavage sites of t h e meture protein.
nation of the polymerase chain reaction (PCR) and discriminant oligonucleotide hybridization.
In this article, we describe the systematic screening of
DNA samples derived from 793 hemophilia A patients for
mutation at eight CpG sites within the factor VI11 gene;
these correspond to codons 336, 372, and 1689 (proteolytic
cleavage sites) and - 5 , 427, 583, 795. and 1696 (CGA
codons).
METHODS
B / d samples. Ten to twenty milliliters EDTA-anticoagulated
blood samples were drawn from 793 unrelated individuals with
hemophilia A. Approval for this study had been granted by the
appropriate local committees; patients were informed that the blood
was being drawn for research purposes and that their privacy would
be protected. Four hundred sixty-five were severe. 64 were moderately severe. and 119 were mild as determined by factor VIII:C
measurement (<I% severe, 1% to 5% moderate, >5% mild). The
remainder of samples (145) were from patients whose condition was
of undetermined severity. The ethnic distribution of the samples
screened was 71 7 whites/Europesns/Israelis. 56 Pakistanis/
Indians. and 20 Chinese.
DNA isolation. DNA was isolated from lymphocyte pellets
made from blood samples by established methods.“.”
In vitro amplification using the PCR. Specific fragments of the
factor Vlll gene were amplified by PCR using the following
oligonucleotides. Codon - 5 : DMS SGCA AAT AGA GCT CTC
CAC CTG C3’ and DM7 SGGC AGC TCA CCG AGA TCA CT
3’. Codons 336.372, and 427: DM494 SATG GCA TGG AAG CTT
ATG TCA A3 and JP37 SCAA CAG TGT GTC TCC AAC TTC
3’. Codon 583: DM496 STCA GAC AAG AGG AAT GTC ATC
CGG T3’ and DM497 SGGA ACT CTG GAT CCT CAA GCT
G3’. Codon 795: DM9 5’GAA CCA AGA AGC TTC TCC CA 3’
and DMlO SGGA GAT CAG ATA AGG ATA G C 3’. Codons
I689 and 1696: DM639 SCCC GGG CAA AGC AAG GTA GGA
CTG3’ and DM642 SGAG CCT CTC CAC TGC AGC A3’.
Oligonucleotides were made or obtained as described below. Between 25 ng and l r g of genomic DNA, 0.5 r g of each oligonucleotide, and 1.5 units Taq DNA Polymerase (Cetus, hconsfield,
England) were added to 95 pL buffer containing 1.5 mmol/L MgCI,,
IO mmol/L Tris-HCI (pH 8.3). 50 mmol/L KCI. 0.01%gelatin, and
200 pmol/L each deoxynucleotide 5’ triphosphate (dNTP). The
reaction mixture was initially incubated at 94% for 2.5 minutes
before undergoing 40 cycles of PCR (denaturation at 94OC for 30
seconds, annealing a t 55% for 15 seconds. extension at 72% for 2
minutes [codons -5. 583. 795, 1689. and 16961 or IO minutes
[codons 336. 372, and 4271) using a Perkin-Elmer-Cetus Thermal
Cycler. Negative controls (PCR mix minus genomic DNA) were
included in all amplification experiments as a check for contamination. Ten microliters of the amplification product was electrophoresed on a 2% agarose gel and visualized under ultraviolet light
after ethidium bromide staining (Fig 2).
Dot/s/ot blotting. The amplified DNA was denatured by adding
525 r L 0.1 mol/L NaOH and incubating at 7OOC for 20 minutes.
M
1
2
3
4
llcbp-
SOObp-
1SObp-
Fig 2. Agarose gel electrophoresisof PCR-amplHied fregm.nt.
of t h e bctor Vlll gene. lane 1: exon Eintron h x o n 9 (approximerely 700 bp) amplified using JP37 and DM494. Lane 2: exon 1
(124bp)cosmplifiedwithS‘endofexonl4(213 bp) usingDMSand
DM7 (exon 1) and DM9 and DMlO (exon 14). Lane 3 exon 12 1123
bp) amplified using D M 4 9 6 and 497. Lane 4 3’ end of exon 14 (244
bpl amplified using D M 6 3 9 and DM842.
From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
PATINSON ET AL
2244
then neutralized with 600 pL 0.1 mol/L HCl/10 x saline sodium
citrate (SSC) (1 x SSC = 150 mmol/L NaC1/15 mmol/L sodium
citrate). Four hundred-microliter DNA samples were applied under
suction and in triplicate to Genescreen nylon membranes (DupontNEN, Stevenage, England) using a Slot-Blot or Dot-Blot manifold
(BRL, Uxbridge, England). After rinsing the slots with 100 pL 5 x
SSC, the DNA was covalently bound to the membrane either by
exposure to 120 mJ ultraviolet radiation (Stratalinker, Stratagene,
La Jolla, CA), or by heating in a vacuum oven at 80°C for 2 hours.
Discriminant oligonucleotide hybridization (DOH). For every
site to be studied, three oligonucleotide probes were designed to
correspond to sequences containing either (a) the wild-type (CpG),
T transition (TpG), or (c) a G -+ A transition (CpA)
(b) a C
within that sequence. Oligonucleotides were synthesized on an
Applied Biosystems (Warrington, England) 380A Synthesizer, and
purified by polyacrylamide gel electrophoresis, elution, and phenol/
chloroform extraction, or were obtained commercially (Oswel DNA
Service, Edinburgh, Scotland).
Oligonucleotide probes (40 ng), were end-labeled by incubation
with 70 pCi y - '*P-adenosine triphosphate (3,000 Ci/mmole;
Amersham, England) and 5 U T4 polynucleotide kinase at 37°C for
30 minutes in 50 mmol/L Tris-HC1 (pH 7.4), 10 mmol/L MgCl,, 5
mmol/L dithiothreitol (DTT), and 10 mmol/L spermidine. Labeled
probes were purified by gel filtration through Sephadex G50 columns
(Nick columns, Pharmacia, Uppsala, Sweden).
Membranes were prehybridized for 4 hours at 33OC in 10 mL
0.12% bovine serum albumin, 0.12% polyvinylpyrolidone, 0.12%
ficoll 400, 0.1% sodium dodecyl sulphate, and 200 pg/mL sheared
herring sperm DNA. Membranes were then hybridized overnight
(16 hours) under identical conditions in the presence of 1 to 2 x lo7
-
disintegrations per minute "P-labeled oligonucleotide probe. Membranes first were washed at 33°C and then at temperatures empirically determined to optimize discriminant hybridization of the
oligonucleotide probes (Table 1). The washing buffer was 3 mol/L
tetramethylammonium chloride (TMACl), 50 mmol/L Tris-HC1
(pH 8.0). 4 mmol/L EDTA (pH 8.0), and 0.1% sodium dodecyl
sulphate. TMACl was used in the hybridization wash because, by
binding A-T base pairs selectively, it avoids the preferential melting
of A-T over G-C base pairs and thereby removes the effect of base
composition on oligonucleotide hybridization.16 Dot blots probed
with wild-type and mutant oligonucleotides were autoradiographed
after washing, both at nondiscriminant and discriminant temperatures. This permitted assessment of initial hybridization and subsequent differential melting of the oligonucleotideprobes and allows us
to be confident that we have detected all the potential mutations in
our patient sample. Wherever possible, known mutant DNAs were
used as positive controls.
Analysis of restriction fragment length polymorphism (RFLP).
Where identical mutations were detected, four intragenic RFLPs
were analyzed. The XbaI and Bcll polymorphismswere amplified by
PCR using oligonucleotides described by Kogan et a1.I' Thirty
microliters of the amplification product were digested with XbaI or
BclI and electrophoresed through 2% agarose or 12% polyacrylamide gels and visualized under ultraviolet illumination. The Bgll"
and MspI'' RFLPs were analyzed by digestion of patient DNA with
the relevant enzyme followed by Southern blotting and hybridization
to probes 131BH (a 1.2-kbBamHI/HindIII fragment of exon 26) or
p625.3, respectively.
Direct sequencing. PCR-amplified material was purified using
Geneclean (Bio 101, La Jolla, CA). Appropriate oligonucleotide
Table 1. Oligonucleotides and Washing Temperatures Used for Discriminant Hybridization
Codon
Number
-5
336
372
427
583
795
1,689
1,689
Exon
1
8
8
9
12
14
14
14
-
Washing
Temperature
("C)
DNA
Sequence
Amino
Acid
CGA
Arg
Term
Gln
JP 1
JP2
JP3
CTGTGCCTITGCGATCTG
CTGTGCCTmGTGATTCTG
CTGTGCCTTTTGCAATCTG
63
Arg
Term
Gln
JP4
JP5
JP6
GGAACCCCAACTACGAATGA
GGAACCCCAACTATGAATGA
GGAACCCCAACTACAAATGA
64
Arg
CYS
His
JP7
JP8
JP9
CCllTATCCAAATTCGCTCAG
CCTITATCCAAATTGCTCAG
CCTITATCCAAATTCACTCAG
63
Arg
Term
Gln
JP 10 AAAGTCCGATITATGGCATAC
JP 1 1 AAAGTCTGATITATGGCATAC
JP 12 AAAGTCCAATITATGGCATAC
63
Arg
Term
Gln
JP 13 TGAGAACCGAAGCTGGTACC
JP 14 TGAGAACTGAAGCTGGTACC
JP 15 TGAGAACCAAAGCTGGTACC
64
Arg
Term
Gln
JP 16 GATGCTCTGCGACAGAGTC
JP 17 GATGCTCTTGTGACAGAGTC
JP 18 GATGCTCTTGCAACAGAGTC
62
Arg
CYS
His
JP 19 CAGAGCCCCCGCAGCm
JP20 CAGAGCCCCTGCAGCTIT
JP2 1 CAGAGCCCCCACAGCIIT
62
Arg
Term
Gln
JP22 GAMACACGACACT
JP23 GAAAACATGACACT
JP24 GAAAACACAACACT
48
CGA
CGC
CGA
CGA
CGA
CGC
CGA
Bases marked in bold type denote the codons screened for C
- T and G
Oligonucleotide (5'
3')
A transitions by the discriminantoligonucleotides.
From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
MOLECULAR GENmC ANAL=
OF HEMOPHILIA A
2246
primm were used for sequencing. either by the method of G m CI
al.?”orby using the T 7 polymerase kit (Phannacia).
RESULTS
Screening of patient DNA. A total of 793 unrelated
patients with hemophilia A were screened by PCR amplification of D N A sequence surrounding the eight codons of
interest (Fig 2). followed by discriminant hybridization with
end-labeled oligonucleotidesto dot/slot blots of the amplified
D N A samples (Fig 3). Six point mutations were detected at
codon 336. two at codon 372, one at codon - 5 , one at codon
427. one at codon 583. one at codon 795. three at codon 1689,
and one at codon 1696 (Table 2). The mutation was
confirmed in all cases by direct sequencing of PCR-amplified
material (data not shown). Mutations were therefore detected in all the codons screened. The phenotypic data for the
16 patients harboring these mutations are also presented in
Table 2. Eleven of the mutations detected gave rise to Severe
hemophilia A as a result of the creation of a termination
codon. Four of the remaining five mutations caused moderately Severe hemophilia A and one caused severe hemophilia
A as a consequence of a substitution within one of the two
thrombin cleavage sites. Only one of the 11 nonsense
mutations (codon 1696) was associated with the production
of antibody to factor VIII.
The Occurrenceof identical mutations may be due either to
recurrent mutations or to the inheritance of identical-bydescent alleles. To exclude the latter we conducted both
RFLP haplotype analysis and genealogical investigations.
RFLP haplotype data is shown in Table 3 and distinguished
three patients with a mutation at codon 336. but none of
those possessing thecodon 1689 mutation. Extremegeographical separation allows patient HI9 to be distinguished from
J91. Therefore, five of six patients with the C to T mutation
at codon 336 harbor independent recurrent mutations.
Whereas the three 1689 mutations are indistinguishable by
RFLP analysis. patient 5403 is a sporadic case in the three
generations available for study, and no common ancestor was
found between this family and those of patients J l S S and
5242 in the four generations examined. Statistical analysis
was therefore performed assuming five recurrent mutations
at codon 336, and two at codon 1689. If all codons were
equally likely to mutate, the probability of observing at least
five mutations in one codon and at least two in another is
12%. Although not statistically significant, this may indicate
that codon 336 is hypermutable compared with the other
codons.
There is also a skewed distribution of point mutations
detected within CpG sites. If we exclude consideration of
codon 336 where G
A substitutions may not give rise to
hemophilia, eight of the remaining nine-point mutations
proven to be independent were C T transitions (P .022).
A disproportionate number of these independent mutations Occurred in individuals of Inda-Pakistani extraction (3
of 56) as compared with the remainder of the European
whites ( I 1 of 7 17) (Fisher’s exact test. P = .074).
-
-
C
J84
C
c
c
J397
J88
JP5
JP6
Figs.
Dotbkad.mp(M.dp.tkmDNA,
336 (A) and 1690 (9). A t both
codonCtheconnds ram& hykiditd adyto tho
wild-typodigonud.otid. UP4 and JP22I. w h u r a
patima’ DNA r W i M hW6dh.d o#tbto 0 l i g O n ~ poad n codon
clootidoa containing tho C-to-1 mutotion (JP6 and
JP23). No DNA. ronuimd hybridhod to tho 0ligOnuclootidoa m a i n i n g tho G-to-A mumtion (JW
ond JP24) at thoso 8ttoa.
JPZ
B
-
DISCUSSION
The detection of high-frequency recurrent C
T and
G A mutations within Taq I restriction sites’” isconsistent
with a model of 5mC deamination. Indeed. such “experiments of nature” have provided indirect evidence for the
Occurrence of 5mC in CpG dinucleotides within the factor
V l l l gene. However,Southern blot analysis permits only a
crude assessment of relative mutation rates at CpG dinucla
otides; only Seven Taq I sites are available for analysis in the
factor V l l l gene, and the reliability of fragment detection
and resolution is variable. Furthermore, without D N A se
quencing. the precise nature of any point mutation detected
by Taq I remains unknown. To date, large-scale systematic
surveys of CpG mutations in specific genes arc lacking.
Moreover. whereas in principle the screening of potentially
hypermutable CpG dinucleotides at functionally important
JP4
A
-
-
JP23
JP24
@
e,
C
Hl9
From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
2246
PATTINSON ET AL
Table 2. Point Mutations in the Factor Vlll Gene Causing Hemophilia A Patient Data
Patient
Facta
Bleeding
Tendency
Facta
VIII:Ag (%I
Inhibitors
<1
-
<1
<1
-
<1
<1
<l
<1
<1
3
3
<1
<1
<1
Severe
Severe
Severe
Severe
Severe
Severe
Severe
Moderate
Moderate
Severe
Severe
Severe
J242
J155
J403
5
4
4
Moderate
Moderate
Severe
tl
<1
100
100
ND
J397
<l
Severe
ND
Codon.
Base-Pair
Substitution
Amino Acid
Substitution
Acronym
-5
336
336
336
336
336
336
372
372
427
583
795
CGA-+TGA
CGA-+TGA
CGA-+TGA
CGA-+TGA
CGA-TGA
CGA-+TGA
CGA+TGA
CGC+TGC
CGC-CAC
CGA-+TGA
CGA-+TGA
CGA+TGA
ArpTerm
ArpTerm
ArpTerm
Arg-Term
Arg-Term
Arg-Term
Arg-Term
Arg-Cys
Arg-His
Arg-Term
Arg-Term
ArpTerm
H541
H19
J84
J88
J9 1
J138
J278
J254
H453
H466
J34
H518
<1
<1
1689
1689
1689
CGC+TGC
CGC+TGC
CGC-TGC
Arg-Cys
Arg-Cys
Arg-Cys
1696
CGAdTGA
Arg-Term
VIII:C
(%I
Abbreviation: ND, not determined.
'Codon numbering based on the system of Vehar et
tEthnogeographic origin denotes country of origin, not necessarily the place of abode.
positions could prove to be valuable in increasing the efficiency of mutation search procedures, no assessment of the
practical utility of this strategy has been reported until now.
There are 36 CpG dinucleotide-containingarginine codons
in the factor VI11 gene. These include 12 CGA codons and
two CGC codons at the two thrombin cleavage sites critical
for factor VI11 activation as depicted in Fig l.''*" One of the
CGA codons (codon 336) occurs at the cleavage site for
activated protein C, which inactivates factor VIII. We have
analyzed the five CGA codons located in Taq 1 restriction
sites by Southern blotting and will present mutations identified in our patient sample elsewhere.
Six of the seven CGA codons that occur outwith Taq I sites
and both CGC codons were selected for mutation screening
using a combination of PCR and DOH, thereby circumventing most of the problems associated with Taq 1 screening.
The remaining CGA codon (1966) is flanked by a repetitive
A + T rich sequence, which is likely to reduce severely
oligonucleotide hybridization specificity and efficiency, and
was for this reason omitted from the study. All sense strand
C -+ T transitions within CGA codons should result in a
severe phenotype by producing a termination codon (TGA).
<1
tl
ND
ND
74
ND
tl
Ethnogeogaphic
origint
Denmark
Germany
Pakistan
Pakistan
Pakistan
USSR
UK
Pakistan
Israel
UK
UK
Denmark
UK
UK
UK
UK
-
-
-
+
-
T transitions (except in codon 336),
Antisense strand C
may cause hemophilia by substitution of glutamine (CAA)
for arginine? A C
T transition in the antisense strand at
codon 336 may result in a thrombophilic phenotype by
preventing factor VI11 inactivation. 5mC deamination in the
two CGC codons located within thrombin cleavage sites
would result in the substitution of cysteine for arginine if the
mutation occurs on the sense strand, and histidine for
arginine if it occurs on the antisense strand. Both are
chemically radical substitutions and would be predicted to
abolish thrombin cleavage and hence factor VI11 activation.
We have screened eight CpG dinucleotides in the factor
VI11 genes of 793 hemophiliac DNAs, and have detected 16
mutations. We have estimated elsewhere'' that an average
CpG dinucleotide (-80% are methylated)" in the human
genome is subject to a 12-fold higher frequency of mutation
than any other dinucleotide. We have calculated on this basis
that a total of 14 mutations would be expected in this survey
(see Appendix). The observed hypermutability of CpG in
factor VI11 is thus comparable with that previously calculated for other genes, and is consistent with the model of
deamination of 5mC.
Table 3. Haplotwe Analvsis of Patients With Recurrent Factor Vlll Mutations
Patient No.
Bcl I
Xbal
Bgl I
Mspl
Origin
5.0
5.0
5.0
20
5.0
5.0
4.3
4.3
4.3
7.5
4.3
4.3
Pakistan
Pakistan
Pakistan
USSRIAustria
UK
FRG
5.0
5.0
5.0
4.3
4.3
4.3
UK
UK
UK
~
Codon 336
J84
J88
J9 1
J138
5278
H19
Codon 1689
J155
J242
J403
+
+
+
+
+
+
From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
2247
MOLECULAR GENETIC ANALYSIS OF HEMOPHILIA A
We have demonstrated four new examples of recurrent
mutation of CpG dinucleotides in the factor VI11 gene
causing hemophilia A. The most dramatic is at codon 336,
where five proven independent cases of a C
T transition
T mutation (H4) has been
were found. Only one C
previously reported at this site.22The second example of
recurrent mutation is provided by codon 1689, where we have
T mutations, one of which
detected two independent C
causes moderate hemophilia A. This moderate phenotype
contrasts with the severe phenotype of patient 5403 and
T
the only other patient (H10) reported with this C
tran~ition.~’,’~
The reason for this discrepancy is unclear and
is under investigation. The G A transition found converting Arg 372 to His has been reported in one other unrelated
family resulting in mild (factor VII1:C 5%) hemophilia A.”
Finally, a further example of recurrent mutation occurred in
the 5’ cytosine of the CpG dinucleotide at codon 372; the
mutation reported here (described in more detail el~ewhere)~’
has been reported once before in a similarly affected patient.26
Hitherto undescribed point mutations were also detected
in codons -5, 427, 583, 795, and 1696; the C
T base
changes in these five patients resulted in the creation of
premature termination codons. Because each of these patients has undetectable factor V1II:Ag in his plasma, or
antibodies to factor VI11 (J397), presumably any truncated
protein translated is either nonsecreted or unstable in the
circulation. All four previously reported point mutations in
patients with inhibitors were termination codons located in
the factor VI11 light chain. Further studies should determine
whether the position of the mutation in the factor VI11 gene
is of relevance to the development of inhibitors in hemophiliacs.
There may be differences in mutation rates between the
CpG dinucleotides examined in this study. We also have
noted a probable difference in the overall mutation rate in the
eight sites studied between patients of Indo-Pakistani and
Caucasian origin. These differences may reflect partial
methylation resulting from differential methylation of CpG
dinucleotides, either between different germ cells of an
individual or between different individuals in a population.
The observed between-site differences in mutation rates
could be caused by a higher level of methylation (and
therefore of 5mC deamination) at some codons than others
(eg, codon 336). Similarly, the racial differences could result
from a higher overall level of methylation in these codons in
Indo-Pakistanis as compared with whites. These racial differences, and the fact that sperm remains the sole tissue in
which differential methylation of particular CpG dinucleotides has not been demon~trated?~
suggests that we are
observing differences in the methylation status of specific
sites between individuals.
There is also a disparity in frequency between C --* T and
G
A mutations within the eight codons studied. Previous
studies also have reported a significantly higher number of
C
T over G
A substitution^.^.^ There would appear to
be three possible reasons to account for this finding: (a)
hemimethylation of CpG sites so that the sense strand
cytosine is more frequently methylated than the antisense
-
-
-
-
-
-
-
strand cytosine, (b) a strand preference in mutation repair
efficiency, or (c) the relative absence of phenotypic effect of
G A substitutions leading to a bias in mutation detection.
These explanations are not mutually exclusive.
Arginine residues encoded by CpG-containing codons are
a common feature of proteolytic cleavage sites in coagulation
factor proteins. Indeed, examples of recurrent mutation at
CpG dinucleotides have already been reported in the two
factor XIa cleavage sites of factor IX at residues Arg 14528-30
and Arg 180;’ in the factor Xa cleavage site of thrombin at
~ . ~in
~ the thrombin cleavage site of
the Arg 271 r e s i d ~ e ? and
protein C.34Potentially hypermutable CpG-containing arginine codons are also present in the proteolytic cleavage sites
of factors V, VII, X, XI, XII, XIIIa subunit, and
B-fibrin~gen.~~
If these CpG dinucleotides are methylated in
the germline, then they may well represent hotspots for
mutation leading to removal of the cleavage sites and a
consequent bleeding disorder. Recurrent mutation at specific
CpG sites has also been shown to be a cause of antithrombin
111 deficiency, heparin cofactor I1 deficiency, and
dy~fibrinogenemia.~~.~’
If it is eventually demonstrated that CpG mutability
correlates with methylation status, it will become advisable
to establish the methylation status of specific CpG sites in a
given gene before any labor-intensive screening of patient
samples is undertaken. Perhaps when the cellular role of
cytosine methylation is understood, we shall be nearer to
predicting patterns of gene methylation and hence of mutability.
--+
ACKNOWLEDGMENT
We thank the following clinicians, who provided individual blood
samples: Drs F. Hill, M. McEvoy, H. Daly, D. Mitchell, R.
Vaughan-Jones, M. Winter, D. Bevan, S. Height, G. Savidge, G.
Scott, D. Prangnell, M. Adelman, J. Hayes, J. Shirley, M. Stevens,
T.K. Lam, and A. Li. We gratefully acknowledge the secretarial
assistance of Margaret Runnicles and the statistical advice of
Caroline Dort.
APPENDIX
An Estimate of the Number of Mutations Expected When
Screening 8 CpG Dinucleotides in 793 Patients
Approximate
Figures
Proportion of hemophilia A cases due to point
mutations’
Relative mutability of CpG dinucleotide”
Numbers of CpG dinucleotidesnot screened
Number of non-CpG dinucleotidesin the factor
VI11 gene
Average proportion of base changes giving rise to
an amino acid substitution
793
0.95
(8
X
8 X 12
12) + 0.66 X [(62
X
12)
0.95
12.6
62
6,982
0.66
+ 6,9821 = 13’9
From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
PATINSON ET AL
2248
REFERENCES
1. Gitschier J, Wood WI, Goralka TM, Wion KL, Chen EY,
Eaton DH, Vehar GA, Capon DJ, Lawn RM: Characterization of
the human factor VI11 gene. Nature 312326,1984
2. Haldane JBS: The rate of spontaneous mutation of a human
gene. J Genetics 31:317,1935
3. Gitschier J, Wood WI, Tuddenham EGD, Shuman MA,
Goralka TM, Chen EY, Lawn RM: Detection and sequence of
mutations in the factor VI11 gene of haemophiliacs. Nature 3 19427,
1985
4. Youssoufian H, Kazazian HH, Phillips DG, Aronis S , Tsiftis
G, Brown VA, Antonarakis SE Recurrent mutations in haemophilia
A give evidence for CpG mutation hotspots. Nature 324:380, 1986
5. Antonarakis SE, Youssoufian H, Kazazian HH: Molecular
genetics of hemophilia A in man. Mol Biol Med 4:81,1987
6. Nussinov R Nearest neighbour nucleotide patterns; structural
and biological implications. J Biol Chem 256:8458, 1981
7. Barker D, Schaefer M, White R Restriction sites containing
CpG show a higher frequency of polymorphism in human DNA. Cell
36:131,1984
8. Cooper DN, Krawczak M:Cytosine methylation and the fate
of CpG dinucleotides in vertebrate genomes. Hum Genet 83:181,
1989
9. Cooper DN, Youssoufian H: The CpG dinucleotide and human
genetic disease. Hum Genet 78:151, 1988
10. Cooper DN, Krawczak M: The mutational spectrum of single
base-pair substitutions causing human genetic disease: Patterns and
predictions. Hum Genet 8555, 1990
11. Cooper DN: Eukaryotic DNA methylation. Hum Genet
64:315, 1983
12. Duncan BK, Miller JH: Mutagenic deamination of cytosine
residues in DNA. Nature 287560,1980
13. Pittman DD, Kaufman RJ: Structure-function relationships
of factor VI11 elucidated through recombinant DNA technology.
Thromb Haemost 61:161,1989
14. Kunkel LM, Smith KD, Boyer SH, Borgaonkar DS, Wachtel
SS,Miller OJ, Breg WR, Jones HW, Rary JM: Analysis of human
Y chromosome-specific reiterated DNA in chromosome variants.
Proc Natl Acad Sci USA 74:1245,1977
15. Cooper DN, Smith BA, Cooke HJ, Niemann S, Schmidtke J:
An estimate of unique DNA sequence heterozygosity in the human
genome. Hum Genet 69:201,1985
16. Wood WI, Gitschier J, Lasky LA, Lawn RM: Base composition-independent hybridization in tetramethylammonium chloride:
A method for oligonucleotide screening of highly complex gene
libraries. Proc Natl Acad Sci USA 82:1585,1985
17. Kogan SC, Doherty M, Gitschier J: An improved method for
prenatal diagnosis of genetic diseases by analysis of amplified DNA
sequences. Application to haemophilia A. N Engl J Med 317:985,
1987
Kiltur SD, Patel AS, Kazazian
18. Antonarakis SE, Waber E,
HH, Mellis MA, Counts RB, S!amatoyannopoulos G, Bowie W,
Fass DN, Pittman DD, Wazney JM, Toole JJ: Haemophilia
A-detection of molecular defects and of carriers by DNA analysis.
N Engl J Med 313:842,1985
19. Youssoufian H, Phillips DG, Kazazian HH, Antonarakis S E
MspI polymorphism in the 3' flanking region of the human factor
VI11 gene. Nucleic Acids Res 15:6312,1987
20. Green PM, Berdley DR, Mibashan RS, Nilsson IM, Giannelli F Molecular pathology of haemophilia B. EMBO J 8:1067,
1989
21. Pittman DD, Kaufman RJ: Proteolytic requirements for
thrombin activation of antihemophilic factor (factor VIII). Proc
Natl Acad Sci USA 85:2429,1988
22. Gitschier J, Kogan S, Levinson B, Tuddenham EGD: Mutations of factor VI11 cleavage sites in hemophilia A. Blood 72:1022,
1988
23. OBrien DP, Tuddenham EGD: Purificationand characterization of factor VI11 1,689-Cys: A non-functional cofactor occurring in
a patient with severe hemophilia A. Blood 73:2117,1989
24. Arai M, Inaba H, Higuchi M, Antonarakis SE, Kazazian HH
Jr, Fujimaki M, Hoyer L W Direct characterization of factor VI11 in
plasma: Detection of a mutation altering a thrombin cleavage site
(arginine-372 histidine). Proc Natl Acad Sci USA 86:4277,1989
25. Pattinson JK, McVey JH, Boon M, Ajani A, Tuddenham
EGD: CRM+ haemophilia A due to a missense mutation (372-Cys)
at the internal heavy chain thrombin cleavage site. Br J Haematol
15:73, 1990
26. Shima M, Ware J, Yoshioka A, Fukui H, Fulcher C A An
arginine to cysteine amino acid substitution at a critical thrombin
cleavage site in a dysfunctional factor VI11 molecule. Blood 74:1612,
1989
27. Silva AJ, White R: Inheritance of allelic blueprints for
methylation patterns. Cell 54145, 1988
28. Noyes CM, Griffith MJ, Roberts HR, Lundblad RL: Identification of the molecular defect in factor IX Chapel Hill: Substitution
of histidine for arginine at position 145. Proc Natl Acad Sci USA
804200,1983
29. Toomey JR, Stafford D, Smith K Factor IX Albuquerque
(Arginine 145 to cysteine) is cleaved slowly by factor IXa and has
reduced coagulant activity. Blood 72:312, 1988 (suppl)
30. Liddell MB, Peake IR, Taylor SAM, Lillicrap DP, Giddings
JC, Bloom AL: Factor IX Cardiff A variant factor IX protein that
shows abnormal activation is caused by an arginine to cysteine
substitution at position 145. Br J Haematol72:556, 1989
31. Huang M-N, Kasper CK, Roberts HR, Stafford CW, High
KA: Molecular defect in factor IX Hilo, a hemophilia Bm variant:
Arg + Gln at the carboxyterminal cleavage site of the activation
peptide. Blood 73:718, 1989
32. Rabiet MJ, Furie BC, Furie B: Molecular defect of prothrombin Barcelona. Substitution of cysteine for arginine at residues 273. J
Biol Chem 261:15045,1986
33. Rabiet MJ, Furie BC, Furie B: Molecular defect in prothrombin Madrid Substitution of arginine 273 by cysteine precludes
activation. Thromb Haemost 58:313,1987
34. Grundy C, Chitolie A, Talbot S, Bevan D, Kakkar V, Cooper
DN: Protein C London 1: Recurrent mutation at Arg 169
(CGG TGG) in the protein C gene causing thrombosis. Nucleic
Acids Res 17:10513, 1989
35. Cooper DN, Tuddenham EGD: The Molecular Genetics of
Coagulation Disorders. New York, NY, Oxford (in press)
36. Blinder MA, Andersson TR, Abildgaard U, Tollefsen DM:
Heparin cofactor I1 Oslo: Mutation of Arg-189 to His decreases the
affinity for dermatan sulfate. J Biol Chem 2645128, 1989
37. Lane DA, Caso R: Antithrombin: Structure, genomic organisation, function and inherited deficiency, in Tuddenham EGD ( 4 ) :
Baillibre's Clinical Haematology, vol 2. London, UK, Bailliere
Tindall, 1989, p 961
38. Vehar GA, Keyt B, Eaton D, Rodriguez H, OBrien DP,
Rotblat F, Oppermann H, Keck R, Wood WI, Harkins RN,
Tuddenham EGD, Lawn RM, Capon DJ: Structure of human factor
VIII. Nature 312337,1984
-
-
From www.bloodjournal.org by guest on June 14, 2017. For personal use only.
1990 76: 2242-2248
The molecular genetic analysis of hemophilia A: a directed search
strategy for the detection of point mutations in the human factor VIII
gene
JK Pattinson, DS Millar, JH McVey, CB Grundy, K Wieland, RS Mibashan, U Martinowitz, K
Tan-Un, M Vidaud and M Goossens
Updated information and services can be found at:
http://www.bloodjournal.org/content/76/11/2242.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.