Download Factor IXHollywood

From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
Factor IXHollywood:
Substitution of ProJ5by Ala in the First Epidermal Growth
Factor-like Domain
By S.G. Spitzer, M.N. Kuppuswamy, R. Saini, C.K. Kasper, J.J. Birktoft, and S.P. Bajaj
Factor IX is a multidomain protein essential for hemostasis.
We describe a mutation in a patient affecting the first
epidermal growth factor (EGFI-like domain of the protein.
All exons and the promoter region of the gene were
amplified by the polymerase chain reaction method, and
sequenced. Only a single mutation (C
G), that predicts
the substitution of Promby Ala in the first EGF domain was
found in the patient’s gene. This mutation leads to new
restriction sites for four enzymes. One new site (Nsrl)was
tested in the amplified exon IV fragment and was shown to
provide a rapid and reliable marker for carrier detection
and prenatal diagnosis in the affected family. The factor IX
protein, termed factor
,,,,X
,I
(X
I ),
was isolated to
homogeneity from the patient’s plasma. As compared with
normal factor IX (IX,), X
I,,
contained the same amount of
7-carboxy glutamic acid but twice the amount of &OH
aspartic acid. Both X
I,
and IX, contained no detectable
free -SH groups. Further, X
I,,
could be readily cleaved to
yield a factor IX,-like molecule by factor Xla/Caz+. How-
ever, X
I a,
(comparedwith IXa,) activated factor X approximately twofold slower in the presence of Ca2+and phospholipid (PL), and 8- t o 12-fold slower in the presence of
Ca2+, PL, and factor Vllla. Additionally, X
I a,
had only
approximately 10%of the activity of IXa, in an aPTT assay.
In agreement with the nuclear magnetic resonance-derived
structure of EGF, the Chou-Fasman algorithm strongly
predicted a B turn involving residues Asn-Pro=-Cys-Leu in
IX,. Replacementof Pros6by Ala gave a fourfold decrease in
the B turn probability for this peptide, suggesting a
change(s1in the secondary structure in the EGF domain of
IX,.,.
Since this domain of IX, is thought t o have one
high-affinity Ca2+ binding site and may be involved in PL
and/or factor Vllla binding, the localized secondary structural changes in X
I,,
could lead t o distortion of the binding
sitels) for the cofactor(s1 and, thus, a dysfunctional molecule.
0 1990 by The American Society of Hematology.
F
nal catalytic domain.’-3The activation of factor IX (either by
factor VIIa/tissue factor/Ca’+ or by factor XIa/Ca’+)
involves the cleavage of two peptide bonds (Arg145-Ala146
and Argl80-Val18 1) to yield two-chain disulfide-linked
factor IXa and a 35-residue activation peptide.’ Factor IXa
is a serine protease, which converts factor X to factor Xa in a
reaction requiring Ca2+,phospholipid (PL) membrane surface, and factor VIIIa.3 The two aminoterminal domains of
factor IX, namely, the Gla domain and the first E G F domain
(containing the @-OHAsp), mediate Ca’+ and P L binding to
the protein.’z3v4The first E G F domain and portions of the
catalytic domain of factor IXa may be involved in binding to
factor V I I I ~ . ~ . ~
Studies of naturally occurring variants of factor IX
molecules present in hemophilia-B patients have contributed
substantially to our understanding of the molecular nature of
the d i ~ e a s eWe
. ~ report a hemophilia-B variant (Pro55to Ala)
in which the first E G F domain is affected. At the DNA level,
we provide a rapid and reliable method for screening the
pedigree for presence of the mutant allele. At the protein
level, we provide evidence that this mutation may partially
impair the interaction of factor IXa with its cofactors,
namely, Ca’+, PL, and factor VIIIa. We have designated the
variant protein factor IXHollyvood
(IXHw)after the birthplace
of the propositus. A preliminary account of this work has
been presented.’
-
ACTOR IX participates in the middle phase of the
intrinsic as well as the extrinsic clotting cascade.’
Absence or reduced activity of factor IX causes an X-linked
bleeding disorder commonly known as hemophilia-B. It is a
multidomain glycoprotein and is synthesized in the liver as a
precursor molecule of 461 amino acids.’ The aminoterminal
28 amino acids constitute the signal peptide domain and the
next 18 amino acids constitute the propeptide domain.’These
two domains are removed before secretion of the molecule.
Also, during biosynthesis, the first 12 glutamic acid residues
in the protein are carboxylated at the y position and the
aspartic acid residue a t position 64 is partially hydroxylated
a t the @ position.’ Circulating factor IX (415 residues)
consists of an aminoterminal y-carboxyglutamic acid (Gla)
domain, two epidermal growth factor (EGF)-like domains, a
connecting activation peptide domain, and a carboxytermiFrom the Departments of Medicine, Biochemistry, and Pathology, St Louis University School of Medicine, St Louis, MO; the
Orthopaedic Hospital, University of Southern California, Los
Angeles. CA; and the Department of Biochemistry and Molecular
Biophysics, Washington University School of Medicine, St Louis,
MO.
Submitted February 9, 1990; accepted June 11,1990.
Supported by National Institutes of Health grants no. HL36365
and 30572 (Project 5) to S.P.B., and the Lucille P. Markey
Charitable Trust supporting The Markey Center for Research in
Molecular Biology of Human Disease at Washington University to
J.J.B. S.G.S. was supported by a training grant (HM7050). S.P.B.
is a recipient of an AHA-Bristol-Myers Thrombosis Grant.
Address reprint requests to S.P. Bajaj, PhD. Hematology Division, St Louis University Hospital, 3635 Vista Ave at Grand Blvd,
PO Box 15250. St Louis. MO 631lO-0250.
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.
0006-4971/90/7608-0007$3.00/0
1530
MATERIALS AND METHODS
Patient data. Factor IX antigen in the patient’s plasma was 60%
and factor IX activity was 7%.* The affected male members in this
pedigree have mild hemophilia-B and manifest abnormal bleeding
after trauma.
Polymerase chain reaction amplification and DNA sequencing.
The set of primers used for polymerase chain reaction (PCR)
amplification of the putative promoter region and each exon of the
factor IX gene have been provided earlier from this laboratory?.”
Genomic DNA was isolated from leukocytes obtained from the
patient and from a normal adult male as outlined by Kan et al.”
Blood, Vol 76, No 8 (October 15). 1990: pp 1530-1537
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
EFFECT OF
~
~
0
1531
TO
5 5ALA CHANGE IN FACTOR IX
Target sequences in the genomic DNA were amplified essentially
according to Saiki et a1.l’ Our procedure has been described in detail
in an earlier publication? Following amplification, the DNA was
purified by ultrafiltration using a Centricon-30 microconcentrator
(Amicon, Beverly, MA), and an aliquot was analyzed on 1.5%
agarose gel in Tris-acetate-ethylenediamine tetraacetate (EDTA)
buffer’’ to ascertain the formation of the desired PCR product. The
remaining DNA sample was digested with the appropriate restriction endonucleases and cloned into M 13 mpl8 or mp19 vectors for
sequencing by the dideoxy chain termination method.I4 Sequencing
was performed using the Sequenase enzyme (USB Biochemical,
Cleveland, OH). In addition to the 17-mer universal primer, 20-mer
primers specific for factor IX gene were synthesized and used to
sequence several of the exons.
Proteins. Human factors IX and X were purified as outlined
earlier.I5 Factor IX,, was purified by the same technique as
described for factor IXNormal
(IX,). Human factor XI was purified by
the method of Kurachi and DavieI6and factor XIa was prepared as
described previo~s1y.I~
Protein concentrations were determined spectrophotometrically using Eo:; of 13.4 for factor XI, 11.6 for factor X,
and 13.2 for factor IX,.15-17A 13.2 value of
was also used for
factor IX,,, since its amino acid composition is the same as that of
factor IX, with one exception, ie, Pro to Ala change (see Results).
Factor VIII/von Willebrand factor preparation was made by gel
filtration of Cutter Koate (Cutter Biologics, Berkley, CA) as
described earlier: It was activated with thrombin before use!
’H-sialyl factor X was prepared as described earlier! It retained
92% of the biologic activity of nonlabeled control. Use of blood from
volunteer donors was approved by the human subjects committee of
St Louis University and of the University of Southern California.
5’ DNA terminus labeling. The terminal 5’ phosphate in the
sample DNA was exchanged with radiolabeled y-phosphate from
[y-32P]adenosine triphosphate by the use of T4 polynucleotide
kinase. The protocol used was that of Maniatis et ai.”
Assays of factors ZX and ZXa. Factor IX activity was measured
by a one-stage assay in which 50 pL of factor IX-deficient plasma, 50
pL of automated aPTT reagent, and 50 pL of test sample were
incubated for 5 minutes at 37°C. Then, 50 pL of prewarmed (37OC)
35 mmol/L CaCl, was added and the clotting time was noted.
Citrated pooled human plasma from 20 healthy donors was used as a
standard (defined as containing 1 U of factor IX activity/mL).
Factor IXa activity was also measured in the same test system.
Electrophoresis. Sodium dodecyl sulfate (SDS) gel electrophoresis was performed according to the method of Weber and Osborn.I8
The protein standards used to determine apparent molecular weights
have been des~ribed.’~
Gla and @-OHAsp analysis. Gla and @-OHAsp content of IX,
and IX,, was determined as outlined by Przysiecki et
Determination of free sulfhydryl groups. The -SH group determination was performed essentially as outlined by Habeeb.” The
buffer used was 0.05 mol/L Tris, 0.15 mol/L NaCl, pH 7.5
(Tris/NaCl), containing 20 mmol/L EDTA. Reagent 5,s’dithiobis(2-nitrobenzoic acid) (DTNB) was obtained from Sigma
Chemical Company (St Louis, MO) and the reaction was performed
both in the absence (native protein) and presence (denatured
protein) of 6 mol/L guanidinium chloride. The reaction was initiated
by the addition of 50 pL of protein (to yield a final concentration of 4
to 5 pmol/L) to 0.7 mL of buffer (+ guanidinium chloride)
containing 100 pmol/L DTNB. Absorbance at 412 nmol/L was
recorded against a blank containing 50 pL of buffer (instead of
protein) for 2 hours.
Activation of factor X by factors ZXa, and ZXa,,
Rates of
activation of ’H-factor X were measured by the activation peptide
release assay of Silverberg et a1,” with minor modifications as
outlined earlier for factor IX.I7
PL preparation. PL used in factor X activation studies was
obtained from Sigma Chemical Company; it was extracted from
rabbit brain by the procedure of Bell and Alton.” Based on
thin-layer chromatography on silica gel and phosphorus content of
individual phospholipid classes, the molar percentage of various
lipids was as followsz’: lysolecithins, 0.6%; phosphatidylcholines,
28%; phosphatidylethanolamines, 29%; phosphatidylserines, 9%;
sphingomyelins, 10%; phosphatidyl inositol, 0.8%; phosphatidic acid,
18%; and unknown and neutral lipids, 4.6%. This PL preparation
was used in the present study since it is representative of the
phosphatidylserine content of many membranes, including that of
platelet^.'^
Sequence alignments and secondary structure predictions. The
sequence alignments and the secondary structure predictions were
performed using the UWSCG Sequence Analysis Software Package
developed at the University of Wisconsin Biology Technology
Center, Madison, WLZ5We used PEPPLOT Program Version 5.3
released in July 1988. This program principally uses the ChouFasman algorithm in predicting the secondary structures of proteins.
RESULTS
Identification of the mutation in factor ZX,,
Since
purified factor IX,, has an apparent molecular weight
similar to that of factor IX, (Table 1 and Fig l), this rules
out any gross deletion or rearrangement of the factor IX,,
gene. Therefore, to identify the mutation resulting in hemophilia-B in this patient, we enzymatically amplified and
sequenced all eight exons and 40 to 60 bases flanking each
coding region, as well as the putative promoter region. The
nucleotide sequence revealed a single point mutation (C G)
in exon IV at position 10,415 (numbering according to
Yoshitake et al’). This nucleotide substitution was confirmed
by sequencing two different clones, each pair coming from
two independent PCR reactions, as well as by sequencing the
opposite strand. The G for C base replacement noted in the
patient’s allele changes a CCA codon to a GCA codon and
predicts the substitution of Pro by Ala at position 55 in the
mature protein (Fig 2). Since this was the only mutation
found in the exon sequences of the patients gene, we conclude
that the ProSSto Ala change results in a dysfunctional
molecule.
-
Table 1. Comparison of Factors IX. and IX,,
Components
Apparent molecular
weight’
Factor IX activity$
Gla$
@-OHASP$
Free -SH groups5
Ix,
* 2.000
61,000
225 f
10.35 f
0.33 k
0.035
(0.03 f
*
15 Ufmg
0.52
0.01
0.015
0.02)
IX”,,
61,000 f 2,000
22 2 Ufmg
9.90 0.1 1
0.63 0.01
0.030 0.010
(0.02 f 0.01)
*
*
*
‘Determined by SDS gel electrophoresis. Average of three determinations.
t A s measured in an aPlT system (see Materials and Methods).
$Average of two determinations.
§Free -SH groups obtained for a control protein, bovine serum albumin
were 0.66 f 0.02 (expected values were approximately 0.7.”).
Numbers in parentheses represent the data obtained in the presence of 6
mol/L guanidinium chloride. The data represent an average of two
determinations.
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
SPlTZER ET Ai
1532
A
6
1 2 34
1 2 3 4
Fig 1. SDS g d dmrophorotk ano)yris of M o r s 1% and 1%
W o r e (AI and a f t r (BI activation with factor Xla and calcium. Ten
micrograms of protein were applied t o each lane, and the anode is
at the bottom of the gels. ( A ) Lane 1 is unreduced and lane 2 is
reduced factor IX,;
similarly, lane 3 is unreduced and lane 4 is
reduced factor X
I .,
(B) Lanes 1 and 2 are factor IX, activated whh
factor Xla/Ca” for 16 and 30 minutes. respectively. Similarly.
lanes 3 and 4 are factor IX,
activated with factor XlalCn’ for 16
and 30 minutes, respectively. The activation was podormod at
37 C in TrisINaCI. pH 7.6, buffer and the reaction mixture
at 200 NglmL. factor Xla at 6
contained factor IX, (or fsctor tX,,.)
pgImL, and Ca” at 6 “ o l / L . IX. naive factor IX, is comprised of
residues 1415: Mu. the heavy chain of factor IX activation
intermediate. is comprised of amino acid residues 148416; HB. the
heavy chain of factor 1x0. is comprised d residuer 1814 1 6: L, the
light chain of factor 1x0. is comprised d residues 1-146 of factor
I X . Activation peptide, w h k h is comprised of residues 148-160.
stains poorly end was not observed in these gels. Dye front is
marked by tho word “dyo” on the bottom right of panel B.
The mutation identified above is located in the first
EGF-like domain of factor IX. This mutation creates recognition sites for four restriction enzymes ( N s i l . Avull. Nsp75241, and NspHI) as determined by the IRI Pustell Sequence Analysis Program. version 2.02 (International
Riotcchnologia. Inc. Uew Haven. CT). The sequence 5’ATCCATGT in the normal gene surrounding the nucelotide
10.415 is changed to ATGCATGT in the Hollywood gene.
The recognition sequence ATGCAT for Nsil and Avull and
A or G. and
the recognition squence RCATGY ( R
Y = C or T) for h‘.rp-75241 and NspHI are present in the
Hollywood gene but not in the normal gene. We tated the
possibility of using the commonly available restriction cn7yme ( N s i l . Rethesda Research Laboratories. Rethesda,
MD) for carrier detection and prenatal diagnosis of hemophilia-R in this family. The amplified exon I V fragments
(10.328 to 10.564 base pairs). each from a normal control
subject. the paticnt. and his mother. were digested with the
A’siI enzyme. labeled with “P at the 5’ end. and analy7ed by
polyacrylamide gel electrophoresis. The results of the autoradiograph obtained are shown in Fig 3. As expected. normal
exon IV fragment was not digested with this enzyme.
whereas patient exon IV fragmcnt was digested into two
smaller fragments of predicted sizes. The DNA from the
mother (a carrier) of the patient revealed three bands, one
corresponding to normal DVA and two corresponding to the
mutated DNA. Thus, it appears that the Nsil restriction site
present in the mutant allele can be used as a marker for
carrier detection and antenatal diagnosis in pedigrees with
this mutation.
Churucferizof ion ofjucfor /X,,” protein. Approximately
5 mg of factor IX,,, protein was isolated from 5 L of the
patient’s plasma. The results of SDS gel electrophoretic
analysis of the purified factor IX,,, are shown in Fig I A . For
comparison, SIX gels of factor IX, arc also depicted in this
figure. Each protein was etfectively homogenous and gave an
apparent molecular weight (IMr) of 61.OOO z 2.000 by this
mcthod. However, IX,,, had only one tenth the specific
clotting activity of IX, as measured in an aPTT assay system
(Table 1). All results were obtained with the preparation of
factor IX,,, shown in gels 3 and 4 of Fig !A.
The Gla residues pcr mole of IX,,, were essentially the
same as for factor IX, (Table I). Factor lXllW, however.
di!Tered from factor IX, in its &OH Asp content. &OH Asp
content was 0.33 mol/mol for factor IX, and 0.63 mol/mol
for factor IX,,, (Table I ) . Since the EGF domain has three
-S-S- bonds and since replacement of Pro-55 by Ala is
predicted to alter the secondary structure (see Discussion),
bond(s) formation may be impaired
we reasoned that -S-Sin IX,,,. However. as for IX,. no free -SH groups were
detected in IX,,,. Moreover. we conjecture that disulfide
bonds in factor IX,,,. in all probability. are correctly paired.
Factor 1XHw
Factor IX
GATC
G A T C
-
-A
-c
‘G’
Factor IX
54
Asn
AAT
Factor IXHw AAT
Asn
55
PrO
CCA
GCA
Ala
Fig 2. MmHwrtion of th. mutation k foetor 1%.
Tho
ooquonco of the coding otrand s w w n d i n g the nuclootido 10,416
of exon IV of tho HW gene is shown on tho I& and that of tho
normal 0.m is shown on tho right. The bar0 change (G for C ) is
highlighted by esterisks. This change (C
6 ) in the first nucleotide of tho codon for amino add 66 cauu. Pro in normal foctor IX
t o be subathut& by Ala in factor lXw. aa shown at the bottom of
the figure.
-
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
EFFECT OF PRO8* TO
ALA
1533
CHANGE IN FACTOR IX
2
Fig 3. N.ll rntrktion digost pottorn of tho
emplitiod oxon IV fragmma (10328 to 10,664
base pairs) from dH1orent individunls of factor IX,
pedigree. The fragments were radiolabeled et the
5' end with "P and electrophoresed on 4% acrylamide gel. The eutoradiograph shown was dweloped aftw a 2-hour exposure M room 1omp.rature
with one intensifying acreen. Lens 1. DNA markora: lane 2. control DNA from a normal subject:
lane 3, pationt DNA: and lano 4, DNA from tho
mother. who is a carrier: bp, b
au v i r . The number
of b e u pairs in the fragments in knea 2 to 4
includa the linkora containing tho r m r i c t b n sitos
used in PCR amptihwtion. The uquence of tho
primors u s d tor amplification is given in an omrlier
publication.'"
3
4
bP
396,
344298-
-
-248
254134d
-155
220,
201
4-93
75
Activation o f factor IX,,
by factor Xla/Ca*' was also
studied. A t saturating concentrations o f Ca" (5 mmol/L).
factor IX,,
was readily activated to a factor IXa-like
molecule as revealed by the SDS gel electrophoretic data
(Fig 1 R). Next. the ability o f factor IXa,,, to activate factor
X in the presence and absence of P L and factor V l l l a was
examined. These data are prcsented i n Table 2. I n the
prcsence o f Ca". but in the absence of P L and factor Vllla.
the ratcs of factor X activation by factor IXa, and factor
IXa,,, were extremely slow. and a reliable comparison of the
r a t a obtained under these conditions by the normal and
mutated en7ymcs could not be made. Inclusion of P L in the
reaction mixture gave augmented ratcs of factor X activation. Under the conditions of our experiments and at two P L
concentrations. factor IXa,,, activated factor X at ratcs that
were approximately twofold slower than those obtained with
factor IXa,. Further. in a complete intrinsic factor Xase
system (Ixa. Ca?.. PL. Vitta). factor IXa,,,
activated
factor X at r a t a that were 8- to 12-fold slower than those
obtained with factor IXa, under similar conditions (Table
2). These data suggcst that factor IX,,, fails to fully function
Tablo 2. Comporhan of Rot-
in hemostasis primarily bec3use the activated molecule is
partially impaired in i t s interaction with the cofactors.
Theability of factor IXa,,, tocorrect the partial thromboplastin time of factor IX-deficient plasma was also examined.
These data arc praented in Fig 4. I t is evident from these
i s only one tenth as active in
data that factor IXa,,,
achieving the same degree o f hemostatic correction as factor
IXa, in this system. Moreover. since factor IX,,
also has
only one tenth the specific activity of factor IX, in the aPTT
assay (Table 1). the data of Fig 4 strengthen our conclusion
that the primary defect in factor IX,,
i s not i n i t s ability to
be converted to a factor IXa-like molecule. but i n i t s ability
to activate factor X under physiologic conditions once i t
converted to the activated molecule.
DISCUSSION
-
I n the praent study. we have identified a mutation
(C G ) i n the factor I X gene at the nucleotide position
10.41 5.' which predicts a replacement o f Pro'' by Ala in the
mature protein and rcsults in hemophilia-R. The Nsil rcstriction site generated by this mutation should provide a reliable
of F.ctor X Activotkwt by Foetan IXm,, and tX.w,,
Rned F a t - X.
bv
*
0.075 2 0.025 pg/mL/h'
0.42 pg/mL/h a 30 pmolA PL
0.45 pglmL/h at 60 rmolA PL
0.055 0.015 pe/mVh'
0.22 pg/mL/h a 30 rmolA PL
0.19 p/mLlh a 60 pmdA PL
0.1 1 pgImLlmin at
30 nglmL IXa
0.25 pg/mL/min a
60 WmL IXa
0.014 pg/mL/min at 30 n g / d IXa
0.02 1 pg/mUmin at
60 na/mL 1%
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
SPITZER ET AL
1534
IXaHW (ng/ml) (0-0)
n
10
20
30
100
50
200
400
0
8
W
E
i=
c,
:
50
40
30
0
zs
20
1
I
I
2
3
I
,
5
I
,
,
,
10
I
Fig 4. A comparison of the ability of factors
IXa, and IXa, to shorten the partial thromboplastin time of factor IX-deficient plasma (see Methodd. The logarithm of clotting time is plotted
20
lXaN (ng/ml) ( 0 - 0 )
marker for presence of the mutant allele (Fig 3). Because two
other families (ref. 26 and Chen S-H, Thompson AR,
unpublished results) have been identified and more may be
found carrying this mutation, the NsiI restriction site should
facilitate carrier detection and prenatal diagnosis in these
pedigrees. The restriction site analysis is preferable over the
factor IX activity to antigen ratios and restriction fragment
length polymorphism (RFLP) analysis since factor IX activity to antigen ratios are often misleading for the detection of
and since RFLP analysis in approximately one
third of the families is not informative.28The new restriction
site in the factor IX,, gene in combination with the PCR
amplification technique provides a fast and reliable screening
alternative for the presence of mutant allele.
The mutation identified in factor IX,, is in the first
EGF-like domain of human factor IX. The other naturally
occurring variants of factor IX mutated in this domain,
which result in hemophilia-B, are A~p47Gly,’~
GlnS0Pr0,~~
Gl~60Ser,~’
and @-OHA~p64Gly.’~
All known mutations in
the EGF domain of factor IX result in a mild or moderate
form of hemophilia-B, except for the GlnS0Pr0~~
mutation,
which results in a severe bleeding diathesis. Factor IXAlabama
(Asp47Gly) and factor IX,, (ProSSAla) do not appear to be
impaired in their conversion to the activated forms. Factor
IXNw Landon (GlnSOPro), however, appeared to be somewhat
impaired in its conversion to the activated form,32while the
data for factor IXDurham
(Gly60Ser) and factor IXLondan
:6 are
not yet available. Moreover, several recombinant factor IX
molecules mutated in the EGF domain were also found not to
be impaired in their conversion to factor IXa by factor
XIa/Ca’+.’ Based on the above data and the observation that
the EGF domain mutated forms of factor IXa in the presence
of Ca’+, PL, and factor VIIIa activate factor X at a
markedly slower rate (10% to 15% of that obtained with
factor IXa,), Rees et al’ have postulated that binding of
Ca’+ to this domain stabilizes the conformation of factor IXa
suitable for PL and factor VIIIa binding. This hypothesis is
consistent with our kinetic data (Table 2) and with the
preliminary observations of Naworth et
who suggested
with the endothean abnormal interaction of factor IXAlabama
lial cell surface. However, it is not yet clear whether the first
EGF domain of factor IX/IXa is directly involved in binding
assay. Note that the concentration axes for IXa,
and IXa,
differ by a factor of 10.
to PL34and factor VIIIa” or whether the binding of Ca2+to
this domain (or other parts of the molecule) induces a
coni6rmational change that allows the heavy chain (catalytic
domain) of factor IXa to bind to factor VIIIa.6 Moreover, it
has been shown that factor IXAlabama
binds to a synthetic PL
preparation containing 30% phosphatidylserine (in contrast
to approximately 5% to 10%present in biologic membranes)
with an affinity indistinguishable from that obtained for
factor IX,.36 Thus, we believe that detailed direct binding
studies of Ca2+,biologic PL membranes, and factor VIIIa to
the isolated wild type and the variously mutated forms of the
EGF domain are required to fully understand the function of
this domain in factor IX. Nonetheless, it is apparent from the
mutations observed thus far in this domain of factor IX that
the failure of these molecules to function in hemostasis does
not result primarily from their ability to be converted to
factor IXa-like molecules, but from the impaired ability of
the factor IXa formed to activate factor X under physiologic
conditions.
Another point of interest in the present study is the fact
that factor IX,, contains twice the amount of &OH Asp
compared with that present in IX, (Table 1). Whether it
results from the increased efficiency of hydroxylase in the
factor IX,, individual or from the structural alteration in
the protein that makes it a better substrate for the enzyme is
not known at present. We are currently expressing factor IX,
and plan to express factor IX,, in a mammalian expression
system. If the recombinant factor IX,, contains markedly
increased contents of &OH Asp compared with the recombinant factor IX,, it may mean that the spatial relationship
between Asp- and Tyr69(a major recognition site for the
8-hydroxylase enzyme3’) may be altered in the variant
protein so as to make it a better substrate for hydroxylation.
On the other hand, factor IX,, may not have the unusual
trisaccharide sugar chain linked to Sers3reported for factor
IX,, since a /3 turn at AsnS4(Table 3) may be necessary for
glyco~ylation.~~
We are currently collaborating with Dr S.
Iwanaga to answer this question.
The three-dimensional structure of factor IX remains
undetermined. However, based on the sequence homology of
segments of factor IX with other proteins of known structure,
a molecular framework of factor IX can be inferred. The
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
EFFECT OF
PRO^^ TO ALA CHANGE IN FACTOR IX
1535
strands are connected by one or more 8 turns. Based on this
inf~rmation,~~-~’
a consensus structure for the first EGF
domain in factor IX is depicted in Fig 5A and can be
summarized as follows: 8 strands include residues Asp-47 to
Gln-50, Ser-61 to Asp-64, Ser-68 to Trp72, Phe-77 to
Glu-78, and Glu-83 to Leu-84. There are two /3 turns
between Cys-5 1 and Ser-61, and a single /3 turn each between
Asp-64 and Ser-68, Trp-72 and Phe-77, and Glu-78 and
Glu-83. These assignments and, in particular, those for the /3
turns are further supported by a secondary structure prediction performed using the Chou-Fasman algorithm (Table 3)
as implemented in the UWSCG program pa~kage.~’
In the
Chou-Fasman analysis, the four most likely turns start at
Asn-54, Asn-58, Cys-73, and Gly-79, respectively, in agreement with the assignments based on comparison with the
NMR-derived structures of EGF39-41and TGF-a4’ This
model of the EGF domain of factor IX depicted in Fig 5A
provides a structural framework, albeit a hypothetical one,
for our discussion of the consequences of amino acid substitutions in several known EGF-domain mutants of hemophilia-B. Since, relative to the human EGF sequence, the
EGF domain of factor IX has two sets of deletions that are
located in the loop regions connecting the 8 strands: we
believe that the modeled structure shown in Fig 5A represents a reasonable framework for this domain in human
factor IX.
Most calcium-binding sites in proteins normally contain
several carboxyl acid side chains, ie, aspartate, glutamate, or
y-carboxyglutamate. In the first EGF domain of factor IX,
these acidic residues have been tentatively identified as the
Table 3. fl Turn Prediction Values for the Selected Regions of the
First EGF-Like Domain of Human Factor IX
Protein (Residues)
Seauenu,
Factor IX (47-50)
Factor IX,
(47-50)
Factor IX (49-52)
Factor IX,
LMdOn (49-52)
Factor IX (54-57)
Factor IX,, (54-57)
Factor IX (58-6 1)
Factor lXouham
(58-6 1)
Factor IX (65-68)
Factor IX (73-76)
Factor IX (79-82)
DGDQ
GGDQ
DQCE
DPCE
NPCL
NACL
NGGS
NGSS
DINS
CPFG
GKNC
B Tvn Probabili
2.20
1.52
1.10
3.30
3.97
1.oo
2.76
1.81
1.20.
4.43
2.87
Prediction values were obtained using the UWSCG sequence analysis
software package.25This program is particularlyuseful in predictingthe j3
turns in proteins. A value of approximately 2.5 is a strong indicator of j3
turn occurrence. The single-letter code for amino acids is used.
*Although this value predicts the absence of a j3 turn in this region, the
two adjacent @ stands in the antiparallel @ sheet are linked by residues
64-68 by chain reversal (for details see Fig 5A and Discussion).
amino acid sequence of the EGF domains of factor IX
displays considerable homology to those of human and
murine EGFs39-41
and transforming growth factor-a (TGFa).‘“ Tentative structures based on two-dimensional nuclear
magnetic resonance (NMR) studies have been proposed for
both of these protein^.^^-^' The structures of EGF and TGF-a
are rather similar,42and both can be described as 8-structure
proteins. Each protein is composed of five /3 strands assembled in an extended antiparallel &sheet structure and the p
A
c
Fa++)-
5
o
\
r
B
IXNew London
IX Alabama
8
o t
IXHoilywood
G9
4
Fig 5. Schematic representations of the secondary and tertiary structures of the first EGF-like domain of factor IX. The standard
single-letter code for amino acids is used throughout. (A) Schematic diagram of the topology of the first EGF-like domain of factor IX. This
The
modeledstructure is based on the description of the NMR-derived structures of human and murine EGFs and that of human TGF-a?arrows represent fl strands assembled in a five-stranded, antiparallel fl sheet. The loop connecting the first two 0 strands contains two fl
turns that are indicated with sharp hairpin bends. The location of this loop cannot be described with any precision, and it might be folded on
top or below the fl sheet. The three disulfide bridges and the amino acid, Tyr”, implicated in @-hydroxylationof Aspa?’ are depicted
systematically. Residues involved in the &strands and the loops connecting them are detailed in the Discussion. (B) Primary structure of
residues 47-69 of the first EGF-like domain of factor IX. The sequence is taken from Yoshitake et a1.2 and residues that are thought t o be
spatially adjacent are indicated by their proximity in the diagram. The two @ turns starting at residues 54 and 58 are shown as hairpin
bends. All five known, naturally occurring, dysfunctional variants of factor IX that result in hemophilia-B are depicted systematically. The
locations of the amino acid replacements for these hemophilia variants are also marked with the single-letter amino acid code in panel A.
The putative calcium-binding site and the &OH Asp are indicated with Cat+ and 8, respectively.
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
1536
SPITZER ET AL
side chains of Asp-47, Asp-49, and Asp-64.’ These residues
are assumed to lie on the same face of the @ sheet and are
juxtapositioned for participation in forming the high-affinity
Ca2+binding site (Fig 5). Since Asp-64 is partially hydroxylated, it is possible that either the y-carboxyl or the @-hydroxyl group, or both groups, provides the ligands necessary
for calcium binding.43Additional ligands may be provided by
other side chains, by main chain atoms, and by solvent
molecules.
In factor IX,,, Pro” is replaced with Ala, while in factor
IXDurham,
Gly6’ is replaced with Ser. In normal factor IX,
both of these residues are indicated to be part of the @ turns
(Fig 5B). The Chou-Fasman algorithm predicts that these
replacements will result in a reduction in the @-turnprobability at these two locations (Table 3). Therefore, some change
in backbone conformation is expected for these two mutants.
Moreover, the change in size of the amino acid side chain
moieties could also cause steric effects. Residues 55 and 60
are not directly adjacent to the Ca2+ binding site (Fig 5).
Thus, while the binding of Ca2+may not be affected in these
two mutants, the effect of these two replacements might be in
the interaction of factor IX/IXa with PL membranes and/or
factor VIIIa.
In factor IXN,, London, Gln” is replaced with Pro. This
residue is believed to be part of the @-sheetstructure (Fig 5)
and the introduction of Pro residue is likely to interrupt the
hydrogen bonding in the putative @ sheet. The introduction of
a Pro residue also restricts the available conformational
space for the polypeptide main chain at residues 49 and 50.
The Chou-Fasman prediction suggests a significant increase
in the @-turnprobability for a turn starting at Asp49,again
suggesting a change in the main chain conformation (Table
3). Recalling that Asp49is part of the calcium binding site, it
is to be expected that the introduction of Pro at position 50 is
likely to disrupt the binding of CaZ+to factor IX,,, London. In
the case of factor IXAlabama,
the Asp4’ to Gly replacement does
not lead to a predicted change in secondary structure
potential (Table 3). In this case and in the case of factor
IXLondon
(@-OHAsp64Gly), the most likely explanation for
hemophilia-B is the loss of one of the putative Ca2+binding
ligands. As a consequence of loss of Ca2+binding in IX,,,
London, IXLondon
6r and IXAlabama,
the interaction of other cofactors (PL and VIIIa) to these mutants may also be impaired.
As mentioned earlier, further studies that examine the direct
interaction of the isolated normal and mutated forms of the
EGF domain with Ca2+,PL, and factor VIIIa are needed to
fully support these projected conclusions.
ACKNOWLEDGMENT
We thank Drs Dean Welsch and Paul Friedman of Merck, Sharp
and Dohme Research Laboratories for performing the Gla and
&OH Asp analysis. We thank Beth Haase for her assistance in
preparation of the manuscript.
REFERENCES
1. Davie EW: Introduction to the blood coagulation cascade and
cloning of blood coagulation factors. J Protein Chem 5:247, 1986
2. Yoshitake S, Schach BG, Foster DC, Davie EW, Kurachi K:
Nucleotide sequence of the gene for human factor IX (antihemophilic factor B). Biochemistry 24:3736, 1985
3. Furie B, Furie BC: The molecular basis of blood coagulation.
Cell 53:505,1988
4. Morita T, Isaacs BS, Esmon CT, Johnson AE: Derivatives of
blood coagulation factor IX contain a high affinity calcium binding
site that lacks y-carboxyglutamic acid. J Biol Chem 2595698, 1984
5. Rees DJG, Jones IM, Handford PA, Walter SJ, Esnouf MP,
Smith KG, Brownlee GG: The role of B-hydroxyaspartate and
adjacent carboxylate residues in the first EGF domain of human
factor IX. EMBO J 7:2053,1988
6. Bajaj SP, Rapaport SI, Maki S L A monoclonal antibody to
factor IX that inhibits the factor VIIICa potentiation of factor X
activation. J Biol Chem 260:11574, 1985
7. Spitzer SG, Katzman D, Kasper C, Bajaj SP Factor IX
Hollywood: Substitution of Pro 55 to Ala in the first EGF domain.
Thromb Haemost 62:203, 1989 (abstr 617)
8. Kasper CK, asterud B, Minami JY, Shonick W, Rapaport SI:
Hemophilia-B: Characterization of genetic variants and detection of
carriers. Blood 50:351, 1977
9. Spitzer SG, Pendurthi UR, Kasper CK, Bajaj SP Molecular
J Biol Chem 263:10545,1988
defect in factor IXs, Lnke
10. Spitzer SG, Warn-Cramer BJ, Kasper CK, Bajaj SP Replacement of 397-Ile by Thr in the clotting protease factor IXa (Los
Angeles and Long Beach variants) affects macromolecular catalysis
but not L-tosyl arginine methyl ester hydrolysis. Biochem J 265:219,
1990
11. Kan YW, Dozy AM, Trecartin R, Todd D: Identification of a
nondeletion defect in a-thalassemia. N Engl J Med 297:1081, 1977
12. Saiki RK, Gelfand DH, Stoffel H, Scharf SJ, Higuchi R,
Horn GT, Mullis KB, Erlich HA: Primer directed enzymatic
amplification of DNA with a thermostable DNA polymerase.
Science 239:487,1988
13. Maniatis T, Fritsch EF, Sambrook J: Molecular Cloning. A
Laboratory Manual. Cold Spring Harbor, NY, Cold Spring Harbor
Laboratory, 1982, p 170
14. Sanger F, Nicklen S, Coulson AR: DNA sequencing with
chain termination inhibitors. Proc Natl Acad Sci USA 745463,
1977
15. Bajaj SP, Rapaport SI, Prodanos C: A simplified procedure
for purification of human prothrombin, factor IX and factor X. Prep
Biochem 11:397,1981
16. Kurachi K, Davie EW: Activation of human factor XI
(plasma thromboplastin antecedent) by factor XIIa (activated
hageman factor). Biochemistry 165831,1977
17. Bajaj S P Cooperative CaZ+binding to human factor IX. J
Biol Chem 257:4127,1982
18. Weber K, Osborn M: The reliability of molecular weight
determinations by dodecyl sulfate polyacrylamide gel electrophoresis. J Biol Chem 244:4406, 1969
19. Przysiecki CT, Staggers JE, Ramjit HG, Musson DG, Stern
AM, Bennett CD, Friedman PA: Occurrence of @-hydroxylated
aspargine residues in non-vitamin K-dependent proteins containing
epidermal growth factor-like domains. Proc Natl Acad Sci USA
84:7856, 1987
20. Habeeb AFSA Reaction of protein sulfhydryl groups with
Ellman’s reagent. Methods Enzymol25:457, 1972
21. Silverberg SA, Nemerson Y, Zur M: Kinetics of the activation of bovine coagulation factor X by components of the extrinsic
pathway. J Biol Chem 252:8481, 1977
22. Bell WN, Alton HG: A brain extract as a substitute for
platelet suspensions in the thromboplastin generation test. Nature
174:880,1954
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
EFFECT OF
~ ~ 0 TO
5 5ALA
CHANGE IN FACTOR IX
23. Bajaj SP Activation of prothrombin by purified factors
(factor Xa, calcium, phospholipid and factor V). Doctoral thesis,
University of Minnesota, Minneapolis, MN, 1974
24. Marcus AJ, Ullman HL, Safier LB: Lipid composition of
subcellular particles of human blood platelets. J Lipid Res 10108,
1969
25. Deverenx J, Haeberli P, Smithies 0: A comprehensive set of
sequence analysis programs for the VAX. Nucleic Acids Res 12:387,
1984
26. Green PM, Bentley DR, Mibashan RS, Nilsson IM, Gianelli
F Molecular pathology of hemophilia B. EMBO J 8:1067,1989
27. Graham JB, Barrow ES, Elston RC: Lyonization in hemophilia: A cause of error in direct detection of heterozygous carriers.
Ann NY Acad Sci 240:141,1975
28. Winship PR, Anson DS, Rizza CR, Brownlee GG: Carrier
detection in hemophilia B using two further intragenic restriction
fragment length polymorphisms. Nucleic Acids Res 12:8861, 1984
29. Davis LM, McGraw RA, Ware JL, Roberts HR, Stafford
DW: Factor IXAlabama:
A point mutation in a clotting protein results
in hemophilia B. Blood 69:140, 1987
30. Lozier JN, Stanfield-Oaklay SA, High KA: Factor IX,,,
London: A point mutation causing hemophilia B. Thromb Haemost
62:162, 1989 (abstr 483)
31. Denton PH, Fowlkes DM, Lord ST, Reisner HM: Hemophilia B Durham: A mutation in the first EGF-like domain of factor
IX that is characterized by polymerase chain reaction. Blood
721407,1988
32. Lozier JN, Monroe DM, Smith KJ, Lin SW, Roberts HR,
High KA: Structural studies of factor IX,,, London. Blood 74:252a,
1989 (suppl 1) (abstr 949)
33. Naworth PP, Wilner GD, Stern DM: The EGF domain of the
factor IX molecule is involved in factor IX endothelial interaction.
Circ Res 74:11-232, 1986 (abstr 929)
34. Derian CK, VanDusen W, Przysiecki CT, Walsh PN, Berkner
KL, Kaufman RJ, Friedman PA: Inhibitors of 2-ketoglutaratedependent dioxygenases block aspartyl @-hydroxylationof recombinant human factor IX in several mammalian expression systems. J
Biol Chem 264:6615,1989
1537
35. Lin S-W, Smith KJ, Welsch D, Stafford DW: Expression and
characterization of human factor IX-factor X chimeras in mouse
C127 cells. J Biol Chem 265:144, 1990
36. Jones ME, Griffith MJ, Monroe DM, Roberts HR, Lentz B R
Comparison of lipid binding and kinetic properties of normal,
variant, and y-carboxyglutamic acid modified human factor IX and
factor IXa. Biochemistry 248064, 1985
37. Stenflo J, Lundwell A, Dahlback B: &Hydroxyasparagine in
domains homologous to the epidermal growth factor precursor in
vitamin K-dependent protein S. Proc Natl Acad Sci USA 84:368,
1987
38. Nishimura H, Kawabata S, Kisiel W, Hase S, Ikenaka T,
Takao T, Shimonishi Y, Iwanaga S: Identification of a disaccharide
(Xyl-Glc) and a trisaccharadie (Xy1,-Glc) 0-glycosidically linked to
a serine residue in the first epidermal growth factor-like domain of
human factors VI1 and IX and protein Z and bovine protein Z. J Biol
Chem 26420320,1989
39. Cooke RM, Wilkinson AJ, Baron M, Pastore A, Tappin MJ,
Campbell ID, Gregory H, Sheard B The solution structure of
human epidermal growth factor. Nature 327:339, 1987
40. Makino K, Morimoto M, Nishi M, Sakamoto S, Tamura A,
Inooka H, Akasaka K: Proton nuclear magnetic resonance study on
the solution conformation of human epidermal growth factor. Proc
Natl Acad Sci USA 847841,1987
41. Mayo H, Cavalli RC, Peters AR, Boelens R, Kaptein R:
Sequence specific 'H-NMR assignments and peptide backbone
conformation in rat epidermal growth factor. Biochem J 257,:197,
1989
42. Kohda D, Shimada I, Miyake T, Fuwa T, Inagaki F
Polypeptide chain fold of human transforming growth factor analogous to those of mouse and human epidermal growth factors as
studied by two-dimensional 'H-NMR. Biochemistry 28:953, 1989
43. Morita T, Kisiel W: Calcium binding to a human factor IXa
derivative lacking y-carboxyglutamic acid: Evidence for two high
affinity sites that do not involve 8-hydroxyaspartic acid. Biochem
Biophys Res Commun 130:841, 1985
From www.bloodjournal.org by guest on June 12, 2017. For personal use only.
1990 76: 1530-1537
Factor IXHollywood: substitution of Pro55 by Ala in the first epidermal
growth factor-like domain
SG Spitzer, MN Kuppuswamy, R Saini, CK Kasper, JJ Birktoft and SP Bajaj
Updated information and services can be found at:
http://www.bloodjournal.org/content/76/8/1530.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.