0 1991 Federation of European Biochemical Societies 00145793/91/%3.50
285. number 2, 248-250
Further wickme of the role of the reactive centre loop in the inhibitory
of the serpins
D.9. Perryl, M. Dalyl, P.L. Harper’, R.C. TaiF, J. Price’, LB. WalkerZ and R.W. t2arrelll
’ DeparIment of Haematology, University of Cambridge, MRC Centre, Hills Road, Cambridge, CR.? 2QH, UK and f&partment
Haemarology, Glasgow Royal Infirmary. Castle Street, Glasgow. UK
Received 7 May 1991
Four unrelated individuals have been identified with an identical antithrombin
thromboses. In each case, the plasma antithrombin
decrease in the heparin-induced thrombin inhibition suggesting a mutation at or near the reactive centre of the molecule. Amplification
sequencing of exon 6 showed a G-*T mutation at nucleotide 1246, which corresponds
This is one of a series of conserved alanincs that form the stalk to the reactive ccntre loop. The observed changes in this variant are compatible
with recent structural studies that infer that mobility of this stalk with partial re-entry into the A-sheet of the molecule is necessary for optimal
variant, associated in one of them with episodes of recurrent venous
was normal and the only functional abnormality concentration was a minor but consistent
to a substitution of a serine for an alanine at residue 384.
Antithrombin; PCR; Active site variant
2. A4ATERIALS AND METHODS
to the Serpin superfamily.
number of serine proteases involved in coagulation,
cluding factor IIa (thrombin)
normal physiological conditions its inhibitory activity is
relatively slow, but in the presence of heparin this ac-
tivity is increased by IOOO-fold [ 11. The importance
amtithrombin in the regulation
is emphasised by the recurrent
dividuals with either a deficient or abnormal antithrom-
bin are prone to develop .
We have used the polymerase
 to amplify exon 6  of the antithrombin
four patients with an antithrombin
hibited heparin binding but defective inhibitory
ty. The identified mutation
proposal that inhibitory activity of the serpins is depen-
dent on a folding of the N-terminal
tive centre loop.
is a 58 kDa plasma glycoprotein
It is the major inhibitor of a
III (here referred to as an-
and factor Xa [l]. Under
of normal haemostasis
thromboses that in-
chain reaction (PCR)
variant which ex-
provides srlpport for the
region of the reac-
Correspondence address: R.W. Carrell,
University of Cambridge, MRC Centre, Hills Road, Cambridge, CB2
ZQH, UK. Fax: 0223-336827
Dept. of Haematology,
2. I . Materials
All reagents unless otherwist stated were obtained from the Sigma
Chemical Co., Poole, Dorset,
polynucleotide kinase were obtained
Biotechnology, Milton Keynes,
polymerase from Perkin-Elmer Cetus, Beaconsfield,
point agarose from BioRad Labs Ltd., Watford, 55:; i’!trz:~~Jre urea,
agarose and restriction enzymes from Gibco-BRL.
DNA sequencing used Sequenase ~2.0 obtained from United States
Biochemical Corporation (Cambridge
Chromogenic substrate Chromozym
inger Mannheim, Lewes. E. Sussex and S2222 from Kabi Vitrum. Ux-
UK; low melting
TH was obtained
Patient 1 was a 68-ye&r-old male who was presented at the age of
35 years with a deep vein thrombosis
gout and subsequently experienced two further spontaneoils
boses. There was no family history of thromboembolic
Patients 2, 3 and 4, were unrelated to the first patient. In each case
there was no history of thrombocmbolic
was detected as part of a screening programme
donors to establish the incidence of antithrombin
following an acute episode of
disease and the abnormality
of normal blood
2.3. Plasma antithrombin ussays
tithrombin progressive activity) or in the presence of heparin (heparin
cofactor activity) according to Abildgaard
substrate Chromozym TH. Anti-Xa activity was performed
to Odegard  in the presence of heparin (anti-Xa heparin cofactor)
using the substrate S2222. Crossed immunoelectrophoresis
antigen was measured by competitive
activity was measured
in the absence (an-
 using the chromogenic
Published by Efsevier S&wee Publishers B. V,.
Volume 28.5, number 2
formed according to Sas [X] with and without the addition of heparin
(25 U/ml) in the first dimension.
DNA was isolated from peripheral blood leucocytes as previously
Synthetic oligonucleotides S’AATGAACACAAGTGTTTGGT-
TTTTAT and 5’-AAGCATTGAGGAATTGCTGTGTCTGT
designed to selectively amplify a 249-base pair (bp) region of the AT
gene encompassing the entire coding region of exon 6 together with
some 5’-intronic and 3 ‘-untranslated
was gel-purified and directly sequenced using a “P-labelled
sequencing primer [lo].
and sequencing of exon 6 of the antilhrombin
sequence. The PCR product
presence of heparin was reduced to 76% in two cases
and to 30% and 59% in the remaining two cases. Cross-
ed immunoelectrophoresis showed an identical mobility
of the normal and abnormal
presence and absence of heparin.
DNA was isolated from peripheral blood leucocytes.
In the four patients studied, amplification
the AT gene generated a single fragment of 249 bp. Se-
quencing of the PCR product demonstrated
at nucleotide position 1246 representing
guanine (G) and the mutant thymidine (T) bases. The
sequences for codon 384 derived from the anti-sense
strand are, therefore, GCA coding for alanine and TCA
coding for serine.
This mutation is associated with the loss of a Pvctll
site, which allows for the rapid screening
members. Testing of related family members has iden-
tified the mutation in a further
assays (Table I) demonstrated
levels of antithrombin,
activity and normal heparin induced
activity in the
components both in the
of exon 6 of
We report here an antithrombin
II, found in four unrelated
heterozygotes, in only one
Ala 384 (Pl 0) Pro AT111 Cambridge 1
Ser AT111 Cambridge 2
la 382 (Pl?) ‘Thr AT111 Hamilton
ovalbumin , avertical view of that shown in Fig. 28, indicating the
position of the N-terminal stalk substitutionsat alanine 384 (PIO), ten
residues from the reactive center (PI) and at alanine 382 (PIZ).
I. Projection of antithrombin on the reactive centre of
of whom there was a history of recurrent venous throm-
boses. The variant 384 Ala to Ser has properties in com-
mon with similar variants having mutations
the active site (Fig. l), i.e. normal heparin binding but
a loss of thrombin inhibitory activity. However, a uni-
que feature of the Cambridge II variant (Table I) is that
it retains anti-Xa activity whereas others, including the
same site 384 Ala to Pro [IO,1 I] lose both IIa and Xa in-
Although these results may seem puzzling, they fit
well both with current proposals
basis of inhibition in the serpin family (reviewed by
Huber and Carrel1 ) and with the structural
for a mobile reactive centre recently presented by Stein
et al. . There is now good evidence that the reactive
centre of antithrombin, Arg-Ser 393-394, lies on an ex-
ternal peptide loop hinged on Glu-377 and extending to
Arg-393. It is believed that a critical contribution
loop is to hold the reactive centre in an available and
A proposed basis for the strain is the inability of the
loop, in the intact antithrombin,
to the adjacent A sheet, where it would form a more
cleavage of the loop . The resultant conformation
of the loop on incomplete m-entry into the sheet is likely
to be close to that of the canonical form found in other
at or near
as to the structural
to move fully back in-
Alanine 382 and 384 variants
Inhibitory Activity (070)
Variant Mutation Ila
Perry et al, 1988 
et nl., 1988 
“Data refers to an AT Hamilton
Charleville - identical variant [II].
variant identified locally.
Volume 285, number 2 Download full-text
FEW LETTERS July 1991
(b) ‘Antithrombin’ mo
PI Au 393
PM Ser 380
Fig. 2. Diagrammatic models of the A-sheet and reactive centre loop
of (a) ovalbumin  and(b) of antithrornbin modelled to show how
partial re-entry of strand 4 into the A sheet would give a loop structure
approximating to the optimal inhibitory conformation.
the loop to give strand 4 (s4a) is incomplete and can be modelled
beyond this point for perhaps one or two residues emphasising the
need for flexibility at PI0 and PI2 (382 and 384 in antithrombin).
families of serine proteinase
ches the active site of the protease (Fig. 2). This need to
maintain mobility provides an explanation
servation throughout the inhibitor
family, of a trio of amino acids with low-bulk
chains, usualIy -Ala-Ala-AIa-,
proposed bend of the loop, 382-384 in antithrombin
Alanine 382 is strongly conserved throughout
hibitor serpins and its replacement
tithrombin Hamilton  results in a loss of inhibitory
activity. Alanine 384 is less strictly conserved and hence
likely to, be less critical for function.
with the observation that a major change such as the
replacement of alanine 384 by proline [IO] causes an ap-
parently complete loss of activity
drastic substitution of the alanine by a sizrine causes on-
ly a selective loss of inhibition.
Although the presence of a row of alanines suggests
a requirement for flexibility, there will also be steric and
polarity requirements in order for these residues to fit in
place as they fold back into the A-sheet. Evidence for
the requirements of fit comes from binary complexing
experiments in which a synthesised peptide with the se-
quence of the active centre
380-392 (P2-P14) can be shown to anneal in the A-
sheet in the intact inhibitors
382 (P12) alanine by threonine as in AT Hamilton gives
a much slower annealing indicating a decrease in the fit
of the threonine in the A-sheet (D. Evans et al., un-
published). The same experiments
inhibitors  that mat-
for the con-
members of the
near the at a position,
in an- by threonine
This is in keeping
whereas the less
loop of antithrombin
. Replacement of the
have not been tried
with the new AT Cambridge I1 mutant but it is difficult
to see from the model (Fig. 2b) how reincorporation
residue 384 could occur in the intact protein even to give
the fully stretched canonical form of the reactive centre.
It is likely then that the AT Cambridge
primarily affect the mobility of the loop rather than its
ability to fit sterically into the A-sheet.
The observed findings with these mutants
tithrombin fit with the changes that follow loosening of
the loop by the insertion of an extra residue as in the
natural variant antiplasmin
its inhibitory activity, and with site-directed
antitrypsin in which the addition of a residue alters the
specificity, as well as activity of inhibition
the overall results are compatible
alanines 382-384 in the maintenance
the reactive centre as well as its ability to fit back into
II mutation will
Enschede  which loses
with a role for
of the flexibility of
Wellcome Trust, The Medical Research Council and The British
Heart Foundation. We are grateful to Dr. C. Marshall for the com-
puter analyses depicted in Fig. 2.
We gratefully acknowledge the support of The
Rosenberg, R.D. and Damus, P.S. (1973) J. Biol. Chem. 248,
Beresford, C.H. (1988) Blood Rev. 2, 239-250.
Saiki, R.K., Scharf, S.. Faloona, F., Mullis, K.B., Horn, C.T.,
Erlich, W.A. and Arnheim, N. (1985) Science 230, 1350-1354.
Prochownik, E.V., Bock, SC. and Orkin, S.H. (1985) J. Biol.
Chem. 260, 9608-0612.
Edgar, P., Jennings, I. and Harper, P. (1988) J. Clin. Path. 42,
Abildgaard, U., Lie, M. and Odegard, O.R. (1977) Thromb.
Res. I I, 549-553.
Odegard, O.R., Lie, M. and Abildgaard,
Sas, G., Pepper, D.S. and Cash, J.D. (1975) Br. J. Haem. 30,
Perry, D.J. and Carrell, R.W.
Perry, D.J., Harper, P.L., Fairham, S., Daly, M. and Carrell,
R.W. (1989) FEBS Lett. 254, 174-176.
Molhosabatier, P., Aiach, M., Gaillard,
Fischer, A.M., Chadeuf, G. and Clauser, E. (1989) J. Clin. In-
vest. 84, 1236-3242.
Huber, R. and Carrell, R.W.
Leslie, G.W., Finch,
McLaughlin, P.J. and Carrell, R.W. (1990) Nature 347,99-102.
Carrel!, R.W. and Owen, M.C. (1985) Nature 317, 730-732.
Bode, W., An-Zhi, W., Huber, R., Meyer, E., Travis, J. and
Neumann, S. (1990) EMU0 J. 5, 2453-2458.
Devraj-Kizuk, R., Chui, D.H.K.,
C.J., Ofosu, F.A, and Biajchnian,
Schulze, A.J., Bauman, U., Knof, S., Jaegc, E., Huber, R. and
Laurell, C.-B., Eur. J. Biochem., in press.
Nieuwenhuis, H.K., Rijken, D.C. and Collen, D. (1987) Science
Avron, A., Reeve, F,, Lickorish, J.M. and Carrel], R.W. (1991)
FEBS Lett. 280, 41-43,
U. (1976) Haemostasis
(1990) Mol. Biol, Med. 6,
I., Fiessinger, J.N.,
(1989) Biochemistry 28,
J.T., Turnell, W.G.,
(1988) Blood 72,
Nelles, L., Kluft, C.,