Protein Expression and Purification 61 (2008) 31–35

Contents lists available at ScienceDirect

Protein Expression and Purification

j o u r n a l h o m e p a g e : w w w . e l s e v i e r. c o m / l o c a t e / y p r e p

A new feasible method for fibrinogen purification based on the affinity

of Staphylococcus aureus clumping factor A to fibrinogen
Chao-Zong Liu a,*, Hui-Ju Cheng b, Ling-Ya Chang a
a
Depart­ment of Phar­ma­col­ogy, Col­lege of Med­i­cine, Tzu Chi Uni­ver­sity, No. 701, Chung-Yung Road, Sec. 3, Hu­alien City, Hu­alien County, Tai­wan
b
Insti­tute of Phar­ma­col­ogy and Tox­i­col­ogy, Col­lege of Life Sci­ences, Tzu Chi Uni­ver­sity, Hu­alien City, Hu­alien County, Tai­wan

a r t i c l e i n f o a b s t r a c t

Article history: Plasma fibrin­o­gen par­tic­i­pates in sev­eral phys­i­o­log­i­cal and path­o­log­i­cal events thus becom­ing a use­
Received 15 February 2008 ful study­ing mate­rial in bio­med­i­cal research. Here we report a new con­ve­nient method for fibrin­o­gen
and in revised form 9 May 2008 puri­fi­ca­tion based on the affin­ity of Staph­y­lo­coc­cus aureus clump­ing fac­tor A to fibrin­o­gen. Clump­ing
Available online 23 May 2008
fac­tor A (ClfA) is a cell wall-anchored sur­face pro­tein of S. aureus bac­te­ria that binds with a high affin­ity
to the fibrin­o­gen c chain C-ter­mi­nus via a seg­ment encom­pass­ing the res­i­dues 221–550. This activ­ity
Key­words: of ClfA (ClfA221–550) was pro­duced in fusion to the C-ter­mi­nus of glu­ta­thi­one–S-trans­fer­ase (GST) with
Clump­ing fac­tor A
recombinant tech­nol­ogy and used as an affin­ity ligand to cap­ture plasma fibrin­o­gen. GST–ClfA221–550
Fibrin­o­gen
Affin­ity puri­fi­ca­tion
fusion pro­tein was immo­bi­lized onto the glu­ta­thi­one-con­ju­gated beads packed in a plas­tic col­umn by its
GST part. Then, this affin­ity col­umn was loaded with cit­rated and hep­a­rin­ized human plasma. After wash­
ing out unbound pro­teins, col­umn-cap­tured fibrin­o­gen was spe­cif­i­cally eluted down with a cit­rate buffer
solu­tion (50 mM, pH 5.6). Puri­fied human fibrin­o­gen exhib­ited the abil­ity to sup­port plate­let adhe­sion
and aggre­ga­tion and formed fibrin clot by throm­bin, indi­cat­ing that ClfA221–550-puri­fied human fibrin­o­
gen is a func­tion­ally active prod­uct. We also found that both the rat and mouse fi­bri­no­gens could be puri­
fied as well as human fibrin­o­gen with this method. By vir­tue of its sim­plic­ity and fea­si­bil­ity, ClfA221–550-
based method would be very use­ful to the inves­ti­ga­tors who need fibrin­o­gen to per­form their stud­ies.
© 2008 Else­vier Inc. All rights reserved.

Fibrin­o­gen is a large plasma pro­tein (»340 kDa) com­posed of
two sets of three dif­fer­ent poly­pep­tide chains, Aa, Bb and c [1].
It plays an impor­tant role in many path­o­log­i­cal and phys­i­o­log­i­cal
events, such as throm­bo­sis, he­mo­sta­sis, inflam­ma­tion, wound heal­
ing and angi­o­gen­e­sis [2], and has become a valu­able mate­rial in
study­ing these fields. Recently, the assem­bled prod­uct of fibrin­o­
gen (fibrin) was suc­cess­fully used as a matrix for cell growth in
tis­sue engi­neer­ing [3,4] as well as a res­er­voir for the bind­ing and
release of cer­tain growth fac­tors [5], sug­gest­ing that the demand
of puri­fied fibrin­o­gen in bio­med­i­cal research will increase.
Today, puri­fied fibrin­o­gen can be pur­chased from lab­o­ra­tory
prod­uct com­pa­nies; how­ever, users often encoun­tered the dif­fi­
culty to dis­solve com­mer­cial fibrin­o­gen powder and a decrease
of bio­log­i­cal activ­ity fol­low­ing recon­sti­tu­tion. For these rea­sons,
inves­ti­ga­tors would like to purify fibrin­o­gen them­selves if there is
a fea­si­ble and afford­able method for them to per­form fibrin­o­gen
puri­fi­ca­tion.
Affin­ity chro­ma­tog­ra­phy is one of the most use­ful tech­niques for
the rapid puri­fi­ca­tion of a tar­get mol­e­cule from a com­plex mix­ture.
The mol­e­cule to be puri­fied is spe­cif­i­cally and revers­ibly adsorbed
by a com­ple­men­tary bind­ing sub­stance (ligand) attached to an insol­
u­ble sup­port matrix, then recov­ered by chang­ing the exper­i­men­tal
* Cor­re­spond­ing author. Fax: +886 3 8561465.
E-mail address: czliu33@mail.tcu.edu.tw (C.-Z. Liu).
con­di­tions to those favor­ing desorp­tion. Fibrin­o­gen has been puri­
fied from the plasma using the gels cou­pled with syn­thetic pep­tide
[6,7], mono­clo­nal anti­body [8] or prot­amine [9]. These affin­ity gels
though have been proved quite use­ful, they have some dis­ad­van­
tages. Pep­tide-based affin­ity chro­ma­tog­ra­phy may result in a reduc­
tion of the bio­log­i­cal activ­ity of fibrin­o­gen due to harsh puri­fi­ca­tion
con­di­tion, while the high cost of mono­clo­nal anti­body lim­its its use
widely. Prot­amine on the other hand is not regarded as a spe­cific
ligand for fibrin­o­gen because the high con­tent of argi­nine res­i­dues
in prot­amine may inter­act with other mol­e­cules bear­ing neg­a­tive
charges. Recently, a hydro­pho­bic inter­ac­tion chro­ma­tog­ra­phy (HIC)
for the rapid puri­fi­ca­tion of plasma fibrin­o­gen was doc­u­mented
[10]. This method allows plasma fibrin­o­gen to be puri­fied to homo­
ge­ne­ity in a sin­gle step by using a set of expen­sive equip­ments to
per­form col­umn chro­ma­tog­ra­phy. From these data, it seems likely
that a fea­si­ble fibrin­o­gen puri­fi­ca­tion method which also pro­vides
spec­i­fic­ity and easy acces­si­bil­ity is still lack­ing.
Clump­ing fac­tor A (ClfA) is a cell wall-anchored sur­face pro­
tein of Staph­y­lo­coc­cus aureus (S. aureus) known for its abil­ity to
1 Abbre­vi­a­tions used: ClfA, clump­ing fac­tor A; GST, glu­ta­thi­one S-trans­fer­ase; ADP,
aden­o­sine 59-diphos­phate; pNPP, p-nitro­phenyl phos­phate; BSA, bovine serum albu­
min; GPIIb/IIIa, gly­co­pro­tein IIb/IIIa; GST, glu­ta­thi­one S-trans­fer­ase; TBS, Tris-buf­
fered saline; ACD, acid cit­rate dex­trose; FXIII, fac­tor XIII; PT, pro­throm­bin time;
aPTT, acti­vated partial throm­bo­plastin time; ClfA, clump­ing fac­tor A.

1046-5928/$ – see front matter © 2008 Else­vier Inc. All rights reserved.
doi:10.1016/j.pep.2008.05.007
32 C.-Z. Liu et al. / Protein Expression and Purification 61 (2008) 31–35
bind with a high affin­ity to the c chain C-ter­mi­nus of fibrin­o­gen
[11,12]. The fibrin­o­gen-bind­ing activ­ity of ClfA has been local­
ized to a seg­ment encom­pass­ing the res­i­dues from 221 to 550
[13]. Our recent works on ClfA revealed that the fibrin­o­gen-bind­
ing seg­ment of ClfA (ClfA221–550) could be mass-pro­duced with
recombinant tech­nol­ogy as a fusion prod­uct with glu­ta­thi­one–
S-trans­fer­ase (GST) [14] and this fusion pro­tein bound plasma
fibrin­o­gen with a high affin­ity and spec­i­fic­ity [15]. These find­ings
inspired us to take advan­tage of ClfA’s fibrin­o­gen-bind­ing activ­ity
(ClfA221–550) to develop a fea­si­ble method for fibrin­o­gen puri­fi­ca­
tion. Recombinant GST–ClfA221–550 fusion pro­tein was first immo­
bi­lized onto the glu­ta­thi­one-con­ju­gated beads by its GST part and
used as an affin­ity ligand to cap­ture plasma fibrin­o­gen, which in
asso­ci­a­tion with ClfA221–550 was spe­cif­i­cally eluted down by cit­
rate buffer. This method allowed plasma fibrin­o­gen to be puri­fied
in less than 2 h with­out using any expen­sive equip­ment to per­
form col­umn chro­ma­tog­ra­phy.
Mate­ri­als and meth­ods
Reagents
Glu­ta­thi­one-con­ju­gated Sepharose 4B beads, bovine throm­
bin and goat anti-GST anti­body were pur­chased from Amersham
Bio­sci­ences Ltd. (Upp­sala, Swe­den). Com­mer­cial fibrin­o­gen was a
lyoph­i­lized powder pur­chased from Cal­bio­chem, a brand of EMD
Bio­sci­ences, Inc. (La Jolla, CA, USA). Aden­o­sine 59-diphos­phate
(ADP), p-nitro­phenyl phos­phate (pNPP) and bovine serum albu­min
(BSA) were from Sigma Chem­i­cal Co. (St. Louis, MO, USA). BSA was
fur­ther puri­fied with anion exchange chro­ma­tog­ra­phy on DEAE
col­umn. Glu­ta­thi­one (reduced form) was a com­mer­cial prod­uct
of Merck Co. (Darms­tadt, Ger­many). The clin­i­cally used gly­co­pro­
tein IIb/IIIa (GPIIb/IIIa) antag­o­nist tir­ofi­ban (Ag­gra­stat®) was man­
u­fac­tured by Merck & Co., Inc. (West Point, PA, USA). Recombinant
GST–ClfA221–550 fusion pro­tein was pro­duced by Esch­e­richia coli
(E. coli) and puri­fied to homo­ge­ne­ity as pre­vi­ously described [14].
In brief, the DNA seg­ment cor­re­spond­ing to ClfA221–550 was cop­
ied from the geno­mic DNA of S. aureus 30326, a clin­i­cal iso­late
from Tzu Chi Gen­eral Hos­pi­tal (Hu­alien, Tai­wan), by PCR using
the primer pair 59CGGGATCCGTAGCTGC­A­GATG­CACC39 (for­ward)
and 59CGGA­ATTCGGCTCATCAGGTTGTT­CAG39 (reverse), both of
which con­tain the restric­tion cleav­age sites Bam­HI and Eco­RI,
respec­tively (under­lined). PCR prod­uct was cloned into the expres­
sion vec­tor pGEX-2T, which allows ClfA221–550 to be pro­duced in
fusion to the C-ter­mi­nus of glu­ta­thi­one–S-trans­fer­ase (GST) in
E. coli fol­low­ing IPTG induc­tion. After dis­rup­tion of IPTG-treated
E. coli, GST–ClfA221–550 fusion pro­tein was puri­fied from the sol­u­ble
frac­tion of bac­te­rial lysate with glu­ta­thi­one–Sepharose 4B col­umn.
Cit­ric acid and tri-sodium cit­rate were pur­chased from Wako Pure
Chem­i­cal Indus­tries, Ltd. (Osaka, Japan). BCA and Bio-Rad pro­tein
assay kits were from Pierce Chem­i­cal Co. (Rock­ford, IL, USA) and
Bio-Rad lab­o­ra­to­ries (USA), respec­tively. Page­Blue pro­tein stain­
ing solu­tion was a com­mer­cial prod­uct of Fer­men­tas Life Sci­ences
(USA). ICR mice (30 g) and WKY rats (250–300 g) were pro­vided by
the Bio­LAS­CO Co., Ltd. (Tai­pei, Tai­wan) and National Lab­o­ra­tory Ani­
mal Cen­ter (Tai­pei, Tai­wan), respec­tively. They were main­tained in
our uni­ver­sity’s ani­mal facil­i­ties. The use of ani­mals was approved
by the Insti­tu­tional Com­mit­tee on the Use and Care of Ani­mals of
Tzu Chi Uni­ver­sity.
Pre­pare affin­ity col­umn for fibrin­o­gen puri­fi­ca­tion
Affin­ity col­umn was freshly prepared by pass­ing GST–ClfA221–550
fusion pro­tein (4 mg in 4 ml of Tris-buf­fered saline) through a plas­
tic col­umn packed with 10 ml of glu­ta­thi­one–Sepharose 4B beads,
which was then washed with Tris-buf­fered saline (TBS).
Plasma prep­a­ra­tion
Plasma was prepared from the freshly drawn whole blood
an­ti­co­ag­u­lat­ed with hep­a­rin-con­tain­ing (10 units/ml) tri-
sodium cit­rate solu­tion (3.2%, w/v) in a vol­ume ratio of 9:1.
After hav­ing spun down blood cells with a high speed cen­tri­
fu­ga­tion (3000g) at 4 °C for 10 min, plasma was col­lected and
passed through a 0.22 lm fil­ter set to remove par­tic­u­late mate­
ri­als. Human bloods were obtained from healthy donors by
ve­ni­punc­ture. Mouse and rat whole bloods were col­lected from
pen­to­bar­bi­tal (50 mg/kg, i.p.)-anes­the­tized ani­mals by heart
punc­ture.
Fibrin­o­gen puri­fi­ca­tion
Affin­ity col­umn was loaded with 8 ml of cit­rated and hep­a­rin­
ized human plasma by grav­ity flow. After wash­ing out unbound
plasma pro­teins with TBS, col­umn was eluted with a cit­rate
buffer solu­tion (50 mM, pH 5.6). Elu­ent was col­lected into var­i­
ous tubes (2 ml/tube) and the pro­tein frac­tion was deter­mined
by mix­ing 10 ll of each col­lected elu­ent with 100 ll of Bio-Rad
pro­tein assay reagent, which pro­duced a blue color eas­ily to be
detected by naked eyes or pho­tom­e­try. Pho­tom­e­try was car­ried
out with a micro­plate reader (VER­SA­max, Molec­u­lar Devices)
set­ting at 595 nm. The purity of eluted fibrin­o­gen was checked
by SDS–PAGE anal­y­sis and the con­tam­i­na­tion of GST–ClfA221–550
fusion pro­tein was exam­ined by Western blot using anti-GST
anti­body as the probe. The fibrin­o­gen-con­tain­ing elu­ents were
pooled, con­cen­trated with a cen­tri­fuge con­cen­tra­tor (Centri­con
plus 20, Mil­li­pore), dia­lyzed against TBS at 4 °C and quan­ti­fied
by BCA pro­tein assay kit using bovine serum albu­min as the
stan­dard. For long-term stor­age, puri­fied fibrin­o­gen (in cit­rate
buffer) was sub­jected to freeze dry­ing and stored at ¡20 °C.
Before use, lyoph­i­lized fibrin­o­gen was recon­sti­tuted with
dis­tilled water and the cit­rate buffer was changed into TBS by
dial­y­sis or pass­ing through a gel fil­tra­tion col­umn (Sepha­dex
G-25) equil­i­brated with TBS. Both the rat and mouse fi­bri­no­
gens were puri­fied as that for human fibrin­o­gen, but in a
smaller scale.
Regen­er­a­tion of glu­ta­thi­one-con­ju­gated Sepharose 4B beads
After elut­ing down fibrin­o­gen, col­umn bound GST–ClfA221–550
fusion pro­tein was removed by glu­ta­thi­one (10 mM in 50 mM Tris
buffer, pH 8.0) elu­tion and the glu­ta­thi­one-con­ju­gated Sepharose
4B beads were sub­se­quently regen­er­ated accord­ing to man­u­fac­
turer’s instruc­tion. These beads were kept in 20% eth­a­nol and
stored at 4 °C until reused.
Prep­a­ra­tion of washed human plate­lets
Washed human plate­lets were prepared from the whole
blood an­ti­co­ag­u­lat­ed with 10% vol­ume of acid cit­rate dex­trose
(ACD) solu­tion. Whole blood was cen­tri­fuged at 626g, 30 °C
for 10 min to spin down leu­ko­cytes and eryth­ro­cytes. Then,
plate­lets were iso­lated from plate­let-rich plasma as pre­vi­ously
described [16] and sus­pended in Ty­rode solu­tion (NaH2PO4,
0.4 mM; NaCl, 136.9 mM; KCl, 2.7 mM; NaHCO3, 11.9 mM; MgCl2,
1 mM; CaCl2, 1 mM; glu­cose, 1 mg/ml; bovine serum albu­min,
3.5 mg/ml; pH 7.35) at a con­cen­tra­tion of 3.75 £ 108 plate­lets/
ml. For adhe­sion assay, washed human plate­lets were sus­
pended in a cal­cium-free Ty­rode solu­tion (NaH2PO4, 0.4 mM;
NaCl, 136.9 mM; KCl, 2.7 mM; NaHCO3, 11.9 mM; MgCl2, 2 mM;
glu­cose, 1 mg/ml; bovine serum albu­min, 3.5 mg/ml; pH 7.35),
in which pros­ta­glan­din E1 (2 lM) was included to inhibit plate­
let acti­va­tion and aggre­ga­tion.
 C.-Z. Liu et al. / Protein Expression and Purification 61 (2008) 31–35 33
Plate­let aggre­gom­e­try
Plate­let aggre­ga­tion assay was per­formed with an aggre­gom­e­
ter (Pay­ton Sci­en­tific, Sto­uff­ville, Ontario, Can­ada) under stir­ring
con­di­tion (900 rpm). Washed human plate­lets, either in the pres­
ence or absence of ClfA221–550-puri­fied fibrin­o­gen (0.2 mg/ml) sup­
ple­ment, were aggre­gated by ADP (8 lM) at 37 °C for 5 min. The
final reac­tion vol­ume and plate­let count for this assay were 500 ll
and 3 £ 108 plate­lets/ml, respec­tively, and the increase of light
trans­mis­sion in plate­let sus­pen­sion rep­re­sents the aggre­ga­tion of
plate­lets.
Plate­let adhe­sion assay
Plate­let adhe­sion assay was per­formed in 96-well micro­ti­ter
plate as pre­vi­ously described [17]. Max­i­sorp plate (Nunc, Den­mark)
was coated with BSA (1 mg/ml, 50 ll) or ClfA221–550-puri­fied human
fibrin­o­gen (1 mg/ml, 50 ll) at room tem­per­a­ture for 4 h, fol­lowed
by TBS wash for 3 times. Then, an ali­quot (50 ll) of human plate­
let sus­pen­sion (1 £ 108/ml), which has been pre­treated with either
TBS, EDTA (5 mM) or tir­ofi­ban (100 ng/ml) at room tem­per­a­ture for
10 min, was added into the well and the plate­lets were allowed
to adhere to coated pro­tein at room tem­per­a­ture for 20 min.
After remov­ing non-adher­ent plate­lets by a gen­tle wash with PBS
(200 ll), adher­ent plate­lets were quan­ti­fied by mea­sur­ing the acid
phos­pha­tase activ­ity of plate­lets using pNPP as the sub­strate [18].
Mea­sure­ment of fibrin clot for­ma­tion
The for­ma­tion of fibrin clot was detected by tur­bi­dim­e­try. In the
pres­ence of CaCl2 (2 mM) sup­ple­ment, ClfA221–550-puri­fied fibrin­
o­gen (1 mg/ml, 0.9 ml) was treated with throm­bin (10 units/ml,
0.1 ml) at room tem­per­a­ture for 30 min. The optic den­sity (600 nm)
was mea­sured by a spec­tro­pho­tom­e­ter (DU650, Beck­man Instru­
ment Inc., USA) at 1-min inter­val.
SDS–PAGE anal­y­sis of fibrin clot
The fibrin clot formed by treat­ing fi­bri­no­gens with throm­bin at
room tem­per­a­ture for 30 min as stated above was sol­u­bi­lized by
add­ing equal vol­ume of 2-mercap­toethanol (8%, v/v)-con­tain­ing
SDS–PAGE sam­ple buffer (2£) and heat­ing at 95 °C for 10 min. After
that, an ali­quot of the solu­tion was sub­jected to SDS–PAGE and the
pro­tein bands were visu­al­ized with Page­Blue stain­ing.
Results and dis­cus­sion
After sequen­tially pass­ing GST–ClfA221–550 fusion pro­tein and cit­
rated and hep­a­rin­ized human plasma through the col­umn packed
with glu­ta­thi­one-con­ju­gated Sepharose 4B beads, plasma fibrin­o­
gen was spe­cif­i­cally cap­tured by the col­umn and then eluted down
with a cit­rate buffer solu­tion (50 mM, pH 5.6), which pref­er­en­tially
dis­rupt the inter­ac­tion between fibrin­o­gen and ClfA221–550 over the
bind­ing of GST to glu­ta­thi­one beads. GST–ClfA221–550 fusion pro­tein
was sub­se­quently eluted down by glu­ta­thi­one (10 mM in 50 mM
Tris buffer, pH 8.0) (Sup­ple­men­tal Fig. 2). Eluted fibrin­o­gen exhib­
ited a high purity since only the typ­i­cal bands of human fibrin­o­gen
were vis­i­ble on the SDS–elec­tro­pho­retic gels (Fig. 1) and less than
1 ng of GST–ClfA221–550 fusion pro­tein was detected with 5 lg of puri­
fied fibrin­o­gen by Western blot (data not shown). If the con­tam­i­na­
tion of GST–ClfA221–550 was con­cerned, pass­ing eluted fibrin­o­gen
through a small glu­ta­thi­one–Sepharose 4B col­umn equil­i­brated
with cit­rate buffer could fur­ther reduce this impu­rity. Eluted fibrin­
o­gen was revealed by the Bio-Rad pro­tein assay reagent, which pro­
duced a vis­i­ble color (blue) eas­ily to be detected by naked eyes or
pho­tom­e­try (Sup­ple­men­tal Fig. 1). This way of detec­tion allowed
n e n e
ma at us ma at us
Hu R Mo Hu R Mo
kDa kDa Fibrinogen
94 94
67 67 chain
chain
chain
43 43
30 30
(NR) (R)
Fig. 1. SDS–PAGE anal­y­sis of ClfA221–550-puri­fied fi­bri­no­gens. Ali­quots of ClfA221–550-
puri­fied human, rat and mouse fi­bri­no­gens were sub­jected to SDS–PAGE in the pres­
ence (R) or absence (NR) of reduc­ing agent (2-mercap­toethanol, 4%). The pro­tein
bands on the slab gel were visu­al­ized with Page­Blue pro­tein stain­ing solu­tion.
the frac­tion of fibrin­o­gen to be iden­ti­fied in only few min­utes, thus
short­en­ing the puri­fi­ca­tion time. It was esti­mated that around 6 mg
of human fibrin­o­gen was puri­fied from 8 ml of plasma in 2 h at the
expense of 4 mg of GST–ClfA221–550 fusion pro­tein. These results
indi­cate that a bulk of highly puri­fied human fibrin­o­gen could be
eas­ily obtained by ClfA221–550-based method.
Plasma fibrin­o­gen is an impor­tant mol­e­cule that medi­ates plate­
let adhe­sion and aggre­ga­tion. It also forms fibrin clot in response
to throm­bin. To see whether puri­fied human fibrin­o­gen is func­tion­
ally active, it was tested for its abil­ity to sup­port plate­let adhe­sion
and aggre­ga­tion and to form fibrin clot by throm­bin. As shown
in the upper panel of Fig. 2, ADP (8 lM) caused acti­va­tion but
not aggre­ga­tion of washed human plate­lets (lower trac­ing). This
is because washed human plate­lets need a strong stim­u­la­tion to
release sig­nif­i­cant amount of fibrin­o­gen from their a gran­ules to
medi­ate aggre­ga­tion whereas ADP is a weak stim­u­la­tor of washed
human plate­lets. By con­trast, a marked aggre­ga­tion of washed
human plate­lets was achieved by ADP with the sup­ple­ment of puri­
fied human fibrin­o­gen (0.2 mg/ml) (upper trac­ing), indi­cat­ing that
ClfA221–550-puri­fied human fibrin­o­gen has the capac­ity to medi­ate
plate­let aggre­ga­tion. We also found that puri­fied human fibrin­o­gen,
when immo­bi­lized onto a sur­face, can inter­act with the diva­lent
cat­ion-depen­dent adhe­sion mol­e­cule GPIIb/IIIa, lead­ing to plate­let
adhe­sion. Washed human plate­lets adhered to the sur­face coated
with puri­fied human fibrin­o­gen rather than the sur­face coated
with BSA and this spe­cific adhe­sion was abro­gated when plate­lets
were pre­treated with EDTA (5 mM), a diva­lent cat­ion che­la­tor, or
tir­ofi­ban (100 ng/ml), a clin­i­cally used GPIIb/IIIa antag­o­nist (Fig. 2,
lower panel). The clot­ta­bil­i­ty of ClfA221–550-puri­fied human fibrin­
o­gen was also dem­on­strated by tur­bi­dim­e­try. A marked increase
of optic den­sity (600 nm) was detected along with a gel-like struc­
ture for­ma­tion after treat­ing ClfA221–550-puri­fied human fibrin­o­
gen (1 mg/ml; 0.9 ml) with throm­bin (10 units/ml; 0.1 ml) at room
tem­per­a­ture for 30 min (data not shown). These results together
con­firm that ClfA221–550-puri­fied human fibrin­o­gen is a bio­log­i­cally
active prod­uct.
Blood coag­u­la­tion fac­tor XIII (FXIII) is a pro­glu­ta­min­ase con­
verted into a glu­ta­min­ase (FXIIIa) by throm­bin in the ter­mi­nal
step of coag­u­la­tion cas­cade [19]. FXIIIa cross-links poly­mer­ized
fibrin, result­ing in an increase of mechanic sta­bil­ity of fibrin clot
and its resis­tance to plas­min deg­ra­da­tion [20]. FXIII has been
shown to cir­cu­late in asso­ci­a­tion with plasma fibrin­o­gen [21] and
this may explain why c–c dimer was observed with com­mer­cial
human fibrin­o­gen fol­low­ing throm­bin treat­ment. ClfA221–550-puri­
fied human fibrin­o­gen was shown with little FXIIIa activ­ity; c–c
34 C.-Z. Liu et al. / Protein Expression and Purification 61 (2008) 31–35
Light transmission
ADP 8 M
16
Adherent platelets (×105)
12
8 cle (TBS) at room tem­per­a­ture for 30 min. At the end of reac­tion, an equal vol­ume of
4 solu­tion was run on 12% SDS–poly­acryl­amide gel and the pro­tein bands were visu­al­
0
n (data not shown). From these data, it is likely that the dif­fi­culty
PB
S
PB
S TA ba
ED M i r ofi /ml
5m T 0ng
10
BSA Fibrinogen
Fig. 2. The sup­port­ing effects of ClfA221–550-puri­fied human fibrin­o­gen on plate­let
adhe­sion and aggre­ga­tion. Upper panel shows the aggre­gat­ing response of washed
human plate­lets to ADP. Washed human plate­lets (3 £ 108/ml) were aggre­gated by
ADP (8 lM) in the pres­ence (+) or absence (¡) of ClfA221–550-puri­fied human fibrin­
o­gen (200 lg/ml). Loss of oscil­la­tion sig­nals and increase of light trans­mis­sion in
the plate­let sus­pen­sion rep­re­sent the acti­va­tion and aggre­ga­tion of plate­lets, respec­
tively. Lower panel shows the results of plate­let adhe­sion assay. Washed human
plate­lets (1 £ 108/ml, 50 ll) were allowed to adhere to the sur­faces coated with BSA
or ClfA221–550-puri­fied fibrin­o­gen at room tem­per­a­ture for 20 min. After that, non-
adher­ent plate­lets were removed and adher­ent plate­lets were quan­ti­fied by com­par­
ing their acid phos­pha­tase activ­ity with a known num­ber of plate­lets. Please refer
to the sec­tion of Mate­ri­als and meth­ods for exper­i­men­tal detail. To deter­mine the
effects of tir­ofi­ban and EDTA on the plate­let adhe­sion to ClfA221–550-puri­fied human
fibrin­o­gen, washed human plate­lets were pre­treated with tir­ofi­ban (100 ng/ml) or
EDTA (5 mM) at room tem­per­a­ture for 10 min before add­ing into pro­tein-coated
wells. All the exper­i­ments were per­formed in trip­li­cate and the data were shown
in means § SEM (n = 3).
dimer was not vis­i­ble with ClfA221–550-puri­fied human fibrin­o­gen
fol­low­ing throm­bin treat­ment (Fig. 3, lane 6) as com­pared with
com­mer­cial fibrin­o­gen (Fig. 3, lane 3). This result not only reveals
that ClfA221–550-puri­fied human fibrin­o­gen is free from sig­nif­i­cant
FXIIIa activ­ity, it may also imply that the purity of ClfA221–550-puri­
fied human fibrin­o­gen is bet­ter than those of com­mer­cial human
fi­bri­no­gens. We think FXIII was dis­so­ci­ated from fibrin­o­gen dur­ing
puri­fi­ca­tion.
Puri­fied pro­teins often are sub­jected to freeze dry­ing for long-
term stor­age and easy deliv­ery, and then recon­sti­tuted with water
or spe­cific buffer before use. In this study, we found that lyoph­i­
lized powder of ClfA221–550-puri­fied human fibrin­o­gen (in cit­rate
buffer) dis­solved very well in dis­tilled water after stor­age at ¡20 °C
for 3 months and its capa­bil­ity to sup­port plate­let aggre­ga­tion was
indis­tin­guish­able from that of freshly puri­fied human fibrin­o­gen
M 1 2 3 4 5 6
kDa
97 - dimer
66 chain
chain
43 chain
30
Ca2+
Thrombin
CommercialFg PurifiedFg
(Calbiochem®)
Fig. 3. SDS–PAGE anal­y­sis of fibrin­o­gen c chain cross-link­ing. In the pres­ence or
absence of CaCl2 (2 mM) sup­ple­men­ta­tion, both ClfA221–550-puri­fied human fibrin­o­
gen and com­mer­cial fibrin­o­gen were treated with either bovine throm­bin or vehi­
2-mercap­toethanol (8%)-con­tain­ing SDS–PAGE sam­ple buffer (2£) was added and
the fibrin clot was sol­u­bi­lized by heat­ing at 95 °C for 10 min. An ali­quot (10 ll) of the
ized with Page­Blue. Lane M shows the molec­u­lar weight mark­ers.
to dis­solve com­mer­cial fibrin­o­gen powder and reduc­tion of bio­
log­i­cal activ­ity fol­low­ing recon­sti­tu­tion will not be a prob­lem for
ClfA221–550-puri­fied human fibrin­o­gen.
As to the cost of using GST–ClfA221–550-based method to purify
plasma fibrin­o­gen, it seems to be more accept­able than that using
mono­clo­nal anti­body-based method since GST–ClfA221–550 fusion
pro­tein could be mass-pro­duced with recombinant tech­nol­ogy and
eas­ily puri­fied to homo­ge­ne­ity. An esti­mated amount of 30 mg of
GST–ClfA221–550 fusion pro­tein was puri­fied from the trans­formed
E. coli grow­ing in 3-liter cul­ture medium, accord­ing to our pre­vi­
ous work [14]. Another finan­cial ben­e­fit of using GST–ClfA221–550
as a tool to purify fibrin­o­gen is that the glu­ta­thi­one-con­ju­gated
Sepharose 4B beads, one of the two essen­tial mate­ri­als for ClfA221–
550-based fibrin­o­gen puri­fi­ca­tion, could be regen­er­ated and reused
more than 20 times with­out marked reduc­tion of its capac­ity
(<20%) to bind GST–ClfA221–550 fusion pro­tein.
Rat and mouse are two widely used exper­i­men­tal ani­mals.
Their fi­bri­no­gens vary sig­nif­i­cantly from humans both in the
molec­u­lar size and amino acid sequence [22,23]. In spite of these
dif­fer­ences, both ani­mals’ fi­bri­no­gens could be puri­fied as well as
human fibrin­o­gen with ClfA221–550-based method (Fig. 1). We sup­
pose that ClfA221–550 binds to a sequence shared by the human,
rat and mouse fi­bri­no­gens, and the c chain C-ter­mi­nal pep­tide
AGDV appears to be the most likely tar­get since ClfA221–550 has
been shown to bind human fibrin­o­gen c chain C-ter­mi­nal pep­
tide (AGDV) [12,13], which was also found in the c chain C-ter­mi­
nus of murine fibrin­o­gen [24]. It was noted that both the rat and
mouse fi­bri­no­gens are sold much more expen­sively than human
fibrin­o­gen in lab­o­ra­tory prod­uct com­pa­nies; the prices for the
rat, mouse and human fi­bri­no­gens in the small­est pack­ages are
US$390/25 mg, US$288/5 mg and US$114.5/500 mg, respec­tively,
accord­ing to Sigma Chem­i­cal Com­pany. As expen­sive study­ing
mate­rial always has a neg­a­tive impact on bio­med­i­cal research,
we believe that ClfA221–550-based method would be mean­ing­ful to
the inves­ti­ga­tors who need both ani­mals’ fi­bri­no­gens to per­form
their stud­ies.
 C.-Z. Liu et al. / Protein Expression and Purification 61 (2008) 31–35 35
Recent clin­i­cal stud­ies revealed a vari­ety of patients with fibrin­
o­gen dis­or­ders, in which either the quan­tity (afi­brin­o­gen­e­mia,
char­ac­ter­ized by the com­plete defi­ciency of fibrin­o­gen, and hypo­fi­
brin­o­ge­ne­mia, char­ac­ter­ized by fibrin­o­gen lev­els <1.5 mg/ml) or the
qual­ity (dys­fi­brin­o­ge­ne­mia) of the cir­cu­lat­ing fibrin­o­gen or both
(hyp­o­dysfi­bri­no­gen­emia) was affected [25]. It was rec­om­mended
that a more spe­cial­ized test with puri­fied fibrin­o­gen should be car­
ried out for accu­rate diag­no­sis of dys­fi­brin­o­ge­ne­mia because the
com­monly used clot­ting time tests, such as pro­throm­bin time (PT)
and acti­vated partial throm­bo­plastin time (aPTT), are sus­cep­ti­ble to
a high rate of false results in deter­min­ing fibrin­o­gen func­tion [25].
Accord­ing to this sug­ges­tion, ClfA221–550-based method may be use­
ful in the diag­no­sis of dys­fi­brin­o­ge­ne­mia, espe­cially for the patients
who also exhibit a low level of plasma fibrin­o­gen (hyp­o­dysfi­bri­no­
gen­emia), in terms of pro­vid­ing highly puri­fied plasma fibrin­o­gen.
In con­clu­sion, we have suc­ceeded in the devel­op­ment of a fea­si­ble
method for the rapid puri­fi­ca­tion of plasma fibrin­o­gen based on the
spe­cific bind­ing of S. aureus clump­ing fac­tor A (ClfA) to fibrin­o­gen.
This method uses a flow­through pro­to­col which allows highly puri­
fied and func­tion­ally active fibrin­o­gen to be iso­lated from plasma in
less than 2 h with­out hav­ing to use a set of expen­sive equip­ments to
per­form col­umn chro­ma­tog­ra­phy. We think the value of this method
is it pro­vides an easy way for inves­ti­ga­tors to purify fibrin­o­gen in lab­
o­ra­tory, even for those who are not good at pro­tein puri­fi­ca­tion and
have no bud­get to set up a chro­mato­graphic unit.
Acknowl­edg­ment
This study was finan­cially sup­ported by the Grant NSC96-2320-
B-320-008 from the National Sci­ence Coun­cil of Tai­wan.
Appen­dix A. Sup­ple­men­tary. data
Sup­ple­men­tary data asso­ci­ated with this arti­cle can be found,
in the online ver­sion, at doi:10.1016/j.pep.2008.05.007.
Ref­er­ences
[1] J.W. Wei­sel, C.V. Sta­uff­ach­er, E. Bul­litt, C. Cohen, A model for fibrin­o­gen:
domains and sequence, Sci­ence 230 (1985) 1388–1391.
[2] M.W. Mos­es­son, Fibrin­o­gen and fibrin struc­ture and func­tions, J. Thromb. Hae­
most. 3 (2005) 1894–1904.
[3] Q. Ye, G. Zund, P. Bene­dikt, S. Joc­ken­hoe­vel, S.P. Ho­erst­rup, S. Saky­ama, J.A.
Hub­bell, M. Turi­na, Fibrin gel as a three dimen­sional matrix in car­dio­vas­cu­lar
tis­sue engi­neer­ing, Eur. J. Car­di­ol­tho­rac. Surg. 17 (2000) 587–591.
[4] A. Mol, M.I. van Lie­shout, C.G. Dam-de Veen, S. Neu­ensch­wan­der, S.P. Ho­erst­
rup, F.P. Baa­ij­ens, C.V. Bo­uten, Fibrin as a cell car­rier in car­dio­vas­cu­lar tis­sue
engi­neer­ing appli­ca­tions, Bio­ma­te­ri­als 26 (2005) 3113–3121.
[5] C. Wong, E. In­man, R. Spa­ethe, S. Helg­er­son, Fibrin-based bio­ma­te­ri­als to
deliver human growth fac­tors, Thromb. Hae­most. 89 (2003) 573–582.
[6] C. Ku­yas, A. Haeb­er­li, P. Wal­der, P.W. Straub, Iso­la­tion of human fibrin­o­gen and
its deriv­a­tives by affin­ity chro­ma­tog­ra­phy on Gly-Pro-Arg-Pro-Lys-Frac­to­gel,
Thromb. Hae­most. 63 (1990) 439–444.
[7] D.B. Ka­uf­man, M.E. Hen­tsch, G.A. Baum­bach, J.A. Bu­ett­ner, C.A. Dadd, P.Y.
Hu­ang, D.J. Ham­mond, R.G. Car­bo­nell, Affin­ity puri­fi­ca­tion of fibrin­o­gen using
a ligand from a pep­tide library, Bio­tech­nol. Bio­eng. 77 (2002) 278–289.
[8] M. Tak­eb­e, G. Soe, I. Ko­hno, T. Sugo, M. Mat­su­da, Cal­cium ion-depen­dent
mono­clo­nal anti­body against human fibrin­o­gen prep­a­ra­tion, char­ac­ter­iza­
tion, and appli­ca­tion to fibrin­o­gen puri­fi­ca­tion, Thromb. Hae­most. 73 (1995)
662–667.
[9] C.E. Demp­fle, D.L. He­ene, Puri­fi­ca­tion of human plasma fibrin­o­gen by chro­ma­
tog­ra­phy on prot­amine-aga­rose, Thromb. Res. 46 (1987) 19–27.
[10] H.P. Jen­nis­sen, A. Demi­rog­lou, Inter­ac­tion of fibrin­o­gen with n-al­kyl­aga­ros­es
and its puri­fi­ca­tion by crit­i­cal hydro­pho­bic­ity hydro­pho­bic inter­ac­tion chro­ma­
tog­ra­phy, J. Chro­ma­togr. A 1109 (2006) 197–213.
[11] T.J. Fos­ter, M. Höök, Sur­face pro­tein adhe­sins of Staph­y­lo­coc­cus aureus, Trends
Micro­biol. 6 (1998) 484–488.
[12] J. Hawi­ger, S. Tim­mons, D.D. Strong, B.A. Cott­rell, M. Riley, R.F. Doo­lit­tle, Iden­
ti­fi­ca­tion of a region of human fibrin­o­gen inter­act­ing with staph­y­lo­coc­cal
clump­ing fac­tor, Bio­chem­is­try 21 (1982) 1407–1413.
[13] D. McDe­vitt, T. Nana­vaty, K. House-Pom­peo, E. Bell, N. Turner, L. McIn­tire, T.J.
Fos­ter, M. Höök, Char­ac­ter­iza­tion of the inter­ac­tion between the Staph­y­lo­coc­
cus aureus clump­ing fac­tor (ClfA) and fibrin­o­gen, Eur. J. Bio­chem. 247 (1997)
416–424.
[14] C.Z. Liu, M.H. Shih, P.J. Tsai, ClfA221–550, a fibrin­o­gen-bind­ing seg­ment of Staph­
y­lo­coc­cus aureus clump­ing fac­tor A, dis­rupts fibrin­o­gen func­tion, Thromb. Hae­
most. 94 (2005) 286–294.
[15] C.Z. Liu, T.F. Hu­ang, P.J. Tsai, P.J. Tsai, L.Y. Chang, M.C. Chang, A seg­ment of
Staph­y­lo­coc­cus aureus clump­ing fac­tor A with fibrin­o­gen-bind­ing activ­ity
(ClfA221–550) inhib­its plate­let-plug for­ma­tion in mice, Thromb. Res. 121 (2007)
183–191.
[16] C.Z. Liu, T.F. Wu, T.F. Hu­ang, D.H. Wu, G.L. Lin, Trim­ucy­tin, a col­la­gen-like snake
venom pro­tein, acti­vates plate­lets inde­pen­dent of I-domain within a2 sub­unit
of a2b1 inte­grin, Thromb. Res. 105 (2002) 153–160.
[17] C.Z. Liu, T.F. Hu­ang, Cro­vi­dis­in, a col­la­gen-bind­ing pro­tein iso­lated from snake
venom of Cro­ta­lus vir­i­dis, pre­vents plate­let–col­la­gen inter­ac­tion, Arch. Bio­
chem. Bio­phys. 337 (1997) 291–299.
[18] P. Bel­la­vite, G. Andr­i­o­li, P. Guz­zo, P. Ari­gli­ano, S. Chi­rum­bo­lo, F. Manz­ato, C.
Santo­nast­aso, A col­or­i­met­ric method for the mea­sure­ment of plate­let adhe­
sion in micro­ti­ter plates, Anal. Bio­chem. 216 (1994) 444–450.
[19] K.R. Sie­ben­lisk, D.A. Meh, M.W. Mos­es­son, Pro­trans­glu­ta­min­ase (fac­tor XIII)
med­i­ated cross­link­ing of fibrin­o­gen and fibrin, Thromb. Hae­most. 86 (2001)
1221–1228.
[20] K.R. Sieb­en­list, M.W. Mos­es­son, Pro­gres­sive cross-link­ing of fibrin c chains
increases resis­tance to fibri­no­ly­sis, J. Biol. Chem. 269 (1994) 28414–28419.
[21] K.R. Sieb­en­list, D.A. Meh, M.W. Mos­es­son, Plasma fac­tor XIII binds spe­cif­i­cally
to fibrin­o­gen mol­e­cules con­tain­ing c9 chains, Bio­chem­is­try 35 (1996) 10448–
10453.
[22] M. Mu­rak­a­wa, T. Okam­ura, T. Kam­ura, T. Shi­buya, M. Ha­ra­da, Y. Niho, Diver­sity
of primary struc­tures of the carboxy-ter­mi­nal regions of mam­ma­lian fibrin­o­
gen Aa chains, Thromb. Hae­most. 69 (1993) 351–360.
[23] R.F. Doo­lit­tle, J.M. Koll­man, Natively unfolded regions of the ver­te­brate fibrin­o­
gen mol­e­cule, Pro­teins 63 (2006) 391–397.
[24] K. Holm­bäck, M.J. Dan­ton, T.T. Suh, C.C. Dau­ht­er­ty, J.L. Degen, Impaired plate­
let aggre­ga­tion and sus­tained bleed­ing in mice lack­ing the fibrin­o­gen motif
bound by inte­grin aIIbb3, EMBO J. 15 (1996) 5760–5771.
[25] M.T. Cunn­ing­ham, J.T. Brandt, M. Lapos­at­a, J.D. Ol­son, Lab­o­ra­tory diag­no­sis of
dys­fi­brin­o­ge­ne­mia, Arch. Pathol. Lab. Med. 126 (2002) 499–505.