The adsorption of thrombin to polypropylene tubes: The effect of polyethylene glycol and bovine serum albumin

THROMBOSIS RESEARCH 37; 201-212, 1985 0049-3848/85 $3.00 + .OO Printed in the USA. Copyright (c) 1985 Pergamon Press Ltd. All rights reserved,

THE ADSORPTION OF THROMBIN TO POLYPROPYLENE TUBES: THE EFFECT OF POLYETHYLENE GLYCOL AND BOVINE SERUM ALBUMIN

McDonald

K. Iiorne III

Hematology Service, Clinical Pathology Department, National Institutes of Health, Bethesda, MD, 20205, USA

(Received 1.8.1984; Accepted in revised form 16.10.1984 by Editor K.M. Brinkhous) ABSTRACT The adsorption of 125 I-thrombin to polypropylene tubes has been studied with thrombin dissolved in tris-buffered saline (TBS), TBS with 0.66% polyethylene glycol (PEG), and TBS with 0.01% or 0.20% .bovine serum albumin (BSA). Adsorption stabilized over time from all three solutions, increased with increasing concentrations of unadsorbed thrombin, and was reversible by repetitive washing. At equilibrium, more thrombin adsorbed from TBS-PEG than from TBS. The difference was attributable to an increased amount of thrombin deposited by evaporation during aspiration of the polypropylene tubes. Adsorption equilibrium and capacity were otherwise relatively unaffected by the presence of PEG. However, PEG was effective in retarding thrombin adsorption by markedly reducing the adsorption rate. Reduction of adsorption in the presence of BSA, on the other hand, was explained by competitive inhibition. Pretreatment of the tubes with PEG or BSA was also shown to be effective in reducing thrombin adsorption from TBS.

INTRODUCTION Adsorption of thrombin to the surfaces of laboratory vessels is believed to account for the disproportionate loss of clotting activity observed as thrombin is diluted to the threshold range of its biologic activity, 10e9 -10-8 M (1,2). Polymers of polyethylene glycol (PEG), HO-CH2-(CH2),-CH2-OH, of average molecular weight 6000-7500 reduce this phenomenon, apparently by preventing adsorption of thrombin to glass and plastic surfaces (2-4). PEG, however, also precipitates proteins by excluding protein molecules from regions of solvent (5). This property restricted its use in experiments

Key Words:

thrombin,

adsorption.

201

202

THROMBIN ADSORPTION

Vo1.37, No.1

we were attempting to perform with desialylated fibrinogen. Although native fibrinogen is stable in the usual PEG concentration included in thrombin dilutions, 0.66%, the modified fibrinogen immediately precipitated in this medium. This problem prompted us to substitute bovine serum albumin (BSA), which proved to be very effective, and to investigate the mechanisms by which PEG and BSA stabilize thrombin in low concentrations. MATERIALS AND METHODS Purified human thrombin was provided b Dr. John Fenton of Albany, NY (6). The protein was radiolabeled with Na l&J I (Amersham Corp., Arlington Heights, IL) and solid state lactoperoxidase (Sigma Chemical Co., St. Louis, MO) (7). After separation from free 1251 by passage over a 0.8 x 25 cm column of Sephadex G25 (Pharmacia, Uppsala, Sweden), the iodinated protein was dialyzed overnight against tris-buffered saline (TBS) (0.10 M NaCl, 0.04 M tris HCl, 0.01 M tris base, pH 7.4). Thrombin biologic activity was measured by its ability to clot purified human fibrinogen (grade L; Kabi Diagnostics, Greenwich, CT) in a fibrometer (BBL FibroSystem; BectonDickinson, Orangeburg, NY) (8). Clotting activity was unchanged by radioiodination, remaining approximately 2 NIH U/pg. Aliquots of radiolabeled thrombin were mixed with unlabeled thrombin (approximately 1 mg/ml), which had also been dialyzed against TBS. The protein concentration of the stock thrombin solutions was determined by the Folin method with BSA as a standard (9). Radioactivity was measured in an automatic gamma counter (Tracer Analytic, Des Plains, IL). When solutions of 1251-thrombin were diluted, subsequent thrombin concentrations were calculated from radioactivity measurements (counts per minute, CPM) and the specific activity of the original undiluted sample (CPM/pg). In all calculations, the molecular weight of thrombin was assumed to be 36,600 (6). PEG 8000 (molecular weight 7000-9000; Fisher Scientific Co., Fair Lawn, NJ) and BSA (fraction V; Miles Laboratories, Elkhart, IN) were dissolved in TBS to make concentrated stock solutions. In all experiments, 1.5 ml conical polypropylene tubes (Sarstedt, Princeton, NJ) were used. When 0.25 ml of liquid was placed in these tubes, the area of the liquid-surface interface was calculated to be 1.9 cm2. The general experimental method was to dilute thrombin (labeled with a trace amount of 1251-thrombin) with either TBS. TBS with 0.66% PEG (TBS-PEG), or TBS with 0.01% BSA (TBS-BSA) and then to deliver 0.25 ml aliquots to the Care was taken conical polypropylene tubes with a polypropylene pipet tip. to avoid unnecessary contact between the tube wall and the thrombin solution. The radioactivity of the entire sample was measured to quantitate the total amount of thrombin present. The thrombin solution was then aspirated to The radioapparent dryness with a thin, flexible tube to avoid scratching. activity of the empty tube was measured. All experiments were performed at room temperature (21-23°C). To determine whether an equilibrium is reached between adsorbed and unadsorbed thrombin in TBS. TBS-PEG, and TBS-BSA, a set of tubes was prepared containing a constant thrombin concentration in the solutions under study. Several thrombin concenTubes were aspirated at varying time intervals. trations were used over a range of l-400 pg/ml (2.7 x 10-8 - 1 x 10-S M).

Vo1.37, No.1

THROMBIN ADSORPTION

203

The initial rates of thrombin adsorption from TBS and TBS-PEG were compared by aspirating tubes after relatively short time intervals (l-3 minutes for TBS, lo-30 minutes for TBS-PEG) and measuring the retained CPM. Adsorption of 125 I-thrombin was compared with adsorption of the unlabeled protein by measuring adsorption over time from thrombin preparations of equal concentration but varying 30-fold in specific radioactivity. To assess reversibility of adsorption, tubes initially equilibrated with 1251-thrombin solutions were repetitively aspirated and refilled with buffer (0.25 ml TBS or TBS-PEG). Residual radioactivity was measured after each wash. Isotherms were generated from measurements of thrombin adsorption after a 24-hour equilibration period. Total thrombin concentration varied between 5 and 370 ug/ml. The isotherms were fitted with weighted least squares regression lines with the LIGAND system of computer programs (10). From these curves were calculated adsorption equilibrium constants and capacities. The relative benefits of including PEG or BSA in thrombin buffers under more-or-less typical laboratory circumstances were studied by allowing thrombin (l-50 ug/ml) to equilibrate with the polypropylene tubes for only 30-60 minutes in TBS. TBS-PEG, and TBS with 0.20% BSA. The tubes were then aspirated and the residual CPM measured. The adsorption of thrombin to tubes previously exposed to PEG or BSA was also assessed. Tubes were filled with TBS-PEG or TBS with 1% BSA. These solutions were then aspirated to dryness before adding 1251-thrombin in TBS to the tubes. Total and residual thrombin after aspiration were quantitated in the usual manner following varying periods of equilibration. RESULTS Adsorption of thrombin to polypropylene stabilized over time (Fig. 1). Maximum adsorption from a given thrombin concentration in TBS or TBS-BSA was This difference reached more quickly than maximum adsorption from TBS-PEG. was highlighted in a comparison of the initial rates at which thrombin adsorbed from TBS and TBS-PEG (Fig. 2). With a total thrombin concentration of 2 pg/ml adsorption was approximately 12-fold faster from TBS than from TBS-PEG over the initial period when adsorption changed linearly with time. Adsorption from 75 uglml thrombin measured after 2 and 4 hours was unaffected by thrombin specific activity, i.e., the calculated values of adsorbed protein were virtually identical when the specific activity was 70 CPM/pg and 2100 CPM/ug. However, when a similar experiment was performed approximately 3 months after iodination, at a time when all thrombin clotting activity had been lost, iodinated thrombin appeared to adsorb more avidly than the native protein. Therefore, the studies reported here were done with freshly labeled thrombin with retained clotting activity. The maximum amount of thrombin adsorbed increased with increasing concentrations of unadsorbed protein, as shown in the isotherms of adsorption after 24 hours (Fig. 3). However, over the concentration range studied, adsorption from TBS-PEG was always greater than from TBS. By comparison, adsorption from TBS-BSA was less than from TBS, but the difference between the two diminished at the highest concentrations of unadsorbed thrombin.

204

THROMBIN ADSORPTION

Vo1.37, No.1

When the adsorption isotherms were fitted with least-squares regression lines and treated as representing equilibrium conditions, the curves were consistent with two mechanisms of adsorption, one saturable (shown as Scatchard plots in Fig. 4) and the other unsaturable. Equilibrium constants, K, and binding capacities, R, calculated from the saturable components of the curves for TBS, TBS-PEG, and TBS-BSA are presented in the Table. Also shown for each medium is the fraction N, representing the proportionality of thrombin adsorbed by the unsaturable mechanism to thrombin remaining in solution (unadsorbed).

TBS-PEG

FIG. 1

ts 9

:

si--

Adsorption of thrombln from TBS (o), TBS-PEG Co), or TBS-BSA (0) to polypropylene tubes over time. Data are expressed as means + 1 standard error (SE). Numbers of replicates at each time point are shown in parentheses. (A) 2 pg/ml thrombin

TBS_BSA ____n _____ _ ____ _-_--n(6) (5) I ’ /II 46 72 96 24 HOURS

(B) TBSI.) TBS-PEG ( 0 )

TBS-BSA

6

12

18

24

(0)

46

HOURS

Adsorbed thrombin could be eluted by repeated washing with TBS or TBS-PEG. Although no consistent differences were demonstrated in the patterns of elutlon with TBS and TBS-PEG. some studies suggested that thrombin adsorbed from TBS could be removed more readily than thrombin adsorbed from TBS-PEG (Fig. 5). The elution profiles typically Included an initial component which was relatively quickly removed and a later component which desorbed more slowly.

Uol.37, No.1

THROMBIN ADSORPTION

205

FIG. 2 Initial adsorption of i251-thrombin (2 Pg/ml) over time from TBS (upper panel) and TBS-PEG (lower panel) expressed as CPM/tube. Note the difference in time axes. The data have been fitted with least-squares regreasion lines. The slope of the TBS line is 52 CPM/minute. and that of the TBS-PEG line is 4.3 CPM/minute. Points are shown as means 5 1 SE. The numbers of replicates at each time interval are shown in parentheses. Points at time 0 were obtained by aspirating the thrombin solution immediately after filling the tubes (within 5 seconds).

MINUTES

MINUTES

TABLE Parameters

Calculated

From the Adsorption Isotherms TBS. TBS-PEG and TBS-BSA

(M -: )

for Thrombin

R (mol/m 2 )

in

N

TBS

1.8 x 106 (32%)

1.2 x 10-7 (23%)

.026 (14%)

TBS-PEG

5.6 x lo6 (21%)

2.0 x 10'7 (10%)

.072 (6%)

TBS-BSA

2.7 x lo5 (81%)

1.2 x 10-7 (99%)

.026 (24%)

R, adsorption capacity based upon a K, equilibrium constant. N, the fraction polypropylene-liquid interface of 1.9 cm2ltube. adsorbedlunadsorbed thrombin, attributable to a non-specific Coefficients of variation (%) are shown in parentheses.

process.

206

THROMBIN ADSORPTION

12

3

4

UNADSORBED

5

6

THROMBIN (M x

Vo1.37, No.1

7

8

9

14)

FIG 3. Isotherms for the adsorption of thrombin from TBS, TBS-PEG and TBS-BSA to polypropylene. Weighted least-squares regression lines have been fitted to the points with the LIGABD system of programs (10). Points are shown as means + 1 SE.

Studies of thrombin adsorption from TBS, TBS-PEG, and TBS with 0.20% BSA after a relatively brief incubation time (30-60 minutes) showed that 5-15% of the thrombin in TBS and TBS-PEG remained after aspiration of the tubes when the total thrombin concentration was lo-50 pg/ml (Fig. 6). Below 5 PgJml,

Vo1.37, No.1

THROMBIN ADSORPTION

207

the adsorbed fraction rose rapidly in TBS but remained at approximately 10% in TBS-PEG. On the other hand, adsorption of thrombin from TBS with 0.20% BSA never exceeded 2% regardless of the total thrombin concentration employed. Thrombin adsorption was markedly reduced in tubes which had been previously exposed to PEG or BSA. With concentrations of thrombin between 5 and 500 Pg/ml adsorption to BSA-exposed tubes was never over 2% of the total thrombin content of the tube. Studies of adsorption to PEG-exposed tubes were performed with 5 Pglml thrombin in TBS. When adsorption was followed for 19 hours and compared with adsorption to control tubes containing thrombin in TBS-PEG, no differences were observed in the rate or extent of thrombin adsorption (Fig. 7).

.15

r.

(A)

FIG. 4

0.5

1.5

1.0

ADSORBED (Mel x 1071m*) (B) .70 .60 -

\ \o \ \ \

50 -

\ \

.4Q -

\ 0’

\ \

30 -

\ \

a

-

o ‘\ ‘\O

.lOI 1.0

,

ho \ Oop 2.0

ADSORBED (Mel x W/m*)

8

Scatchard plots (adsorbedlunadsorbed thrombin vs. adsorbed thrombin) derived from the data shown in Fig. (A) Curves for TBS ::, and TBS-BSA (a). (B) Curve for TBS-PEG. Non-specific adsorption (2.6% for TBS and TBS-BSA; 7.2% for TBS-PEG) has been subtracted.

208

THROMBIN ADSORPTION

Vo1.37, No.1

DISCUSSION Thrombin adsorption to polypropylene behaves as an equilibrating phenomenon. This concept is supported by our observations that adsorption stabilizes over time (Fig. l), is proportional to the concentration of unadaorbed thrombin (Figs. 1 and 3). and is reversible (Fig. 5). Additional evidence that protein adsorption to hydrophobic surfaces reaches equilibrium is discussed by Macritchie (11). One component of thrombin adsorption to polypropylene appears to result The other apparently arises from a non-saturable from a saturable process. process, probably the deposition of protein by evaporation of buffer during aspiration of the tubes.

0

‘1 g

0

FIG 5

10,ooo~ O”O 000~

$

a ??

8

!iow-

?? e*.

o”ooooo oo”ooooooooo .~e~~.a~e.*~.~*.~~

TBS TBS-PEG

Elution of 1251thrombin adsorbed from TBS and TBS-PEG, plotted as the logarithm of CPM/tube vs. number of washes.

Because of the equilibrating and saturable nature of thrombin adsorption, the differences in adsorption from TBS, TBS-PEG and TBS-BSA were described in terms'of equilibrium constants (K) and binding capacities (R). It is seen from the Table that R is approximately the same in TBS, TBS-PEG, The increased thrombin retained in the and TBS-BSA (1.2-2.0 x low7 mol/m2). TBS-PEG tubes after aspiration (Fig. 3) is due largely to a higher contribution from the non-saturable component of adsorption (Na.072). K for adsorption in the presence of TBS-BSA is almost lo-fold less than This difference without a difference in R is K in the presence of TBS alone. indicative of competitive inhibition of thrombin adsorption by BSA.

Vo1.37, No.1

THROMBIN ADSORPTION

209

-

TBS

*--

TBS.

0.66%

PEG

O---Q

TBS,

0.20%

BSA

.m

IHTB *--_-_ ,,I,,

__--_ v ______--__________ ____________ ________ _____ ,,,, ~,,,(,,,l_*~ ,(,,, &,,,I

-l- ,,,,

,,),,,,,,

30

m

10 THROMBIN

CONCENTRATION

40

50

lrglmll

FIG. 6 Fraction of thrombin adsorbed to polypropylene tubes from TBS, TBS-PEG, BSA, after 30-60 minutes equilibration. See text for and TBS with 0.20% details.

K for adsorption from TBS-PEG, however, is 3-fold higher than K with TBS alone. Therefore, the apparent usefulness of PEG in retarding thrombin adsorption (Fig. 6) is related not to an effect in binding equilibrium or capacity, but to a reduction in adsorption rate, as indicated in Fig. 2. The apparent decrease in adsorption rate in the presence of PEG despite a modest increase in equilibrium constant must be explained by a decrease in desorption rate as well. The elution studies described here (Fig. 5) were, nevertheless, unsuccessful in demonstrating reproducible differences in desorption with and without PEG. The experimental method, however, may have been at fault since between washes the adsorbed protein was exposed to air and may have undergone changes related to the solid-gas interface which altered its behavior at subsequent solid-liquid interfaces. Of more practical value to laboratory workers are the data presented in Fig. 6 and the observations with tubes previously exposed to BSA or PEG. Both 0.20% BSA and 0.66% PEG were very effective in reducing adsorption over

210

THROMBIN ADSORPTION

Vo1.37, No.1

relatively brief time intervals (30-60 minutes). As explained above, however, adsorption progresses over time, and the effectiveness of 0.66% PEG in reducing adsorption is completely lost if the exposure period is sufficiently long to allow thrombin to equilibrate with the tube wall. Other workers have previously taken advantage of the fact that pre-treating tubes with PEG reduces thrombin adsorption (12). Our studies indicate that such pre-treatment is as effective as including 0.66% PEG in the thrombin buffer itself. The same conclusion can be drawn for BSA. Although the total concentration of PEG in the pre-treated tubes was much less than in tubes containing TBS-PEG, the effectiveness of PEG in retarding adsorption was equivalent in both circumstances (Fig. 7). This suggests that PEG coating the tubes does not elute significantly when the thrombin solution is added. The effectiveness of BSA-coating can be similarly explained and is consistent with the described tight adsorption of albumin to surfaces (11,131.

HOURS

FIG. 7 Adsorption of thrombin from TBS to tubes pre-treated with PEG (01 and from TBS-PEG to tubes which had not been pre-treated (01, over time. Data are expressed as means f 1 SE, based upon 4 measurements at each time point. Thrombin concentration is 5 ug/ml (1.3 x 107 M).

Vo1.37, No.1

THROMBIN ADSORPTION

211

In conclusion, it should be emphasized that the studies reported here are more important for the principles they demonstrate than for the quantitative data they provide. All of this work pertains only to adsorption of thrombin from 0.25 ml volumes of the designated solutions in conical polypropylene tubes at room temperature. Tube surface-to-volume ratio will have a definite bearing on the degree to which adsorption affects the thrombin available for biologic reactions in a given situation, and solution variables such as pH, ionic strength, and temperature may also influence adsorption rates and capacities (13).

ACKNOWLEDGMENTS The author wishes to acknowledge many helpful discussions with Dr. Elemer Mihalyi during the performance of this work and to thank Drs. Mihalyi, Allen Minton, and Harvey Gralnick for reviewing the manuscript. REFERENCES 1.

SEEGERS, W-H., MILLER, K.D., ANDREWS, E.B. and MURPHY, R.C. Fundamental interactions and effect of storage, ether, adsorbants and blood clotting on plasma antithrombin activity. Am. J. Physiol., 169, 700-711, 1952.

2.

WASIEWSKI, W., FASCO, M.J., MARTIN, B.M., DETWILER, T.C. and FENTON, J.W. Thrombin adsorption to surfaces and prevention with polyethylene glycol 6,000. Thromb. Res., 8, 881-886, 1976.

3.

FENTON, J.W. and FASCO, M.J. Polyethylene glycol 6,000 enhancement of the clotting of fibrinogen solutions in visual and mechanical assays. Thromb. Res., 4, 809-817, 1974.

4.

HAWK, G.L., CAMERON, J.A. and DUFAULT, L.B. Chromatography of biological materials in polyethylene glycol treated controlled-pore glass. Prep. Biochem., 1, 193-203, 1972.

5.

ATHA, D.H. and INGHAM. K.C. Mechanism of precipitation of proteins by polyethylene glycols. Analysis in terms of excluded volume. J. Biol. Chem ., 256, 12108-12117, 1981.

6.

FENTON, J.W., FASCO, M.J., STACKROW, A.B., ARONSON, D.L., YOUNG, A.M., and FINLAYSON, J.S. Human thrombins. Production, evaluation and properties of a-thrombin. J. Biol. Chem., 252, 3587-3598, 1977.

7.

Protein iodination with solid state DAVID, G.S. and REISFELD, R.A. lactoperoxidase. Biochemistry, 12, 1014-1021, 1974.

8.

LATALLO, Z.S. Thrombin clotting assays. In: Thrombosis and Bleeding Disorders. N.U. Bang, F.K. Beller, E. Deutsch, and E.F. Mammen (Eds.) New York: Academic Press, 1971, pp. 183-186.

9.

Miscellaneous analytical methods. In: Techniques BAILEY, J.L. Protein Chemistry. New York: Elsevier, 1962, pp. 340-352.

10.

MUNSON, P.J. and RODBARD, D. LIGAND: for characterization of ligand-binding 220-239, 1980.

a versatile computerized systems. Anal. Biochem.,

in approach 107,

212

THROMBIN ADSORPTION

F.

Vo1.37, No.1

11.

MACRITCHIE, 1978.

Proteins at interfaces. Adv. Protein Chem., 32, 283-326,

12.

LOTTENBERG, R., BALL, J.A., FENTON, J.W. and JACKSON, C.M. The action of thrombin on peptide p-nitroanilide substrates: hydrolysis of tos-gly-pro-arg-pNA and D-phe-pip-arg-pNA by human CL and y and bovine o and 8-thrombins. Thromb. Res. 2, 313-332, 1982.

13.

MORRISSEY, B.W. The adsorption and conformation of plasma proteins: physical approach. Ann. N.Y. Acad. Sci., 283, 50-64, 1977.

a