ANALYTICAL

BIOCHEMISTRY

200,156-162

(1992)

Spectrophotometric Method to Quantify arA Discriminate Urokinase and Tissue-Type Plasminogen Activators IU

Jijrg Schnyder, Roland Marti, Philip H. Cooper, and Trevor

G. Payne

Sandoz ResearchInstitute Berne Ltd., P.O. Box, CH-3001 Berne, Switzerland

Received

August

5,199l

Plasminogen activator and urokinase are often used as biological markers of cell activation. However, the methods currently used are cumbersome, make no discrimination between tissue-type plasminogen activator and urokinase, and do not allow expression of the results of the overall reaction in International Units. The one-step method described in this paper lacks these drawbacks. Moreover, we propose use of H-D-Val-PheLys-4-nitroanilide as substrate which has a lower K,,, than the standard H-D-Val-Leu-Lys-4-nitroanilide which is commercially available. Low concentrations of sodium dodecyl sulfate in the reaction mixture dramatically and preferentially accelerate the reaction catalyzed by tissue-type plasminogen activators. Identical results are obtained under kinetic or fixed-time assay conditions using either a photometer or 96-well plate reader. The corresponding formulae are provided. 0 1992

Academic

Press,

Inc.

Plasminogen activators are widely distributed throughout the tissues and fluids of the body and play important roles in physiological and pathological processes such as hemostasis, inflammation, and malignancy (reviewed in (1)). As such, plasminogen activator (PA)’ is a useful marker in many biological systems. For example, macrophages release PA upon activation either in uiuo or in vitro (2-3), secretion of PA from rabbit chondrocytes is decreased following addition of interleukin-1 (4), calcii Abbreviations used: BMDM, bone marrow-derived macrophages; BSA, bovine serum albumin; IU, International Units; NaClIP,, phosphate-buffered saline; PA, plasminogen activator (used in this paper for all enzymes which catalyze the conversion of plasminogen to plasmin); PLG, plasminogen; PUK, prourokinase; SBTI, soybean trypsin inhibitor; SDS, sodium dodecyl sulfate; t-PA, tissue-type PA; UK, urokinase.

tonin stimulates PA production by renal tubular cells (5), and tumor necrosis factor induces PA secretion by endothelial cells (6). The methods currently used to quantify PA andurokinase (UK) in biological samples are generally cumbersome and inherently inaccurate because no account is made of the nonlinear reaction kinetics (7-9). In this paper, a simple, rapid method is described that allows quantification of these enzymes in a kinetic or fixedtime (end point) assay. The data obtained further allow the expression of activity in International Units (IU). Moreover, inclusion of small amounts of sodium dodecyl sulfate (SDS) in the reaction mixture allows a simple discrimination between different PAS, since SDS, like fibrin(ogen) accelerates the activity of tissue-type PA dramatically. MATERIALS

AND

METHODS

Cell cultures. Resident and thioglycollate-elicited mouse peritoneal macrophages were collected and cultured in medium containing 1% acid-treated serum as described earlier (2). Bone marrow-derived macrophages (BMDM) were grown in the presence of colonystimulating factor for 10 days in Teflon bags (10). The cells were then incubated at 37°C in 5% CO,/95% air for 16 h in medium containing no serum in the presence of 100 pg/ml concanavalin A. The media were then removed from the monolayers and clarified by centrifugation at 1OOOg for 10 min, and the cells were lysed in 0.01% digitonin. Rabbit chondrocytes were prepared and cultured as described elsewhere (4). Reagents and preparations. Plasminogen (PLG) was isolated from human serum by affinity column chromatography on lysine-Sepharose by a modification of the procedure described by Deutsch and Mertz (11). By increasing the concentration of the plasmin inhibitor tamino-n-caproic acid (Serva Feinbiochemica, Heidelooo3-2697/92

156 All

Copyright 0 1992 right8 of reproduction

$3.00

by Academic

Press, Inc.

in any

reserved.

form

DETERMINATIONS

OF PLASMINOGEN

berg, Germany) in the elution buffer to 0.5 M, a stable preparation of PLG was obtained. Because of the high concentration of c-amino-n-caproic acid (which mayprecipitate out in the cold), the whole procedure was performed at room temperature. No attempt was made to separate “Glu-PLG” from “Lys-PLG”. Fibrinogen was purified from Cohn fraction I (Swiss Red Cross, Berne, Switzerland) by the method of Laki (12). The calorimetric substrates for plasmin, H-D-ValLeu-Lys-4-NA, H-D-Ile-Phe-Lys-4-NA, and H-D-ValPhe-Lys-4-NA were synthesized by conventional peptide chemistry and used as the trifluoracetate salts. In initial experiments, commercially available substrates were used, i.e. H-D-Val-Leu-Lys-4-NA (Kabi Diagnostica, Stockholm, Sweden or Ortho Diagnostics, Raritan, NJ), and H-D-Ile-Phe-Lys-4-NA (Bachem, Bubendorf, Switzerland). Urokinase, which was used as a standard, was obtained from Leo Pharmaceutical Products (Ballerup, Denmark) (1700 Plough units) or as Ukidan from Serono Pharmazeutika (Basel, Switzerland) (100,000 CTA units). Sodium dodecyl sulfate (electrophoretic purity reagent) was obtained from Bio-Rad (Richmond, CA) and soybean trypsin inhibitor was from Sigma (St. Louis, MO). Lys-PLG and Glu-PLG were purchased from American Diagnostica Inc. (American House, New York), one- and two-chain t-PA from Biopool AB (Umea, Sweden), and prourokinase (PUK) was obtained from Sandoz (Basel, Switzerland). Preparation of fibrin(ogen)-coated wells and vials. To prepare fibrin-coated wells, a stock solution (10 mg/ml) of purified fibrinogen in 0.3 M KCl, containing 60 mM Tris-HCl, pH 7.6 (Fluka, Buchs, Switzerland), was diluted with water to 20 pg/300 ~1. This volume was added to 24-well culture plates (Flow Laboratories, Irvine Ayrshire, Scotland), which were kept for 3 days in a 45°C incubator and then stored at 4’C. Before use, 1% fetal calf serum in Dulbecco’s saline (NaClIP,; Fakola AG, Basel, Switzerland) was added and the plates were incubated overnight at 37°C to convert fibrinogen to fibrin. The plates were then rinsed several times with NaClIP,. Fibrinogen-coated vials were prepared by adding 20 pg of purified fibrinogen to 1.5-ml vials (Milian, Geneva, Switzerland) which were kept for 2-4 days at 42-45°C and then stored at 4°C. Assay system. time or a kinetic

PA is determined using either a fixed method. Fixed-time assay: 25 ~1 of cell extract, medium, or enzyme were mixed with 500 ~1 of 50 mM Tris-HCl, pH 8.5, containing 1 mM H-D-Val-Leu-Lys-4-NA or 0.6 mM H-D-Val-Phe-Lys-4-NA, 2.5 pg plasminogen and, where indicated, SDS. Following incubation at 37”C, the reaction was stopped by the addition of 400 ~1 of 50 mM Tris-HCl, pH 8.5, containing 4 pg of soybean trypsin inhibitor. Absorbance was then measured at 405 nm

ACTIVATORS

(Photometer many).

AND UROKINASE

6120, Eppendorf

157 GmbH,

Hamburg,

Ger-

Kinetic assay: samples of 20 ~1 were pipetted into a 96-well plate (Inotech AG, Wohlen, Switzerland) and the reaction was started by addition of 200 ~1 of 50 mM Tris-HCl, pH 8.5 containing PLG, SDS, and the chromogenic substrate at the concentrations indicated above or in the legends to the tables and figures. The plate was then put into an eight-channel photometer (Twinreader, Flow Laboratories) connected to a desk computer. For kinetic analysis, the optical density of each of the 96 wells was measured at fixed-time intervals. The data were stored for calculation of the activity by the unweighted linear regression analysis described below. In both assays, control mixtures were run by omitting the sample, PLG, SDS, or all three. Determination of units. The quantification of PA is indirect, since it is the second step of the reaction, plasmin-dependent hydrolysis of a synthetic substrate, which is monitored. The overall reaction rate is a function of time squared (13,14). A straight line with a slope corresponding to half the acceleration rate is obtained when the absorbance is plotted against the square of the incubation time. In order to express enzyme activity in IU (pm01 of substrate hydrolyzed per min), which depends on the rate of hydrolysis rather than on the acceleration rate, the velocity curve was calculated as the first differential of the quadratic function (Fig. la). The initial rate of hydrolysis equals twice the acceleration rate. The formulae for both horizontal and vertical light path readers are: Photometer (horizontal light path)

PUlml

=

2 X AA X assay volume t X d X t2 X sample volume

96-well plate reader (vertical rUlm1

=

x log

[l]

light path)

2 X AA X area f X t2 X sample volume

x 106,

PI

where d = 1 cm, E= 10,054 M-’ X cm-‘, area = 0.343 cm2, A = absorbance. From both formulae it can be seen that in 96-well plate readers the value for enzyme activity is independent of the assay volume. For Eq. [2] the standard formula for readers with vertical light path is included: A = t X 1 X c, 1 = vo&/area, mole =

c = mole/vol,

A X area X vol, t x vol, *

It should be kept in mind that vol, is in milliliters [cm31 and vol, is in liters or [dm3].

131 or

158

SCHNYDER

ET AL.

Time(h21

Time(h)

FIG. 1.

Theory vs Praxis. (a) Plot of tion (squares) and the first differential thioglycollate-elicited macrophages (2) mM Tris-HCl, pH 8.2, containing 1 mM with 400 ~1100 mM Tris-HCl, pH 8.2,

the mathematical function y = 0.1 f. The parabolic quadratic curve (circles), the curve after linearizaof time Aylhr (line) are given. (b) Linearization of the quadratic time curve. Conditioned media from were used as a source of PA; 25 gl PA was incubated at 37Y! for 0 to 4.5 h with prewarmed 500 ~1100 H-n-Val-Leu-Lys-4-NA, 0.003% SDS, and 2.5 pg PLG. At indicated times, the hydrolysis was stopped containing 4 rg SBTI.

When kinetic readings instead of end-point measurements are made, AA/t2 will be the slope of the linear regression curve.

mally have higher affinity for the enzyme (17), we used H-D-Val-Phe-Lys-4-NA in all further experiments. Linearization

RESULTS

Substrates for Plasmin

The determination of PA is indirect because the enzyme converts PLG’ to plasmin, and it is the subsequent plasmin-dependent hydrolysis of a second substrate which is monitored. Fibrin or casein are often used, but are unspecific substrates. Low-molecular-weight substrates for plasmin are commercially available and the Kabi substrate S 2251 is a well-recognized standard (14,15). We compared the kinetic parameters of S 2251 with two other substrates, H-o-Val-Phe-Lys-4-NA and D-Ile-Phe-Lys-4-NA, which had been proposed by the group of Eliidi (16). Using a standard UK as a source of PA, it was shown by use of the Hanes-Woolf plot (17) that the K,,, values of these substrates are about five times better than that of the Kabi standard substrate, whereas the V,, values are similar (Fig. 2 and Table 1). Since substrates with lower apparent K, values nor-

of the Absorbance Curves

As mentioned above, the method of measurement of the activity of PA is indirect, since plasmin-dependent hydrolysis of a chosen peptide substrate is monitored. The interpretation of the kinetics is also complicated by the fact that autocatalytic reactions can take place. A plot of absorbance vs time obtained in an assay of PA present in the culture medium of elicited macrophages is parabolic and a straight line is obtained by plotting the absorbance vs square of time (Fig. lb). The slope of the latter curve corresponds to half the acceleration rate of the reaction, which was shown by others (14) and by us (4) to be a linear function of the enzyme concentration. As shown in Table 2, almost identical activities (expressed in IU) were found, using either kinetic or fixedtime readings, and were independent of the method or apparatus used to determine the activity of the PA. From Table 2, one can see that the standard deviations of five determinations within a single experiment from

DETERMINATIONS

OF PLASMINOGEN

ACTIVATORS

159

AND UROKINASE TABLE

Kinetic s/:

Parameters

Val-Leu-Lys-4-Na

of

1

Three Selected Substrates

Val-Phe-Lys-4-NA

Ile-Phe-Lys-4-NA

K,,, values 0.23 f 0.05 (4)

2x10-3-

V,,

0.062

f 0.021

(4)

717 2 81 (4)

Note. The experiments were performed as described to Fig. 2. Means and SE from N experiments (numbers triplicates are given. K,,,, Michaelis constant (mM); velocity of enzyme reaction.

10-b

5x10-h

10-3

s

Hanes-Woolf plot: 500 pl 50 mM Tris-HCI, pH 8.2, conFIG. 2. taining different concentrations of the three chosen substrates (0: H-D-Val-Leu-Lys-4-NA, A: H-D-Ile-Phe-Lys-4-NA; 0: H-n-ValPhe-Lys-4-NA) for plasmin in the presence or absence of 2.5 pg PLG was prewarmed in 1.5-ml vials. The reaction was started by adding 25 al culture medium-0.1% BSA, containing 1 Plough U/ml UK, and after 60 to 90 min the hydrolysis was stopped with 400 al 50 mM Tris, pH 8.2, containing 4 pg SBTI. The reaction rate is expressed in pU/ml using the formula given under Materials and Methods. One selected experiment is shown. For means and SE of K,,, and V,, see Table 1.

Acceleration of the Overall Reaction by SDS and Establishment of Standard Assay Conditions Using a diffusion assay, Saksela (18) showed that SDS dissociates PA from inhibitors. We tested the effect of this detergent in our incubation system and found that it strongly enhanced the hydrolysis of the chromogenic substrate by PA in conditioned media from BMDM, but activity in the lysates, which was already detectable at a high level, was only slightly stimu-

322 in the legend in brackets) in V-, maximum

lated (Fig. 3). In the same figure, the various parameters used to establish optimal conditions for PA in conditioned media or cell lysates of stimulated BMDM are shown. In a further series of experiments, we compared the effect of SDS on PA with that of fibrinogen. We coated the vials normally used in the endpoint assay with 20 pg fibrinogen. As shown in Table 3, the hydrolysis of the synthetic substrate by PA from both stimulated mouse peritoneal macrophages and BMDM was enhanced in fibrinogen-coated vials, while there was no effect on UK activity. Moreover, the effect of 0.003% SDS added either to coated or uncoated vials on PA from macrophages was much more dramatic and the combination of fibrinogen and SDS led only to a minor increase over that observed by SDS alone. Since it is known that UK and t-PA have different affinities to fibrin(ogen) and their kinetic parameters are changed by fibrin(ogen) (19), further experiments were performed using commercially available UK and

TABLE

both methods are low (l.l-1.4%). When only three readings were performed, which for capacity reasons are normal, the corresponding value rises, as expected, to 2.5%. With this lower number of readings, we observed a standard error between three different experiments of 3.1%.

(2)

values

544 f 148 (4)

10-3-

0.050

Fixed-Time

Measurements

Fixed-time assay (&J/ml)

Reader Photometer 96-Well plate

vs Kinetic

2

reader

1268 + 17.7 1307 * 14.7

Kinetic assay (pU/ml)

1231 + 16.6

Note. For the fixed-time and kinetic assays in 96-well plates, 20 al UK (6 Plough U/ml) was added to 200 al50 mM Tris-HCl, pH 8.5, containing 0.6 mM H-D-Val-Phe-Lys-4-NA and 1 pg PLG. The end points were measured at 44 min, whereas 8 cycles per 4 min were performed in the kinetic assays. For the fixed-time readings using a conventional Eppendorf photometer, the assay volume was doubled and the reaction in 5-ml vials was stopped with 400 ~1 50 mM TrisHCl pH 8.5, containing 3.6 ag SBTI. All the experiments were run at the same time to avoid changes due to variation in room temperature (see section on the effect of temperature). The formulae given under Materials and Methods were used to calculate the units. Means (pU/ ml) and SD from five parallel readings are given.

160

SCHNYDER

0.6-

ET

AL. TABLE

a

3

Acceleration of Plasminogen Activator-Dependent Hydrolysis of H-Val-Leu-Lys-4-NA by Fibrin or SDS Uncoated Sample

b:

5 to PLASHINOGEN

15 20 &ml

PA from peritoneal macrophages PA from BMDM Urokinase

SOS pglml

O6.5

7.5

6.5

9.5

PH 2.5

Plus

Fibrin-coated

SDS

No SDS

wells Plus

SDS

E-405

1

0

No SDS

wells

f

0.029 0.055 0.299

1.087 2.370 nd

0.351 0.327 0.259

1.403 2.400 0.344

Note. Conditioned media from BMDM (10) or peritoneal macrophages, which had been stimulated in vitro with phagocytosis of zymosan particles (24), were used as a source of PA. Urokinase, 0.1 CTA U/ml, was used as a reference. A 50-~1 sample and 500 ~1 of 100 mM Tris-HCl, pH 8.2, containing 1 mM H-D-Val-Leu-Lys-4-NA, 2.5 pg PLG, and 0.003% SDS (where indicated), were added to uncoated or coated wells (see Materials and Methods). After 4 h at 37”C, the hydrolysis was stopped by adding 400 pl of 100 mht Tris-HCl, pH 8.2, containing 4 gg SBTI. Means (E405 nm) from triplicate determinations are given.

1

2.0 1.5 1.0 0.5 0 I

0

FIG.

10

I

,

20 30 40 SAMPLE pL

1

50

K 0

10

20

t2

30 40 (HOURS)2

to commercially available Lys-PLG, indicating that at least this form was present in the mixture. Effect of Temperature and Sensitivity

of the Assay

50

3.

Establishment of optimal conditions for the assay of BMDM PA. BMDM were cultivated and monolayers prepared as described under Materials and Methods. The standard assay for PA consists of H-D-Val-Phe-Lys-4-NA (0.5 mM for the medium, 1.0 mM for the cell lysates) plasminogen 5 pg/ml, SDS 10 pg/ml, 50 mM TrisHCl, pH 8.5, in a final volume of 250 ~1 and the reaction was started by addition of 10 ~1 sample. After 60 min of incubation, the reaction was stopped by addition of 200 ~150 mM Tris-HCl, pH 8.5, containing 10 pg/ml SBTI and the absorption measured at 405 nm. The curves show the effect of varying (a) substrate, (b) SDS and (c) PLG concentration, (d) pH, (e) sample volume, and (f) incubation time. (0) activity in cell lysates, (0) activity in medium.

Since the incubation temperature within the Twinreader can vary between 25 and 3O”C, two fixed-time experiments were performed using a water bath to de-

TABLE Effect

of Glu-PLG, Lys-PLG,

and SDS on Reaction

Glu-PLG

Enzyme

t-PA, both of which occur either as single-chain or twochain enzymes (1). Single-chain UK is called either prourokinase (PUK) or single-chain urokinase-type PA. All these enzymes act on plasminogen, of which two forms exist: Glu-PLG and the shorter form Lys-PLG, which has a high binding affinity to fibrin (19). The results in Table 4 show that in the absence of SDS, GluPLG is a poor cosubstrate for UK, PUK, and single- or two-chain t-PA, and Lys-PLG is a good cosubstrate for UK and PUK but not for single- and two-chain t-PA. SDS was shown to stimulate the activity of all four enzymes, but to a different extent depending on whether Glu- or Lys-PLG was used as cosubstrate. PLG isolated by affinity chromatography, but not further purified to separate the Glu-PLG and Lys-PLG, behaved similarly

4

No SDS

PLG

Lys-PLG Plus SDS

No SDS

Rate

Plus SDS

No SDS

Plus SDS

167

3363

~Ulilll Single-chain PA 250 U/ml Two-chain PA 50 U/ml UK 3 U/ml Pro-UK 200 nglml

226

3417

34.7

3170

0.0

2729

63.6

2530

921

2829

909

2593

91.8

1660

1317

2719

960

2474

56.8

2676

71.7

2641

Note. Singleand two-chain t-PA, UK, and PUK were used as source of enzymes. A 20-/J culture medium-0.1% BSA with enzyme was added to 200 ~1 of 50 mM Tris-HCl, pH 8.5, containing 0.6 mM H-D-Val-Phe-Lys-4-NA, 1 pg Glu-PLG, Lys-PLG or PLG purified with affinity chromatography (see Materials and Methods) and 0.002% SDS (where indicated). Eleven cycles per 4 min were measured and the units (pU/ml) calculated as described under Materials and Methods. Means from one to three similar experiments with five

parallelreadingsare given.

DETERMINATIONS

OF

PLASMINOGEN

ACTIVATORS

AND

UROKINASE

161

directly to calculate units. However, the first differential of time of the quadratic function gives the velocity of the hydrolysis, which mathematically equals twice the value of the acceleration rate (Fig. la). For that reason, all formulae given to express pU/ml for the classical photometer or 96-well plate readers with vertical light path contain the factor 2. To our knowledge this is the first description of a method which allows the direct expression of PA activities in IU. The type of PA secreted by monocytes and macrophages is still debatable, since UK (21), PUK (22), and DISCUSSION t-PA (23) have all been described by different laboratoH-D-Val-Leu-Lys-4-NA is widely used as a substrate ries. In our hands, macrophage PA appears to bind to for plasmin in clinical chemistry and biology to monitor fibrin, since the PA in conditioned medium cannot be transferred from fibrin-coated wells (not shown). MorePA (8,9,15,20). However, Eliidi synthesized tripeptidyl4-NA substrates for plasmin and proposed other strucover, the activity is dramatically stimulated by fibrin tures, which he predicted should exhibit much improved using a synthetic substrate for plasmin. However, when kinetic parameters (16). We synthesized two of these we first tried to measure PA using the low-molecularweight substrate H-D-Val-Leu-Lys-4-NA, no activity substrates, H-D-Val-Phe-Lys-4-NA and H-D-Ile-PheLys-4-NA, and compared their kinetic parameters with could be observed, whereas a standard UK preparation, those of the standard substrate II-D-Val-Leu-Lys-4which was equally as active in the fibrin-plate method as NA, since no such direct comparison had been made by PA from macrophages, cleaved the synthetic substrate ElSdi. In our hands, the two new substrates displayed rapidly. By adding the H-D-Val-Leu-Lys-4-NA to consimilar kinetic parameters (Fig. 2, Table 1). In contrast ditioned medium on fibrin-coated wells, PA in the conditioned medium from macrophages amidolyzed the to the speculations of Elodi, H-D-Val-Phe-Lys-4-NA and D-Ile-Phe-Lys-4-NA have similar K,,, values, both substrate efficiently (Table 3). All these observations being about five times lower than that for H-D-Valindicated that PA released from mouse macrophages is Leu-Lys-4-NA, and the V,, values were similar for all different from UK. three substrates. Due to its favorable kinetic parameSaksela (18) reported that SDS might dissociate PA ters and ease of chemical synthesis, H-D-Val-Phe-Lysfrom inhibitors, which are normally present in condi4-NA was chosen as the standard substrate for plasmin. tioned media from phagocytes (21,22), and the results In addition, it was found that using H-D-Val-Leupresented here show that SDS is indeed a very powerful Lys-4-NA as substrate, the temperature coefficient Q1,, stimulator of the overall reaction. However, it is not for PA has a value of 5, indicating that the reaction rate possible to deduce whether SDS at low concentration is more sensitive to temperature change than most en- dissociates enzyme-inhibitor complexes or whether it zymes which have a Q10 in the range l-2 (17). This indiinduces conformational changes in the activator (t-PA cates that accurate control of the incubation temperaor UK) or cosubstrates (PLG, plasmin). ture is important in order to obtain reproducible results To validate our assumption that t-PA, but not UK(Table 2). like PA, is stimulated by SDS, we could show that when Having chosen the appropriate substrate to deteraffinity chromatography-purified PLG or Lys-PLG are mine PA, it was necessary to find a way of transforming used as substrate, SDS stimulates one- and two-chain the nonlinear kinetics of the overall reaction comprising t-PA dramatically, whereas little effect is seen on UK or the activation of the latent proenzyme PLG to plasmin PUK activities which already exhibit high activity in the by the activators and the reaction of plasmin with both presence of this cosubstrate. However, Glu-PLG is a its substrate and with PLG (autocatalysis). As shown by very poor cosubstrate for all PAS in the absence of SDS. others (13,14) and ourselves (4, and the present paper), Hence it is possible to deduce whether t-PA or UK are the overall reaction rate is a function of time squared present in the test samples by examining the effects of and linear curves can be constructed either by plotting SDS on the enzymic activity using Lys- and Glu- PLG the square root of extinction vs time or by extinction vs substrates. time squared (Fig. la). In the first case one is left with To summarize, the method we propose in this paper the problem of how to substract blanks, i.e. fl has the following advantages over existing procedures: VBlank or YE?%& In the second preferred case, the (i) The use of a chromogenic substrate for plasmin acceleration rate for both the blanks and the test sammakes it possible to circumvent the classical fibrinples of the reaction can easily be determined. Since plate method, which is both cumbersome and has no standard IU are expressed as micromoles of substrate means to express units of activity directly. (ii) Of the cleaved per minute, the acceleration rate cannot be used three substrates tested, H-D-Val-Phe-Lys-4-NA is the

termine the effect of temperature on the reaction rate. Using the Arrhenius plot (not shown), the temperature coefficient Q1,, was determined (17) and was estimated to have a value of 5. The reaction rate was also shown to be linear over a range of four log units of UK (0.01 to 100 Plough units/ml, not shown); 0.01 Plough unit/ml was found to be the detection limit when 11 cycles over 30 min were run in a kinetic assay system. For higher units of UK, the cycle time was reduced to 1 min.

SCHNYDER

162

best substrate since it has the highest V,,/K,,, ratio and is also more economical to use since its Km value is less than that of the standard H-D-Val-Leu-Lys-4-NA, and hence less of the expensive substrate needs to be included in the reaction mixture. (iii) The results can be expressed as IU, since the first differential of time of the quadratic function represents the velocity of the reaction. (iv) Low concentrations of SDS accelerate dramatically and preferentially the activity of t-PA using Lysor Glu-PLG as cosubstrate, whereas UK- or PUK-dependent hydrolysis is stimulated only markedly when Glu-PLG is used as cosubstrate. Therefore, by using combinations of different PLG and SDS, one can infer whether t-PA or UK is present in the test samples. (v) Identical activities are obtained by fixed-time or kinetic readings by using either a classical photometer or a 96well plate reader with vertical light path. The corresponding formulae are given. Using the kinetic method, where the computer calculates the initial velocity from 11 readings, very precise measurements can be obtained which are comparable to those obtained by one fixedtime measurement with a conventional photometer. When different time intervals are chosen, the method allows determination of UK over a four-log range of Plough units. Considering all these advantages, we hope that by using the method described in this paper biologists and clinicians will have less problems measuring, discriminating, and quantifying UK or t-PA.

J., and Baggiolini,

Proc. Natl. Acad. Sci. USA

M. (1980)

77,414-417. 4. Evtquoz, Fleisch,

V., Schnyder, J., H. (1984) Bioc/rem.

Trechsel,

U.,

Baggiolini,

M.,

and

J. 219,667-677.

5. Dayer,

J. -M., Vassalli, J. -D., Bobbitt, and Krane, S. M. (1981) J. Cell Biol.

L., Hull,

R. N., Reich,

E.,

91, 195-200.

6. van Hinsbergh, V. W. M., van den Berg, E. A., Fiers, W., and Dooijewaard, G. (1990) Blood 75,1991-1998. 7. Marsh, N. A., and Gaffney, P. J. (1977) Thromb. Haemostasis 38, 545-551. 8. Searls,

D. B. (1980)

9. Leprince,

P., Rogister,

Anal. Biochem. 107,64-70. B., and Moonen, G. (1989)

Anal. Biochem.

177,341-346. 10. Cooper,

P. H., Mayer,

P., and Baggiolini,

M. (1984)

J. Zmmurwl.

133,913-922. 11. Deutsch, 12. Laki,

D. G., and Mertz,

K. (1951)

13. Christensen,

E. T. (1970)

Science

170,1095-1096.

Arch. Biochem. Biophys. 32,317-324. U., and Mtillertz, S. (1977) Biochim. Biophys. Acta

480,275-281. 14. Drapier,

J. C., Tenu,

J. P., Lemaire,

G., and

Petit,

J. F. (1979)

Biochimie 61,463-471. 15. Campbell, E. E., Shifman, Pizzo, S. V. (1982) Clin. 16. Cs.-Szabo, 199-206.

G., Pozsgay,

M. A., Lewis, M., and Elodi,

P. (1980)

17. Dixon, M., and E. C. Webb (Eds.) (1979) pp. 47-206, Academic Press, San Diego. 18. Saksela, 19. Collen,

0. (1981) D. (1980)

J. G., Pasqua,

J. J., and

Chem. 28, 1125-1128. Thromb. Res. 20,

The Enzymes,

third

ed.,

Anal. Biochem. 111,276-282. Thromb. Haemostasis 43,77-89. M., Monsigny,

M.,

and Mayer,

R. (1991)

Anal.

21. Saksela, 125-132.

O., Hovi,

22. Vassalli,

J.-D.,

T., and Vaheri,

Dayer,

J.-M.,

A. (1985)

Wohlwend,

J. Cell. Physiol. 122,

A., and Belin,

D. (1984)

J. Exp. Med. 159,1653-1668.

REFERENCES

450.

3. Schnyder,

Biochem. 193,248-255.

We thank Ms. Agathe Wild, Mr. Eugen Jenni, Mr. Bruno Streit, and Mr. Roman Bischofberger for their excellent technical assistance and Dr. A. MacKenzie for help in preparing the manuscript.

0. (1985) Biochim. J., and Baggiolini,

AL.

20. Harnois-Pontoni,

ACKNOWLEDGMENTS

1. Saksela, 2. Schnyder,

ET

Biophys. Acta. M. (1978) J.

823, 35-65. Erp. Med. 148, 435-

23. Hart, P. H., Vitti, G. F., Burgess, D. R., Singleton, Hamilton, J. A. (1989) J. Exp. Med. 169,1509-1514. 24. Schnyder, 1457.

J., and Baggiolini,

M. (1978)

D. K., and

J. Exp. Med. 148,1449-