Low-Molecular-Weight Heparin and Dermatan Sulfate End Group-Labeled with Tyramine and Fluorescein. Biochemical and Biological Characterization of the Fluorescent-Labeled Heparin Derivative Job Harenberg, M.D.,1 Benito Casu, Ph.D.,2 Marco Guerrini, Ph.D.,2 Reinhard Malsch, Ph.D.,1 Annamaria Naggi, Ph.D.,2 Lukas Piazolo, Ph.D.,1 and Giangiacomo Torri, Ph.D.2

ABSTRACT

To improve the understanding of the biological functions and pharmacology of heparin and dermatan sulfate, low-molecular-weight heparin (LMWH) and lowmolecular-weight dermatan sulfate (LMWDS) were labeled with tyramine (T) by covalently linking T to the terminal residue of 2,5-anhydromannose (or 2,5-anhydrotalose for dermatan sulfate). The covalent labeling was demonstrated by nuclear magnetic resonance spectroscopy. The tyramine-labeled LMWH (LMWH-T) was also labeled with fluorescein (F) by further reacting it with fluorescein isothiocyanate. The fluoresceinated LMWH-T (LMWH-T,F ) was used to analyze biological functions on blood coagulation and binding to leukocytes. The biological activities on factor Xa and thrombin inhibition remained unchanged compared with the parent compound. Flow cytometric analysis of leukocytes demonstrated binding of the modified heparin to granulocytes, monocytes, and lymphocytes, the half-live being twice as long as the antifactor Xa activity. F-labeled heparin was displaced by unlabeled heparin from all three populations of leukocytes. Binding of heparin to leukocytes may play an important role in inflammation and atherosclerosis. KEYWORDS: Glycosaminoglycans, heparin, dermatan sulfate, tyramine- and

fluorescent-labeled low-molecular-weight heparin, leukocytes

Objectives: Upon completion of this article, the reader should be able to (1) describe some of the anticoagulant properties of fluorescein-labeled low-molecular-weight heparin and (2) explain the binding of this compound to leukocytes. Accreditation: Tufts University School of Medicine is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. TUSM takes full responsibility for the content, quality, and scientific integrity of this continuing education activity. Credit: Tufts University School of Medicine designates this education activity for a maximum of 1.0 hours credit toward the AMA Physicians Recognition Award in category one. Each physician should claim only those hours that he/she actually spent in the educational activity.

Glycosaminoglycans: Anticoagulant and Nonanticoagulant Actions; Editor in Chief, Eberhard F. Mammen, M.D.; Guest Editors, Job Harenberg, M.D. and Benito Casu, Ph.D. Seminars in Thrombosis and Hemostasis, volume 28, number 4, 2002. Address for correspondence and reprint requests: Job Harenberg, M.D., IV. Department of Medicine, University Hospital Mannheim, Theodor-Kutzer-Ufer, D-68167 Mannheim, Germany. Email: [email protected]. 11st Department of Medicine, Medical University Clinic, Mannheim, Germany; and 2Istituto di Chimica e Biochimica “G. Ronzoni”, Milan, Italy. Copyright © 2002 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662. 0094-6176,p;2002,28,04,343,354,ftx,en;sth00810x.

__ ls __ le 343

344

SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 28, NUMBER 4 2002

G

lycosaminoglycans (GAGs) are naturally occurring polysaccharides, the one with the most clinical relevance being heparin, which consists of repeating units of variously sulfated hexuronic acid (d-glucuronic of l-iduronic acid) and d-glucosamine.1,2 The molecular mass of unfractionated heparins ranges from 3000 to 30,000 daltons and of low-molecular-weight heparins (LMWH) from 1200 to 8000 daltons.3 Heparins exert their anticoagulant actions by enhancing the inactivation of several serine proteases of the coagulation system and by potentiating the activity of antithrombin (AT).4 They also exhibit a variety of AT-independent significant anti-inflammatory and antimetastatic activities.5,6 The antithrombotic potency is established in postoperative7,8 and general medicine9,10 as well as for the treatment of acute thromboembolic diseases.11,12 However, the AT-independent actions of GAGs are less known and currently under investigation. Other iduronic acid–containing GAGs and especially dermatan sulfate and its low-molecular-weight derivative (LMWDS) are also being considered for clinical use as antithrombotic agents with low anticoagulant activity. Labeling of heparins and dermatan sulfates by radioactivity and fluorescence is a major problem because the biological activity and receptor-mediated binding of the GAGs may be substantially modified by the labeling reactions. Therefore, we have developed a method for labeling GAGs through a tyramine residue, which was linked to the anhydrohexose group of LMWH and LMWDS by endpoint attachment,13 thus not modifying the structure of the GAG chains and permitting them to interact with biological receptors as for unlabeled GAGs. Further labeling of the tyramine moiety of LMW-T with iodine has been used to develop a sensitive binding assay14 and to study the renal and liver metabolism in animals.15 Fluorescein-5-isothiocyanate (FITC) has been specifically tagged to the tyramine group of LMWH without modifying the biological activity as compared with LMWH-T.13,16 The binding of the fluoresceinated product (LMWH-T,F ) to human leukocytes presented evidence of specific binding.17

MATERIALS AND METHODS Materials Unfractionated porcine intestinal mucosa heparin was obtained from Medac, Hamburg, Germany. LMWH with terminal 2,5-anhydromannose residues (batch 20006000), average molecular mass in daltons (Mw = 5275 Da, range 1200–10215 Da), was generously provided by Novartis Pharma GmbH, Nürenberg, Germany. Reference low-molecular-weight dermatan sulfate, heparin pentasaccharide, and LMWH dalteparin were from ls __ Alfa Wassermann, Bologna, Italy; Sanofi-Synthelabo le __ Recherche, Toulouse, France; and Pharmacia GmbH, ll __

Erlangen, Germany, respectively. Rabbit phycoerythrin conjugated antimouse CD11c was obtained from Daco (Hamburg, Germany). Tyramine (no. T 7255), fluorescein-5-isothiocyanate (no. F 7250), sodium cyanoborhydride (no. S 8628), and protamine (no. P 3880) were from Sigma GmbH (Deisenhofen, Germany). Acetonitrile was from Fison (Loughbourough, England). Antithrombin was from Behringwerke (Marburg, Germany). The chromogenic substrates S2222 (N-benzoyl-isoleucyl-L-glutamyl-(OR)-glycyl-L-arginine-p-nitroaniline hydrochloride) and its methyl ester and S2238 (H-Dphenylalanyl-L-pipecolyl-L-arginine-p-nitroanilide dihydrochloride) were from Chromogenix (Mölndal, Sweden) and Pharmacia AB (Stockholm, Sweden), respectively.

Preparation of Tyramine-Conjugated and Fluorescent-Labeled Glycosaminoglycans Tyramine (T) was bound to LMWH by reductive amination in the presence of sodium cyanoborohydride, as previously described.18 The product (LMWH-T, Mw = 6,361 Da, range 660–10,340 Da) was purified by highperformance size exclusion chromatography (HPSEC). and reacted with FITC in solution buffered with sodium hydrogen carbonate (pH 8.5) to afford the fluoresceinated derivative LMWH-T,F. The product was precipitated by ethanol, purified by HPSEC, dialyzed against distilled water, and freeze-dried (Mw = 5,775 Da, range 660–7600 Da). It was excited at 480 nm and emitted fluorescence at 515 nm. Nuclear magnetic resonance (NMR) spectra indicated that tyramine was endpoint attached to LMWH and FITC to the secondary amino group of LMWH-T. Averages of 1 tyramine molecule and of 0.8 FITC molecule per 20 disaccharide units of heparin were calculated from quantitative 1HNMR analysis.16 Tyramine-labeled LMWDS was prepared from partially N-deacetylated DS. N-Deacetylation was performed with hydrazine as previously described for the preparation of intermediates of deuterium-labeled GAGs. The resulting LMWDSs terminating with 2,5-anhydrotalose residues were reacted with tyramine in the presence of sodium cyanoborohydride under conditions similar to those used for preparation of LMWH-T to afford LMWDS-T. The product had a mean molecular weight of 3600 daltons and contained 80 antifactor Xa units/mg and 5 AT units/mg using the first international low-molecular-weight heparin standard.18

Coagulation Assays Antifactor Xa and AT activities were determined using a pool of human (Behringwerke, Marburg, Germany) or rat plasma, respectively, loaded with the first international standard for LMWH and the specific chromogenic substrates S2222 and S2238, respectively. The

TYRAMINE AND FLUORESCENT LABELED GLYCOSAMINOGLYCANS/HARENBERG ET AL

tests were performed according to standard laboratory procedures.19,20 The specific activities were for LMWH-T 110 aXa and 41 aIIa and for LMWH-T,F 66 aXa and 5 aIIa units mg-1 with human plasma.

Neutralization of Endpoint-Attached LMWH The neutralization of the LMWH preparations was performed using protamine as antagonist. The method was described before.21 Briefly, heparin and LMWHs at a concentration of 1 mg mL1 were incubated with protamine ranging from 1 to 10 mg mL1. The antifactor Xa activity of these mixtures was assayed with the chromogenic antifactor Xa assay, as described earlier.

Animal Experiments The pharmacokinetic and pharmacodynamic properties of LMWH-T,F and LMWH-T (150 aXa IU per rat corresponding to 2.1 and 1.5 mg, respectively) were compared with those of the parent compound (LMWH) and of heparin. Eight male Sprague Dawley rats (350– 550 g) were anesthetized by intramuscular administration of 0.3 mg kg1 ketamine hydrochloride and 0.04 mg kg1 diazepam (Hoffmann La Roche, Basle, Switzerland). Thereafter, ether narcosis was performed for blood sampling. Blood (0.45 mL) was taken by puncturing the retro-orbital sinus of the rats. Blood samples were drawn at 0, 10, 30, 60, 120, 240, and 360 minutes. They were collected into syringes containing 0.05 mL of 0.13 M sodium citrate solution and mixed immediately. Within 30 minutes, the samples were centrifuged at 3000 g, 20 minutes, and the plasma was shock frozen and stored at 70°C until assayed.

Calculation of the Pharmacodynamic and Pharmacokinetic Parameters The following parameters were calculated from the plasma antifactor Xa, AT activity, and the concentration of LMWH-T,F . The maximal concentration (Cmax) after intravenous application was calculated by extrapolation of the beta-elimination phase to the y-axis and the instantaneous volume of distribution was estimated as Cmax/dose. The time versus concentration curve was calculated by a linear trapezoidal method with extrapolation to infinity. The mean total clearance was obtained by the ratio of the injected dose to the area under the curve (AUC). The volume of distribution Vd was estimated as the ratio of the clearance to the slope of the terminal phase of elimination.22 The relative bioavailability was the ratio of the intravenous AUC of LMWH-T and LMWH-T,F to the intravenous AUC of LMWH. Statistical significance of the differences between the drugs was analyzed using the MannWhitney U test. The level of significance was set at p =

345

.01 and is referred to in the text as significantly different or as significant.

Reversed Phase High-Performance Liquid Chromatography (RP-HPLC) An aliquot of the plasma samples (100 L) was precipitated by bentonite (Roche Diagnostics, Mannheim, Germany). Twenty microliters of the supernatant was used for RP-HPLC analysis in duplicates. A system consisting of a Waters multisolvent delivery system with a Waters 600 E system controller, an injector, a 20-L loop, a column resolve 5 m spherical C18 3,9 .150 mm and guard-pak module with HPLC precolumn inserts C18 from Millipore Waters (Milford, MA)23 was connected between the injector and the pump. A fluorescence detector model Waters 420 AC was used, and the fluorescence signals with emission at 510 nm were recorded and integrated by PAD software from Millipore Waters. For the elution, a linear isocratic 25/75% acetonitrile-water mixture and a flow rate of 0.5 mL/min were used. The area under the absorption time curve (AUC) was integrated with the linear trapezoidal method and extrapolated to infinity. Plotting the area under the concentration time curve (AUC) versus the concentration of LMWH-T,F, a linear correlation was found. The concentration of LMWH-T,F, which is directly available for the anticoagulation of the blood, was measured by emission of fluorescence. A small amount of background fluorescence was due to the autofluorescence of the supernatant. The lower detection limit of this method was 50 g mL1 LMWH-T,F and the recovery of LMWH-T,F was 89%.

AQ1

Blood Sampling Venous blood was obtained from healthy volunteers, who did not take any medications for the preceding 10 days. All volunteers had given conformed consent prior to blood sampling. Blood was obtained by puncturing an antecubital vein with an 18-gauge butterfly without tourniquet to minimize platelet activation during blood collection. After the first 2 mL of blood was discarded, 5 mL was collected in plastic tubes containing EDTA for anticoagulation.

Preparation of Samples for Flow Cytometry One hundred microliters of anticoagulated whole blood was incubated with 10 L of LMWH-T,F at concentrations ranging from 0.01 to 100 g/mL for 15 minutes in the dark and at room temperature. Then 200 L of lysis reagent (Becton Dickinson, Heidelberg, Germany) was added and incubated for 10 minutes at room temperature and in the dark, at which time eryth- __ ls rocytes were lysed. Samples were centrifuged at 500 g __ le __ ll

346

SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 28, NUMBER 4 2002

for 5 minutes. The pellet was washed twice with 2 mL of 0.05 m tris-NaCl buffer, pH 7.4.

Flow Cytometry Analysis The fluorescence on the leukocytes was analyzed by a FAC-Scan cytometer (Becton Dickinson) equipped with a 15-mW air-cooled 488-nm argon laser. Forward and side scatters as well as green (FITC) signals were acquired by logarithmic amplification with a 585  21 nm filter for collection of FITC signals. Acquisition processing of data from 10,000 cells was carried out with the Consort 30 software (Becton Dickinson) on a Hewlett Packard 300 computer. The percentage of positive cells and the mean fluorescence intensity were calculated as median, mean, and standard deviation. Mean values and standard deviations were calculated from the data given in the tables and figures.

Displacement Experiments Dilutions of the unfractionated heparin, dalteparin, dermatan sulfate, and pentasaccharide were made from to 0.1 g to 1 mg/mL. Samples were mixed with 10 g LMWH-T,F/mL and incubated with leukocytes from 1 mL of EDTA-anticoagulated blood in the dark for 30 minutes. Fixation of cells was performed with 5% paraformaldehyde as described.24 All experiments were carried out threefold and in duplicate.

RESULTS Synthesis and Structural Characterization of Labeled LMWH and LMWDS In order to prepare tyramine- and fluorescent-labeled LMW GAGs, the strategy of end-group derivatization was followed in order to preserve the structure of the original polysaccharide chains. LMW heparin and dermatan sulfate (LMWH and LMWDS) to be conjugated were prepared by partial nitrous acid depolymerization according to established methods, but avoiding the final step of reduction commonly used to stabilize the terminal 2,5-anhydrohexose residues. To prepare LMWDS, dermatan sulfate was partially N-deacetylated with hydrazine. Different times of the hydrazinolTable 1 Molecular Weight (dalton)* of Four Samples of LMW Dermatan Sulfate

ls __ le __ ll __

Samples

Mw

Pd

1 2 3 4

3100 2400 1800 1450

1.21 1.13 1.13 1.20

*Determined by GPC-HPLC.

ysis reaction were used to afford dermatan sulfate fragments of different molecular weights (Table 1), and a fragment with Mw 3100 was chosen for conjugation with tyramine, which was performed as previously described for LMWH. The reaction schemes for preparation of LMWH-T, LMWDS-T, and LMWH-T,F are shown in Figure 1. In order to characterize the structure of LMWDS and LMWDS-T by NMR spectroscopy, model disaccharides were prepared by exhaustive nitrous acid depolymerization of both heparin and dermatan sulfate. The 1H and 13C signals of these models and of the corresponding tyramine conjugates, as assigned by two-dimensional homo- and heterocorrelation methods, are shown in Table 2a to 2c. Signals are fully compatible with the structure of L-iduronic acid 2-sulfate -linked to 2,5-anhydromannose 6-sulfate for the heparin derivative and of L-iduronic acid -linked to 2,5-anhydrotalose 4-sulfate for the dermatan sulfate derivative. For LMWDS-T, conjugation of LMWDS with tyramine is revealed by a shift of the C-1 signal of 2,5-anhydrotalose 4-sulfate from 91.8 to 52.5 ppm and of the tyramine C-8 signal from 44.0 to 51.7 ppm, confirming covalent attachment of tyramine to the reducing anhydrohexose residue.

Anticoagulant Activities The antifactor Xa and IIa activities of heparin and LMW heparins were determined in plasma. All lipophilically substituted heparins could be determined with these test systems because they were water soluble. LMWH-T exhibited antifactor Xa and IIa activity in plasma and in aqueous solution similar to that of LMWH (Table 3). The antifactor Xa and antifactor IIa activity of LMWH-T,F, however, was slightly decreased because of the further substitution.

Neutralization of Endpoint-Attached LMWH Heparin was neutralized by protamine on an equigravimetric level. Neutralization of LMWH and its derivatives was achieved with a 1.6- to 4.1-fold excess of protamine. Substituted low-molecular-weight heparins required different amounts of protamine for neutralization independent of the lipophilicity of the substitution (Table 4).

Pharmacokinetic and Pharmacodynamic Parameters of Endpoint-Labeled LMWH The endpoint-labeled low-molecular-mass heparins, LMWH, and heparin were injected into eight rats each. Table 5 summarizes the pharmacokinetic and pharmacodynamic data after intravenous administration. The antifactor Xa dose administered showed only minor differences (Fig. 2). The AT doses, however, varied between 19 to 150 units/kg1 for all LMWH species but

Figure 1 Reaction schemes of LMWH with tyramine, LMWHDS with tyramine, and LMWH-tyramine fluorescein-5-isothiocyanate.

TYRAMINE AND FLUORESCENT LABELED GLYCOSAMINOGLYCANS/HARENBERG ET AL

347

__ ls __ le __ ll

348

SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 28, NUMBER 4 2002

Table 2 Chemical Shifts of (A) IdoA-At4SO3 and IdoA-Atol4SO3 Disaccharides (ppm referred to internal standard 13C: MeOH 51.7 ppm; 1H TSP: 0 ppm), (B) IdoA-Atol4SO3-Tyrm Disaccharide (ppm referred to internal standard 13C: MeOH 51.7 ppm; 1H TSP: 0 ppm); and (C) IdoA2SO3-Amol6SO3-Tyrm Disaccharide (ppm referred to internal standard 13C: MeOH 51.7 ppm; 1H TSP: 0 ppm) (A) 1H

13C

At1

5.07

91.8

At2 At3 At4 At5 At6 At6

3.95 4.59 4.98 4.29 3.84 3.76

73.1 78.0 79.3 82.2 62.7

1H

13C

Ido1

5.02

102.4

Ido2 Ido3 Ido4 Ido5

3.62 3.70 3.94 4.59

71.6 73.8 84.8 72.8

Atol1 Atol1 Atol2 Atol3 Atol4 Atol5 Atol6 Atol6

1H

13C

3.85 3.73 4.14 4.55 5.04 4.33 3.84 3.76

63.6 82.7 77.3 80.0 81.8 62.9

1H

13C

I1

4.95

102.4

I2 I3 I4 I5

3.64 3.73 3.97 4.59

72.4 74.1 73.1 73.1

(B)

AtolT1 AtolT1 AtolT2 AtolT3 AtolT4 AtolT5 AtolT6 AtolT6

1H

13C

3.05 2.98 4.21 5.02 4.21 4.28 3.88 3.75

52.5 79.5 79.1 79.6 81.3 62.5

Ido1 Ido2 Ido3 Ido4 Ido5

1H

13C

4.88 3.62 6.64 3.94 4.56

10.2 72.7 74.0 72.96 73.0

1H

13C

5.12 4.22 4.01 4.00 4.52

101.5 77.2 71.2 71.6 71.3

Tyr2/6 Tyr3/5 Tyr7 Tyr8

1H

13C

7.22 6.89 2.84 3.0527

132.2 117.8 34.6 51.7

(C)

AmolT1 AmolT2 AmolT3 AmolT4 AmolT5 AmolT6 AmolT6

1H

13C

3.27 4.21 4.15 4.21 4.37 4.28 4.21

50.4 81.0 79.2 86.6 82.6 69.9

Ido1 Ido2 Ido3 Ido4 Ido5

showed only minor variation for the single LMWH (Fig. 3). The respective data for the extrapolated Cmax are given in Table 4. The Cmax values of LMWH and LMWH-T showed only minor differences. LMWHtyr-FITC was about twofold higher. In contrast, the Cmax did not show significant differences for the LMWH. The concentration of LMWH-tyr-FITC was also measured by RP-HPLC using fluorescence detection. The Cmax was somewhat lower (25 vs. 39 g/mL) when compared with the concentration obtained by the antifactor Xa determination (Table 5 and Fig. 4). Comparing the in vitro with the ex vivo data, it is noteworthy that LMWH-tyr-FITC displayed lower anticoagulant activities in vitro than ex vivo. Some explanations of these phenomena are mentioned later (see discussion). The AUC of the antifactor Xa activity was ls __ le __ about three- to fourfold higher after administration of ll __ LMWH-tyr-FITC (82 g/mL/h) compared with

Tyr2/6 Tyr3/5 Tyr7 Tyr8

1H

13C

7.22 6.89 2.94 3.27

132.3 117.8 32.8 50.9

LMWH (27 g/mL/h). The AT activity expressed in g/mL was rather high for the low-molecular-mass heparin because of the low aIIa activity in vitro. When calculated from the AT activity, the AUCs were similar for LMWH and LMWH-tyr (35 and 39 g/mL/h) and 10-fold higher for LMWH-tyr-FITC (340 g/ mL/h). The AUC of the concentration measured by RP-HPLC (40 g/mL/h) was similar to the antifactor Xa activity and 10% of the AUC of the AT activity expressed in g/mL/h (Table 4). The plasma clearances differed after bolus injection of heparin, LMWH, and modified LMWHs. Based on the antifactor Xa assay; the clearances of the endpoint-attached heparins were 59 to 64% lower compared with LMWH. The respective clearances based on the AT assay were 20 to 87% lower. The clearance of LMWH-tyr-FITC measured by RP-HPLC was 2-fold higher compared with the antifactor Xa clearance and 10-fold higher than for the AT clearance (Table 4).

TYRAMINE AND FLUORESCENT LABELED GLYCOSAMINOGLYCANS/HARENBERG ET AL

349

Table 3 Chemical and Anticoagulant Properties of Endpoint-Attached Labeled LMM-Heparin (LMWH-tyramine and LMWH-tyramine-FITC)*

Compound Substitution (%) Mw (Da) aXa (IU/mg) aIIa (IU/mg) Neutralization by protamine (w/w) Absorbance (nm) Sensitivity (M) Excitation Emission wavelength (nm) Color

LMWH

LMWHtyramine

LMWHtyramineFITC

5275 100 40 2.2

50 6361 110 41 4.1

40 5775 66 5 1.6

— —

200, 220 280 n.d.† n.d.

White

White

206, 220 380, 480 1.3  109 480 515 Yellow

203

*The LMWH s have a comparable average molecular mass (Mm equivalent to Mw) and anticoagulant activity. Neutralization of the antifactor Xa activity was performed by different amounts of protamine. The absorbance of heparin and LMWH was the same and differed from the modified endpoint-attached LMWH. The sensitivity of the LMWH-T,F and its maximum wavelengths for excitation and emission are shown. The color of the compounds was different. †n.d., not done.

The distribution volumes (Vd) did not differ to the same extent as the clearances. When calculated from the antifactor Xa activity, Vd did not differ substantially between LMWH and the N’-alkylamine LMWHs. In contrast, when calculated from the AT activity, Vd was lower for LMWH-tyr (41%) and LMWH-tyr-FITC (94%) compared with LMWH , respectively. The Vd measured by the concentration (RP-HPLC), however, showed a value similar to the Vd of the AT activity (274 vs. 259 mL/kg) and was about 100% higher than the antifactor Xa activity Vd (Table 5). The elimination half-lives of the endpointattached heparins were calculated. Two elimination phases were identified for LMWH and the modified endpoint-attached LMWH preparations. The alpha half-life of LMWH-tyr was significantly longer compared with the other LMWH preparations. The beta half-life of both endpoint-attached LMWH preparations was significantly longer compared with LMWH. When calculated from the AT activity, the alpha half-lives of the LMWHs did not differ. The betaelimination did not differ significantly between the LMWH preparations (Table 5). Based on the RPHPLC method, two elimination phases were calculated for LMWH-tyr-FITC. The alpha phase of elimination was 55 minutes and the beta phase was 132 minutes. The relative bioavailabilities of the endpointattached LMWH preparations were calculated versus LMWH. Calculated from the antifactor Xa activity, LMWH and LMWH-tyr displayed the same and LMWH-tyr-FITC a twofold relative bioavailability. Calculated from the AT activity, the relative bioavailabilities of the LMWH preparations did not differ.

Dose-Dependent Binding of LMWH-T,F to Leukocytes LMWH-T,F was incubated in increasing amounts from 0.01 to 100 g in 1 mL of human blood containing 6  106 leukocytes. The relative fluorescence intensity of LMWH-T,F bound to lymphocytes, monocytes, and granulocytes was analyzed by flow cytometry. The number of cells was plotted against the relative fluorescence intensity. Figure 5 shows the binding of LMWH-T,F to human granulocytes, lymphocytes, and monocytes. As can be seen, the relative fluorescence intensity was highest on granulocytes.

Displacement of LMWH-T,F by Different GAGs The displacement of fluorescent-labeled LMM-heparin has been studied using unfractionated heparin, LMMheparin, dermatan sulfate, and pentasaccharide. Increasing concentrations of these compounds were incubated with 10 g of LMWH-T,F and thereafter incubated with 1 mL of human blood. The displacement of LMWH-T,F was calculated from the mean of the relative fluorescence intensity and expressed as a percentage of the relative fluorescence intensity of 10 g of LMWH-T,F (Table 6). The percent displacements of 10 g of LMWHT,F plotted against the various amounts of unfractionated heparin and LMM-heparin are given in Figure 6. As can be seen, there are minor differences for unfractionated and LMM-heparin. Dermatan sulfate displaces to a lower extent LMWH-T,F from all three leukocyte populations (data not shown). __ ls The pentasaccharide was less effective in displac- __ le ing LMM-heparin-tyr-FITC. The dependence of the __ ll

350

SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 28, NUMBER 4 2002

Table 4 Pharmacokinetic Parameters (n = 8, mean  SD) Calculated after Intravenous Administration of 150 aXa Units of Heparin, LMWH, LMWHTyramine and LMWH-Tyramine-FITC*

AQ3

Parameter Dose injected (aXa IU ml1) aXa IU  kg1 aIIa IU  kg1 aXa g kg1) aIIa (g kg1) Cmax aXa IU  ml1 aIIa IU  ml1 aXa (g ml1) aIIa (g ml1) RP-HPLC (g ml1) AUC aXa IU  ml1h aIIa IU  ml1h) aXa (g ml1h) aIIa (g ml1h) RP-HPLC (µg ml1h) Clearance aXa (mlxkg1 xh1) aIIa (mlxkgxh1) RP-HPLC aXa (ml kg1) aIIa (ml kg1) RP-HPLC Half-life (min) Alpha-phase aXa aIIa RP-HPLC -phase aXa aIIa RP-HPLC Relative bioavailability (%) versus LMWH aXa aIIa

LMWH (n = 8)

LMWH-T (n = 8)

LMWH-T, F (n = 8)

150 303.7 ± 52.6 121.5 ± 21.0 3.0 ± 0.5 3.0 ± 0.5

150 402.1 ± 57.7 149.8 ± 21.5 3.7 ± 0.5 3.7 ± 0.5

150 249.9 ± 26.7 18.5 ± 1.5 3.8 ± 0.4 3.7 ± 0.3

1.5 ± 0.4 1.5 ± 0.8 15.0 ± 4.0 37.5 ± 20.0

1.5 ± 0.6 1.4 ± 1.0 13.6 ± 5.4 34.1 ± 24.3

2.6 ± 1.1 1.6 ± 0.9 39.4 ± 16.7 320.0 ± 180.1 25.1 ± 11.0

2.7 ± 0.9 1.4 ± 0.5 15.0 ± 4.0 37.5 ± 20.0

2.6 ± 1.2 1.6 ± 0.7 13.1 ± 5.4 34.1 ± 24.3

5.4 ± 0.8 1.7 ± 0.3 39.4 ± 16.7 320.0 ± 180.1 40.0 ± 14.2

124.2 ± 41.1 101.8 ± 33.2

51.0 ± 17.7 81.9 ± 24.7 103.2 ± 29.4 150.7 ± 43.1 152.8 ± 68.8

44.8 ± 6.7 10.9 ± 2.1

139.2 ± 53.7 259.0 ± 97.0

138.9 ± 25.5 15.3 ± 7.8 274.0 ± 54.7

51.0 ± 13.0 53.0 ± 6.0

55.0 ± 15.0 54.0 ± 27.0 55.0 ± 34.0

70.0 ± 13.0 42.0 ± 12.0

69.0 ± 26.0 67.0 ± 26.0

141.0 ± 48.0 77.0 ± 33.0 132.0 ± 88.0

25.0 ± 31.0 66.0 ± 16.0

97.9 ± 45.8 116.0 ± 51.5

199.9 ± 29.9 123.5 ± 19.8

*Cmax is the maximal extrapolated plasma level; AUC is the area under the activity and concentration time curve extrapolated to infinity. The clearance is obtained by the dose Cmax1. The Vd is the distribution volume of the beta-phase. RP-HPLC is the fluorescence measurement using reversal phase chromatography, nd: not done.

Table 5 Ratios by which 50% and 20% Displacement of LMWH-Tyr-FITC Were Obtained by Different Glycosaminoglycans on Lymphocytes, Monocytes, and Granulocytes Displacement 50% on

AQ4

ls __ le __ ll __

Displacement 20% on

Agent

Lymphocytes

Monocytes

Granulocytes

Lymphocytes

Monocytes

Granulocytes

UF-heparin Dalteparin Pentasaccharide Dermatan sulfate

0,51,0 0,6 25 2010

1,10,04 0,9 2,9 5,1

0,090,07 0,9 0,43 1,8

0,03 0,45 1,7

0,04 0,42 1,5

0,02

TYRAMINE AND FLUORESCENT LABELED GLYCOSAMINOGLYCANS/HARENBERG ET AL

Figure 2 The ex vivo antifactor Xa activity of low-molecularweight endpoint-attached heparins expressed in U mL1. They exhibit different pharmacodynamic profiles from heparin and LMWH. A bolus of 150 aXa units was administered to eight rats and the mean for eight rats plotted against time up to 360 minutes. Lipophilically, LMWH and LMWH-T,F exhibit higher antifactor Xa levels and area under the activity time curve than LMWH.

molecular weight of di-, tetra-, penta-, hexa-, octa-, and decasaccharides, LMM, and LMH-heparin from the displacement of LMWH-tyr-FITC from granulocytes is shown in Figure 6.

DISCUSSION This study compared the pharmacokinetic and pharmacodynamic properties of endpoint-attached heparins (LMWH-tyr and LMWH-tyr-FITC) with LMWH . As heparin is a heterogeneous compound concerning size, structure, and activity,25 which might have an impact on its pharmacokinetic properties, standardized doses of heparin, low-molecular-mass heparin, and endpoint-attached low-molecular-mass heparins (LMWHtyr and LMWH-try-FITC) 150 U kg1 were administered to rats in our study. Modifications of heparin by endpoint attachment are considered to differ from those obtained by substitution at the saccharide backbone.26,27 The selec-

Figure 3 The antithrombin activity of low-molecular-weight endpoint-attached heparins expressed as U mL1. They exhibit pharmacodynamic profiles similar to those of heparin and LMWH. A bolus of 150 aXa units was given to eight rats and the mean of eight rats plotted against time (360 minutes).

351

Figure 4 The concentration of LMWH-T,F measured with RPHPLC is shown in comparison with the concentration calculated from the antifactor Xa and antithrombin activity after an intravenous bolus of 150 aXa units in rats (n = 8, mean values).

tively tagged N-alkylamine low-molecular-mass heparins were anticoagulantly active in vitro and were used for metabolic and pharmacokinetic investigations. The pharmacokinetic-pharmacodynamic findings for LMWH-tyr-FITC resulted from the cleavage of heparin, forming heparin fragments with lower antifactor IIa activity. A minimal binding chain length of 17 or 18 saccharide units is required for the AT activity.28 The molecular mass distribution of LMWH, analyzed by polyacrylamide gel electrophoresis, suggested only a small proportion of material above 6000 daltons. A critical molecular mass and affinity toward thrombin may be required for the prolongation of the antifactor Xa activity because LMWHs, which are covalently

Figure 5 Binding of increasing amounts of LMWH-T,F to human granulocytes, monocytes, and lymphocytes. The relative fluorescence intensity is plotted against the concentrations of LMWH-T,F (g).

__ ls __ le __ ll

352

SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 28, NUMBER 4 2002

Figure 6 Relation of the molecular weight of different LMMheparin preparations (daltons) plotted against the molar excess to displace 50% of LMWH-tyr-FITC from granulocytes. The lowest molecular weight preparation is a disaccharide.

bound to proteins of larger molecular size, showed a prolongation of the antifactor IIa activity.29 The binding of heparin and LMWH to plasma proteins was analyzed by Young et al.30 Low-molecularmass heparins bind less to plasma proteins than unfractionated heparin.31 A butyryl derivative of heparin was reported to bind more tightly to albumin and in this way to reduce the catalytic activity of the antifactor Xa and the AT activity. The interaction of modified polyanionic compounds with protamine as a polycationic agent may provide information on structural changes during the synthesis. Therefore, neutralization experiments with the modified LMM-heparins were performed with the polycationic agent protamine. Protamine, which is extracted from fish sperm, possesses free arginine residues, giving a positive net charge to the protein surface. Neutralization of heparin by protamine occurred at a ratio of 1.1. A minimal size of 20 saccharide units is required for optimal binding. Thus, low-molecular-weight compounds show decreased binding to protamine and a larger amount is needed for neutralization.32 Lipophilic compounds can enforce the binding to proteins (protamine) by hydrophobic interactions. The compounds measured needed different amounts of protamine, ranging from a 2.1- to 4.4-fold excess of protamine for neutralization, which might be due to steric hindrance by FITC in the LMWH-tyr-FITC molecule. In the present studies, the pharmacodynamic propls __ erties of UFH and LMWH were compared with those le __ of the N’-alkylamine derivatives of LMWH. In addill __

tion, the pharmacokinetics of the LMWH-tyr-FITC derivative were analyzed by RP-HPLC with fluorescence detection. The pharmacodynamic properties of LMWH differed from those of unfractionated heparin in many respects, such as area under the concentration time curve, half-life, and bioavailability.33 For LMWH a biphasic and longer elimination was measured for the antifactor Xa and AT activity. The half-life of the antifactor Xa activity of LMWH was longer than that of heparin, in accordance with the decreased molecular mass. The results may also reflect more rapid clearance of longer chain molecules with antifactor Xa activity in rat, as suggested previously. The clearance of the AT activity of LMWH was smaller than that of the antifactor Xa activity, which may result from the differences in the binding to albumin and the fact that the AT doses differed between the heparins administered. After bolus injection of the endpoint-attached modified LMWH, the antifactor Xa profile showed similar elimination phases compared with LMWH. The half-lives of the AT activity, however, were alike for all LMWH preparations. The area under the antifactor Xa activity-time curve of endpoint-attached heparins increased, the clearances decreased, and the half-life values for the antifactor Xa activity were prolonged. This might be explained by a blocking of the scavenger receptor in the liver by lipophilically labeled heparins. When calculated from the AT activity, the elimination did not differ between LMWH and the endpointattached heparins. This finding is in accordance with data from the literature. The relative bioavailability allows comparison of the pharmacokinetic profile of an agent with that of a reference or parent substance.34 Thus, LMWH was compared with LMWH-T and LMWH-t,F. LMWHT and LMWH displayed equivalent relative bioavailability of the antifactor Xa and AT activity. LMWHT,F showed twofold higher bioavailability of antifactor Xa activity and similar bioavailability of AT activity compared with the parent compound. These results indicate that antifactor Xa activity might be liberated in vivo, which cannot be measured by in vitro test systems. These results may also be correlated with the liberation of endogenous “heparin-like” substances.35 Another explanation for the increased pharmacological response of the ex vivo antifactor Xa activity is stronger binding to AT. However, the present data suggest that the binding of modified LMWHs to albumin may also contribute to the enhanced antifactor Xa activity. The measurement of the pharmacokinetics of GAGs depends on labeling these materials with either radioactivity or fluorescence. Labeling with fluorescence did not alter the biological activity of heparan sulfates.36 The measurement of the concentration of fluorescent-labeled heparins by the emission of fluorescence in aqueous solution and in plasma enables comparison

TYRAMINE AND FLUORESCENT LABELED GLYCOSAMINOGLYCANS/HARENBERG ET AL

of the pharmacokinetic and pharmacodynamic properties of LMWH. The concentration of LMWH-T,F in plasma and the pharmacokinetic profile after intravenous administration were measured by RP-HPLC combined with a fluorescence detector. The Cmax, AUC, and the half-lives were comparable to those of the antifactor Xa activity of LMWH-T,F. The clearance and distribution volume were similar to those of the parent agent. Thus, the pharmacokinetic data of endpoint-attached fluorescent LMW-heparin supports the assumption that the long-lasting antifactor Xa activity is mainly due to unspecific binding to plasma proteins or to AT. So far, this explanation seems to be more likely than that of the release of endogenous substances with antifactor Xa activity. Several studies have shown that LMWH is a valuable anticoagulant and antithrombotic drug with higher bioavailability than heparin. The answer to these findings is not yet fully explained. The endpointattached heparins may be helpful in solving some aspects of the pharmacodynamic and pharmacokinetic differences between heparin and LMWH and their interactions with the endothelium, blood cells, and plasma proteins. They might also be of therapeutic interest because of their improved anticoagulant properties. Nonanticoagulant activities of heparins have been reemphasized37 on the basis of reports of a reduction of exercise-induced asthma with inhaled heparin.38 The mechanisms of action of this effect are unknown so far. Heparin is assumed to be phagocytized by mast cells, and this releases endogenous heparin. Direct inhibition of the stimulation of the complement system or the leukotriene pathway is also discussed. When administered by inhalation, heparin comes in direct contact with cell surfaces of the bronchoalveolar system and of mast cells. However, binding of heparin to these cells has not been demonstrated. The effect of systemically administered heparin on inflammatory and other anticoagulant-related properties of heparin are being investigated. In this respect, the present study demonstrates for the first time that heparin, low-molecular-mass heparin, and the pentasaccharide as well as the fluorescentlabeled modified low-molecular-mass heparin are bound to lymphocytes, monocytes, and granulocytes. The clinical relevance of lymphocytes, monocytes, and granulocytes in the pathogenesis of autoimmune diseases and the development of arteriosclerosis and inflammation is well documented. The beneficial effects of these cells in arteriosclerosis, inflammatory, and autoimmune diseases are well documented. The consecutive healing process leads to the formation of collagen and fibrin fibers and tissue repair. These processes destroy the original cell system and lead to secondary destruction of the organ. Inhibition of these processes may be beneficial. It can be speculated from the present results that heparin may be beneficial in

353

these senses by inhibiting the fibrous repair and promoting angiogenesis.39 Equimolar amounts of unfractionated heparin and LMW-heparin were required to displace the LMWH-T,F from the surface of leukocytes. This indicated specific binding of the negatively charged polysaccharide backbone to positively charged groups on the surface of leukocytes. Alternatively binding may also have been mediated by the lipophilic fluorescein-5isothiocyanate bound to LMWH-T,F. However, these data clearly indicate that the lipophilic substitute does not influence the binding of the synthesized LMWH to the surface of leukocytes. Dermatan sulfate has a lower degree of sulfation than heparin. Much higher amounts of dermatan sulfate are needed to displace LMWH-T,F from leukocytes. To interpret this finding, it must be considered that FITC-labeled dermatan sulfate binds about 100fold less to leukocytes than LMWH-T,F (data not shown). Therefore, it can be concluded from these results that GAGs with a low degree of sulfation interact less with the binding sites on the surface of leukocytes. The rather low rate of displacement of LMWHT,F by the synthetic pentasaccharide may be explained in two ways. First, the lower molecular mass compared with heparin results in a higher number of cellular binding sites. Second, the molecular structure of the binding sites is different from that of the AT binding sites of the pentasaccharide. Therefore, the positively charged peptide sequences differ from those of the AT binding site for the pentasaccharide. Further studies are under way to identify the sequences of binding of GAGs to the cell surfaces of lymphocytes, monocytes, and granulocytes. ACKNOWLEDGMENTS

This study was supported by the Deutsche Forschungsgemeinschaft (DFG), grant Ha 1164/3-1 and 3-2, Fakultät für Klinische Medizin Mannheim, D. Hopp Stiftung, Rosiny Stiftung, Ministerium für Wisssenschaft und Kunst, Baden-Württemberg, Germany.

REFERENCES 1. Lam LH, Silbert JE, Rosenberg RD. The separation of active and inactive forms of heparin. Biochem Biophys Res Commun 1976;69:570–577 2. Casu B, Oreste P, Torri G, et al. The structure of heparin oligosaccharide fragments with high antifactor Xa activity containing the minimal antithrombin III–binding sequence. Biochem J 1981;197:599–609 3. Harenberg J, de Vries JX. Characterization of heparins by high performance size exclusion liquid chromatography. J Chromatogr 1983;261:287–292 4. Rosenberg RD, Damus PS. The purification and mechanism of action of human antithrombin-heparin cofactor. J Biol __ ls Chem 1973;248:6490–6505 __ le

__ ll

354

SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 28, NUMBER 4 2002

5. Linhardt RJ, Loganathan D. In: Gebelein G, ed. Biomimetic Biopolymers. New York: Plenum; 1990:135–175 6. Lane DA, Lindahl U, eds. Heparin: Chemical and Biological Properties, Clinical Applications. London: Arnold; 1989 7. Kakkar VV, Cohen AT, Edmonson RA, et al. Low molecular weight versus standard heparin for prevention of venous thromboembolism after major abdominal surgery. Lancet 1993;341:259–265 8. Nurmohamed MT, Rosendaal FR, Büller HR, et al. Low-molecular-weight heparin versus standard heparin in general and orthopaedic surgery: a meta-analysis. Lancet 1992;340:152–156 9. Harenberg J, Heene DL. Pharmacology and special clinical applications of low molecular weight heparins. Am J Hematol 1988;29:233–240 10. Harenberg J, Roebruck P, Heene DL on behalf of the Heparin Study in Internal Medicine Group. Subcutaneous lowmolecular-weight heparin versus standard heparin and the prevention of thromboembolism in medical inpatients. Haemostasis 1996;26:127–139 11. Merli GJ. Low-molecular-weight heparins versus unfractionated heparin in the treatment of deep vein thrombosis and pulmonary embolism. Am J Phys Med Rehabil 2000;79:9–16 12. Harenberg J, Schmidt JA, Koppenhagen K, Tolle A, Huisman MV, Büller HR, and the EASTERN-Investigators. Fixeddose, body weight-independent subcutaneous LMW heparin versus adjusted dose unfractionated intravenous heparin in the initial treatment of proximal venous thrombosis. Thromb Haemost 2000;83:652–656 13. Harenberg J, Malsch R. German Patent P4217916.5–43, 1992 14. Harenberg J, Löhr G, Malsch R, et al. Magnetic bead protamine-linked microtiter assay for detection of heparin using iodinated low-molecular-mass heparin-tyramine. Thromb Res 1995;79:207–216 15. Stehle G, Friedrich EA, Sinn H, et al. Hepatic uptake of a modified low molecular weight heparin in rats. J Clin Invest 1992;90:2110–2116 16. Malsch R, Guerrini M, Berti C, et al. Synthesis and biological effects of N-alkylamine-labeled low-molecular-mass dermatan sulfate. Semin Thromb Hemost 1997;23:99–107 17. Harenberg J, Malsch R, Piazolo L, Huhle G, Heene DL. Preferential binding of heparin to granulocytes of various species. Am J Vet Res 1996;57:1016–1020 18. Malsch R, Guerrini M, Torri G, et al. Synthesis of N-alkylamine anticoagulant active low molecular mass (LMM) heparin for radioactive and fluorescent labeling. Anal Biochem 1994;217:255–264 19. Harenberg J. Modified anti–factor Xa chromogenic substrate assay for heparin and low molecular weight heparins. Ärztl Lab 1987;33:39–41 20. Harenberg J, Giese C, Knödler A, Zimmermann R. Comparative study of a new one-stage clotting assay for heparin and its low molecular weight derivatives. Haemostasis 1989;19:13–20 21. Alban S, Harenberg J. Neutralization of heparins and heparin-related oligosaccharides by protamine and polybrene. Pharm Pharmacol Lett 1991;1:37–40 22. Colas-Linhart N, Berthelot JL, Ducret A, Petiet A, Bok B. Technetium 99m labeled heparin: pharmacokinetics and tissue distribution in rats after vascular surgery. Biomed Pharmacother 1987;41:189–191

ls __ le __ ll __

23. Malsch R, Harenberg J. Analysis of acid mucopolysaccharides by HPLC techniques. Krankenhauspharmazie 1992;13:67– 75 24. Smith CW, Marlin SW, Rothlein R, Toman C, Anderson DC. Cooperative interaction of LFA-1 and Mac-1 with intercellular adhesion molecule-1 in facilitating adherence and transendothelial migration of human neutrophils in vitro. J Clin Invest 1989;83:2008–2017 25. Malsch R, Harenberg J. Considerations of the correct nomenclature of glycosaminoglycans. Thromb Haemost 1993;70: 718–719 26. Linhardt RJ, Hakim AA, Liu J. Acidic polysaccharides: their modifications and potential uses. In: Gebelein CG, ed. Biotechnology and Polymers. 1991:155–165 27. Hoffmann AS, Larm O, Scholander E. A new method for covalent coupling of heparin and other glycosaminoglycans to substances containing primary amino groups. Carbohydr Res 1983;117:328–331 28. Sie P, Ofosu F, Fernandez F, et al. Respective role of antithrombin III and heparin cofactor II in the in vitro anticoagulant effect of heparin and of various sulphated polysaccharides. Br J Haematol 1986;64:707–714 29. Piazolo L, Harenberg J, Malsch R, Casu B, Heene DL. Heparin-protein conjugates specifically increase thrombin inhibition in animals. Ann Hematol 1993;66(Suppl 1):Abstr 26 30. Young E, Cosmi J, Weitz B, Hirsh J. Comparison of the nonspecific binding of unfractionated heparin and low molecular weight heparin (enoxaparin) to plasma proteins. Thromb Haemost 1993;70:625–630 31. Young E, Cosmi J, Weitz B, Hirsh J. Comparison of the nonspecific binding of unfractionated heparin and low molecular weight heparin (enoxaparin) to plasma proteins. Thromb Haemost 1993;70:625–630 32. Sugiyama T, Itoh M, Ohtawa M, Natsuga T. Study on neutralization of low molecular weight heparin (LHG) by protamine sulfate and its neutralization characteristics. Thromb Res 1992;68:119–129 33. Dawes J. Comparison of the pharmacokinetics of enoxaparin (Clexane) and unfractionated heparin. Acta Chir Scand 1990; 556(Suppl):68–74 34. Stricker H, ed. In: Martin A, Swarbrick J, Cammarata A. Physikalische Pharmazie Wissenschafliche. Stuttgart: Verlagsgesellschaft; 1987:350–352 35. Vanucchi S, Pasquali F, Chiarugi V, Ruggiero M. Internalization and metabolism of endogenous heparin by cultured endothelial cells. Biochem Biophys Res Commun 1986;140: 294–301 36. Gervasi GB, Farina C, Bartoli C, et al. Disposition of a new heparan sulfate with fibrinolytic activity in rat. Drug Res 1993;43:445–449 37. Lane DA, Adams L. Non-anticoagulant uses of heparin. N Engl J Med 1993;329:130–131 38. Ahmed T, Garrigo J, Danta I. Preventing bronchoconstriction in exercise-induced asthma with inhaled heparin. N Engl J Med 1993;329:90–95 39. Folkman J, Langer R, Linhardt RJ, Haudenschild C, Taylor S. Angiogenesis inhibition and tumor regression caused by heparin or a heparin fragment in the presence of cortisone. Science 1983;221:719–725

AQ2

AQ1: Au: Correct as shown? Please clarify. AQ2: Au: Please fill in publisher and location. AQ3: Au: Ok for Column 1 heading? AQ4: Au: Ok for column heading? Au: Does this mean 0.02? Please clarify all numbers.