Polymeric inhibitors of platelet aggregation. II. Copolymers of dipyridamole and related drugs with N-vinylpyrrolidone

Bmchimica et Biophysica Acta 886 (1986) 109-118 109 Elsevier

BBA11687

Polymeric inhibitors of platelet aggregation. Synergistic effects and proposals for a new m e c h a n i s m

C l e m e n t H. Bamford , f an P. M i d d l e t o n a n d K a d e m G. A I - L a m e e

The Institute o/Medical and Dental Bioengineering, University of Liverpool, Duncan Building, Royal Liverpool Hospital, P.O. Box 147, Liverpool, L69 3BX (U.K.)

(Received August 1st, 1985) (Revised manuscript received November 15th, 1985)

Key words: Platelet aggregation; Aggregation inhibition; Polymer coupling; Prostaglandin analog

We have studied the inhibition of ADP-induced platelet aggregation in sheep platelet-rich plasma by water-soluble polymers bound to the prostaglandin analogue 5-(6-carboxyhexyl)-l-(3-cyclohexyl-3-hydroxypropyl)hydantoin ('BW 245' C, (I)). The use of unambiguous modes of binding this antiplatelet drug to polymers has enabled us to study some structural features which influence inhibitory activity. Evidence is adduced which indicates that the chemical mechanisms responsible for inhibition by free and coupled BW 245 are similar. The most important observation is a remarkable synergism demonstrated by the greatly enhanced activity of a mixture of a polymer coupled to BW 245 with the uncoupled parent polymer. In some cases (e.g., with high-molecular-weight dextran) the effect may reach (and possibly exceed) two orders of magnitude. The influence of polymer molecular weights and 'cross-polymer' effects have both been examined. A mechanism has been proposed to account for these phenomena, involving adsorption of the added (inactive) polymer on to the platelet membranes, facilitating interaction of the polymer-bound drug with receptors, made more accessible by alteration to the surface geometry. This mechanism is based on physical processes, unlike other explanations of synergistic behaviour, e.g., that of prostaglandins used in conjunction with non-polymeric drugs. The observed dependences of synergistic effects upon polymer molecular weight and type and distribution of drug molecules along chains are typical 'polymer' phenomena which are all consistent with the proposed mechanism.

Introduction

Recently, we have prepared macromolecular species in which the prostaglandin analogue 5-(6- carboxyhexyl)-l-(3-cyclohexyl-3-hydroxypropyl)- hydantoin ('BW 245', (I)) [1], which has potent platelet anti-aggregatory activity [2], is coupled to polymer chains through the carboxyl groups.

Abbreviations: BW 245 (C), 5-(6-carboxyhexyl)-l-(3-cyclohexyl-3-hydroxypropyl)hydantoin; PEG, poly(ethylene glycol); D, dextran; PNVP, poly(N-vinyl pyrrolidone); NVP, N-vinyl pyrrolidone.

Water-soluble macromolecules used include poly(N-vinyl pyrrolidone), poly(ethylene glycol) and dextran; details of syntheses, purification and characterisation are presented elsewhere

o n c ..... 3 , , ~ coon

II " o oH

(Refs. 3-7 and our unpublished data). BW 245 has been incorporated in side-chains in

the polymers by copolymerization of appropriate

0167-4889/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

110

methacrylate esters of BW 245 of structures (II) and (III) with N-vinyl pyrrolidone.

I c =CH 2 C'CH2 I I CH 3 CH3

(II) (lid

These esters provide spacer arms of different lengths and hydrophilicities between the coupled units of BW 245 and the polymer backbones, so that we were enabled to assess [3,6,7] the influence of these variables on the antiplatelet activities of the polymers.

BW 245 has also been attached in terminal positions of vinyl polymer chains with the aid of the halo-esters (IV) and (V); the halogen-containing groups of these have subsequently been used to initiate polymerization of N-vinyl pyrrolidone by reaction with a suitable transition metal carbonyl (Refs. 5, 6, and our unpublished data).

@ -COOCH2CCI3 @ - COOCH2CN2Br

(iv) (v )

Terminal attachment has also been carried out by direct coupling of the carboxyls of BW 245 to the terminal hydroxyls of poly(ethylene glycol) [6,7].

The relative antiplatelet activities of these polymers have already been reported [3,6,7] and are summarised in Tables I and II. In the tables, and elsewhere, activities of coupled BW 245 are ex- pressed relative to that of free BW 245, on a molar basis.

Conversion of the carboxyl of BW 245 into the ethyl ester is accompanied by a reduction in anti- aggregatory activity to about 0.17 of the original, [3]. Tables I and II reveal similar behaviour with more complex esterifying groups; in all cases ester activities lie in the range (3.5-12). 10 -z compared to unsubstituted BW 245. When these esters are copolymerized with N-vinyl pyrrolidone so that the polymer backbones carry esters of BW 245 in side-chains there is a further loss of activity to approx. 10 -4 to 10 -3 (Table I). Since the nature of the chemical bonds near the carboxyl is not greatly affected by the polymerization, it is likely that the relative low activities arise mainly from steric interferences caused by the polymer chains, attainment of the spatial requirements necessary

TABLE I

ANTIPLATELET ACTIVITIES OF MONOMERIC ESTERS OF BW 245 AND THEIR COPOLYMERS WITH N-VINYL PYRROLIDONE (NVP) [3,5-7]

Ester Activity of monomer Copolymer with NVP

relative to free tool ratio activity BW 245 NVP : ester relative to (× 102) free BW 245

(XlO 4)

( / / ) n = l X (III) m = 2 / 3.7 65 1.2

(II) n = 2 9.4 64 2.1 (II) n = 4 11.0 51 10.0 (III) m = 5 3.8 66 1.1 (III) m = 12 - 65 6.5

for drug-receptor interactions being subject to obstruction by contacts between bulky chains and platelet surfaces. This is consistent with the observation (Tables I, II) that BW 245 is more highly active when attached to a polymer chain as a terminal group, in which case the activities are in the range 10 -2 to 10 -3 . Again the chemical changes involved in synthesizing the polymer from the esters (IV) and (V) are minimal but the steric obstructions offered by the chains could well be less for terminal attachment.

In attempts to elucidate further the important features in the design of polymers with antiplatelet activities, we have studied the behaviour of BW 245 coupled to dextran and also investigated in

TABLE II

ANTIPLATELET ACTIVITIES OF ESTERS (IV), (V) (INI- TIATORS), AND OF TERMINALLY-COUPLED POLY- MERS OF N-VINYL PYRROLIDONE, AND OF DI- RECTLY-COUPLED POLYMERS [6,7]

Species Activity of Content Activity initiator of BW 245 relative relative to (% w / w ) to free free BW 245 BW 245 (XlO 2) ( x l 0 4 )

(IV) 6.5

(V) 12.2

PEG (M r 4000) - PEG (M, 10000) -

5.4 53 0.74 160 1.0 7.8 0.16 31 0.7 350 0.2 514

more detail the activities of BW 245 coupled to poly(ethylene glycol). This work was led to unex- pected results which are the subject of this paper.

In the following we denote a polymer, P, coupled to BW 245 by (P-BW 245); thus for com- pounds of dextran, poly(ethylene glycol) and poly(N-vinyl pyrrolidone) we write (D-BW 245), (PEG-BW 245) and PNVP-BW 245), respectively. When necessary, the content of BW 245 is given as % w/w.

Methods and Materials

Assessment of anti-aggregatory activity in vitro Sheep blood was collected via a jugular vein,

using a 19-gauge butterfly needle, into plastic syr- inges containing trisodium citrate (3.13%, 10% v /v with blood), mixed, then put into 10-ml plastic containers. The citrated blood was centrifuged (250 x g for 25 min) at room temperature, and the platelet-rich plasma was withdrawn. The remain- ing blood was spun at 800xg for 15 min to obtain platelet-poor plasma. The platelet-rich plasma was diluted with platelet-poor plasma, if necessary, to obtain a platelet count in the range (250-300). 109 platelets/litre, determined in a Coulter Counter (Coulter Electronics Ltd., U.K.).

Platelet aggregation was measured in a Born aggregometer by incubating 0.5 ml of platelet-rich plasma at 37°C, stirred at 9000 rpm, with (1-2 /~g) ADP for 5 min, to produce a non-reversing control aggregation. Inhibition of platelet aggregation was determined by incubating the inhibitor with platelet-rich plasma for 2 min, prior to addition of ADP. The percentage inhibition was calculated as the maximum extent of aggregation recorded in the presence of inhibitor as a percentage of the maximum extent of aggregation without inhibitor.

Materials We are grateful to Sir John Vane and Dr.

Norman Whittaker of the Wellcome Research Laboratories for generous gifts of BW 245 C.

Dextran of 'average molecular weight' (M r 71 200 and 9000) from Sigma Chemical Company, poly(ethylene glycol) (M r 10000 and 4000), from BDH, poly(N-vinyl pyrrolidone) (M r 44000 from BDH and 160 000 from Polysciences) were used as supplied,

111

ADP (Sigma) was dissolved in phosphate- buffered saline (0.2 mg/ml) and frozen, being thawed immediately prior to use.

Theophylline (Sigma) was dissolved in saline to • give a solution containing 782 #g/ml before use.

Trisodium citrate (A.R.) from BDH was used as supplied.

Results and Discussion

(1) Synergistic effects

(a) Highly active polymer species The relative anti-platelet activities of two speci-

mens of D-BW 245 coupled to different extents given in Table III show that the polymer with the lower content of BW 245 is the more active. Perusal of Table II reveals similar behaviour for PNVP-BW 245 and PEG-BW 245. We were thus led to speculate on the generality of this type of behaviour, which could easily be tested by making observations after dilution of the coupled polymers with their inactive parents.

Fig. 1 presents aggregometer measurements of the percentage inhibition of platelet aggregation as a function of concentration of BW 245 (in the platelet-rich plasma) for D-BW 245 (25%) (as coupled) (curve 1) and for diluted specimens with 2.5% and 0.025% of BW 245 (curves 2 and 3, respectively) obtained by mixing the original material with parent dextran. Evidently, dilution gives rise to greatly enhanced activity of coupled BW 245, the activity of D-BW 245 being increased by the presence of free dextran in solution. The effect is quite large; the relative activities deduced from the 50% inhibition values in curves 1, 2, 3 are

TABLE III

ANTIPLATELET ACTIVITIES OF BW 245 COUPLED TO DEXTRAN (D-BW 245)

Content of Average number Weight of Activity BW 245 of units of polymer relative to in polymer BW 245 required free BW 245 (% w/w) per chain of for 50% ( × 104)

dextran inhibition (mg/ml)

0.1 0.2 2.9 9.0 25 64 0.03 3.3

112

100-

c- O_ c~

~50.. (33 C~

0 4 -J

-o_ c-

- C 10-1

4 I

1 10 i I i

102 I ' I

103 104 Concentrotion ol BW 245 in PRP (ng/ml)

, !

105

Fig. 1. Synergism between D-BW 245 (M r 71200) and polymeric diluents in the inhibition of ADP-induced platelet aggregation in sheep platelet-rich plasma (PRP). For the experiments in curves 2-5, the initial polymer D-BW 245 (25%) was diluted with other polymers as shown below. Results are mean 5: S.E. of four experiments.

Curve Diluent Final concentration of BW 245 in polymer (% w/w)

1 2 dextran, M r 71200 3 dextran M r 71200 4 dextran, M r 9000 5 PEG, M r 10000

6 free BW 245 7

25 (as coupled) 2.5 0.025 2.5 2.5

BW 245 diluted by dextran (M r 71200) to give a concentration of dextran of 42 #g/ml in platelet-rich plasma

3.0.10 -4, 1 .2 .10-e and 3.2.10 -2, respectively, so that the increase in activity amounts to about two orders of magnitude. D-BW 245 (0.1%), as coupled, behaved similarly on dilution with dextran, the relative activity increasing from 9.0- 10 -4 (D-BW 245 (0.1%)) to 3.2- 10 -2 (D-BW 245 (0.01~)).

Fig. 1, curve 6, shows corresponding data for free BW 245. The activity is increased only slightly by addition of dextran or other polymers (compare curves 6 and 7).

Dextran itself, in the absence of BW 245, inhibits platelet aggregation initiated by ADP in sheep plasma, but this is significant only at concentrations far in excess of those used in the present experiments. We found that concentrations up to 5 mg/ml were required to give about 20% inhibition, whereas in the experiments of Fig. 1 and others to be reported, the concentrations

(even for high degrees of inhibition) were gener- ally much less than 1 mg/ml (cf. Fig. 4). The highest concentrations were those with D-BW 245 (0.1%) in Table III.

PEG-BW 245 (0.2%) of M r 10000 shows a similar effect when diluted, as may be seen from Fig. 2, in which curves 1 and 2 refer to the original PEG-BW 245 (0.2%) and to PEG-BW 245 (0.01%) obtained by dilution with poly(ethylene glycol) (Mr 10 000), respectively. The relative activities of the original and diluted specimens, estimated from Fig. 2, are 4.9-10 -2 and 0.47. Although the enhancement here is only 10-fold, the diluted polymer (PEG-BW 245 (0.01%)) has the highest activity (per mol BW 245) recorded for any derivative of BW 245, including non-polymer derivatives.

Like dextran, poly(ethylene glycol) is a weak inhibitor of platelet aggregation, but is only effec- tive at relatively high concentrations. We found

that a concentration of 1 mg/ml gave about 13% inhibition.

We have also examined the behaviour of PNVP-BW 245. Upon diluting the initial PNVP- BW 245 (0.2%) with poly(N-vinyl pyrrolidone) (M r 160000), an enhancement in activity was observed (Fig. 3, curves 1 and 2), from 2.2.10 -3 to 4.4- 10-3. The 2-fold increase is less than that found with dextran and poly(ethylene glycol).

The parent poly(N-vinyl pyrrolidone) was found to have negligible effect on inhibition of platelet aggregation.

These findings clearly suggest that the anti-aggregatory activity of BW 245 bound to a water- soluble polymer is likely to depend strongly on the weight-fraction of BW 245 in the polymer, the lower weight-fractions corresponding to higher activities. We shall see later that other 'polymer-type' factors, notably polymer molecular weight, are also important.

In addition to the concentration of bound BW

113

245 in solution, the total concentration of polymer required to give significant inhibition of aggregation is also of interest, particularly from a practical standpoint. Data are given for the dextran systems in Fig. 4. The total concentrations of polymer required to give 50% inhibition are not very different for D-BW 245 with contents of BW 245 in the range 0.1-25%, approximately, varying somewhat erratically between 15 and 70 /~g/ml. At the lower BW 245 contents, the concentration required begins to increase sharply, and the curve presumably ascends to a value approaching many thousands of Fg/ml for [BW 245] = 0.

The corresponding relative activities of BW 245 presented in Fig. 4 reveal clearly the rapid increase with decreasing weight-fraction of BW 245 in the polymer. The limiting value of the activity for low weight-fractions is not clear from these results. At very high weight-fractions the curves approach the values for fully coupled dextran, which unfortunately are not known.

100-

c- O ~d O3 © b

D 50- ~D1 (3

c-

23 c- c -

- 0

10 -1 ' t ' I I ~ I

1 10 102 103

ConcentrQtion of BW 245 in PRP (ng/ml) Fig. 2. Synergism on dilution of PEG-BW 245 (0.2%) (M r 10000) by various polymers. Results are mean + S.E. of six experiments.

Curve Diluent Final concentration of BW 245 in polymer (% w / w )

1 - 0.2 (as coupled) 2 PEG, M r 10000 0.01 3 PEG, M r 4000 0.01 4 dextran, M r 71200 0.01 5 poly(1,1,1 -trishydroxymethyl N-methylmethacrylamide) 0.01 6 PNVP, M r 44000 0.1

7 free BW 245

114

100-

s_

1'o 3 Concentrubon of BW 245 in PRP (ng/rnl)

1~4

Fig. 3. Synergism with PNVP-BW 245 and poly(N-vinyl pyrrolidone). Results are mean + S.E. of three experiments. Curve 1, PNVP-BW 245 (0.2%) (as coupled); curve 2, after dilution to PVNP-BW 245 (0.01%) by poly(N-vinyl pyrrolidone), M r 160000.

(b) Molecular-weight effects We have examined the influence of molecular

weight changes both in the coupled polymers and in the polymeric diluents. The most detailed study has been made with PEG-BW 245.

Fig. 2, curve 3, refers to PEG-BW 245 (0.2%) ( M r 10 000) diluted with poly(ethylene glycol) ( M r 4000) to give PEG-BW 245 (0.01%). This dilution produces a significant increase in activity (approx. 2.8-times) (compare Curves 1 and 3), but less than dilution with the parent polymer of higher molecular weight (approx. 10-times) (curve 2).

Fig. 5, curve 1, presents results for PEG-BW 245 (0.7%) ( M r 4000). Dilution to 0.07% by addition of polymer with the same molecular weight (curve 2) enhances the activity (about 2.3- times) and a further small increase is obtained by dilution to 0.035% (curve 3). However, the effects

c)

3 0 0 -

©

d 3 : 1

-6 ~ ~oo- C o

© f i - U ~ g~ ( j 10 -2

X

10 -1 I 10 BW 245 in po lymer (°low/W)

- 3

8 - 2 0

x >.

-1 5

0

102

Fig. 4. Dependence on % BW 245 in polymer of: total concentration of D-BW 245, (25%)+diluent for 50% inhibition (curve 1); activity relatively to free BW 245 (curve 2). Results are mean+S.E, of four experiments.

are smaller than those observed with the polymer of M r 10000 (Fig. 2). Further, dilution of PEG-BW 245 (0.7%) to PEG-BW 245 (0.035%) by addition of poly(ethylene glycol), M r 10000, leads to an activity similar to that arising from addition of lower-molecular-weight poly(ethylene glycol) (Fig. 5, curve 4).

These results suggest that high molecular weights favour high enhancements of activity on dilution, and show that both the molecular weight of the initial coupled polymer and that of the added polymer may become limiting.

An experiment with dextran supporting these conclusions is shown in Fig. 1, curve 4. Here the initial D-BW 245 (25%) ( M r 71200) has been diluted with dextran of M r 9000 to give D-BW 245 (2.5%). The resulting small increasing in activity (about 1.5-times) may be compared with the much larger effect (about 35-times) obtained on addition of dextran M r 71 200 (curve 2).

(c) Cross-polymer effects It was of obvious interest to ascertain whether

the activity of BW 245 attached to one species of polymer chain is changed by addition of a different kind of polymer.

Fig. 1, curve 5, shows that little effect on the activity of D-BW 245 (25%) (M~ 71 200) is produced by addition of poly(ethylene glycol), M r

100-

E .co

P

8

f

j y i i

10 102 103 ConcentrcrLion ot BW 245 in PRP (ng/rnl)

Fig. 5. Synergism with PEG-BW 245 (M r 4000) and poly(ethylene glycol) of different molecular weights. Results are mean + S.E. of six experiments.

Curve Diluent Final concentration of BW 245 in polymer (% w/w)

1 - 0.7 (as coupled) 2 PEG, M r 4000 0.07 3 PEG, M r 4000 0.035 4 PEG, M r 10000 0.035

10000. In the reverse situation, the activity of PEG-BW 245 (0.2%) (M r 10000) is somewhat increased (about 3-times) by addition of dextran (Fig. 2, curves 1 and 4). Other cross-additions are shown in Fig. 2, in which PEG-BW 245 (0.2%) is diluted with poly(1,1,1-tr ishydroxymethyl N-methylmethacrylamide) (curve 5) and poly(N- vinyl pyrrolidone) (curve 6).

Thus, in the cases examined, cross-addition produces enhancements in activity, but the effects are relatively small. It should be noted that in our systems no incompatible polymer combinations, leading to precipitation, were used. This was checked by direct experiment.

To summarise, the results described demon- strate that polymers (notably poly(ethylene glycol) and dextran) coupled to BW 245 can be greatly increased in antiaggregatory activity by admixture with uncoupled polymers, particularly the parent macromolecule. This is true synergism with respect to the macromolecular species, since both coupled and uncoupled macromolecules have relatively weak activity.

(2) Mechanism of polymer synergism

Synergism in platelet aggregation processes has, of course, been reported previously. With human platelets, mixtures of prostaglandin E 1 and theophylline or, to a small extent, aspirin, have a

Theophyll ine Theophyll ine ' l Theophylline

(PEG-E, W 2 4 5 BW245 0.2*/*) 0.01%)

ADP ADP , 1_ I I ADP

115

Synergistic action on inhibition of aggregation induced by collagen [8]. Whittle et al. [9] demonstrated a powerful augmentation of the inhibitory action of prostacyclin on ADP-induced aggregation by concentrations of theophylline which alone had no effect on aggregation of human platelets. The latter workers also reported similar observations with some other systems in sheep, horse and rabbit plasmas, pointing out that the behaviour of theophylline, which inhibits phosphodiesterase, supports the view that prostacyclin inhibits platelet aggregation by raising cyclic AMP level in platelets. We believe that these phenomena differ fundamentally from the synergism we have reported in that they are accompanied by modifica- tions in chemical processes associated with inhibition. By studying the influence of theophylline on inhibition of aggregation of free BW 245 and BW 245 coupled to macromolecules we have obtained evidence that the process is associated with elevated cyclic AMP levels in the platelets, so that inhibition probably follows the same chemical route with each. Potentiation by theophylline of inhibition of ADP-induced platelet aggregation by free BW 245, PEG-BW 245 (0.2%) and PEG-BW 245 (0.01%) (M r 10000) is shown in Fig. 6, the similarity in behaviour indicating that the influence of PEG does not extend to processes following interaction with receptors.

In our opinion, the effects we observe arise

i I 1rain

a b c

j2 Fig. 6. Effect of theophylline (43/~g/ml) on ADP-induced platelet aggregation inhibited by: (a) free BW 245, (4.2 #g/ml); (b) (PEG- BW 245, 0.2%) (42 #g/ml); (c) (PEG-BW 245, 0.01%) (60/~g/ml). In each case, curve 1 represents the control aggregation induced by ADP. Curves 2a-c represent inhibition by (a) free BW 245, (b) PEG-BW 245 (0.2%) and (c) PEG-BW 245 (0.01%). Curves 3a-c indicate the enhanced inhibition in the presence of theophylline.

116

from polymer adsorption on to platelet membranes. This general process is well-authenticated in the literature, from both in vivo and in vitro observations. For present purposes we note here the work of Weiss [10], who proposed that adsorption of dextran directly alters the platelet membrane, as inferred from the change in electro- phoretic mobility. The observations of Pontecorvo [11] and Maggio et al. [12] on cell fusion induced by poly(ethylene glycol) and the interactions of this polymer with lipid bilayers are also of particu- lar interest. Note that much of the work described in the literature has been carried out with polymer concentrations much higher than those we have used, and in some cases with polyelectrolytes [13-17], which interact more strongly.

We propose that adsorption of (neutral) polymers can disorganize the lipid bilayers of the membranes, to an extent which is greatest in the vicinity of the glycoprotein components (Fig. 7), on which receptors are located. Such disorganization makes the receptors more accessible by reduc- ing steric .obstruction arising from the lipid molecules. The influence of the increased accessibility is likely to be minimal for reactions of free BW 245, a relatively small molecule, but very significant in reactions of BW 245 coupled to a polymer chain, which necessarily involve movements of several chain segments to accommodate the required geometry of the transition state. This mechanism can, therefore, account for the observed differences in behaviour of free and coupled BW 245 in the presence of added uncoupled polymers.

The general type of mechanism is algo consistent with the other experimental data already presented. First, according to Fig. 4, which refers to the dextran-BW 245 system, the total weight of D-BW 245 required to produce 50% inhibition is approximately independent of the BW 245 content of the polymer over the range 25-0.1% (w/w). Correspondingly, the relative activity of the coupled BW 245 increases from 0.5.10 -2 to 2.5. 10 -2 . The results suggest that interaction of this species with receptors always requires an adequate polymer concentration, free or bound, sufficient to modify the membrane surface. If the content of BW 245 in the material is high, the measured relative activity is low, and vice versa.

Secondly, it is well known that high polymers are more strongly adsorbed than their lower- molecular-weight analogues. This property is doubtless partly responsible for the outstanding activities of high-molecular-weight dextran and poly(ethylene glycol) in our experiments.

Thirdly, the distribution of coupled BW 245 along the macromolecular chains appears to have an important influence on activity. Comparison of Table III and Fig. 4 shows that, for polymer designated D-BW 245 (0.1%), the material prepared by diluting D-BW 245 (25%) with dextran is much more active (about 12-times) than D-BW 245 (0.1%) made directly by coupling. This is all the more striking because D-BW 245 (0.1%) directly prepared is more active than D-BW 245 (25%) (Table III). The distributions of BW 245 in the two materials are very different; in that prepared from D-BW 245 (25%), molecules of BW

: : : : : : : : : : : : : : : : : : : : :

0 0 0 0 0 • 0 0 0 • 0 • 0 0 • 0 0 0 • • • b • • 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

o o ~"o • • o , o . o o o . ~ ' ~ J g : O 0 0 0 0 0 0 0 0 ~ 0 ~ ~ ~ • , . .

" ' ' ' " • " • • • " ' " " i i ' i i i i o o o o o o o o { % ~ o o o 0 ~ a o - | 0 0 0 0

• • • • • , 1 " / ~ . 5 ~ . , ~ . . ~ ' ~ ' ,

: : ' " " " " " " . : " : c . . . . . ~ ~ ' ~ ' i . . . . .o o . . . o . o o o , . e o o . • . - . o Fig. 7. Diagrammatic representation of mechanism of polymer synergism. (a) Surface of portion of cell membrane (plan) with glycoprotein components embedded in lipid bilaycr, carrying receptors for BW 245. (b) Disorganization of lipid bilayer in the neighbourhood of giycoprotein molecules, brought about by adsorption of a macromolecule (not shown). (c) Different areas of disorganization produced by adsorption of different macromolccules.

245 are bunched together in a few chains, most chains being free from BW 245, whereas in D-BW 245 (0.1%), as coupled, very few chains carry more than a single molecule of BW 245 and the free chains are a smaller fraction of the total. These findings may be explained if the length of free chain between the adsorbed segments of a coupled polymer and the attached BW 245 molecule can become critical in determining a suitable geometry for reaction.

Fourthly, as we have seen, cross-polymer effects are relatively small, even with 'active' polymers such as poly(ethylene glycol) and dextran (Figs. 1 and 2). This may indicate that adsorption of different macromolecules can occur in different regions of the membrane, possibly partly determined by the character of the interacting groups of glycoproteins in the neighbourhood. We therefore suggest that areas of the membrane in the vicinity of different proteins may have altered adsorption characteristics; an added polymer increases the accessibility of receptors for polymer- bound BW 245 only if it is adsorbed in the same region as the coupled polymer. For example, dextran and poly(ethylene glycol), having different types of interacting group, are probably adsorbed in different locations on the membrane. Dilution of D-BW 245 with poly(ethylene glycol) therefore produces no change in activity (Fig. 1), while dilution of PEG-BW 245 with dextran gives a small enhancement (about 3-times) (Fig. 2), probably attributable to some spread of the influence of the more strongly bound dextran into a poly(ethylene glycol) area.

Fig. 7 illustrates the proposal of increased accessibility to receptors arising from local membrane disorganisation produced by adsorption of different macromolecules.

An alternative suggestion is that the cross-polymer phenomenon is a manifestation of polymer incompatibility, i.e., an adsorbed polymer impedes the adsorption of a second polymer of different type in the same locality. Nevertheless, a strongly adsorbed polymer should displace a more weakly bound one. This is not consistent with observations on the PEG-BW 245/dextran system referred to above, since dextran, by displacing PEG-BW 245, would be expected to lower the activity.

117

(3) Comments on other mechanisms

We propose that mechanisms be classified as 'positive' or 'negative'. In a positive mechanism, the added polymeric diluent gives rise to a new species or process, or potentiates one already ex- isting, which inhibits platelet aggregation. For example, liberation of BW 245 from D-BW 245 by enzymic hydrolysis facilitated by the additive would constitute a positive mechanism. In a negative mechanism, an intermediate in the aggregation process is deactivated by added polymer. An additive blocking the arachidonate pathway to thromboxane by removing a necessary intermediate would function by a negative mechanism. Thus, a positive mechanism stimulates inhibition and a negative mechanism removes a stimulator of aggregation.

Since the mechanism advocated in part 2 includes polymer adsorption - a process not encountered in inhibition by free BW 245 - it is classified as positive.

Practical distinction between positive and negative mechanisms for inhibition of aggregation by polymer-bound prostaglandins is difficult. Sinha and Colman [18] found that human plasma incubated with prostaglandin E 1 coupled through a hexyl spacer to agarose inhibits aggregation, but they were unable to detect any elution of prostaglandin E 1 by radiotracer measurements or any change in cyclic AMP levels. They explored the possibility of a negative mechanism, viz the re- moval by prostaglandin El-hexylagarose of a plasma cofactor necessary for ADP-induced platelet aggregation. However, they could not find direct evidence for this and suggested that some form of positive mechanism of inhibition, independent of cyclic AMP, might operate.

Ebert et al. [19] reported the immobilisation of prostacyclin on cross-linked polystyrene beads and observed powerful anti-aggregatory and anti-ad- hesive behaviour. The remarked that "a question that remains unanswered is whether these antiplatelet effects are due to a direct interaction between the immobilized prostaglandin and the platelets or through a plasma co-factor". In a latter paper [20], Kim et al. reject the latter alternative and suggest (on the basis of unpublished

118

data) that Sinha and Colman's findings might arise from "microlevel leakage of prostaglandin El-diaminoalkane conjugates" in view of the in- stability of the chemical link employed. Ebert et al. [19] present data for the leakage of prostacyclin from their specimens; in our view, these data do not support the claim [19] that leakage is insuffi- cient to prevent ADP-induced aggregation.

In our systems, derivatives such as D-BW 245 and PEG-BW 245 contain ester links, hydrolysis of which facilitated by added polymer with liberation of BW 245 could constitute a positive mechanism. On chemical grounds, such a process, oc- curring in homogeneous liquid phase, is extremely unlikely to make a major contribution in view of the combination of properties required to be com- patible with our findings. In addition to the im- probable premises, it would be necessary to as- sume that catalysis of hydrolysis is effectively confined to added polymers of high molecular weight acting on polymers bound to BW 245 of the same type, and also of high molecular weight. Chemically identical ester linkages between BW 245 and other polymers (of different type, or low molecular weights) must be assumed to be unaffected.

Negative mechanisms, such as the inactivation of Factor VIII [21], appear to be excluded, since free BW 245 is relatively unaffected by added polymers (Fig. 1). Otherwise it would follow that free and coupled BW 245 operate through different chemical mechanisms.

We intend to investigate the possible occur- rence of similar antiplatelet phenomena with in- soluble polymers. The possible rfle of plasma proteins in our experiments has not so far been examined directly, but we hope to take up this matter by studying 'pure' systems, with washed platelets. An extension to other types of cells using appropriate polymer-drug combinations, would also be of interest.

Acknowledgement

We are indebted to Dr. Y. Satake for his in- valuable work in the early stages of this project.

References

1 Caldwell, A.G., Harris, C.J., Stepney, R. and Whittaker, N. (1979) J. Chem. Soc. Chem. Commun. 55, 561-562

2 Caldwell, A.G., Harris, C.J., Stepney, R. and Whittaker, N. (1980) J. Chem. Soc. Perkin Trans. 1, 495-505

3 Satake, Y. (1985) Thesis, University of Liverpool 4 Bamford, C.H., Middleton, I.P. and Satake, Y. (1984) Poly-

mer Preprints 25(2), 27-28 5 Bamford, C.H. Middleton, I.P., Satake, Y. and AI-Lamee,

K.G. (1986) in Advances in Polymer Synthesis (Culbertson, B.M. and McGrath, J.E., ed.), pp. 291-320, Plenum Press, New York

6 Bamford, C.H., Middleton, I.P., Satake, Y. and Al-Lamee, K.G. (1985) Biomaterials in Blood Therapy, European Re- search Co-ordination on Development, Testing and Appli- cation, Siena, Italy

7 Bamford, C.H., Middleton, I.P., Satake, Y. and A1-Larnee, K.G. (1985) 2nd International Conference on Polymers in Medicine, Capri, Italy, June 1985, (Abstract C5).

8 Ball, G., Brereton, G.G., Fulwood, M., Ireland, D.M. and Yates, P. (1970) Biochem. J. 120, 709-718

9 Whittle, B.J.R., Moncada, S. and Vane, J.R. (1978) Pros- taglandins 16, 373-388

10 Weiss, H.J. (1967) J. Lab. Clin. Med. 69, 37-46 11 Pontecorvo, G. (1975) Somatic Cell Genet. 1, 397-400 12 Maggio, B., Quet Fall Ahkong and Lucy, J.A. (1976) Bio-

chem. J. 158, 647-650 13 Ikeda, T., Ledwith, A., Bamford, C.H. and Hann, R.A.

(1984) Biochim. Biophys. Acta 769, 57-66 14 Kabanov, V.A. (1986) Macromol. Chem., Macromol. Symp.

1, 101-124 15 Kirby, E.P. and Mills, D.C.B. (1975) J. Clin. Invest. 56,

491-502 16 Kirby, E.P. (1982) J. Lab. Clin. Med. 100, 963-976 17 Tirrell, D.A., Turek, A.B., Wilkinson, D.A. and Mclntosh,

T.J. (1985) Polymer Preprints 26(2), 196-7 18 Sinha, A.K. and Colman, R.W. (1978) Science 200, 202-203 19 Ebert, C.D., Lee, E.S. and Kim, S.W. (1982) J. Biomed.

Mater. Res. 16, 629-635 20 Kim, S.W., Ebert, C.D., McRea, J.C., Briggs, C., Byun,

S.M. and Kim, H.P. (1983) Ann. NY Acad. Sci. 416, 513-524

21 Aberg, M. Hedner, U. and Bergentz, S.-E. (1979) Ann. Surg. 189, 143-247

Documents

Polymeric inhibitors of platelet aggregation. II. Copolymers of dipyridamole and related drugs with N-vinylpyrrolidone