Download - In vitro contact phase activation with haemodialysis membranes: role of pharmaceutical agents

i?iomoteria/s 16 (1995) 1305-1312 Elsevier Science Limited

Printed in Great Britain 0142-9612/95/$10.00

In vitro contact phase activation with haemodialvsis membranes: role of pharmaceitical agents

B.M. Matata, S. Wark, S. Sundaram, J.M. Courtney, J.D.S. Gaylor, SK. Bowry*, J. Vienken* and G.D.O. Lowet Bioengineering Unit, University of Strafhctyde, Wolfson Centre. Glasgow, G4 ONW, UK; ‘Akzo Nobel, Membrana, Germany; tDepartment of Medicine, Glasgow Royal Infirmary, Glasgow, UK

Contact phase activation was investigated in vitro using flat sheet type of haemodialysis membranes, Cuprophan (Akzo, Faser, Germany) and AN69S (Hospal, France), and a negatively charged polyamide Ultipor NR 14225 membrane as a control. The investigation focussed on the determination of factor XII- like activity (FXIIA) as an indicator of contact phase activation in the supernatant phase and at the membrane surface after plasma-membrane contact using an incubation test cell. The findings were compared with the observations from a plasma-free system utilizing purified unactivated factor XII. ‘The plasma FXIIA bound to the membrane surface was significantly different between the membranes, Jvhile the supernatant phase FXIIA exhibited no significant differences. In contrast, the plasma-free system exhibited significant differences in the supernatant FXIIA and membrane-bound FXIIA for all ::he materials used and the magnitude of the activity was significantly greater for negatively charged materials. This finding demonstrated the strong influence of the interaction of other plasma constituents on the membrane surface and as such the binding and subsequent activation of factor XII may be altered possibly due to competitive binding and steric hindrance. On the addition of anticoagulants such as heparin, low-molecular-weight heparin, citrate and hirudin, no significant differences were observed in plasma supernatant phase FXIIA. However, each anticoagulant appears to have a distinct influence on the magnitude of plasma membrane-bound FXIIA. On the addition of aprotinin (a kallikrein inhibitor), no significant differences were observed in the plasma supernatant FXIIA. In contrast, aprotinin appears to significantly reduce membrane-bound FXIIA on Cuprophan and polyamide NR, but significantly increase the magnitude of the membrane-bound FXIIA on AN69S.

Keywords: Membranes, haemodyalisis, contact phase activation

Received 17 February 1995; accepted 23 March 1995

The fundamental understanding of blood-biomaterial compatibility requires the study of various parameters such as protein adsorption, platelet adhesion and activation, blood coagulation, fibrinolysis, kallikrein- kinin system, complement activation and cellular interactions. Despite th.e advantage of a multiparameter assessment, there are limiting factors such as sample collection and analysis. Inevitably, there is a need for effectively choosing to cover the more fundamental aspects of the blood response. With respect to blood coagulation, activation of the intrinsic pathway induced by artificial surfaces is initiated by the contact activation phase. This involves the interaction of proteins, e.g. Hageman factor (FXII), prekallikrein (PK), high-molecular-weight kininogen (HMWK), and factor XI. Since the discovery of the coagulation protein FXII (Hageman factor] in the blood of Mr Hageman in 1955 by Ratnoff and Calopy, studies by numerous research

Correspondence to Dr B.M. Matata.

groups since that time have revealed that FXII either plays a major role in, or is associated with, the four plasma defence systems (coagulation, fibrinolysis, kallikrein-kinin and complement]. It might also partici- pate in the renin-angiotensin and protein C systems’. A current focus is on the measurement of contact phase activation in relation to blood-biomaterial interaction. The parameter representative of contact phase activation has been the determination of factor XII-like activity (FXIIA) and has commonly been assayed using chromogenic substrates.

FXII is a single-chain glycoprotein with a molecular weight of approximately 80 000, a sedimentation coeffi- cient of approximately 4.5 s and an isoelectric point of approximately 6.1-6.5. It migrates as a /?-globulin on electrophoresis’. Plasma levels’ of FXII antigen range between 23 and 40pgml-‘. FXII is composed of three domains, with approximate molecular weights of 40000, 30 000 and 12 000. FXII can be cleaved by various proteases into two fragments, an amino-

1305 Hiomaterials 1995. Vol. 16 No. 17

1306 Contact phase activation with haemodialysis membranes: B.M. Matata et al.

terminal portion of approximately 52 000 molecular weight and a carboxyl-terminal portion of approximately 28 000 molecular weight. The surface binding property of FXII resides in the 52 000 molecular weight component, while the 28000 component contains the enzymatically active region’. The properties of FXII cannot be discussed independently of the other three proteins of the contact system. Those proteins circulate in inactive forms in plasma and are converted to active enzymes or liberate active peptides during activation on contact surfaces. The proenzymes FXII PK (88000 molecular weight) and FXI (160 000 molecular weight) are converted by limited proteolysis into the active serine proteases a-FXIIa, and fl-FXIIa, plasma kallikrein (KK) and FXIa. PK and FXII circulate in the blood as bimolecular complexes with HMWK3, which has a positively charged histidine-rich region. This binds HMWK together with complexed PK and FXI to negatively charged surfaces. A portion of the FXII molecule also binds to negatively charged surfaces, so that when blood comes into contact with such a surface, the four proteins of the contact system are assembled on the surface. The major inhibitor of a- and P-FXIIa in plasma is the complement 1 esterase inhibitor (Cl-INH). This has been shown to contribute up to more than 90% of plasma inhibition of a- and /?- FXIIa4. Other inhibitors include &-glycoproteins, al- antitrypsin, protein C inhibitor and the so-called ‘plasminogen activation inhibitor”. Antithrombin III (AT-III) is a weak inhibitor of a- and /I-FXIIa, FXIa and KK, but heparin has been reported as an activator of FXII and the contact phase5.

Although the precise mechanism for the critical activation of FXII is a subject of controversy, it is believed that when FXII is bound to a negatively charged surface, autoactivation occurs, followed by the rapid generation of a- and /I-FXIIa, and KK. Proteolytic cleavage of a-FXIIa generates a 25 000 fragment named /I-FXIIa’. a-FXIIa binds to negatively charged surfaces while P-FXIIa does not. Both a- and /?-FXIIa activate PK, while a-FXIIa is a much better activator of FXI than /I-FXIIa’. a- and B-FXIIa have also been shown to activate other proenzymes including FVI17, plasminogen* and Cl’. Since KK is such a powerful activator of surface-bound FXII and because a- and /I-FXIIa convert PK to KK, a self-amplification occurs with explosive activation of FXII. The resulting generation of KK not only leads to further activation of FXII but also the liberation of bradykinin (BK) enhances from HMWKl’. BK enhances vascular permeability, produces hypotension, contracts smooth muscles, causes pain and releases tissue plasminogen activate?. KK converts plasminogen to plasminl’, releases a- and /I-FXIIa from FXII13, and liberates renin from prorenin14.

The plasma defence system is made up of a series of interrelated biochemical pathways, i.e. complement, coagulation, fibrinolysis and the kinin systems15. This relationship has been strengthened by the recent observations of adverse clinical reactions following blood-membrane interaction, Cases of life-threatening anaphylactoid reactions occurring within the first 5 min of haemodialysis characterized by oedema of the face, tongue and lips, bronchospasms, hypotension, vomiting and abdominal cramps have been reported’620.

The reactions described were only associated with the use of PAN (AN69)-membranes and the simulta- neous administration of angiotensin-converting enzyme (ACE) inhibitor. The reactions were not observed with other membranes such as Cuprophan and Hemophanl’ or polysulphone F601*. The reactions were believed to be due to an enhanced generation of BK. AN69, a negatively charged membrane, was believed to stimulate the activation of FXII of the contact phase followed by the activation of the KK and kinin systems leading to the release of BK”.

BK is normally degraded to its peptide components by the ACE, but in the presence of ACE inhibitors the degradation does not take place and BK levels may rise leading to the manifestation of a biological effect characteristic of excess BK22. The hypothesis was supported by in vitro studies23 and the data were later confirmed in an animal model using sheep24. It is possible that surfaces containing negative charges in distinct concentration give rise to contact phase activation with subsequent BK generation. The adverse effects observed during low density lipoprotein (LDL)- apheresis25*26 appear to support this hypothesis, in which some perfusion columns for LDL-apheresis contain negatively charged dextran sulphate particles and as a result give rise to BK generationz5.

Although enhanced BK release during dialysis with AN69 has been demonstrated23, the mechanism involved at the membrane surface has yet to be elucidated. The events leading up to BK release appear to be the surface aggregation of FXII followed by its autoactivation. An attempt was made using in vitro experiments to link the magnitude of surface-bound FXII by ionic forces to AN69 membrane and BK release27. However, the hypothesis of increased magnitude of surface-bound FXII, its autoactivation and the link with the plasma KK-kinin system was not demonstrated. This hypothesis would be demonstrated if the magnitude of the activation of the surface-bound FXII was estimated.

The central role of FXII in initiating the contact phase has been well established and emphasizes the relevance of quantifying this protein in the assessment of the blood response to biomaterials. When FXII is bound to a surface, autoactivation occurs followed by the rapid generation of a- and P-FXIIa. a-FXIIa ‘possesses surface binding -properties while /I-FXIIa is released into the plasma fluid and most of it becomes inactivated by plasma inhibitors such as the Cl-INH. The use of the chromogenic substrate method to evaluate FXII activation following blood-biomaterial contact has been extensively evaluated2*-30. Measure- ment of plasma /I-FXIIa using the chromogenic substrate assay, however, appears not to be a sensitive index of FXII activation as indicated by some in vitro studies28-30. As a consequence, a new approach to the use of a chromogenic substrate assay has been adapted to quantify FXII activation induced by polymeric biomaterials at the membrane surface. Recent studies, however, by measuring FXII in the supernatant (plasma), reported a lack of significant differences for a range of different membranes. The objective has since been the modification of the assay method of measuring FXIIA. Estimation of FXIIA on the surface of the

Biomaterials 1995. Vol. 16 No. 17

Contact phase activation with haemodialysis membranes: B.M. Matata et al. 1307

was necessary, in view of the tendency of the contact factors to bind to artificial surfaces. In order to achieve a broader understanding of contact phase activation, an experimental approach that compared a simplified system (purified unactivated FXII) to a complex system (plasma) was adapted.

The successful application of haemodialysis membranes depend:3 on the use of drugs such as heparin, low-molecular-weight heparin, antiplatelet agents and recombinant hirudin31. New approaches have been to extend .the traditional role of antithrombotic agents by the introduction of multienzyme inhibitors (nafamostat mesilate)3z or KK inhibitors (aprotinin)33 to inhibit contact phase activation. This leads to an important overall relationship between blood, the biomaterial and the pharmaceutical agent34, with the possibility of an altered blood response resulting from changes to t;he biomaterial, the pharmaceutical agent or both. Another point for consideration is the adsorption by biomaterials of the pharmaceutical agent and the possible influence on biological activity. This study investigated the role of pharmaceutical agents such as anticoagulants and KK inhibitors on the initiation of contact phase activation.

EXPERIMENTAL STRATEGY

The investigation focussed on three aspects of contact phase activation in relation to blood-membrane interactions:

(1) The comparison between FXIIA in a simplified plasma-free system (purified unactivated FXII, 25 pg ml-’ of Tris buffer and sodium acetate buffer) to a complex system (heparinized plasma, 1 IU ml-‘).

(2) The difference between surface-bound and supernatant phase FXIIA.

(3) The influence of various anticoagulants and KK inhibitors on FXIIA.

MATERIALS AND METHODS

The flat sheet membranes selected for evaluation were Cuprophan (Akzo .Faser, AG, Germany), AN69S (Hospal, France) and a control polyamide membrane Ultipor NR 14225 with a surface negative charge and of known zeta potential (-18.0 mV) at pH 7 (Pall Filtra- tion, Portsmouth, UK).

The purified FXII was reconstituted in buffer as follows: 0.5mg purified FXII in lml sodium acetate saline (4m~ sodium acetate, 0.15~ NaCl pH5.3) was diluted with 19ml of Tris buffer (0.025 M Tris-HCl, 0.025 M NaCl pH 8).

Chromogenic FXIIa substrate (2-AcOH-H-D-CHT-Gly- Arg-pNA) and KK inhibitor were purchased from Channel Diagnostics (Walmer, Kent, UK). Sodium acetate and Tris-HCl buffer (pH8) were obtained from Sigma Chemicals (F’oole, Dorset, UK). Negatively charged and uncharged polystyrene six-well tissue culture’ plates were purchased from Linbro/Titretek, ICN Biomedicals, Oxon, UK.

Blood was obtained from normal healthy volunteers (after an informed consent approved by the Glasgow Royal Infirmary ethical committee) and anticoagulated with the desired antithrombotic agent and centrifuged at 3000rpm for 12 min at room temperature. Plasma samples were collected and, after contacting with the membranes, aliquots were snap frozen and stored at -70°C ready for assay. Purified FXII was purchased from Enzyme Research Laboratories Ltd (Swansea, UK).

The antithrombotic agents used were as follows: Heparin (Heplok, Leo Laboratories, Buck, UK), low- molecular-weight heparin (Fragmin LMWH, Chromo- genix, Quadratech, Surrey, UK), hirudin (CGP39393, Ciba Geigy, Basle, Switzerland), and trisodium citrate (Sigma, Dorset, UK).

The KK inhibitor used was aprotinin (Trasylol, Bayer, Newbury, UK).

Procedure for measurement of activated FXII in a plasma-free system using purified F’XII

An incubation method was also used to achieve FXII solution-material contact. Membranes were fixed onto a silicone rubber sandwiched between a polystyrene plate and a six-well tissue culture polystyrene plate with holes drilled through the wells. The area of the membranes within the tissue culture wells is approximately 9.6cm2. Reconstituted FXII (1 ml diluted to 25 pg ml-’ of sodium acetate and Tris-buffered saline) was added to the materials of interest and incubated for 2 h at room temperature. Incubation of 2 h was selected based on the fact that autoactivation of FXII in the absence of other contact proteins (HMWK, PK, FXI) proceeds at a slower rate35. Adequate mixing was obtained by using an orbital shaker. Prior to contact all the materials were soaked in 0.9% saline for at least 24 h.

At the end of the incubation period, 100~1 of FXII solution were withdrawn for assay (designated assay of supernatant). The remaining FXII solution was removed and the membrane washed three times in l.Oml of Tris buffer (Tris-HCl 0.025 M, 0.025M NaCl, pH 8) for 1 min. After the removal of the wash buffer, 1.0 ml of substrate (2-AcOH-H-D-CHT-Gly-Arg-pNA) was added to the material and incubated for 20min. A 900 ~1 sample of substrate was withdrawn into polystyrene cuvettes (BDH Ltd, Poole, Dorset, UK) and absorbance read at 405nm against suitable blanks (9OOpl of Tris buffer). A polystyrene tissue culture plate coated with a layer of oxygen ions, therefore negatively charged, was used as a positive control material (Titrek UK Ltd).

Assay of supernatant

To 100~1 of supernatant obtained earlier, 400~1 of buffer were added (prewarmed to 37°C). A 200~1 sample of substrate (prewarmed to 37°C) was added and the mixture incubated in a polystyrene cuvette (BDH) at room temperature for 10min. The reaction was stopped by addition of 200~1 of 50% acetic acid. Absorbance was read at 405 nm against suitable blanks (700bl of buffer and 200~1 of 50% acetic acid) and results expressed as optical density units.

Biomaterials 1995, Vol. 16 No. 17.

1308 Contact phase activation with haemodialysis membranes: B.M. Matata et a/.

Table 1 List of pharmaceutical agents and their properties

Pharmaceutical agent Nature/properties Target Plasma concentraGon’

1. Heparin (standard unfractionated) Anticoagulant

2. Low-molecular-weight heparin Anticoagulant

Catalyses thrombin and FXa inhibition by antithrombin III

Catalyses the inhibition of FXa by antithrombin Ill

0.5 and 2.5 U ml -’

0.5 and 2.5 anti-Xa U ml -’

3. Trisodium citrate (3.2% solution)

4. Recombinant hirudin

5. Aprotinin

Anticoagulant

Antithrombotic agent

Inhibits contact activation via the kallikrein pathway

A chelating agent

A thrombin inhibitor

Kallikrein inhibitor

1:9 plasma ratio

10 ATU ml- ’

50, 100 and 200 KIU ml-’

‘ATU and KIU refer to antithrombin units and kallikrein inhibltor units, respectively.

Procedure for assay of plasma FXIIA

Freshly collected platelet-poor plasma was pooled. A six-well incubation test cell method was employed to achieve plasma-material contact. A l.Oml sample of plasma was then added to fie materials of interest and incubated for 10min at rgom temperature. Adequate mixing was obtained by using an orbital shaker. Prior to contact, all materials were soaked in 0.9% saline for at least 24 h at room temperature to remove preserva- tives.

At the end of the incubation period, 2004 of plasma were withdrawn for assay (designated assay of supernatant). The remaining plasma was then removed and the membrane washed twice in l.Oml of buffer (Tris-HCl O.O25M, NaCl 0.025 M, pH8) for 1 min. The wash buffer, was removed and l.Oml of KK inhibitor (KI) (Channel Diagnostics), diluted at l/50 in Tris-HCl buffer, was added to the material and incubated for 10 min. The KI was removed and 1.0 ml of substrate (2- AcOH-H-D-CHT-Gly-Arg-pNA, l.O~molml-’ * distilled water) was added to the material ai: incubated for 20 min. Substrate and KI were prewarmed to 37°C for 0.5 h prior to use.

A 9OOpl sample of the substrate was withdrawn into polystyrene cuvettes (1.0cm3) and absorbance read at 405 nm against a suitable blank (900 ~1 of buffer).

Assay of supernatant

A 2004 sample of supernatant obtained earlier was diluted 1:s in buffer and assayed using a microtitre format (Dynatech Ltd, 96-well plates). To 25~1 of diluted plasma, 754 of KI in buffer were added and incubated for 1 min. The reaction was started by addition of 50~1 of substrate, incubated for 10 min, and terminated by the addition of 504 of 50% acetic acid. Absorbance was read at 405 nm using a microtitre plate reader (Dynatech Ltd, Billingshurst, UK) against suitable blanks (200~1 of 50% acetic acid and 7004 buffer). The absorbance readings expressed as optical density units represented the magnitude of FXIIA.

The effect of pharmaceutical agents on FXIIA

There is a general requirement for the understanding of the nature of the interaction between a pharmaceutical agent and the membranes, how this association

influences the overall blood response. This study investigated a number of pharmaceutical agents whose general properties are shown in Table 1 and the emphasis was on their overall effect on the initiation of the contact phase activation.

Statistics

Statistical analysis was performed using the Minitab package (version 8.0). In order to represent the pattern of the polymer influence on FXII activation, the actual mean values were used. Comparison of the differences between two means was carried out using the two-way analysis of variance and two-sample Student’s t-test and these were reported at 95% confidence intervals (P < 0.05) as well as the Mann-Whitney test for difference in the medians of the values.

RJNJLTS

Plasma-free system FXIIA (purified FXIIA)

Supernatant FXIIA assay The supernatant phase FXIIA (j?-FXIL4 equivalent) was significantly lower for the neutral Cuprophan compared to negatively charged AN69S and polystyrene, whereby the latter was significantly lower than the former (Figure 1). This appears to indicate that the magnitude of supernatant FXIIA in a plasma-free system increased with the strength of the negative charge.

Membrane sux$ace FXIIA Cuprophan exhibited a significantly lower surface- bound FXIIA compared to either AN69S or polystyrene but the latter exhibited significantly lower FXIIA values than the former. This agrees with the general view that negatively charged surfaces are much stronger activa- tors of FXIIGV2’. The results also suggest that autoactivation of FXII to a- and /?-FXIIa can occur spontaneously at an appropriate surface, and does not require the involvement of other contact proteins.

Plasma FXIIA

Plasma supernatant FXIZA The mean value of the FX.IIA in the plasma supernatant (Figure 2) which is equivalent to fl-FXIIa in the plasma

Biomaterials 1995, Vol. 16 No. 17

Contact phase activation with haemodialysis membranes: B.M. Matata et a/. 1309

1.2 -

z ‘.O r 0 SupWnatmt FXliA

g 0.8 - I Membrane-bound FXIIA

r

Q) 0.6 -

Cuprophan Ineutral charged)

ANSQS Lnrgrtivrly charged1

Membranes

Polyrtyrrnr lnegltlvrly charged1

Figure 1 Determination of factor XII-like activity (FXIIA) in a plasma-free system using purified unactivated FXll on Cuprophan, AN69S and a negatively charged polystyrene. A significant difference in supernatant phase and membrane-bound FXIIA was observed for all the materials; P < 0.05.

was not significantly different (P > 0.05) in all the membranes evaluated, and this observation was in agreement with previous findings using whole blood without anticoagulant2822Q. This would appear to suggest that the assay method for the determination of supernatant FXIIA (,kFXIIa equivalent) in plasma may not be sufficiently sensitive to allow the response by the different material be detected. The alternative hypothesis would be the rapid inhibition of FXIIA in plasma by plasma inhibitors such that the extent of the influence of the m’embrane characteristics on FXII activation could not be distinguished.

Plasma membrane-bound FXIIA In contrast to the plasma supernatant FXIIA observed, the membrane-bounld FXIIA appeared to be highly discriminative, and the activity was material dependent (Figure 2). Against expectations, Cuprophan (neutral charged) exhibited higher FXIIA than the negatively charged AN69S membrane, which indicated the least FXIIA. The reason for this pattern

1.0

1 2 0.6

i

0 Hepuinisnd plasm8 I1 U/ml) tupernatant FXIIA

VI m Heparinirt8d plwnl (1 U/ml) membrtne-bound FXIIA T 0 2 0.6 -

0” 2 0.4 -

k n

9 0.2-

0.0 - n+ A

Cuprophan Ineutral charged1

AN69S lnrgltirely chuged)

Membranes

polyamide NR InegatIvely charged1

Figure 2 Factor XII-like activity (FXIIA) during plasma- membrane contact. No significant differences in plasma FXIIA were observed. This contrasts sharply with the significantly higher membrane-bound FXIIA on the control polyamide NR 14225 compared to Cuprophan and AN69.S. The latter indicated the least FXIIA; P < 0.05.

is not entirely clear, but it may be fair to speculate that this is due to the affinity of AN69S membrane for certain plasma proteins, and as a consequence conipetitive binding minimized the amount of FXII bound to the surface. The alternative hypothesis would be that as a result of the high affinity of AN69S for FXI127 the surface binding results in an unfavourable configuration and conformation with minimal autoactivatior?. In contrast, the negatively charged polyamide NR control exhibited the highest membrane surface FXIIA, although the supernatant FXIIA remained similar to the other membranes. This may suggest that there is a difference in the nature of the surface charge between AN69S and polyamide NR and that this may be the overall surface distribution. This appears to have an effect on the magnitude of FXIIA. An alternative hypothesis could lie in the differences in the ability to adsorb heparin from plasma onto the membrane surface. Heparin with its strong anionic centres can activate FXII and, from our earlier studies on heparin binding (unpublished), AN69S appears to have a very low affinity for heparin.

Effect of pharmaceutical agents on FXIIA

Heparin No significant differences were observed for plasma supernatant FXIIA between the membranes (Figure 3). However, membrane-bound FXIIA was significantly higher for the control polyamide NR, with AN69S exhibiting the least activity (Figure 4). Only the polyamide NR control exhibited a significant increase in FXIIA with increasing heparin concentration and this may indicate that polyamide NR adsorbs more heparin than the other polymers (shown by the gradual rise in FXIIA with increasing plasma heparin concentration),

Fragmin No significant differences were observed for plasma supernatant FXIIA between the membranes. Membrane-bound FXIIA exhibited a similar pattern to standard heparin, but the values were significantly higher for Cuprophan and polyamide NR with increasing Fragmin concentrations.

Trisodium citrate No significant differences were observed for plasma supernatant between the membranes. The magnitude of the membrane-bound FXIIA was variable between the membranes. Cuprophan FXIIA values were higher than when using heparin or Fragmin. Polyamide membrane-bound FXIIA values for citrate were significantly reduced compared to heparin or Fragmin.

Recombinant hirudin No significant differences were observed for plasma supernatant FXIIA between the membranes. The magnitude of membrane-bound FXIIA was found to be similar to that of heparin.

Aprotinin The aprotinized plasma supernatant FXIIA values were similar to the values observed using the other pharmaceu-


1310 Contact phase activation with haemodialysis membranes: B.M. Matata et al.

Figure 3 Effect of pharmaceutical agents on plasma supernatant factor XII-like activity (FXIIA). No significant differences in plasma supernatant FXIIA were observed between the pharmaceutical agents or materials. This illustrates a minimum influence of pharmaceutical agents on plasma fluid phase contact activation; P < 0.05.

1.2

1

0.0 - Cuprophan AN& Polyamide NR ,nwa., chww, In*p”h.l” chrwd, I”.~tlr.I” &.,l.dl

Membranes

Figure 4 Effect of pharmaceutical agents on plasma membrane-bound factor XII-like activity (FXIIA). The control polyamide NR indicated a significantly higher FXIIA than Cuprophan and AN69S in all the membranes. The magnitude of membrane-bound FXIIA appeared to vary significantly with the pharmaceutical agent used for polyamide NR. However, only aprotinin expressed a significant increase in membrane-bound FXIIA on AN69S membrane; P < 0.05.

tical agents. However, for aprotinin, the membrane- bound FXIIA on AN69S was significantly higher relative to the other pharmaceutical agents. In contrast, aprotinin appeared to significantly reduce the membrane-bound FXIIA for Cuprophen and polyamide NE.

DISCUSSION

The findings in this study offer many suggestions regarding contact phase activation with respect to the in vitro evaluation of membranes and these can be summarized as follows:

(1) FXII activation/binding on membrane surfaces depends on the milieu used, i.e whether in plasma or buffer/saline solution (Figures z and 2).

(2) Plasma proteins may prevent FXII binding due to competitive binding.

(3) Some negatively charged membranes may bind plasma FXII in an unfavourable orientation or conformation for activation.

(4) Negatively charged membranes bind and activate more FXII in plasma-free saline solutions than in plasma.

(5) Anticoagulants do not appear not to play a significant role in FXII binding or activation of plasma FXII.

(6) The KI aprotinin appears to have a significant effect on activation of plasma FXII by AN69S.

These observations indicate that the magnitude of autoactivation mechanism of FXII may be strongly related to the presence of negatively charged components of a surface, whereas the superseding interaction with other plasma proteins, particularly HMWK, PK, KK, FXIa, thrombin, Cl-INH and other plasma inhibitors, influences the overall reaction kinetics of FXII activation in plasma or on the membrane surface. This may help to support the hypothesis that FXIIa in plasma (j?-FXIIa) is rapidly inactivated by other plasma constituents. In contrast, surface-bound activated FXII may be protected from inhibition by other plasma proteases, by the membrane surfaces. The magnitude of this protection may be dependent on the surface characteristics of membranes, i.e. presence of leached membrane ingredients, presence of antithrombotic agents, surface distribution of charge, hydrophilicitylhydrophobicity, polarity and porosity.

The lack of significant differences of plasma supernatant FXIIA between the various pharmaceutical aeents indicates that their effect in the plasma milieu was minimal. Membrane-bound FXIIA was different but not significantly influenced by the type of anticoagulant used. The most significant observation was the effect of aprotinin on membrane-bound FXIIA on AN69S, where the values increased instead of diminishing as observed for Cuprophan and polyamide NE. It is possible that the observations were due to increased surface binding of aprotinin on AN69S membrane leading to diminished inhibitory properties of surface-bound aprotinin on KK. This may be due to a change in conformation or the presence of only a single binding site on aprotinin. As a consequence generated KK activates more surface-bound FXII. The magnitude of KK generated becomes amplified leading to an enhanced release of BK into the plasma milieu, the consequence of which might be the onset of anaphylactoid reactions. This observation may suggest that the use of AN69S membranes in dialysis in combination with the KI aprotinin enhances the risk of adverse reactions because of the increased generation of BK. A similar observation has been reported when using ACE-inhibitor therapy (captopril) with serious implications in the development of anaphylaxis during haemodialysis with AN69 membranes’620*23~24 Further studies are therefore necessary in an effort to understand the clinical implications of using aprotinin in combination with AN69 membranes.

In the assessment of FXIIA as a parameter representative of contact phase activation in plasma, it is therefore important to evaluate the surface activity, since protein


Contact phase activation with haemodialysis membranes: B.M. Mat&a et al. 1311

adsorption (including contact factors) occurs within microseconds of bslood-biomaterial contact. The clinical significance of such an observation could be in the ominous risk from microthrombi formation on the membrane surface during dialysis in a way that affects dialyser efficiency and possible subsequent hypersensi- tivity reactions.

7

8

Chromogenic substrate assays of activated plasma clotting factors are relatively cheap, rapid and offer a moderately sensitive screening procedure. However, chromogenic substrate assays, by the nature of the mechanism involved in splitting of the substrate, are not without disadvantages as they are designed to assess proteases that have many properties in common. This also alpplies to the substrate used for this study. Although the enzyme selectivity has been enhanced by the inclusion of specific coagulation protease inhibitors, other plasma proteases such as elastase could split the substrate.

9

10

11

As an alternative, measurement of enzyme-inhibitor corn lexes by immu:noassay36*37 or radioimmunoas-

say 3f may accurately reflect contact phase activation

induced by biomaterials. It is also relevant to extend the investigation sol as to include the role of membrane-bound FXIII activity in the BK generation after plasma contact with AN69S in combination with some pharmaceutical agents such as aprotinin. A current focus is in the development of a novel immunoassay method that can detect both plasma supernatant FXIIa and FXIIa-inhibitor complexes.

12

13

24

15

ACKNOWLEDGEMIZNTS 16

The authors wish to express their gratitude to Akzo Nobel, Membrana, G.ermany, for financial support. Many thanks to the Engineering and Physical Sciences Research Council (UK) for providing a studentship to B.M.M. and the Glasgow Royal infirmary for the laboratory facilities. Many thanks to Mrs Ann Romley for her assistance in the laboratory.

17

18

19 REFERENCES

Fuhrer G, Gallimorfe MJ, Heller W, Hoffmeister, HE. FXII BZut1990; 61: 256-266. Revak, SD, Cochrane, CG, Johnson AR, Hugli TH. Structural changes accompanying enzymatic activation of human Hageman factor. Z Clin Invest 1974; 54: 619- 627. Mandle R Jr, Colm,an RW, Kaplan AP. Identification of prekallikrein and high molecular weight kininogen as a circulating complex in human plasma. Proc Nat1 Acad Sci USA 1976; 73: 41744163. De Agostini A, Li,jnem HR, Pixley RA, Colman RW, Schapira M. Inactivation of factor XII active fragment in normal plasma: predominant role of Cl-inhibitor. Z Clin Invest 1964; 73: 1542-1549. Hojima Y, Cochrane CG, Wiggins RC, Austen KF, Stevens RL. In vitro activation of the contact activation (Hageman factor) system of plasma by heparin and chondroitin sulphate. Blood 1984; 63: 1453-1459. Revak SD, Cochranse CG, Bouma BN, Griffin JH. Surface and fluid phase activities of two forms of activated

20

21

22

23

24

Hageman factor produced during contact activation of plasma. J Exp Med 1978; 147: 719-729. Radcliffe R, Bagdasarin A, Colman RW, Nemerisin Y. Activation of FXII by Hageman factor fragments. Blood 1977; 59: 611-617. Goldsmith, G, Saito H, Ratnoff OD. The activation of plasminogen by Hageman factor (FXII) and Hageman factor fragments. Z Chin Invest 1978; 62: 54-60. Ghebrewit B, Randazzo BP, Dunn JT, Silverberg M, Kaplan AP. Mechanism of activation of the classical pathway of complement by Hageman factor fragment. J CZin Znvest1983; 71(Z): 1450-1456. Nakayasha T, Nagasaw S. Studies on human kininogen I. Isolation, characterisation and cleavage by plasma kallikrein of high molecular weight kininogen. J Biochem 1979; 85: 249-258. Egberg N, Gallimore MJ, Green K, Jacobsson J, Vester- quist 0, Wiman B. Effects of plasma kallikrein and bradykinin infusions into pigs on plasma fibrinolytic variables and urinary excretion of thromboxane and prostacyclin metabolites. Fibrinolysis 1988; 2: 101-106. Gallimore MJ, Cochrane CG, Griffin JH. The purification of a plasma kallikrein with weak plasminogen activator activity. Thromb Res 1987; 12: 409420. Revak SD, Cochrane CG, Griffin JH. Multiple forms of active Hageman factor (coagulation factor XII) produced during contact activation. Fed Proc 1977; 36: 329-338. De&x FHM, Bouma BN, Schalekamp MPA, Schale- kamp MADH. An intrinsic factor XII-prekallikrein- dependent pathway activates the human plasma renin- angiotensin system. Nature 1979; 280: 315-316. Sundsmo J, Fair D. Relationships among the complement kinin, coagulation and fibrinolytic systems in the inflammatory reaction. Clin Physiol Biochem 1983; 1: 225-284. Tielemanns C, Madhoun P, Lenaers M, Schandene L, Goldman M, Vanherweghem J. Anaphylactoid reactions during hemodialysis on AN69 membranes in patients receiving ACE inhibitors. Kidney Znt 1990; 36: 982-984. Verresen L, Waer M, Vanrenterghem Y, Michielsen P. Angiotensin-converting enzyme inhibitors and anaphylactoid reactions to high-flux membranes. Lancet 1990; 336: 1360-1362. Parries E, Shapiro W. Anaphylactoid reactions in haemodialysis patients treated with AN69 dialysers. Kidney Znt 1991; 40: 1148-1152. Bronet P, Jaber K, Berland Y, Baz M. Anaphylactoid reactions during hemodialysis and hemofiltration: role of associating AN69 membrane and angiotensin converting enzyme inhibitors. Am J Kidney Dis 1992; 19: 444447. Jadoul M, Struyven J, Stragier A, van Ypersele de Strihou C. Angiotensin-converting enzyme inhibitors and anaphylactoid reactions to high-flux membrane dialysis. Lancet 1991; 337: 112. Matsuda T. Biological responses at non-physiological interfaces and molecular design of biocompatible surfaces. Nephrol Dial Transplant 1989; 4(supplement): 59-66. Vienken J. Biocompatibility: what role does the haemodialysis membrane play? Biocompatibility Monograph-EDTNA/ERCAJ 1993; 19 (Suppl. 1): 29-34. Lemke H, Fink E. Generation of bradykinin in human plasma using AN69 and PAN-17DX membranes in the presence of ACE-inhibitor in vitro. Nephrol Dial Transplant 1992; 7: 728-729 (abstract). Krieter D, Lemke H, Fink E et al. Hemodialysis related anaphylactoid reactions in sheep are associated with AN69 and not with SPAN membrane, Abstract XII.


1312 Contact phase activation with haemodialysis membranes: B.M. Mafata et a/.

25

26

27

28

29

30

31

Congress International Society of Nephrology, Jerusa- lem, 1993: 319. Kojima S, Harada-Shiba M, Nomura S et al. Effect of nafamostat mesilate on bradykinin generation during LDL apheresis using a dextran sulfate cellulose column. Trans ASAIO 1991; 37: 644-648. Olbricht C, Schaumann D, Fischer D. Anaphylactoid reactions, LDL apheresis with dextran sulphate, and ACE inhibitors. Lancet 1992; 340: 908-909. Mahiout A, Shaldon S, Koch KM. Increased adsorption of Hageman factor (FXII) to AN69 dialyser membrane: possible initial mechanism involved in AN69 adverse reactions. Nephrol Dial Transplant 1993; 8: 996-997. Sundaram S. Contact activation phase of blood coagulation during in vitro blood-biomaterial interactions and clinical applications. PhD Thesis, University of Strath- Clyde, Glasgow, UK, 1992. Yu J. Modification of polymeric biomaterials for improved biocompatibility: investigation of protein adsorption and in vitro blood response. PhD Thesis, University of Strathclyde, Glasgow, UK, 1993. Irvine L. Blood-material interactions: investigations into granulocyte elastase release and contact phase activation. PhD Thesis, University of Strathclyde, Glasgow, UK, 1989. Talbot MD, Ambler’ J, Butler KD et al. Recombinant desulphatohirudin I(CGP 393931 anticoagulant and antithrombotic properties in vivo. Throm Haemost 1989; 61: 71-80.

32 Miyamoto H, Hirose H, Matsuda H. Analysis of comple-

33

34

35

36

37

38

ment activation profile during cardiopulmonary bypass and its inhibition by FLIT-175. Trans ASAIO 1985; 31: 508-510. Heller W, Fuhrer G, Gallimore MJ, Michael M, Hoffme- ister H-E. Changes in the kallikrein-kinin system after different dose regimen of aprotinin during cardiopulmonary bypass. In: Abe K, Moriya H, Fujii S, eds. Kinins V part B. Advances in Experimental Medicine and Biology. New York Plenum, 1989: 4348. Courtney JM, Irvine L, Jones C et al. Biomaterials in medicine - a bioengineering perspective. Int J Artif Organs 1993; 16: 165-171. Tans G, Rosing J. Structural and functional characteriza- tion of factor XII. Semin Thromb Hemost 1987; 13(l): l- 14. Kaplan AP, Gruber B, Harpel PC. Assessment of Hageman factor activation in human plasma: quantification of activated Hageman factor-Cl inactivator complexes by an enzyme linked differential antibody immunosorbent assay. Blood 1985; 66: 636-641. Nuijens JH, Huijbregts CCM, Cohen M. Detection of activation of the contact system of coagulation in vitro and in viva: quantification of activated Hageman factor-Cl-inhibitor and kallikrein-(:l-inhibitor complexes by specific radioimmunoassays. Thromb Haemost 1987; 58: 778-785. Van der Kamp KWHJ, Van Oeveren W. Contact, coagulation and platelet interaction with heparin treated equipment during heart surgery. Int J Artif organs 1993; 16: 836-842.
