Figure 2 shows isotherms for AT adsorption from buffer to modified and unmodified polyurethanes. Heparin-modified catheters (CBAS) were used as a commercial heparinized surface for comparison with the heparin-containing ATH-coated catheters. Adsorption increased with increasing protein concentration and saturation was not reached at the highest concentration used (0.15 mg/mL). Adsorption levels were substantial for all three surfaces, with the heparin-coated surface showing the highest adsorption, followed by the unmodified polyurethane (commercial control for comparison to the commercial heparin-coated catheters) and the ATH surface, which showed the lowest adsorption. At the highest concentration, binding ranged from 6 to 9 pmol/cm2. For a protein of MW ∼60,000 Da, these amounts are in the order of those expected for monolayers.
Figure 2. Isotherms for AT adsorption from buffer to modified and unmodified polyurethanes (PU) at room temperature for 2 h. The heparin-modified surface was used as a reference. PU-ATH, polyurethane-antithrombin-heparin covalent complex surface; PU, uncoated polyurethane surface; PU-H, polyurethane-heparin surface. Values are mean ± SD (n = 3).
Download figure to PowerPoint
Adsorption of AT from plasma was expected to show differences between the heparinized and nonheparinized surfaces because of the relatively high affinity of AT-heparin interactions. The other much more abundant plasma proteins were expected to reduce low affinity binding (similar to blocking of low affinity binding in ELISAs by albumin), as opposed to the high affinity AT-activation sites on heparins. To exhibit surface activity, the heparin molecules must be in an orientation that favors their contact with fluid-phase AT.44–48 Thus, any increase in binding of AT on heparin or ATH surfaces may also be an indication of the functional availability the heparin moiety on the surface.
Figure 3 shows AT adsorption from plasma. The polyurethane (PU)–ATH surface showed the highest adsorption, reflecting availability of greatest number of high AT-affinity sites. The heparinized surface also showed significant AT adsorption from plasma, while the unmodified PU surface showed a value close to background. Comparison of Figures 2 and 3 shows that when physiological concentrations of plasma protein was present, while total binding of AT to all surfaces was reduced, (PU)–ATH showed a significant trend towards greater AT-binding relative to uncoated surfaces (p < 0.05, Fig. 3). Although binding of AT from a plasma-milieu by heparin-coated surfaces was measurable, the amount was not statistically different from that of uncoated PU.
Figure 3. Adsorption of AT from plasma to polyurethane-based surfaces (room temperature, 2 h). PU-ATH, polyurethane-antithrombin-heparin covalent complex surface; PU-H, polyurethane-heparin surface; PU, uncoated polyurethane surface. Values are mean ± SD (n = 3). * denotes a significant difference (p < 0.05) in AT binding compared to the PU control surface.
Download figure to PowerPoint
SDS-PAGE and Western blots
As a complement to the measurement of adsorption from plasma using radiolabeled AT, SDS-PAGE and immunoblotting analysis were performed on the proteins eluted from coated catheters after contact with rabbit plasma. The data from these experiments provided a more comprehensive picture of the overall composition of the protein layer deposited onto the surfaces. It should be noted that for these experiments: identical catheter areas were used for all analyses, identical volumes of elutant were used, and identical volumes of eluate were loaded on the gels. Thus, the intensity of the bands on the gels and blots gave an indication of the relative quantities of the proteins adsorbed to the different surfaces. However, interpretation of relative protein binding between the commercial heparin surface and ATH–PEO surface is confounded by the potential difference in heparin chains, amounts of surface-bound heparin, base coats, and PEO linker. Thus, the discussion below is guarded on comparisons between catheter types.
Some inherent difficulties that exist regarding the semiquantitative analysis of eluted proteins on SDS PAGE/ Western immunoblots should be noted. For example, high molecular weight and less polar molecules may have slightly reduced yields as more points of detergent contact are required for suspension.49 Thus, a micellular interaction or entrapment mechanism may be involved for removal of these proteins. In fact, for some strongly hydrophobic surface–protein interactions (such as for membrane proteins), a combination of chaotropic agents + detergents has been commonly used,50–52 as well as organic solvents in some cases.53, 54 Nevertheless, SDS is a strong ionic, amphiphobic detergent that has been recommended as one of the most efficient surfactants52 and is commonly used for plasma protein solubilization prior to immunoblotting.55 In fact, as in the present study, we have previously reported data that carefully detailed significant differences between newborn and adult plasma protein binding to a variety of surfaces using this method.41 Polyacrylamide gels of ATH (or heparin) mixed with plasma, followed by buffered SDS at room temperature, show no binding of Coomassie stained proteins to the ATH or heparin. Furthermore, as a proof that washing with SDS is complete, repeat SDS washings of the SDS-treated surfaces did not remove any further detectable protein. Still, variation of elution conditions, detergents, or solvents might theoretically have helped to remove any small amounts of surface-bound protein still remaining. Although analysis by elution and immunoblotting might not be preferred to direct antibody detection on the catheter surface, design and standardization of a direct method is difficult. However, we are in the midst of developing such a technique56 which will hopefully confirm results reported here.
It appeared from the blots shown in Figure 4 that the heparinized and uncoated polyurethane control catheters adsorbed greater amounts of protein than the catheter coated with ATH–PEO (note: right lanes without immunodevelopment, where total protein profiles are shown). Thus, the ATH–PEO surface may have reduced overall affinity for plasma proteins (densitometric analysis of total protein band staining indicated that the ATH–PEO-coating adsorbed 36% and 31% of the amount of protein taken up on heparin-coated and uncoated surfaces, respectively). Antibodies against prekallikrein, fibrinogen, plasminogen, AT, albumin, α-2-macroglobulin, thrombin, heparin cofactor II, and vitronectin were used in the blotting experiments. In general, the protein deposition patterns were complex, and most of the proteins probed for were found to be present on all three materials. However, some differences were observed in the types of proteins bound by the three surfaces, for example, no AT was detected on the uncoated control. Also, fragments of certain proteins were variously detected, for example, fibrin(ogen) degradation products were observed on uncoated and heparin-coated surfaces.
Figure 4. Western blots of proteins eluted from the polyurethane (PU) catheters, heparin-coated catheters (CBAS), and ATH-coated catheters after contact with plasma. Reduced SDS-PAGE (12% gels) was used to separate proteins eluted from catheter segments incubated with rabbit plasma, blotted onto membranes, and the blots cut into 3-mm wide strips. The strips were separately incubated with primary antibodies against individual rabbit plasma proteins, followed by alkaline phosphatase-conjugated anti-IgG. Bands were then visualized by incubation with NBT/BCIP substrate. The lanes at the extreme left in each of the three blots are molecular weight (MW) standards while the lanes to the right are gold-stained SDS-PAGE gels of the eluates.
Download figure to PowerPoint
The behavior of AT was of particular interest since the anticoagulant activity of heparin is mediated by this protein. Both the ATH-coated catheter and the heparinized catheter showed significant adsorption of AT (Fig. 4, circled regions). However, no AT was detected on the uncoated control catheter. Multiple bands were seen in the AT region for the heparinized catheter, possibly due to the well-known heterogeneity in AT glycosylation.57 The ATH modified surfaces showed only a single AT band, suggesting that only one form of AT was adsorbed; the lower molecular weight of this band being suggestive of β-AT isoform, which has a higher affinity for heparin binding sites.57
The high molecular weight AT-containing band from the heparinized surface is typical of the 90 kDa thrombin–antithrombin (TAT) complex58–60 (generated either in vitro or in the animal prior to blood collection). The presence of TAT revealed that thrombin generation had taken place in a procoagulant environment. Since release of TAT complex is required for the continued catalytic function of the heparinized surface, retention of TAT on the heparinized surface (CBAS, Fig. 4) would limit the heparin's ability to catalyze further thrombin inhibition (hinder approach of more incoming AT to the surface). In the case of the ATH-coated surface, interactions with thrombin were complicated. As previously demonstrated,32, 33 two separate reactions can occur involving thrombin and ATH. The first reaction is the rapid direct inhibition of thrombin by attaching covalently to the AT moiety in ATH to form an inert thrombin–ATH complex.31 This reaction occurs only once, since ATH can only form an irreversible complex with one thrombin molecule. Unfortunately, since the covalent thrombin–ATH complex cannot dissociate, the bound thrombin could not be eluted by buffer and observed on the Western blot (Fig. 4). The second reaction involves the fact that the heparin chain in either ATH or thrombin–ATH has the ability to bind AT from plasma and catalyze the reaction of that absorbed AT with incoming thrombin to form thrombin–AT (TAT).32, 33 Once formed, TAT can dissociate from the ATH heparin so that the heparin chain is free to bind and activate more plasma AT for reaction with thrombin. Evidence for both direct and catalytic inhibition of thrombin by ATH has been shown with ATH-coated grafts43 and proof that the plasma AT-activating reaction can occur was given in the present study by the fact that catalytic anti-Xa activity was measured on the ATH catheters (Table I, see below).
Table I. Antifactor Xa (FXa) Activity and Antithrombin (AT) Adsorption on Different Polyurethane Catheters
| ||Anti-Xa Activity (pmol/cm2)||AT Adsorption from Buffer (pmol/cm2)||Ratio (Activity/Adsorption)|
|PU-Control||0.00||7.68 ± 1.77||0|
|PU-Heparin||8.42 ± 10.00*||9.55 ± 2.06||0.91 ± 0.20*|
|PU-ATH||10.36 ± 5.87*||6.37 ± 0.84*||1.65 ± 0.22*|
Fibrinogen (molecular weight, 340 kDa) plays a critical role in the formation of fibrin clots. It is converted to fibrin monomer by the action of thrombin, which then polymerizes to form a fibrin clot. Intense bands of intact fibrinogen, as well as fibrin(ogen) fragments, were found on the unmodified control and heparinized catheters, suggesting that both fibrinogen deposition and denaturation/degradation (thrombin activity) had occurred on those surfaces. Relatively small amounts of intact fibrinogen and fibrin(ogen) fragments were present on the ATH-coated catheter, suggesting a reduced surface procoagulant activity (thrombin generation). A potential reason for coagulant activity on the heparin surface is that, since partial depolymerization of heparin is required to get active end groups for heparin coating,39 some chains may be too short to simultaneously bind both the AT and thrombin to neutralize thrombin activity via a bridging mechanism.61
Data on anti-FXa activity for the surfaces studied are listed in Table I. Both the polyurethane heparin-coated (PU-H) and the polyurethane ATH-coated (PU-ATH) materials showed significant anti-FXa activity, reflecting the presence of functional heparin on these surfaces. The control polyurethane material (PU-control) showed no activity. Previous characterization of the ATH-coated and heparin-coated catheters using periodate has verified that the anticoagulant activity in ATH-coated and heparin-coated catheters is entirely due to the pentasaccharide sequence.36 In the case of ATH, exogenous AT can noncovalently bind (and be activated by) the heparin pentasaccharide site on the heparin chains of ATH, even though the conjugated AT in ATH remains covalently attached to the aldose terminus of the heparin moiety.33 Thus, although FXa reacts to produce a FXa-ATH inhibitor complex (or thrombin-ATH in the case of thrombin assays), the ATH heparin chain still retains the ability to potently catalyze reaction of more FXa with exogenous AT.31, 33 The rationale for analysis in the activity assay relies on the validity for use of the standard heparin in the kit as a reference. ATH is a covalent 1 mol:1 mol complex of AT and heparin,31, 33 where the catalytic anti-FXa activity resides only in the heparin moiety31 (thus making it justifiable to refer to standard heparin in the kit). Consequently, because the assay measures the amount of heparin-like activity in the sample that catalyzes inhibition of FXa by added AT, the activity could be calculated as the pmol of active surface-bound ATH in terms of equivalent pmol of standard heparin from the kit. In the case of PU-heparin, a similar calculation was done to get pmol of standard heparin anti-FXa activity on the surface.
Anti-FXa activity and AT adsorption, as determined in this study, provide measures of different aspects of heparin activity. The assay used for anti-FXa activity on a surface measures the number of heparin sequences that can activate AT to give a conformation which reacts rapidly with FXa.62 AT adsorption from plasma measures the number of AT molecules that interact with a minimum level of affinity for the surface (especially found for pentasaccharide-containing heparin molecules).63, 64 Therefore, it is of interest to compare anti-FXa activity and AT adsorption. The anti-FXa assay gave the “turnover” rate of heparin related to the catalysis of inhibition by bound AT. For the heparin-modified surface, the anti-FXa assay showed a value similar to the AT adsorption (Table I). For the ATH-modified surface, the anti-FXa assay gave a higher value than the AT adsorption, a result significantly different to that of the heparin-coated surface (p = 0.023 (2-sample students t-test) for activity/adsorption ratios of PU-heparin versus PU–ATH in Table I). With respect to AT adsorption, it is highly likely that adsorbed AT will remain bound if no further interactions occur with either FXa or thrombin.64 In that case, the maximum quantity of AT that could be adsorbed would simply be that of a monolayer (assuming the binding site density is sufficient). To exhibit anti-FXa activity, however, the bound AT must be converted to its activated conformation, interact with FXa, and the resultant inhibitor complex must be released from the surface.27 If some molecules on the surface could bind AT but not effectively activate the AT to inhibit FXa, compared to the reference heparin, then the ratio of pmol of surface heparin that could inhibit FXa (by AT) to the total pmol of AT that the surface can absorb would be less than 1. In the case that the ratio of pmol of active heparin that catalyze FXa inhibition to pmol of AT bound is greater than 1 (as with the ATH surface), one implication might be that the surface contains molecules that can activate a higher proportion of adsorbed AT for FXa inhibition than the heparin used as a standard. Such a finding was not surprising for ATH since, during synthesis of the conjugate, AT molecules select for heparin chains with the highest AT-affinity and pentasaccharide number to generate the anticoagulant complex.31, 32
Several factors may account for the decreased ratio of anti-FXa activity/bound AT in PU-heparin compared to ATH surfaces. Only one-third of the heparin chains in heparin have active AT-binding sequences,65 whereas in ATH all heparin chains contain at least one pentasaccharide sequence for activation of AT molecules taken up from plasma.32 Furthermore, heparin in the CBAS catheters was covalently bound to an amine-rich coating on the polyurethane catheters via end-point attachment66 through aldehydes generated by partial depolymerization of UFH with nitrous acid.39 Thus, although high substitutions of heparin up to 90.7 ± 4.9 μg/cm2 has been obtained by such oxidative methods,66 AT-activating pentasaccharide sequences may have been degraded. However, as stated above, since PEO is used in the ATH but not the heparin coating and the commercial heparin and ATH surfaces have different base coats, many other potential explanations exist for variation in AT-binding site activities. To resolve this issue, it will be necessary to coat heparin and ATH to the same catheter type with the same linking agent.