Elastin as a Nonthrombogenic Biomaterial

Thrombosis of Vascular Devices

Cardiovascular disease is the leading cause of mortality worldwide and is expected to remain so in this century. Vascular implants and devices manufactured from metal alloys and various polymers have become increasingly common treatments. Components of stents, pacemakers, and heart valves are made using metal alloys such as stainless steel. Synthetic grafts, catheters, and pacemaker leads are made from a wide variety of polymers such as polystyrene. ¹ These devices are frequently associated with significant adverse clinical outcomes, including inflammation, hyperproliferation, and thrombosis, due to the vascular incompatibility of the implant materials. While inflammation and cellular hyperproliferation are problematic, the primary cause of short-term failure of heart valves, ² coronary stents, ³ and small-diameter synthetic vascular grafts ⁴ is thrombosis. For example, the highly thrombogenic nature of metals ⁵ and polymers ⁶ used to make stents translates to an alarming mortality rate of 25% in the case of thrombotic stent occlusion. ⁷ This is currently managed with anti-platelet medication, which increases bleeding risks and induces variable pharmacological responses in patients, while serving only to minimize rather than prevent thrombosis caused by these implants. ⁸ Poor hemocompatibility has plagued the development of vascular devices and there remains a pressing need for nonthrombogenic materials that can reduce the reliance on pharmacological intervention.

Biomaterials and Thrombogenicity

The development of a truly nonthrombogenic biomaterial has been elusive due to the complex nature and interplay of coagulation, thrombosis, and inflammation that occurs after implantation of foreign materials. Understanding these processes will aid the development of compatible materials.

Classical coagulation pathways

Thrombus formation in vivo results from two enzyme cascades, the extrinsic and intrinsic pathway (Fig. 1). The extrinsic pathway is initiated by the activation of tissue factor following the physiological response to diseased or damaged endothelium and exposure of blood to the subendothelium (Fig. 1, green section, bold text). The intrinsic pathway is activated by contact with implanted materials and subsequent conformational changes in plasma proteins (Fig. 1, blue section, regular text). Ultimately, these pathways converge on a common pathway, which is defined by the proteolytic conversion of prothrombin to active thrombin (Fig. 1, yellow section, italic text). Thrombin catalyzes the conversion of fibrinogen to fibrin, which in turn spontaneously forms crosslinked fibrils resulting in a fibrin clot. ⁹

OPEN IN VIEWER

FIG. 1.

Modified coagulation cascade. An overview of the intrinsic, extrinsic (bold), and common (italics) coagulation pathways. The factors that platelets activate are denoted by circles (), and the factors that various complement proteins activate are denoted by open squares () (modified from Refs.^13,14,22). Color images available online at www.liebertonline.com/teb.

Elastin in the Vasculature

Elastin is a dominant component of the vasculature, having both important mechanical and cell signaling roles. Tropoelastin, the precursor of elastin, is produced by endothelial cells and vascular smooth muscle cells (VSMCs) in blood vessels. In the extracellular space, tropoelastin is exposed to lysyl oxidase-mediated oxidation at exposed lysines and then self-associates by coacervation for cross-linking. ¹⁶ The growing elastic polymer is incorporated with microfibrillar proteins to form mature elastic fibers. ¹⁷ Elastin accounts for ∼50% of the dry weight of the arterial wall where it is confined to specific locations based on its function. ¹⁸ Elastic arteries are comprised of three layers, the tunica intima, media, and outermost adventitia (Fig. 2). The adventitia is characterized by elastic fibers interspersed with collagen, providing mechanical strength, recoil, and resilience. The media is predominantly comprised of alternating layers of elastin-rich lamellae interspersed with smooth muscle cells, which are responsible for maintaining blood pressure during pulsatile flow.^19–21 Of particular relevance to the potential for blood contact, a distinct continuous sheet of elastin forms the internal elastic lamina, providing a surface on which endothelial cells form a monolayer and separating the intima from the tunica media (Fig. 2). ²²

OPEN IN VIEWER

FIG. 2.

Elastic artery schematic. Schematic representation of an elastic artery showing the three layers and major components of these layers (modified from Ref. ²² ). Color images available online at www.liebertonline.com/teb.

Elastin interacts with many cell types and causes a variety of cellular responses.^23–25 The major cell types in the vasculature are endothelial cells present in the tunica intima, and VSMCs present in the tunica media. Studies of the elastin knockout mouse revealed that elastin is essential for arterial morphogenesis and for controlling smooth muscle cell proliferation. The null mutation is lethal with the mice dying from obstructive arterial disease, caused by the uncontrolled proliferation of VSMCs, by day 4 postpartum. ²⁶ Elastin inhibits VSMC proliferation by inducing a quiescent contractile phenotype through its VGVAPG motif.^26–29 In contrast, various forms of elastin interact favorably with endothelial cells and support their growth.^30–33 Elastin peptides are chemotactic for endothelial cells ³⁴ and are consequently pro-angiogenic. ³⁵ These properties are not readily utilized because elastin peptides are also chemotactic for additional cell types, including fibroblasts and monocytes and can lead to proinflammatory responses. ³⁶ However, intact elastic fibers have an opposite effect, demonstrating anti-inflammatory properties. ³⁷

In addition to modulation of cell signaling, tissue-engineered elastin constructs and elastin-based materials have the potential to mimic the essential mechanical properties of major vessels. Synthetic elastin scaffolds have been shown to be structurally similar to native elastic fibers (Fig. 2 in Ref. ²⁵ ) and to have comparable elasticity, demonstrating a youngs modulus of 220–280 kPa compared with 300–600 kPa for native elastin. ²⁵ Further incorporation of synthetic elastin into a vascular conduit facilitated the close matching of the conduits compliance to that of a native artery (320 ± 25 kPa vs. 267 ± 46 kPa for the native). ³⁸ In addition to this work elastin and elastin-like materials have been coated onto surfaces, made into hydrogel materials and spun into fibers.^21,39 However, the broad use of naturally occurring elastin is limited due to the insolubility of the mature cross-linked biopolymer and the necessity to purify it from animal sources. Elastin-rich tissues such as large arteries and bovine ligamentum nuchae can be treated with solvents, cyanogen bromide, denaturing agents and selected proteases to remove the cellular and extracellular matrix components, leaving the insoluble elastic network intact. However, to utilize elastin in biomaterial construction it must be rendered soluble by acid hydrolysis or enzyme degradation. ⁴⁰ Alternatively, intact full length material in the form of recombinant human tropoelastin can be purified from substantial overexpression ⁴¹ and has been used to construct a diverse range of materials. ⁴² Additionally, synthesized elastin-like polypeptides are increasingly being used for construction of biomaterials ⁴³ as well as for a range of medical applications. ⁴⁴ This review assesses the existing knowledge of these various forms of elastin and their interactions with blood.

Thrombogenicity of Elastin

The interactions of elastin with components of the blood have not been fully characterized; however, key research points to elastin being nonthrombogenic. ⁴⁵ Research before the 1980s revealed that elastin isolated directly from blood vessels had anti-thrombogenic properties.^46,47 Soluble bovine elastin and elastin directly purified from rabbit aorta had no reaction with platelets from heparinized platelet-rich plasma (PRP). ⁴⁶ This was surprising given the profound thrombogenicity of other vascular matrix components such as collagen. ⁴⁷

Continuing these initial findings, an in situ perfusion system was used to study platelet adhesion and platelet thrombus formation on components of the vessel wall. ⁴⁸ Rabbit arteries were either intact or decellularized to varying degrees to reveal different vessel components on the surface. Inverted arteries displaying adventitial collagen, or endothelium denuded arteries displaying the basement membrane, caused the most platelet adhesion. Arteries digested with collagenase, exposing the microfibrils of the subendothelium were less thrombogenic. Arteries displaying elastin, remaining intact after collagenase and trypsin digestion, showed the least amount of platelet adhesion. ⁴⁸ Further studies showed that elastin isolated from rabbit or pig aorta caused minimal platelet adhesion and no spreading, aggregation, or degranulation, which is characteristic of platelet activation.^49,50 Later research confirmed that the elastin component of elastic fibers is responsible for this lowered thrombogenicity, whereas the microfibrillar components were shown to cause platelet activation, though to a much lower extent than collagen. ⁵¹

These findings were extended in studies of exposed canine femoral arteries in vivo. The arteries were treated proteolytically to remove the endothelium. The now-exposed elastin layer caused a substantial delay in thrombus formation, with thrombus occurring from 7 to 35 days compared to arteries that were injured to destroy the elastic lamina that thrombosed within 48 h. ⁵² Overall, this research demonstrated that elastic fibers from decellularized vessels are the least thrombogenic protein of the major subendothelium components tested (Table 1).

Elastin as a Blood-Contacting Biomaterial

Given the compelling nature of the experiments to date, researchers increasingly sought to adapt elastin for use as a biomaterial. There is great potential to utilize its unique elastic mechanical properties to construct synthetic vascular grafts or take advantage of its signaling properties as a coating for vascular devices. Examples of elastin being used as a biomaterial are discussed below with emphasis on the thrombogenicity of the materials.

The first indication that elastin had the potential to be used to construct nonthrombogenic biomaterials came from in vivo implantation of elastin scaffolds. Elastin purified from porcine common carotid arteries was used to construct a stent graft that placed elastin at the site of an injured vessel to reduce VSMC hyperproliferation. Although hemocompatibility was not the main end point of this study, there was no observed thrombotic or inflammatory response. ²⁹ Additionally, scaffolds of elastin purified from rat aortas were implanted back into rat as abdominal grafts. These had been prepared by treatment with alkali to reveal distinct vascular wall surfaces. This approach showed that the elastin matrix caused the least leukocyte and platelet adhesion compared to basal lamina or adventitial collagen. ⁵³

There have been two studies in which elastin scaffolds were constructed into vascular grafts to directly assess their in vivo performance. The first study utilized purified elastin and collagen scaffolds obtained from porcine aortas and common carotid arteries. The elastin scaffolds showed decreased platelet adhesion and aggregation compared to collagen scaffolds in vitro using PRP. When implanted in a rabbit carotid artery, the elastin-based scaffolds displayed substantially reduced platelet adhesion. ⁵⁴ The second study utilized grafts constructed from purified porcine arterial elastin and fibrin-bonded layers of acellular small intestinal mucosa, used in a porcine carotid artery interposition model to directly test acute thrombogenicity. The elastin-containing grafts remained patent for significantly longer and only had thrombosis associated with the suture line, compared to commercially available expanded polytetrafluoroethylene, which had thrombus along the full length of the graft. ⁵⁵ This research was particularly significant because it was the first to show a reduction in acute thrombus formation in vivo (Table 1). These animal studies support the logic of using elastin in the construction of vascular grafts, taking advantage of the properties of elastin to provide essential mechanical properties to mimic native arteries structurally and provide a nonthrombogenic luminal surface. Indeed, this approach has been employed to construct a vascular graft incorporating recombinant human tropoelastin and polycaprolactone, with mechanical properties matching a human mammary artery and a 100% elastin luminal layer. The elastin-rich lumen enhanced endothelial cell attachment and proliferation and had low platelet interactions in vitro. ³⁸

While the mechanism for elastin's nonthrombogenicity has not been fully elucidated, research into the thrombogenicity of other matrix proteins is illustrative of potential mechanisms. Fibronectin and collagen, both thrombogenic matrix proteins, specifically bind multiple platelet integrins.^10,56 Elastin and derivatives do not contain an integrin binding arganine, glycine, aspartic acid (RGD) sequence ²⁴ and this lack of an RGD sequence could also account for its low platelet binding ability. A second consideration is that platelets can bind through carbohydrate modifications on proteins. ⁵⁷ Elastin is not post-translationally modified with carbohydrates. ⁵⁸ Similarly, albumin displays low thrombogenicity and contains very little carbohydrate modification.^57,59 It may be that a paucity of glycosylation works to elastin's advantage by reducing its platelet binding capacity.

Another potential mechanism involves inhibition of thromboxane B₂-induced platelet aggregation. Soluble elastin and elastin fragments inhibited thromboxane B₂ in nonactivated and thrombin-activated platelets, though high concentrations of elastin were used. ⁶⁰ A mechanism of thromboxane B₂ inhibition is suggested to be by a decrease of free cytosolic Ca²⁺ in platelets. ⁶¹ However, how elastin could decrease intracellular platelet calcium was not explored.

As it is known that insoluble elastin binds calcium, ⁶² calcification could be a concern in the context of vascular biomaterials. Elastin abnormality in diseased arteries is one of many factors that contribute to calcification in atherosclerosis ⁶³ ; however, this is likely due to microfibrillar components that are known to bind calcium. ⁶⁴ There is evidence that partially purified elastin isolated from vessels can calcify^39,65,66; however, this can also be explained by contaminating microfibrillar components.^40,64 In contrast, when purified components such as tropoelastin and intact purified elastic fibers are used, they inhibit calcification in smooth muscle cells.^63,67 The importance of purity is further confirmed by the recognition that elastin-based peptides inhibit calcification. ⁶⁸ Therefore, for use as a biomaterial, impure elastin should be avoided due to potential contaminants. Instead, the use of pure, soluble material would be preferred; this could be in the form of tropoelastin, or elastin-based peptides.

Elastin-Based Peptides and Tropoelastin

Elastin-based peptides bound to surfaces have already been shown to impart reduced thrombogenicity (Table 1). In the first such study to look at the thrombogenicity of elastin peptides, polyGVGP was photochemically immobilized onto silicone rubber; this surface showed decreased fibrinogen adsorption, ⁶⁹ indicating a reduced capacity for fibrin clot formation. Elastin peptides containing exons 20–24 of the human protein were also coated on to catheters and vascular conduits. When incubated with PRP these showed decreased platelet adhesion and aggregation. Reduced platelet activity was further confirmed by decreased sP-selectin and microparticle release, which are markers of platelet activation. In vivo, these elastin peptide-coated catheters remained patent for a longer time period than uncoated controls and showed decreased fibrin accumulation and thrombosis. ⁷⁰ Further research revealed that increasing the molecular weight of these elastin-like peptides reduced fibrinogen adsorption and platelet adhesion. ⁷¹ Of clinical relevance, expanded polytetrafluoroethylene grafts coated with recombinant elastin-mimetic triblock polymers were assessed in a baboon extracorporeal femoral arteriovenous shunt model, in which a section of an exposed artery was replaced with the graft. These grafts demonstrated increased patency, accompanied by decreased platelet adhesion and lower fibrin deposition. ⁷²

Similarly, we have recently shown that recombinant human tropoelastin ⁴¹ decreased the time to thrombus formation upon exposure to static and flowing whole blood when bound to metallic ⁷³ and polymeric substrates (Bax et al. in preparation) (Table 1). For optimal use of elastin and its derivatives in biomaterials, pure tropoelastin and elastin-based peptides should be used to specifically exploit their biological and mechanical features. Further, determining the mechanism responsible for elastin's hemocompatibility would allow these goals to be accomplished, while avoiding the risk of calcification.

Conclusion

There is mounting evidence for the hemocompatibility of elastin and its derivatives. This is demonstrated by examination of its limited interactions with platelets and by a reduction in thrombus formation both in vitro and in vivo. Elastin displays lowered thrombogenicity in various forms, including elastic fibers, purified elastin, tropoelastin, and elastin-based peptides. Additionally, it is a promising candidate for use as a vascular biomaterial or in tissue-engineered constructs due to persistent mechanical properties, favorable association with endothelial cells, and inhibition of smooth muscle cell proliferation. The next generation of elastin-based constructs for use as biomaterials should encompass knowledge of the functional regions of tropoelastin, including delineated nonthrombogenic and cell interactive regions of the molecule, to generate constructs that elicit the desired biological responses.