Purinergic Signaling in Early Inflammatory Events of the Foreign Body Response: Modulating Extracellular ATP as an Enabling Technology for Engineered Implants and Tissues

Abstract

Purinergic signaling is a ubiquitous and vital aspect of mammalian biology in which purines—mainly adenosine triphosphate (ATP)—are released from cells through loss of membrane integrity (cell death), exocytosis, or transport/diffusion across membrane channels, and exert paracrine or autocrine signaling effects through three subclasses of well-characterized receptors: the P1 adenosine receptors, the P2X ionotropic nucleotide receptors, and the P2Y metabotropic receptors. ATP and its metabolites are released by damaged and stressed cells in injured tissues. The early events of wound healing, hemostasis, and inflammation are highly regulated by these signals through activation of purinergic receptors on platelets and neutrophils. Recent data have demonstrated that ATP signaling is of particular importance to targeting leukocytes to sites of injury. This is particularly relevant to the subject of implanted medical devices, engineered tissues, and grafts as all these technologies elicit a wound healing response with varying degrees of encapsulation, rejection, extrusion, or destruction of the tissue or device. Here, we review the biology of purinergic signaling and focus on ATP release and response mechanisms that pertain to the early inflammatory phase of wound healing. Finally, therapeutic options are explored, including a new class of peptidomimetic drugs based on the ATP-conductive channel connexin43.

Introduction

The injury response, including sterile surgical wounds, is an integral consideration for any tissue-engineered repair. It has been suggested that in tissue engineering, many problems can be traced back to scarring at the culmination of the healing process.¹ Injury produces a complex and highly orchestrated sequence of events as the body attempts to restore tissue to its preinjury state.² Wound healing is divided into three overlapping, well-choreographed, event-driven phases: hemostasis-inflammation, proliferation, and scarring-remodeling.^3,4 Normal resolution of inflammation in small defects results in low levels of scarring and essentially restored tissue function and histology. Larger wounds can transition to chronic inflammatory states, which in turn tends to produce more scarring and a concomitant loss of normal tissue histoarchitecture and function.^5,6

Surgical engraftment of implanted devices, tissues, and engineered constructs into the body will elicit an injury response, including acute inflammation, tissue remodeling, and scarring at the site of implantation. This multifaceted response involves the participation of a number of inflammatory and fibrotic cell types including mast cells, neutrophils, monocytes/macrophages, lymphocytes, and fibroblasts.⁷ In particular, long-term, chronic inflammation mediated by macrophages has been an active area of recent research (see McNally and Anderson for an excellent review⁸). Activated macrophages adhere to implanted biomaterials and fuse to form multinucleated foreign body giant cells that persist at the implant site and promote granulation tissue formation, eventually leading to fibrous encapsulation.^7,8

Complicating this is the issue of macrophage “polarization.” Macrophages can assume multiple phenotypes that can be broadly characterized as M1 “classically activated” or M2 “alternatively activated” phenotypes. In general, M1 macrophages are proinflammatory, while M2 macrophages are anti-inflammatory and promote regeneration (this topic is extensively reviewed by Brown et al.⁹). Uniquely, this review will focus on the equally important initial events that lead to chronic inflammation and the formation of a fibrotic capsule–namely, hemostasis and the neutrophil response.¹⁰

A number of paracrine/autocrine signaling pathways are known to initiate, regulate, and dampen the inflammatory response—primarily (anti-)inflammatory cytokines such as interleukin-6 (IL-6), IL-8, macrophage inflammatory protein-2 (MIP-2), and IL-10. However, recent research has brought to light the key role that an alternative signaling pathway plays in the inflammatory response. Namely, purinergic signaling, or signaling through extracellular nucleotides/sides, is involved in hemostasis and recruitment of inflammatory cell types.

The primary mediator in purinergic signaling is adenosine triphosphate (ATP), which is a biological molecule most commonly thought of for its roles in intracellular energy transfer and as a precursor for deoxyribonucleic acid (DNA). Further, ATP is at the forefront of biology as a substrate for kinase reactions in signal transduction pathways, and adenylate cyclase, which converts ATP to the second messenger cyclic adenosine monophosphate (AMP). However, evidence accumulated over the past 40 years has established that ATP can also act as a neurotransmitter and paracrine/autocrine signaling molecule.¹¹

In mammalian systems purinergic signaling is a crucial component of numerous (patho)physiological phenomena, including synaptic transmission,¹² regulation of vascular tone,¹³ inner ear function and hearing,¹⁴ cell volume regulation,¹⁵ ischemic responses in the heart¹⁶ and brain,¹⁷ propagation of intercellular calcium waves,^18,19 the spread of apoptotic cell death,²⁰ kidney function,²¹ hypoxic preconditioning,²² Alzheimer's Disease,²³ cell proliferation,²⁴ memory,²⁵ cancer,²⁶ embryonic development,²⁷ nociception,²⁸ and inflammatory diseases.²⁹ In addition to these many roles for ATP, recent data indicate a crucial role for extracellular ATP in inflammatory cell targeting to sites of injury.^30–33 The focus of this review will be purinergic signaling in the context of injury and the resulting acute inflammation, and how this affects the host response to implanted devices and engineered tissues. Although little data exist on the role of extracellular ATP in macrophage polarization, for readers interested in ATP signaling in chronic inflammation Lemaire et al. provide a first-rate review of purinergic signaling in the process of giant cell formation.³⁴ The purpose of this review is to communicate the importance of this largely ignored signaling mechanism to the tissue engineering and surgical implant community.

Basic Biology of Purinergic Signaling

ATP release mechanisms

Cell death

ATP exists in the cytoplasm of cells at high concentrations, ranging from 1–10 mM.³⁵ Therefore, one of the most common and important mechanisms of ATP release is simply cell rupture. Given that some purinergic receptors have an EC₅₀ for ATP as low as 1 μM it is clear that rupture of even a small number of cells can elicit a response in nearby viable cells.^13,36

Exocytosis

ATP has also been shown to be present in synaptic vesicles.^37,38 Evidence that ATP was released by exocytosis of synaptic vesicles first came from experiments in which exogenously applied quinacrine, a fluorescent antimalarial known to bind ATP,³⁹ was observed to decrease in the cytoplasm of nerves undergoing depolarization.⁴⁰ Recent work by Sawada et al. has identified a protein termed the vesicular nucleotide transporter (VNUT) that they demonstrated to mediate vesicular ATP uptake.⁴¹ Importantly, they showed that PC12 cells, reported to secrete ATP through exocytosis,⁴² displayed reduced ATP release when treated with VNUT siRNA,⁴¹ strongly indicating that exocytosis is a viable mechanism of ATP release. Importantly, a number of other cell types, including endothelial cells, display exocytosis of ATP.^43,44 Endothelial cells are of particular interest for this review because of their access to the blood vessel lumen from which inflammatory cells can be recruited to sites of injury, implanted devices, or tissues.

ABC transporters

ATP-binding cassette (ABC) transporters are a superfamily of membrane proteins that hydrolyze ATP to power the unidirectional efflux of a large variety of molecules.⁴⁵ Perhaps the most well-known of the ABC transporters is the chloride ion transporter, cystic fibrosis transmembrane conductance regulator (CFTR), mutations of which are responsible for the cystic fibrosis genetic disorder.⁴⁶ In addition to other ABC transporters, the CFTR has also been shown to mediate outward ATP currents.⁴⁷ This mechanism of ATP release potentially plays an important role in injured tissues as the CFTR is expressed in a number of relevant cell types including vascular endothelial cells,⁴⁸ red blood cells,⁴⁹ platelets,⁵⁰ and skeletal muscle.⁵¹

Connexin hemichannels

Connexins are a family of integral membrane proteins that form intercellular channels located in plaque-like structures called gap junctions (GJs) that form at points of cell–cell contact.⁵² These channels facilitate the diffusion of small molecules (<∼1000 Da) and ions between cells. The intercellular channels are formed by the interaction of two half-channels (one from each cell), called hemichannels.

Hemichannels newly delivered to the membrane are not immediately incorporated into the GJs of contacting cells, but reside in an adjacent region of the membrane called the perinexus where they await incorporation into the periphery of the GJ plaque.^53–55 While at the perinexal plasma membrane, hemichannels create a conduit for passive diffusion between the cytoplasm and extracellular space.⁵³ ATP release through connexin hemichannels has been demonstrated in a number of cell types, and under a number of conditions, displaying both physiological and pathological properties,⁵⁶ and in particular, this review will focus on Cx43-medited ATP release from hemichannels in neutrophils and endothelial cells, with relevance to inflammation.

Pannexin channels

Pannexin channels are vertebrate homologues of the invertebrate GJ proteins, innexins, but are phylogenetically unrelated to connexins.⁵⁷ Unlike connexins, pannexins do not form intercellular channels or GJs; rather, they exist in the plasma membrane as channels that can connect the cytoplasm to the extracellular space.⁵⁸ Early patch clamp studies demonstrated the permeability of pannexin channels to ATP, and indicated the potential for mechanosensitive ATP release.⁵⁹

Breakdown of extracellular ATP

It may seem counterintuitive to discuss the removal of extracellular ATP prior to reviewing the receptors for ATP, but the breakdown products of ATP (adenosine diphosphate [ADP], AMP, and adenosine) are themselves agonists of purinergic receptors. ATP released into the extracellular space by any of the above mechanisms is rapidly degraded by ecto- and exo-nucleotidases, which are cell-surface attached and soluble in the extracellular space, respectively. These enzymes catalyze the breakdown of ATP to ADP and AMP, and further catalyze the breakdown of AMP to adenosine and inorganic phosphate.

Of the ectonucleotidases, CD39 and CD73 are most salient to this discussion. CD39 (also known as ectonucleotide triphosphate diphosphohydrolase 1) is expressed by a number of cell types including vascular endothelial cells,⁶⁰ leukocytes,⁶¹ and lymphocytes.⁶² It catalyzes the breakdown of ATP to ADP, and further catalyzes the hydrolysis of ADP to AMP.⁶⁰ AMP is further hydrolyzed to adenosine by ecto-5′- nucleotidase, or CD73.⁶³ Like CD39, CD73 is expressed on a number of cell types including endothelial cells⁶¹ and neutrophils.⁶⁴ Its expression appears to modulate endothelial cell barrier function⁶³ and leukocyte adhesion,³⁰ mediated through the generation of adenosine. Finally, adenosine is deaminated to inosine by adenosine deaminase, effectively removing the molecule from purinergic signaling pathways.⁶⁵

Purinergic receptors

Once ATP is released from cells, it can have paracrine/autocrine signaling effects through purinergic (P) receptors. These receptors can be broadly grouped into P1 and P2 families, which are activated by adenosine and primarily ATP/ADP, respectively.¹¹ The P2 receptors can be broken down again into 2 main subtypes, P2X and P2Y, which are ligand-gated ion channels and G-protein coupled receptors (GPCRs), respectively.¹¹

P1

P1 receptors are GPCRs that are sensitive to the nucleoside adenosine. There are currently four known P1 receptors: A₁, A_2A, A_2B, and A₃.⁶⁶

P2X

Unlike the P1 and P2Y GPCRs, P2X receptors are ligand-gated trimeric ion channels.²⁸ Once activated by ATP, the channel opens within milliseconds and forms a selective cation pore with a strong preference for Ca²⁺ followed by Na⁺ and K⁺.²⁸ Pore opening generally results in membrane depolarization, but also an increase in intracellular Ca²⁺ levels, activating differing second messenger systems depending on cell type.⁶⁷ For example, in macrophages, P2X₇ expression is important for release of the proinflammatory cytokine IL-1β.⁶⁸ It has been demonstrated that this effect is mediated by ATP activation of P2X₇ receptors, and subsequent activation of the Nlrp3 inflammasome, a multiprotein complex that activates caspase-1 resulting in the processing and secretion of the neutrophil chemoattractant IL-1β.⁶⁹

P2Y

Like P1 receptors, P2Y receptors are structurally similar GPCRs. There are currently 8 known subtypes: P2Y₁, P2Y₂, P2Y₄, P2Y₆, and P2Y_11–14.⁶⁶ Interestingly, P2Y receptors not only respond to ATP, but also display mixed sensitivities to ADP, UTP, and UDP depending on cellular and tissue context.⁷⁰

Purinergic Signaling and Early Inflammatory Events in Injured Tissues

The sequence of events following injury to any tissue—whether the cause is burns, lacerations, infarct, or surgical placement of an implant—is basically the same.^71–73 The initial response includes hemostasis (if there is injury to blood vessels) and infiltration of the wounded tissue by neutrophils.⁷¹ The neutrophil response in particular has been associated with poor outcomes, and several lines of evidence, including studies in neutropenic mice,⁷⁴ embryonic healing,⁷⁵ and with modulation of Cx43,^4,76,77 indicate that reducing neutrophil invasion leads to improved restoration of original tissue architecture (or reduction of fibrous capsule formation, in the case of implants⁷⁸) as opposed to replacement by scar. Therefore, we review these events, and the role purinergic signaling plays in them, in the following sections (also see Fig. 1).

FIG. 1.

Purinergic signaling in inflammation. This illustration depicts some of the key pathways in which purinergic signals mediate the initial inflammatory response. Adenosine triphosphate (ATP) can be released in a number of ways: damaged cells in injured tissue spill ATP and adenosine diphosphate (ADP) into the extracellular space through loss of membrane integrity; tissue compression can stimulate ATP release from endothelial cells through connexin hemichannels by direct mechanical stimulation and changes in shear stress; mechanically stimulated myocytes release ATP through the cystic fibrosis transmembrane conductance regulator and pannexin channels; endothelial cells release ATP through pannexin channels in response to thrombin. Once released, ATP acts through P2Y₂ and P2X₁ receptors on neutrophils to simulate migration. ATP also acts at P2X₇ receptors on resident macrophages, which activates the Nlrp3 inflammasome resulting in macrophage inflammatory protein-2 (MIP-2) and interleukin (IL)-8 release, which stimulate neutrophil migration and promote neutrophil–endothelial cell interactions respectively. Platelet aggregation is limited by ATP, while ADP from dead cells and activated platelets stimulates aggregation through P2Y₁ and P2Y₁₂. Finally, ATP is converted to adenosine monophosphate (AMP) by CD39, which is in turn converted to adenosine by CD73. Adenosine at low concentrations promotes neutrophil migration through A₃, but at higher concentrations acts through A_2A and A_2B to inhibit migration. This may be a mechanism to prevent extravasated neutrophils from returning to the vasculature.

Early inflammatory events

Hemostasis

Following an injury that damages a blood vessel, platelets adhere and aggregate forming a plug enmeshed in a fibrin clot formed from fibrinogen cleavage by thrombin.⁷⁹ The clot serves several purposes. First, the clot restores hemostasis, preventing further blood loss.⁸⁰ Second, the platelets secrete cytokines with chemotactic and mitogenic properties—platelet-derived growth factor, epidermal growth factor, and transforming growth factors β1 and β2.⁸⁰ The neutrophils, macrophages, and fibroblasts that respond to these signals are able to migrate into the wound because of the third function of the thrombus, which is to provide a temporary extracellular matrix for cell migration.⁸⁰

Purinergic signaling is important for the process of platelet aggregation. ADP generated from damaged tissue has long been recognized to induce platelet aggregation.^81,82 ADP is also secreted by platelets themselves by degranulation,^83,84 potentially forming a positive feedback loop.⁸⁵

Purinergic signaling also functions to contain coagulation to the site of injury. ATP acts as an antagonist of platelet P2Y receptors and functions to limit thrombus formation.⁸⁶ Recent work has demonstrated ATP release from vascular endothelial cells in response to thrombin.⁸⁷ Using pharmacological inhibitors of connexins and pannexins and shRNA technology, the authors provide compelling evidence that pannexins are the channel responsible for ATP release in response to thrombin. This response to thrombin in vascular endothelial cells suggests that ATP may be released in an injury-localized manner that serves to locally limit platelet aggregation to areas of vessel damage and simultaneously attract inflammatory cells such as neutrophils. Taken together, these data indicate that purinergic signaling plays an important role in both activating and regulating hemostasis.

The neutrophil response

Neutrophils are the next cellular component to enter the wound area, and arrive within minutes of injury.⁸⁸ Their main role is to cleanse the area of debris and bacteria.⁸⁰ They also serve as a source of proinflammatory cytokines activating local fibroblasts and keratinocytes.⁸⁹ Ironically, neutrophils can also cause extensive damage to the tissue through the release of reactive oxygen species (ROS).⁹⁰ Neutrophils are eventually removed from the wound with the eschar and phagocytosed by macrophages.⁸⁴

Recent data have implicated a role for injury-generated ATP in the production of chemoattractants, and subsequent targeting of neutrophils to sites of sterile inflammation. McDonald et al. generated a thermal injury on the liver, and found that neutrophils migrated directionally to the site of injury.³¹ Their data suggested that ATP in the injury area activated P2X₇ receptors on resident macrophages, which in turn activated the Nlrp3 inflammasome. This led to the production of MIP-2 and IL-8, targeting neutrophils to the wound site and promoting endothelial cell/neutrophil interaction through integrin α_Mβ₂ (MAC1) on neutrophils and intercellular adhesion molecule-1 on endothelial cells. Once neutrophils were near the wound locus, migration into the injury appeared to be mediated by formylated peptides released from dead cells.

Interestingly, McDonald et al. also found that ATP did not directly mediate neutrophil migration, in contrast with the published literature.^31,33 Specifically, Chen et al. have shown that neutrophils directionally migrate toward fMLP in an ATP-dependent manner.³³ They found that fMLP elicited polarized ATP release from neutrophils, generating an ATP concentration gradient that the neutrophils could follow. More recent work from this group has demonstrated that the route of ATP release was through pannexin channels.⁹¹ However, others have demonstrated that Cx43 hemichannels are a conduit for ATP in activated neutrophils.⁶¹

In addition to ATP, adenosine also modulates a number of neutrophil responses. A₁ and A₃ receptors generally act through G_i/o signaling while the A_2A and A_2B receptor subtypes signal through the counteracting G_s system. Importantly, A₁ and A₃ receptors have a high affinity for adenosine (EC₅₀<500 nM), while A_2A receptors have lower affinities, and A_2B receptors have EC₅₀'s>10 μM.⁹² These opposing pathways play an important role in early inflammatory events. Specifically, A₁ and A₃ receptor activation in neutrophils largely lead to proinflammatory actions: enhanced chemotaxis, adhesion, phagocytosis, and superoxide production. Conversely, pathways activated by A_2A and A_2B inhibit adhesion, migration, degranulation, inflammatory cytokine release, and ROS production.⁹²

Stimulus and sources of ATP and adenosine in the wound healing response

In the setting of an injury, cell necrosis and apoptosis likely represent an acute infusion of ATP into the system.³¹ This source of ATP appears to be important for recruiting neutrophils, but this alone may be insufficient to stimulate a full response. Extracellular ATP is quickly degraded by ectonucleases,⁹³ and therefore secondary sources of ATP are necessary for continued signaling. These secondary sources appear to be leakage from connexin hemichannels^19,94 and pannexin channels⁵⁹ in response to mechanical¹¹ and flow induced shear stress.^95,96 For example, in a brain injury model, work by Davalos and colleagues demonstrated that insertion of an electrode containing ATP into the cerebral cortex was enough to elicit a microglial response, but the response abruptly ended upon removal of the pipette.³² In an elegant experiment, they further showed that ATP-induced ATP release was necessary for the microglial response. This was accomplished by applying the P2Y agonizing nonhydrolyzable ATP analogue AMPPNP in the presence of apyrase, an exonucleotidase. In this condition, or with connexin inhibitors, no microglial response was observed, indicating that continuing microglial activation was due to P2Y-stimulated ATP release through connexin channels.

Implanted devices and ATP

The initial hemostatic and inflammatory response to the surgical placement of an implant, biomaterial, or transplanted tissue are key factors mediating the foreign body response,^72,97 or rejection of transplants and grafts.^98,99 In the case of implants, those early inflammatory events recruit macrophages and fibroblasts that encapsulate the implant and create contractile tissue, potentially leading to physical disfigurement, discomfort, device malfunction, and/or extrusion of the implant; for allo- or xenogeneic transplanted tissues, changes in the microvasculature and/or vascular thrombosis can cause ischemia of the introduced tissue leading to cell death and rejection.

Little data currently exist directly linking purinergic signaling to negative outcomes in the foreign body response to implants. However, we have observed that application of apyrase to the implant pocket of silicone discs placed submuscularly markedly reduces infiltration of inflammatory cells into the surrounding musculature (Fig. 2). Further, the initial inflammatory response to an implant is very similar to that of normal tissue injury, with the exception that encapsulation and chronic inflammation can occur with an implant.⁷² Therefore, much can be extrapolated from the studies discussed above relating purinergic signaling to the hemostatic and neutrophil responses in wounded tissue. We propose that neutrophils, blood clots, and associated platelets are activated by and targeted to implant sites by purinergic signals (Fig. 1). Further, these cells and the thrombus itself are known to release factors that attract macrophages and fibroblasts, which contribute to the formation of a contraction capsule around the implant and induce further inflammation in a chronic cycle.⁷²

FIG. 2.

Apyrase reduces inflammation in an implant model. Silicone implants were submuscularly placed in the backs of Sprague-Dawley rats, and tissue was harvested from the implant site 24 h postsurgery. The tissue was sectioned and stained by hematoxylin and eosin. In (A) and (C) 200 μL of vehicle (phosphate-buffered saline) was pipetted into the implant pocket along with the silicone disc—“Control”; in (B) and (D) 200 μL of apyrase (25 U/mL) was pipetted into the implant pocket with the disc—“Apyrase.” Asterisks indicate areas with necrosed bundles of muscle tissue. Note the preservation of muscle tissue in apyrase treated (B) animals compared to control (A) animals. Arrowheads point to areas where inflammatory cells have invaded the musculature and have begun to cause coagulative necrosis. White arrows show the direction of the implant. Scale bars=500 μm in (A) and (B), and 100 μm in (C) and (D). Color images available online at www.liebertpub.com/teb

Evidence supporting this hypothesis comes from work by Luttikhuizen et al.¹⁰⁰ The authors studied purinergic receptor expression following subcutaneous implantation of a collagen disk in rats. They found that P2X₇, P2Y₁, and P2Y₂ receptors displayed increased expression in the vasculature around the implant until day 2. Infiltrating macrophages and giant cells reacted differently, with expression of P2X₇ and P2Y₁ receptors increasing over the full period studied of 21 days. Importantly, these data suggest that in the setting of an implanted material the ability of cells to respond to ATP may be increased, potentially exacerbating negative outcomes.

Scaffold-based materials, cellularized constructs, and ATP

As is the case with synthetic implants, there is a dearth of data concerning the role of extracellular ATP in the response to scaffold-based materials. However, like other implantable materials, hydrogels based on collagen,^101,102 hyaluronic acid,¹⁰¹ agarose,¹⁰¹ polyethylene glycol,¹⁰³ and alginates¹⁰⁴ generally elicit an innate immune response in the form of robust infiltration of neutrophils into the surrounding tissue, with subsequent encapsulation. Exceptions to this phenomenon are rare in the literature, but some cross-linked polyethers have been shown to have mild innate immune responses.¹⁰⁵ Further, we have found that aligned collagen constructs become populated by host cells and vascularized, as opposed to inducing the formation of a fibrotic capsule.¹⁰⁶

As with synthetic implants, the similarity of the acute inflammatory response to scaffold materials compared to normal tissue injury strongly suggests that purinergic signaling is involved. In support of this notion, recent work from David Becker's lab has shown that placement of a collagen scaffold in full thickness mouse wounds impeded reepithelialization and increased the number of infiltrating neutrophils.¹⁰² Importantly, they found that Cx43 was upregulated at the leading edge of keratinocytes, suggesting the potential for increased release of proinflammatory signals such as ATP.

A major goal for implanting scaffold-based materials has been the delivery of cells—often pluripotent or stem cells—for the regeneration of injured tissue. It is a well-known phenomenon that delivery of “naked” cells into an animal is generally associated with poor survival of the implanted cells.^107–109 Indeed, a recent study in which cardiac derived stem cells were intramyocardially injected into rats shows that a commonly used magnetic resonance imaging (MRI)-based method for detecting injected cells grossly overestimates cell survival, with few stem cells actually surviving at 3 weeks.¹¹⁰ Surprisingly, few studies have characterized cell survival in protocols using scaffold-seeded or -encapsulated cell delivery systems. Mayfield et al. delivered cardiac stem cells encapsulated in agarose to help repair an induced myocardial infarction in mice.¹¹¹ While delivery of encapsulated stem cells did improve cardiac function and reduce infarct size, and survival of encapsulated cells was significantly greater than nonencapsulated cells, the number of transplanted encapsulated cells remaining was only 20% at 1 h and 10% at 3 weeks postdelivery. In combination with the data on the innate immune response to scaffold-based materials, it can be inferred that just as the neutrophil response generates collateral damage to host tissues, neutrophils may also destroy a large number of implanted cells. Therefore, it is strongly suggested that loss of transplanted cells to the initial inflammatory response is a key issue for all researchers developing tissue-engineering methodologies that has been largely overlooked, and the signaling mechanisms involved (e.g., extracellular ATP) should be investigated.

Therapeutic Options

A standard response to nondegradable synthetics is capsule formation and resultant capsular contracture.¹¹² This process can result in implant extrusion, device failure, disfigurement, and chronic pain.^112,113 The interface of these materials and the host tissue never completes the wound healing cycle forming in essence a chronic granuloma.^78,113 Further, prosthetic materials remain vulnerable to infection from bacteremia or direct seeding by infectious organisms. Therapeutic interventions that could temporize, lessen, or allow better engraftment of synthetic materials would be of tremendous clinical benefit saving millions in health care costs and more importantly, patient well being.

Apyrase as an ATP scavenger

Soluble apyrase converts ATP to AMP and is found throughout nature. A number of plant species express soluble apyrases,¹¹⁴ and apyrase isolated from potatoes is commercially available as a relatively inexpensive lyophilized powder. Animal studies show that application of apyrase improves outcomes in situations where the initial inflammatory response is undesirable. For example, infarct volumes, inflammatory cytokines, and numbers of infiltrating neutrophils were significantly greater in CD39-deficient mice compared with wild-type mice in a model of cerebral ischemia.⁶⁴ Application of soluble potato apyrase to CD39-deficient mice prior to ischemia significantly reduced infarct volumes and neutrophil counts compared with saline-injected controls. Although apyrase administration to wild-type mice was not performed, these data indicate that apyrase is a potential therapeutic in cases of stroke, especially given its platelet disaggregating effects.

The utility of apyrase treatment is also supported in skin and liver injury. In a model of epidermal injury, Takahashi et al. used an innovative agarose bead-attached green-emitting luciferase to image subcutaneous ATP release.¹¹⁵ They found that following tape-stripping of mouse skin, ATP release was increased and its bioluminescent detection could be quenched by application of apyrase. Importantly, subepidermal neutrophil infiltrate was observed after injury, and could be significantly reduced by apyrase application. Similarly, in a mouse model of acetaminophen-induced hepatotoxicity, overdose resulted in liver necrosis and massive neutrophil infiltration; again, apyrase administration significantly attenuated this response.¹¹⁶ Apyrase treatment has also shown improvement of outcome and reduction of inflammation in a number of other models including lipopolysaccharide-induced lung inflammation,¹¹⁷ laryngeal airway hyperreactivity,¹¹⁸ asthma,¹¹⁹ mechanical ventilation-induced acute lung injury,¹²⁰ and cardiac xenograft.¹²¹

In the case of implants, we have observed reductions in inflammatory cell infiltrate and coagulative necrosis of muscle tissue when apyrase was applied in a rat submuscular implant model (Fig. 2). Taken together with the above discussed studies, it is suggested that inclusion of apyrase with implanted materials and devices could reduce the incidence of capsular contracture and extrusion.

Blocking or manipulating Cx43 hemichannels

As discussed above, connexin hemichannels and pannexin channels open in response to a number of pathological stimuli and are a conduit for ATP release in diverse scenarios. Manipulation of connexins has proved beneficial in wound healing and in the response to implants. Pioneering work by Qiu et al. demonstrated that application of Cx43 antisense to incisional and excisional skin wounds in mice accelerated wound closure, reduced scarring, and limited neutrophil infiltration of the wound area.⁴ Similar results have been produced using the Cx43-mimetic peptide αCT-1.⁷⁶ Notably, work from the Yost lab has shown that application of αCT-1 to submuscularly placed silicone implants in rats significantly reduced neutrophil infiltration in muscle tissue adjacent to the implant.⁷⁸ Additionally, there was reduced collagen content, increased vascularity, and a reduction of contractile myofibroblasts around the implant in αCT-1 treated rats—suggesting a significantly attenuated foreign body response. Importantly, αCT-1 has been shown to reduce hemichannel activity,⁵³ indicating that these effects could be partially mediated by a reduction in ATP release through Cx43 hemichannels.

Concluding Remarks

The field of purinergic signaling encompasses a wide range of phenomenon—from synaptic transmission to renal function to inflammation. This diversity of function is nothing short of amazing considering the completely separate and central functions of intracellular ATP in energy metabolism and DNA structure. Yet even within the narrow scope of early inflammatory events the role that purinergic signaling plays is not easily defined, having multiple mechanisms of release and acting through different receptors depending on context and cell type, and at times acting through synergistic pathways or antagonistic pathways. We are only beginning to unravel the complex web, which is the cellular and molecular biology of purinergic signaling, but already we are beginning to see the development of therapeutics out of basic science research on extracellular ATP signaling. With this in mind, the prospects for enhancing healing and integration of foreign organs, synthetic tissues, and implanted devices are improving daily.

Acknowledgment

This work was supported by a grant from the National Institutes of Health NIH-NIDCR RO1 DE019355 (to M.J.Y.).

Disclosure Statement

No competing financial interests exist.