Published ahead of print on December 9, 2005, doi:10.1164/rccm.200508-1243PP
© 2006 American Thoracic Society doi: 10.1164/rccm.200508-1243PP
The Inflammatory Role of Platelets in Cystic FibrosisDepartment of Pediatrics, UMass Memorial Health Care; Department of Pediatrics; and Center for Platelet Function Studies, University of Massachusetts Medical School, Worcester, Massachusetts Correspondence and requests for reprints should be addressed to Brian P. O'Sullivan, M.D., Department of Pediatrics, UMass Memorial Health Care, 55 Lake Avenue, North Worcester, MA 01655. E-mail: osullivb{at}ummhc.org ABSTRACT Platelets are an important, albeit generally underappreciated, component of the inflammatory cascade. Platelets are known to contribute to inflammation in atherosclerosis, stroke, and asthma. They produce a large number of proinflammatory lipid mediators and cytokines, and play a vital role in recruitment of leukocytes into inflamed tissue. We review the role of platelets in inflammation, how they assist in the recruitment of leukocytes into lung tissue in asthma, and evidence of their dysfunction in cystic fibrosis (CF). Platelet dysfunction in CF could contribute to pulmonary inflammation and tissue destruction. We hypothesize that platelet activation is important in CF lung disease and suggest research avenues that might help elucidate the role of activated platelets in CF.
Key Words: platelets cystic fibrosis inflammation arachidonic acid aspirin This article summarizes the increasing evidence for an important role of platelets in inflammatory lung disease, presents evidence of their specific role in cystic fibrosis (CF) lung disease, and discusses the possible use of antiplatelet therapy in CF. Airway inflammation in concert with chronic infection by bacterial pathogens accounts for the progressive, suppurative pulmonary disease that leads to respiratory failure and premature death in patients with CF. There is much debate about whether there is a basic defect in the host immune response that causes pulmonary inflammation without infection, or if the inflammation is a response to protracted, intense bacterial infection. A primary dysfunction in inflammation is implied by the presence of inflammatory changes in the pancreas and liver, which are uninfected, and increased inflammatory mediators and polymorphonuclear cells (PMNs) in uninfected CF airway explants (1). However, there is no doubt that pulmonary inflammation is more intense in the presence of chronic infection (2). Whether it is the "chicken" or the "egg," treating this exuberant inflammatory response has become a top priority for CF caregivers. Antiinflammatory drugs commonly used for this purpose include corticosteroids, ibuprofen, and even the antibiotic azithromycin. Drugs that inhibit aspects of platelet function could become an important adjunct to antiinflammatory therapy in CF. PLATELETS AND INFLAMMATION Platelets, anucleate cytoplasts that have a life span of 8 to 12 d, are well known to have a key role in hemostasis and thrombosis. More recently, however, it has become apparent that they play a major role in inflammation (3). Platelets are produced in the bone marrow, but also in peripheral blood and in the pulmonary circulation (4). The pulmonary circulation may be seen as the birthplace of the platelet because megakaryocytes entering the lung release their platelets there. As a result, pulmonary venous blood has a much larger number of platelets than pulmonary arterial blood (4). Coupled with the fact that the pulmonary circulation is the major reservoir for marginated PMNs, there is ample opportunity for plateletPMN interactions to affect lung health. Platelets serve as classical inflammatory cells. They can undergo chemotaxis, contain and release adhesive proteins, activate other inflammatory cells, release vasoactive substances, and have the capacity to express or release proinflammatory mediators such as thromboxane A2 (TXA2), prostaglandins, platelet activating factor (PAF), brain-derived neurotrophic factor (BDNF), and platelet factor 4 (PF4; also known as CXCL4), as well as a host of other chemokines and chemokine receptors (Table 1) (58). Furthermore, direct cellcell contact by specific adhesion molecules facilitates transcellular metabolism of arachidonic acid (AA) by PMNs (9). Platelets provide free AA to PMNs, which enhances production of PMN-derived leukotriene B4 (LTB4) and the cysteinyl leukotrienes LTC4, LTD4, and LTE4. Furthermore, platelet-derived enzymes, such as 12-lipoxygenase, act on PMN eicosanoid products to produce other lipid mediators, including lipoxins (10). Although platelets do not have nuclei, they do contain limited amounts of mRNA and are capable of de novo synthesis of proteins, including adhesion molecules and chemokines (11).
Platelets are a linking element between hemostasis, inflammation, and tissue repair (7). Not only are they activated via traditional pathways (thrombin, adenosine diphosphate [ADP], TXA2) but they can be stimulated by antigens, antigenantibody complexes, microorganisms, and bacterial endotoxins, including lipopolysaccharide from Pseudomonas aeruginosa (12, 13). Release of mediators stored in platelet granules and de novo platelet production of other mediators enhance the inflammatory response. Histamine and serotonin increase vascular permeability; ADP increases the agonist-induced oxidative burst in PMNs; platelet-derived growth factor stimulates chemotaxis for monocytes and primes eosinophils to produce superoxide anion; PF4 induces PMNs to adhere to unstimulated vascular endothelium, induces the release of histamine from basophils, and stimulates the adherence of eosinophils to vascular walls (6). Furthermore, platelets release RANTES (regulated upon activation, normal T-cell expressed and secreted), which plays an important role in recruitment and adhesion of monocytes to activated endothelium (14). After leukocytes are recruited to tissue, platelets potentiate the inflammatory process by inhibiting apoptosis of PMNs, monocytes, and eosinophils (6, 15, 16). Platelets metabolize AA via the cyclooxygenase and lipoxygenase pathways to produce inflammatory mediators, the most abundant of which are TXA2, via the cyclooxygenase pathway, and 12-S-hydroxyeicosatetraenoic acid, via the 12-lipoxygenase pathway (17). Platelet 12-lipoxygenase interacts with the products of AA metabolism by other cells (notably PMNs) to produce lipoxins (10). In addition, PMNs stimulated by endotoxin release PAF, which activates platelets (18), which in turn recruit more PMNs to the inflamed area. Platelets also provide positive feedback mechanisms for their own activation. ADP secreted from dense granules and TXA2 formed from AA bind to specific receptors on the platelet surface, complete initial platelet activation, and recruit additional platelets into the activated fraction. Thus, once the inflammatory cascade has been initiated, leukocytes and platelets combine to propagate and amplify it.
For white blood cells to invade inflamed or infected tissue, the leukocyte must first roll along the endothelial surface, then attach firmly, and finally migrate through the endothelium into the tissue. The final two steps, firm attachment and diapedesis, are the result of up-regulation of integrin molecules, particularly integrin
The Role of Platelets in Asthma and Other Inflammatory Diseases Platelets and their mediators have long been known to play a role in the pathogenesis of asthma. However, it has only been with the advent of molecular biology techniques and use of gene knockout mice that the important role they play in leukocyte recruitment to the lung has been elucidated. In 1997, De Sanctis and coworkers (22) demonstrated that sensitized, P-selectindeficient mice that were challenged with ovalbumin had fewer eosinophils and lymphocytes in bronchoalveolar lavage fluid than did similarly challenged wild-type mice. At about the same time, Moritani and colleagues (23) showed that humans with asthma had increased numbers of circulating platelets expressing P-selectin as compared with control subjects. Others have shown that when blood from subjects with asthma is used in an ex vivo model of leukocyte adhesion, there is increased binding and clustering of eosinophils, an effect that is abrogated if the blood is pretreated with P-selectinblocking antibodies (24). These studies provide strong evidence for an important role for activated, P-selectin surfacepositive platelets in the recruitment of inflammatory cells into the lungs of patients with asthma. Pitchford and coworkers, in a series of elegant experiments (2527), have demonstrated the importance of platelets and platelet-derived P-selectin in the recruitment of leukocytes into the lungs in allergic asthma. Plateletleukocyte aggregates, also known as heterotypic aggregates, are caused by binding of platelet P-selectin to leukocyte PSGL-1 and are a more sensitive measure of platelet activation than is measurement of P-selectin expression alone (28). Pitchford and colleagues (25) observed increased numbers of plateletleukocyte aggregates in the blood of human subjects with allergic asthma compared with healthy control subjects. In addition, platelets from these patients augmented adhesion of human PMNs to cultured vascular endothelial cells, suggesting that platelets participate in cell recruitment and transmigration into tissue (25). This group also showed decreased infiltration of leukocytes into the lungs of platelet-depleted allergic mice compared with mice with normal platelet numbers; leukocyte recruitment was restored by infusing platelets from sensitized mice into the platelet-depleted mice (25). Platelet depletion of allergic mice also markedly reduced changes in airway architecture ("remodeling") compared with animals with normal platelet numbers (26). Finally, these investigators showed that activation of leukocytes via platelet binding leads to increased expression of leukocyte CD11b and very late antigen (VLA)-4 in mice (27). Expression of these adhesion molecules allows for firm attachment of leukocytes to vascular endothelium before diapedesis. Leukocyte CD11b expression was suppressed in thrombocytopenic mice, but was restored with infusion of fixed, stimulated platelets (27). The above series of studies demonstrate that P-selectin expression on the surface of platelets is a major requirement for pulmonary eosinophilia in asthma, but platelets play a similar role in other inflammatory diseases, too. Mulligan and colleagues (29) showed that antiP-selectin antibodies decreased neutrophil recruitment in a rat model of acute lung injury similar to acute respiratory distress syndrome. Kamochi and coworkers (30) were able to demonstrate decreased PMNs (as measured by myeloperoxidase activity) in the lungs of P-selectin/intercellular adhesion molecule-1deficient double-mutant mice after endotoxin-induced injury. In nonpulmonary lesions, it has been shown that not only is P-selectin important in the genesis of postischemic renal failure but that it is specifically platelet-derived P-selectin (not endothelial-derived P-selectin) that is important (31). In addition, there is evidence for an increased number of activated platelets in the blood of patients with inflammatory bowel disease (32). In summary, platelets fit the definition of traditional inflammatory cells (6), capable of phagocytosis and elaboration of proinflammatory cytokines, chemokines, and lipid mediators, and are vital for the process of leukocyte margination (rolling), which is a necessary first step in recruitment of leukocytes to areas of inflammation. Animals made deficient in platelets or in which platelet P-selectin is blocked or deficient are less able to mount an inflammatory response and have fewer white blood cells in target organs. Humans with inflammatory processes have increased numbers of plateletleukocyte aggregates and increased expression of platelet P-selectin, both markers of platelet activation (28). PLATELET DYSFUNCTION IN CF There have been a number of studies of platelet function in CF, which are summarized in Table 2 (3347). Platelet abnormalities in patients with CF were reported as early as 1969 (33) when it was noted that patients with CF had thrombocytosis. An inverse correlation between the arterial partial pressure of oxygen (PaO2) and peripheral blood platelet count was described. Six years later, Samuels and colleagues (34) reported the first functional abnormality in platelets from patients with CF. Prostaglandin E1 (PGE1) normally inhibits ADP-induced platelet aggregation, but these investigators found a much lesser degree of inhibition in platelets from patients with CF than from control subjects. In 1982, Permin and coworkers (35) demonstrated that a factor in CF patient serum, P. aeruginosa immune complexes, could stimulate platelets to release a vasoactive substance, serotonin. These studies established that patients with CF have an increased number of circulating platelets, increased platelet activation in response to gram-negative infection, and increased release of platelet-derived mediators, many of which have the potential to cause negative pulmonary consequences.
Cross-sectional studies showing an inverse relationship between platelet activation and pulmonary function provide evidence for an association between platelet activation and CF lung disease. These studies have shown the following: (1) increased platelet numbers correlate with decreased PaO2 (33); (2) an increase in urinary concentration of thromboxane metabolites (a marker of platelet activation in vivo) correlates with decreased FEV1 (41, 43), and (3) increased plasma concentrations of soluble CD40 ligand (sCD40L), a member of the tumor necrosis factor (TNF) family, 95% of which is derived from platelets, correlate with decreased pulmonary functions (46). Taken together, these studies suggest a link between platelet activation and progressive impairment of CF lung function. However, there have been no longitudinal studies examining platelet function changes over time or in association with clinical exacerbations of CF pulmonary disease. Whether platelet abnormalities are the cause or consequence of inflammation in CF lung disease is unclear. Platelet dysfunction in CF could be the result of platelet abnormalities intrinsic to CF, plasma factors unique to CF, and/or the nonspecific consequence of generalized inflammation.
Intrinsic Platelet Dysfunction in CF Dysfunction of the chloride channel coded for by the CF gene, the CF transmembrane conductance regulator (CFTR), could explain this intrinsic abnormality if it were present in platelets and if chloride channel function were crucial to platelet regulation. Agam and colleagues (42) looked for functional evidence of the presence of CFTR in platelets by size change due to chloride flux. They found a diminished response to PGE1 in five of nine patients with CF, indicating down-regulation of a cAMP-regulated chloride channel. Because CFTR is a cAMP-regulated chloride channel, by inference they proposed that platelets had CFTR, and that CFTR dysfunction was the cause of abnormal chloride transport in platelets. O'Sullivan and coworkers (47) also observed a decreased response to PGE1 in patients with CF, that is, reduced inhibition of agonist-induced platelet surface P-selectin expression. However, these investigators found neither message for CFTR nor CFTR protein itself in healthy donor platelets or in CF platelets. Therefore, if there is an intrinsic abnormality in chloride transport in CF platelets, it must be due to a non-CFTR, cAMP-regulated channel or to a CFTR-related abnormality that is expressed in megakaryocytes but not platelets. Ulane and colleagues (40) demonstrated an intrinsic increase in turnover of platelet membrane phospholipids (specifically, phosphatidylcholine) in CF platelets. These investigators postulated that this was due to CFTR dysfunction, because CFTR is responsible in part for plasma membrane recycling through endo- and exocytosis (48). Increased phospholipid turnover, whether due to CFTR dysfunction or some other mechanism, could have relevance for CF platelet function. Increased turnover of the large AA pool found in platelet cell membranes could account for increased endogenous production of TXA2 by CF platelets. This hypothesis seems particularly attractive given the finding of increased urinary TXB2 in subjects with CF, an indirect measure of platelet activation (41, 43, 44, 46). As discussed above, O'Sullivan and coworkers (47) demonstrated that CFTR is not present in control or CF platelets; thus, it is possible that CF platelets produce excess TXA2 as indicated by measurement of urinary metabolites (although platelets are not the only source of TXA2 [49]), but this is not the result of CFTR dysfunction per se. Although platelet AA and docosahexaenoic acid (DHA) content is normal in CF platelets, another fatty acid, Mead acid (20:3 n-9), is present in excess (47). Interestingly, when Mead acid is metabolized by platelet 12-lipoxygenase an end product is generated with PGE2-like activity, which enhances platelet aggregation (50, 51). Thus, an intrinsic fatty acid abnormality is present in CF platelets and, through the metabolism of Mead acid, may contribute to the observed increase in CF platelet activation. However, whether or not this platelet fatty acid abnormality can explain CF platelet activation has been questioned (47). In summary, there is strong evidence of intrinsic platelet dysfunction in CF, but the precise mechanism(s) remains to be defined. CFTR is not present in platelets and thus is not directly responsible for these abnormalities (47). However, a role for CFTR mutations in megakaryocytes that affects later platelet activation cannot be excluded.
Plasma-mediated Platelet Dysfunction in CF Another secondary consequence of CFTR dysfunction that could contribute to platelet activation in CF is an elevation in plasma ATP (55, 56). The increased plasma level of ATP found in patients with CF is most likely a direct reflection of the fact that the CFTR protein acts as an ATP transporter as well as a chloride channel. Dysfunction of CFTR results in poor transport of ATP and increased plasma ATP concentrations. ATP has counterbalancing effects on platelet function; it can directly inhibit P2Y1 and P2Y12 receptors, thereby inhibiting platelet activation and aggregation, or it can be acted on by nucleotidases (e.g., CD39) to form ADP, a potent agonist for these same receptors (57). ATP also serves as an agonist for the platelet P2X1 receptor. Activation of P2X1 by ATP leads to Ca2+ influx from the exterior and platelet shape change. On balance, ATP is an agonist of platelet aggregation and activation in vivo (57). Another plasma abnormality that is a direct consequence of CFTR dysfunction and contributes to platelet dysfunction is malabsorption of the fat-soluble vitamin E. Approximately 90% of patients with CF have pancreatic insufficiency and are at risk for malabsorption of vitamin E and other fat-soluble vitamins (58). Davis and colleagues (36) and Ciabattoni and coworkers (43) have suggested a relationship between low vitamin E levels and increased platelet activation in patients with CF. Ciabattoni and colleagues (43) postulated that low vitamin E levels lead to increased fatty acid oxidation and production of isoprostanes, which can activate platelets. Davis and coworkers (36) found that platelets from patients with CF had an abnormal response to PGE1 that normalized after supplemental vitamin E therapy. However, not all investigators have observed a correlation between vitamin E status and platelet activation in CF (38, 47). Decreased serum vitamin E concentrations may contribute to platelet activation, but seem to be neither necessary nor sufficient to cause it. In summary, CFTR dysfunction can indirectly increase platelet activation through its effect on plasma factors such as fatty acids, ATP, and vitamin E.
Platelet Dysfunction as a Nonspecific Consequence of Generalized Inflammation NOVEL THERAPEUTIC AND RESEARCH APPLICATIONS Given their multiple important roles in inflammation, platelets are a potential therapeutic target in patients with CF. The literature on atherosclerosis, asthma, adult respiratory distress syndrome, inflammatory bowel disease, and renal disease clearly demonstrates that once platelets are activated they contribute to leukocyte recruitment, adhesion, and activation (27, 2932, 63). Furthermore, platelets elaborate proinflammatory eicosanoids and cytokines that directly provoke and augment inflammation, regardless of leukocyte recruitment. These platelet mediators cause bronchospasm, airway edema, and increased mucus secretion, and contribute to fibroelastic changes in airway architecture (6). Combined with the poor self-regulatory antiinflammatory mechanisms observed in patients with CF (decreased DHA, decreased lipoxins, decreased interleukin-10 [53, 64, 65]), the inflammatory cascade is unchecked and leads to pulmonary tissue destruction.
Recognition of the important role for platelets in inflammation (66, 67) and, in particular, recruitment of PMNs into lung parenchyma potentially opens new doors for CF therapy (68). Platelets can have a sustained response to the initial activating signal and remain functional for many hours (3). Ongoing production of interleukin-1 One weak antiplatelet agent, ibuprofen, has already proven useful in CF, although its mechanism of action remains unclear (69). In addition, the revelation of the importance of platelet function in CF in tandem with the fact that oxidative stress contributes to platelet activation (36, 43) highlights the need to maintain normal levels of vitamin E in patients with CF, as is currently recommended. The antiplatelet agent aspirin has been clearly shown to be beneficial in coronary artery disease and could be of use in other inflammatory disorders, including CF. The antiinflammatory effects of aspirin are mediated through irreversible acetylation of cyclooxygenase-1, which inhibits TXA2 production and platelet activation and aggregation. In two reports of platelet dysfunction in CF, a small subset of patients with CF were treated with 50 mg/d of aspirin for 1 wk (41, 43). In both studies, there was a statistically significant reduction in urinary thromboxane metabolites after aspirin therapy, indicating that platelet inhibition by aspirin in patients with CF can decrease production of the proinflammatory lipid TXA2. Aspirin also reduces production of reactive oxygen species, increases nitric oxide synthase, and may be a more potent protector of endothelium from oxidation than vitamin E (70). Aspirin has been shown to decrease the risk of vascular events in men with elevated serum C-reactive protein to a greater extent than in men with lower C-reactive protein levels, implying a beneficial antiinflammatory effect separate from inhibition of platelet aggregation (71). Patients with CF have similar elevation of nonspecific markers of inflammation, including C-reactive protein (72). Aspirin would be contraindicated in patients at increased risk for hemorrhage due to hemoptysis or esophageal varices and would not be appropriate for very young children who could be at risk for aspirin-induced Reye syndrome. However, aspirin is inexpensive and generally well tolerated. There are a number of newer antiplatelet agents available that have more specific antiplatelet effects than nonsteroidal antiinflammatory agents like ibuprofen and aspirin. For example, the Food and Drug Administrationapproved P2Y12 ADP receptor antagonists (clopidogrel [Plavix] and ticlopidine [Ticlid]) and the GPIIb-IIIa receptor antagonists (abciximab [ReoPro], eptifibatide [Integrilin], and tirofiban [Aggrastat]) have proven clinical benefit in coronary artery disease (73, 74). There is good evidence that all of these ADP receptor antagonists and GPIIb-IIIa antagonists inhibit systemic inflammation in the doses used to treat adult cardiovascular disease (7577). Measurement of sensitive markers of platelet activation such as P-selectin expression and heterotypic aggregates (monocyteplatelet and neutrophilplatelet aggregates) provide novel means by which to assess antiinflammatory therapy (7779). Because platelet activation appears to be an important component of CF lung inflammation and because this activation may be either a nonspecific response to generalized inflammation or a sign of CFTR-mediated up-regulation of systemic inflammation, incorporating platelet function studies into future clinical trials of antiinflammatory medications could lead to a better understanding of the mechanism of action of these agents and serve as a specific endpoint with clinical relevance. At this time, it is unclear what the most useful surrogate markers for trials of new drugs for CF are. Observation of down-regulation of platelet activation could validate the antiinflammatory activity of a novel therapy, be it a platelet-specific medication or a more general antiinflammatory agent. Measurement of platelet activation in clinical trials is therefore worthy of consideration. CONCLUSIONS Platelets have an underappreciated yet very important role in inflammatory lung disease and are therefore a potential therapeutic target in CF. However, much work remains to be done before antiplatelet therapy can be recommended for patients with CF. Further animal studies, perhaps with double knockout mice lacking CFTR and P-selectin (or other important platelet receptors), or platelet-depleted CFTR/ mice or similar mice treated with antiP-selectin antibodies (as done with mouse models of asthma [2527]) could be performed. The utility of such studies could be limited because CF knockout mice have predominantly gastrointestinal disease rather than lung disease. Longitudinal studies in patients with CF need to be performed, including following the degree of platelet activation over time as patients' health deteriorates and looking for changes in platelet activation in the face of acute pulmonary exacerbations. Evaluation of new therapies, such as small molecule inhibitors of selectins and specific antiselectin antibodies, are in early trials in non-CF diseases, including asthma (8082). The CF community may be able to learn much from these studies regarding the potential applicability of antiplatelet agents to inflammatory diseases. FOOTNOTES Originally Published in Press as DOI: 10.1164/rccm.200508-1243PP on December 9, 2005 Conflict of Interest Statement: B.P.O. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.D.M. has received research grants (unrelated to CF or other respiratory diseases) from Eli Lilly and Company, McNeil Consumer and Specialty Pharmaceuticals, and Bristol-Myers Squibb/Sanofi Aventis. Received in original form August 10, 2005; accepted in final form December 7, 2005 REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||