Cystic fibrosis: a mucosal immunodeficiency syndrome (original) (raw)

. Author manuscript; available in PMC: 2013 Feb 20.

Published in final edited form as: Nat Med. 2012 Apr 5;18(4):509–519. doi: 10.1038/nm.2715

Abstract

Cystic fibrosis transmembrane conductance regulator (CFTR) functions as a channel that regulates the transport of ions and the movement of water across the epithelial barrier. Mutations in CFTR, which form the basis for the clinical manifestations of cystic fibrosis, affect the epithelial innate immune function in the lung, resulting in exaggerated and ineffective airway inflammation that fails to eradicate pulmonary pathogens. Compounding the effects of excessive neutrophil recruitment, the mutant CFTR channel does not transport antioxidants to counteract neutrophil-associated oxidative stress. Whereas mutant CFTR expression in leukocytes outside of the lung does not markedly impair their function, the expected regulation of inflammation in the airways is clearly deficient in cystic fibrosis. The resulting bacterial infections, which are caused by organisms that have substantial genetic and metabolic flexibility, can resist multiple classes of antibiotics and evade phagocytic clearance. The development of animal models that approximate the human pulmonary phenotypes—airway inflammation and spontaneous infection—may provide the much-needed tools to establish how CFTR regulates mucosal immunity and to test directly the effect of pharmacologic potentiation and correction of mutant CFTR function on bacterial clearance.

Cystic fibrosis, which affects approximately 30,000 individuals in the United States, is caused by mutations in CFTR, a cAMP-regulated epithelial chloride channel (Box 1). Cystic fibrosis was originally recognized in babies with pancreatic insufficiency who failed to thrive and often succumbed to pulmonary infection in infancy or early childhood1, 2. Although there has been enormous progress made in understanding the basic biology of the CFTR chloride channel, it remains enigmatic how CFTR mutations cause enhanced susceptibility to pulmonary infection and how this susceptibility might be prevented.

BOX 1 CFTR deficiency in cystic fibrosis.

CFTR is a large glycoprotein consisting of two membrane-spanning regions and a cytoplasmic regulatory R domain178 and is expressed primarily in epithelial cells but also in many other cell types, including lymphocytes and PMNs81, 82. In addition to its role as a Cl− channel, CFTR is crucial in the regulation of ion transport, particularly Na+ and HCO3− (ref. 179). A lack of CFTR function results in sodium absorption through ENaC180, and mice with overexpression of ENaC develop lung pathology that, in some ways, mimics cystic fibrosis173, 181, although these mice do not spontaneously develop infection. There are several classes of CFTR mutations182 that correlate well with pancreatic disease, which is also a key component of cystic fibrosis, but these mutations are associated with more variable pulmonary phenotypes183. The most common CFTR mutation, ΔF508/ΔF508, results in a misfolded protein that is improperly glycosylated, is targeted for endosomal degradation and fails to reach the apical surface of the epithelium. Other CFTR mutations, such as G551D, form a partially functional channel whose activity can be potentiated165. The expression of a large number of modifier genes markedly affects the clinical manifestations of the disease184. For example, mannose-binding lectin 2 (MBL2) protein concentrations, especially in combination with high amounts of transforming growth factor β (TGF-β) production, is associated with severe pulmonary disease185. The central role of CFTR in regulating the hydration of the airways has become the focus of therapies, which seek to potentiate partially functional CFTR as well as correct the defective CFTR function attributed to specific mutations166–168.

Cystic fibrosis pulmonary disease is the most challenging problem in the management of cystic fibrosis and is the major determinant of life span and quality of life in affected individuals. Substantial clinical data have linked the recognition of bacterial infection in the lung, usually caused by Staphylococcus aureus or Pseudomonas aeruginosa, with the onset of symptomatic lung disease, which is marked by excessive airway inflammation and the eventual loss of pulmonary function. Aggressive and even prophylactic antimicrobial treatment often eradicates these infections and slows the deterioration of pulmonary function. However, the adaptive responses of the infecting organisms eventually result in isolates that are highly resistant not only to antimicrobial agents but also to the immune response that would readily eradicate these bacteria in hosts. CFTR dysfunction has long been associated with viscid mucus that causes the entrapment of bacteria in airway secretions. The prevailing hypothesis is that CFTR dysfunction, a lack of transport of chloride and accompanying water across the airway epithelium and excessive sodium reabsorption lead to dehydrated airway surface fluid, impaired mucociliary clearance, infection and inflammation3, 4. However, as we discuss below, there is accumulating evidence to suggest that CFTR dysfunction affects several components of innate immunity and that the initial predisposition to infection in infants with cystic fibrosis may represent a primary defect in local mucosal immunity.

Cystic fibrosis pulmonary disease is initiated in airways that are essentially normal at birth but that later become obstructed with mucus plugs. Early pathological findings of cystic fibrosis pulmonary disease include goblet-cell and mucus-gland hyperplasia5, 6. Even in the absence of clinically apparent infection, either viral or bacterial, there is often evidence of inflammation in cystic fibrosis airways, as evidenced by polymorphonuclear neutrophil (PMN) accumulation and excessive concentrations of interleukin-8 (IL-8) and free proteases7–11. Airway colonization and infection occur with diverse flora, which is followed by clinically apparent infection with the typical pathogens S. aureus, P. aeruginosa or both, even in infants at a very young age. Bacterial shedding of immunostimulatory pathogen–associated molecular patterns (PAMPs), such as cell-wall components, lipopolysaccharide (LPS), flagella and DNA, activate a brisk proinflammatory response. Bacterial adaptation to the airway milieu ensues, with a shift from a planktonic to a biofilm mode of growth, followed by the selection of mutants with abundant exopolysaccharide production that are resistant to phagocytosis. The intense inflammatory reaction to this airway infection consists of chemokine and cytokine expression (IL-8 and tumor necrosis factor (TNF)) and mucin secretion12, as well as PMN accumulation13 and the associated release of serine proteases14, 15, which are themselves proinflammatory stimulants16, 17. It is clear that increased airway inflammation does not result in enhanced bacterial clearance18. Airway obstruction results initially in hyperinflation, destruction of the airway walls and fibrosis, leading to decreased lung function as measured by forced expiratory volume and vital capacity.

Effects of CFTR mutation on epithelial innate immune function

The many roles of the airway epithelium in the host defense of the lung are well appreciated19. As CFTR is highly expressed in the airway epithelium, it is logical that defective CFTR function should affect the contribution of the epithelium to innate immunity. CFTR mutations have been associated with both constitutive activation of proinflammatory signaling in the absence of apparent microbial stimuli as well as exaggerated responses to bacterial products (Fig. 1). Endogenous activation of NF-κB (nuclear factor κ light-chain enhancer of activated B cells) and substantial sequestration of leukocytes has been observed in human cystic fibrosis fetal tracheal explants10 obtained before exposure to microbial flora; these observations were confirmed in pathological studies of cystic fibrosis fetuses11. Analyses of bron-choalveolar lavage fluid indicated that the induction of IL-8 and TNF in the airways of infants with cystic fibrosis is higher, often in the absence of known pathogens and elevated out of proportion to the bacterial load in airways contaminated with potential pathogens20–22 compared with normal infants. Long-term primary cultures of cystic fibrosis airway epithelial cells do not reproducibly show constitutive activation of proinflammatory signaling23, although this may be a function of clonal selection in vitro. These cultures are also influenced by epigenetic factors that are specific to the individual patients and that may influence inflammatory signaling. Whether there is, in fact, constitutive activation of proinflammatory signaling directly linked to CFTR dysfunction in humans is still unresolved and is a difficult issue to address experimentally.

Figure 1.

Figure 1

Altered TLR expression and signaling in the cystic fibrosis (CF) epithelium. Expression of TLR2 and TLR5 at the apical surface is increased, whereas TLR4 expression is restricted to the endosome. NF-κB in cystic fibrosis airway epithelial cells is constitutively activated, resulting in the production of inflammatory cytokines, such as IL-8 and GM-CSF, and the recruitment of PMNs independently of TLR’s interaction with the adaptor protein MyD88. After infection, bacterial PAMPs further increase NF-κB signaling through activation of TLR-MyD88 signaling. Intracellular TLR4 activation of Trif from the endosome is impaired, preventing interferon regulatory factor 3 (IRF3) translocation to the nucleus and the activation of type I IFN gene products, which are required for the activation of dendritic cells (DCs) and the clearance of some cystic fibrosis–related pathogens.

In vitro evidence of the effects of CFTR on NF-κB signaling was shown in Chinese hamster ovary cells expressing CFTR with a deletion of Phe508 (ΔF508 CFTR), which resulted in dose-dependent increases in NF-κB activation; this effect was not seen in cells expressing wild-type CTFR8. Numerous in vitro studies using human epithelial cystic fibrosis and control cell lines confirmed increased NF-κB signaling8, 24–26 and the involvement of several proinflammatory cascades, including activation of Ca2+-dependent signaling27, 28, increased nuclear factor of activated T cells (NFAT) transcription29 and mitogen-activated protein kinase (MAPK)-dependent activation of activator protein 1 (AP-1), in cystic fibrosis epithelia compared to control epithelia30, 31. It is crucial to recognize that the construction of wild-type and mutated cell lines with alterations in CFTR gene dosage and expression, as well as the use of CFTR inhibitors to mimic channel failure, could influence the outcome measurements32.

Additional CFTR-dependent effects on epithelia influence bacterial clearance. Glycosylation of both surface-associated and secreted epithelial glycoconjugates are affected by CFTR33, 34. Glycosylation status influences bacteria-host interactions34–36, mucus composition and rheology37–39, all properties that are key in mucociliary clearance. Despite in vitro evidence that CFTR-mediated alterations in glycosylation affect bacterial attachment, it is clear that few intact bacteria actually adhere to airway epithelial cells in vivo. Instead, organisms enmeshed in airway mucin, growing within biofilms, shed PAMPs that are recognized either by epithelial receptors on the apical surface, such as Toll-like receptors (TLR2, TLR4 and TLR5), within the cytoplasm (nucleotide-binding oligomerization domains (NODs) and retinoic acid inducible gene (RIG)) or in endosomal compartments (TLR3, TLR4 and TLR9). CFTR-associated effects on the production or distribution of these pattern-recognition receptors could be crucial in the initial pathogenesis of pulmonary infection.

Cystic fibrosis–related increases in the amount of NF-κB–dependent gene products and decreases in the amount of Trif-dependent gene products have been well documented. Although several studies suggested that cystic fibrosis and non–cystic fibrosis epithelia have similar expression of TLRs40, 41, alterations in receptor localization could result in key differences in TLR-initiated signaling between the two types of epithelia. Increased surface expression of TLR2 and TLR5 on human cystic fibrosis epithelial cells correlates with increased inflammatory responses to bacterial products40, 41. However, knockdown of myeloid differentiation primary response gene (88) (MyD88) in cystic fibrosis cells prevented bacterial-induced signaling but did not reduce baseline NF-κB signaling to the level seen in non–cystic fibrosis cells, which is consistent with some degree of constitutive activation of the NF-κB pathway in unstimulated cystic fibrosis cells. As this baseline proinflammatory status, with increased IL-8 production and neutrophil accumulation, is documented in clinical studies20, 22, 42, 43, it is unclear exactly why bacteria present in the cystic fibrosis lung evade eradication. PMNs do not phagocytose as well when bacteria are suspended in fluid as when they are associated with the cell surface44. Thus, PMNs that have accumulated in excess airway mucin may not be optimal in pathogen elimination.

Intracellular innate immune signaling is also affected by dysfunctional CFTR. Activation of toll-like receptor adaptor molecule (Trif)-dependent effectors contributes to the resolution phase of an acute inflammatory response45–47 and seems to be decreased in cystic fibrosis cells. TLR4 signals through MyD88 at the cell surface and, after internalization to the early endosome, activates Trif-dependent pathways that are linked to the type I interferons (IFNs)48. In cystic fibrosis epithelial cells, the surface expression of TLR4 is reduced49, resulting in decreased activation of both MyD88 signaling (ref. 49) and Trif signaling50. These observations were corroborated in clinical studies showing reduced expression of Trif-dependent gene products in cystic fibrosis airway fluid51–53. Limited Trif signaling would interfere with the resolution of the epithelial NF-κB–dominated inflammatory response and minimize the dendritic-cell–mediated activation of the adaptive immune system50, 54.

Epithelial cells are especially key in the activation of the type I IFN response (involving IFN-α and IFN-β), a signaling cascade that is central to the clearance of diverse airway pathogens45–47, 55, including viruses and extracellular bacteria such as Streptococcus pneumonia, P. aeruginosa and S. aureus. There are limited data detailing how CFTR deficiency affects primary viral clearance mechanisms56–58, which is in contrast to the abundant clinical data showing that viral infections exacerbate coexisting bacterial infections and cystic fibrosis lung disease59–61. Patients with cystic fibrosis seem to have no particular problems handling nonrespiratory viral infections, suggesting that the major mechanisms of viral immunity in these individuals are intact. However, cystic fibrosis epithelial cells produce reduced amounts of type I interferon in response to P. aeruginosa infection and, accordingly, are less capable of activating dendritic cell populations, which initiate the adaptive immune response50. Autocrine activation of signaling downstream of interferon production is subsequently reduced in cystic fibrosis epithelial cells in part because of increased concentrations of the protein inhibitor of activated signal transducer and activator of transcription 1 (PIAS1)52, 62, further adding to the dysregulated immune response.

CFTR dysfunction contributes to oxidative stress in the airway

Whether or not the exaggerated signals for proinflammatory signaling are caused by exogenous (microbial) stimulation, CFTR-associated alterations in signal transduction or both, the net result is an excessive accumulation of dysfunctional PMNs and their products. Constitutive activation of NF-κB signaling results in increased amounts of reactive oxygen species being generated by neutrophils accumulated in the airway. Increased amounts of reactive oxygen species (O2, H2O2 and HOCl) are associated with increased IL-8 production, increased IL-6 production in response to P. aeruginosa, defective autophagy and reduced CFTR expression63–67. Under normal conditions, CFTR contributes to the downregulation of NF-κB signaling during oxidative stress by controlling the degradation of nuclear factor of κ light polypeptide gene enhancer in B cells inhibitor, α (I-κB-α)68, a response that is lacking in the cystic fibrosis lung (Fig. 2).

Figure 2.

Figure 2

Increased oxidative stress in the cystic fibrosis airway. Constitutive NF-κB–mediated production of chemokines, including IL-8, leads to PMN recruitment, which persist in the airway, increasing the oxidative burden in the lung. Oxidative stress activates MAPK signaling pathways in the cystic fibrosis epithelium, amplifying the production of IL-8 and, therefore, recruiting additional PMNs. Mutant CFTR in the epithelial cells is unable to channel the antioxidants GSH and SCN− into the airway, limiting its ability to counteract the oxidative stress. Because SCN− also has antimicrobial properties, bacterial killing in the airway is diminished as well.

Epithelial cells control oxidative damage through the production of antioxidants, such as glutathione (GSH) and thiocyanate (SCN−), that are trafficked into the epithelial-lining fluid69 and may help to modify mucus viscosity70–72. GSH secretion is markedly reduced in patients with cystic fibrosis73 (as well as in the cystic fibrosis mouse74) as a result of a protein trafficking defect in the cystic fibrosis epithelial cells75. CFTR channel dysfunction impairs the trafficking of GSH into the airway surface fluid from epithelial cells but does not impair the production of GSH76, 77. Increased oxidative stress further contributes to alterations in IL-8 gene expression64 and defects in bacterial killing. Diminished trafficking of SCN−, a potent antioxidant78, which is also associated with mutations in CFTR, similarly contributes to the diminished killing of the cystic fibrosis pathogens P. aeruginosa and S. aureus79. Thus, deficient CFTR function in the epithelium results in a diminished ability to counter the oxidative stress that is induced by both constitutive and exogenously activated inflammation.

CFTR affects immune cell function

Effective mucosal immunity is the result of the coordinated participation of epithelial cells and resident and recruited immune cells80. Particularly in the airway, in which excessive PMN accumulation interferes with pulmonary function, proinflammatory signaling is tightly regulated. The influx of immune cells, dendritic cells, macrophages and T cells is crucial in this process. Immunohistological and physiological data indicate that CFTR is both transcribed and functional in several different types of leukocytes81, 82. How CFTR dysfunction in immune cells might interfere with the regulation of innate immunity has not been well studied. Whether or not CFTR dysfunction in these cells has any biological role is unclear given the large amount of clinical data indicating that patients with cystic fibrosis have normal immune cell function at sites other than the lung. There are no clinically apparent defects in the function of cystic fibrosis macrophages or T cells, as would be suggested by the increased prevalence of infection caused by organisms usually controlled by T cells, such as Salmonella or Mycobacterium tuberculosis, in individuals with cystic fibrosis. There is also no increased incidence of abscesses or granulomatous reactions to organisms that would be suggestive of a primary defect in PMN killing in cystic fibrosis. Murine models, although lacking characteristic cystic fibrosis lung disease, nonetheless have been shown to have alterations in macrophage signaling that contribute to the exaggerated inflammatory response in the cystic fibrosis lung83. Excessive proinflammatory signaling by murine and human cystic fibrosis macrophages may caused in part by prolonged TLR4 signaling from the early endosome, resulting in the elevated activation of the NF-κB and interferon pathways84.

A phagocytic defect associated with CFTR mutations that would explain the failure to clear inhaled bacteria has long been sought85–87. Unfortunately, analyses of PMN function in cells harvested from the peripheral blood of patients with cystic fibrosis are complicated by the exposure of these patients to circulating concentrations of LPS and to activated lymphocytes and their products. PMNs85, 86 and macrophages87 isolated from patients with cystic fibrosis have a modest reduction in the ability to kill phagocytosed bacteria in vitro, but such data have not been uniformly reproduced88. In contrast, a recent study of PMNs harvested from newborns with cystic fibrosis identified by neonatal screening studies did not find appreciable differences in PMN phagocytic function between these newborns and normal controls89. Less controversial are studies showing that PMNs isolated from patients with cystic fibrosis have a slower rate of apoptosis than PMNs from control individuals without cystic fibrosis90–92 as a result of a resistance to the proapoptotic cytokine TNF-α and the effects of anti-apoptotic granulocyte-macrophage colony-stimulating factor (GM-CSF), an NF-κB–dependent cytokine that is produced in greater quantity by cystic fibrosis epithelial cells than in non–cystic fibrosis cells93. As a result, cystic fibrosis PMNs persist in the airway, further contributing to airway inflammation. The abundance of PMN-derived proteases also thwarts efficient phagocytosis94, 95, resulting in the cleavage of macrophage receptors and causing inefficient opsonization and impaired bacterial killing96. Thus, even if CFTR does not directly contribute to phagocytic function, it may influence the kinetics of PMN degradation, protease and DNA release to further exacerbate the hyperinflammatory state in the cystic fibrosis airway.

T cell function is also affected by defects in CFTR, although the specific mechanisms involved in this effect have not been defined. Diminished recruitment of T cells in response to S. aureus9 by cystic fibrosis epithelial cells, as well as altered T cell responses to Aspergillus fumigatus in both human97 and murine98 infections, have been documented. Several studies have suggested that CFTR dysfunction results in a T helper type 2 (TH2) cell bias in response to respiratory pathogens99, 100, although more recent data reflect the key role of TH17 signaling in the host response to extracellular pathogens in the lung, including P. aeruginosa101. IL-17 participates in the regulation of neutrophil recruitment by several mechanisms, including inducing the expression of IL-8, and is necessary for an effective host response to P. aeruginosa102–107. Increased TH17 signaling contributes to cystic fibrosis lung pathology, and cystic fibrosis sputa have elevated amounts of IL-17, further enhancing neutrophil recruitment in the cystic fibrosis lung108, 109. Both T cell and neutrophil production of IL-17 are increased in cystic fibrosis compared to control lung specimens110, 111, but there is no evidence that increased production of IL-17 is a direct consequence of CFTR dysfunction.

Effective clearance of inhaled bacteria from the respiratory tract involves the integrated activities of both the epithelium and the immune cells. Clinical observations suggest that induction of the PMN-dominated inflammatory responses in the cystic fibrosis airway is not normally regulated; there is a relative lack of IL-10 production112, diminished lipoxin production113–115 and a failure of the normal progression from PMN predominance to monocytes with the expected resolution phase of acute inflammation in the cystic fibrosis airway95, 116. Thus, even if the primary CFTR-dependent alteration in innate immunity is primarily an epithelial defect, there is a failure of the recruited and resident leukocytes to adequately control the hyperinflammatory state in the airway, and there is ample documentation that the presence of activated PMNs in the airway does not correlate with bacterial eradication.

Bacterial adaptation to the host

Having blamed the failure to eradicate inhaled bacteria on a presumed CFTR-associated defect in mucosal immunity, it is also crucial to recognize the prodigious capabilities of typical cystic fibrosis pathogens to adapt to and flourish within the cystic fibrosis lung. The vast majority of patients with cystic fibrosis eventually become infected with opportunistic pathogens, often P. aeruginosa and S. aureus, in addition to a growing list of both cultivatable and non-cultivatable bacterial species, including anaerobes117, 118. Infection is caused not only by host defects in innate immunity but also by the selection of bacteria that are able to evade immune clearance. The successful cystic fibrosis pathogens share a genetic flexibility and an ability to adapt to the pressures imposed by mucosal immunity (Fig. 3). Even within individual clones from one patient, there is tremendous heterogeneity, such that numerous different gene expression profiles are expressed by the bacteria colonizing the lung, allowing for the selection of optimally fit clones119, 120. Planktonic organisms aspirated from the environment upregulate gene expression for motility, proteolytic activity and carbon utilization, whereas the pathways involved in immune evasion and iron scavenging are increasingly expressed during infection in vivo. Sequential P. aeruginosa strains isolated from chronically infected patients with cystic fibrosis have alterations in the expression of numerous virulence factors (O-antigen biosynthesis, type III secretion and mobility) and of multidrug-resistance genes121, as well as of genes required for survival within the nitrogen-rich, nutrient-deficient cystic fibrosis lung122–124.

Figure 3.

Figure 3

Adaptation of inhaled bacteria to the cystic fibrosis airway. Inhaled bacteria expressing flagella, pili and a type 3 secretion system (T3SS) aggregate within the cystic fibrosis lung, resulting in the formation of biofilm. Within the biofilm, bacteria lose flagella, pili and the T3SS, increase alginate production, release CpG DNA and express a diverse range of virulence factors promoting evasion of the host immune system. P. aeruginosa also releases outer membrane vesicles containing Cif, a protein that inhibits the recycling of CFTR in the host. Furthermore, the lipid A structure of the LPS is altered through the addition of palmitate and aminoarabinose, resulting in increased antibiotic (Ab) resistance and increased induction of IL-8 production by host cells.

Changes in LPS structure contribute to the proinflammatory milieu in the cystic fibrosis lung. LPS from P. aeruginosa isolated from chronically infected patients with cystic fibrosis (cystic fibrosis LPS) is structurally distinct from the environmental strains that initiate infection. The lipid A portion of cystic fibrosis LPS is typically penta-acylated, as opposed to the more common hexa-acylated structure, and contains substantially more palmitate and aminoarabinose than non–cystic fibrosis LPS125, 126. These alterations correlate with increased induction of IL-8 from human endothelial cells. Worth noting is that human TLR4 responds to these changes in LPS structure, whereas murine TLR4 does not127, a factor that may be relevant to the failure of the murine model of cystic fibrosis to reflect human pulmonary pathology.

Once a critical mass of bacteria is present in the airways, in vivo data indicate that they form biofilms128, 129. This then facilitates the coordinate expression of numerous genes throughout the microbial population through secretion of highly soluble quorum sensors, such as the Pseudomonas homoserine lactones and the quinolones, which act in concert with specific transcriptional activators130. The switch to the biofilm phenotype does not necessarily indicate a reduction of proinflammatory stimulation. P. aeruginosa in biofilms is associated with the expression of more inflammatory LPS131, as mentioned above, and the enrichment of extracellular DNA132, both of which contribute to the activation of TLR-mediated proinflammatory signaling in the airways. When serially isolated cystic fibrosis strains of P. aeruginosa were instilled into mice, isolates from late-stage infection did not cause lethal infection, as the early isolates did, but were equally capable of inducing excessive lung inflammation and establishing infection133. Thus, the adaptation to the host does not necessarily imply a failure to activate local mucosal signaling but does suggest a combined failure of first the innate, and then the adaptive, immune response to clear infection.

From within this large bacterial population there is also the spontaneous selection of MucA mutants that overexpress the exopolysaccharide alginate134, 135. These mucoid organisms elude phagocytosis and have been virtually pathognomonic for cystic fibrosis136. Although alginate is immunogenic and elicits antibody production, this amplification of the inflammatory immune response further contributes to oxidative stress without resulting in the clearance of the organisms137, 138. An indication of well established P. aeruginosa infection, the predominance of alginate-producing organisms represents the end result of failed innate immunity and an effective bacterial adaptation to the host.

Bacterial products further contribute to the adaptation of organisms to the milieu within the cystic fibrosis airways. Pyocyanin, a phenazine pigment produced by all P. aeruginosa, particularly those growing in biofilms139, activates proinflammatory signaling140 and inhibits the activity of the antioxidants GSH and _N_-acetylcysteinie by blocking the dual-oxidase–based antimicrobial system141. Pyocyanin may also affect CFTR function by blocking Cl− transport in human bronchial cells142. The P. aeruginosa toxin CFTR inhibitory factor (Cif) redirects CFTR from recycling endosomes to the lysosome, where it is degraded143, compounding the deficiency associated with the major CFTR mutations; however, the in vivo role of this activity in cells that already have mutant CFTR targeted for degradation is not clear. These direct modifications of CFTR function in the host further potentiate the local immune defects associated with cystic fibrosis.

S. aureus has long been recognized as a major cystic fibrosis pathogen, and infection with the epidemic community-acquired methicillin-resistant S. aureus strains has been associated with decreased survival in individuals with cystic fibrosis144. Staphylococci adapt to the cystic fibrosis lung, proliferate as a biofilm and modulate proinflammatory activity, as monitored by cytokine and MAPK activation145. Similar to P. aeruginosa, S. aureus adapt their gene expression during colonization to enhance their fitness in the cystic fibrosis lung146. Within the S. aureus biofilm, the slower growing small-colony variants are more persistent or become resistant to antibiotics and evade host recognition by downregulating numerous virulence factors147–149. Subpopulations of the biomass revert to planktonic growth, further stimulating inflammation in the airways. Culture-independent as well as traditional methodologies have identified numerous other cystic fibrosis pathogens, especially streptococci and anaerobes that have successfully adapted to the cystic fibrosis lung and whose presence has been correlated with clinical symptomatology150. Thus, the major cystic fibrosis pathogens share key characteristics, one of which is the genetic flexibility to adapt to the milieu within the human airway, including resistance to antibiotic pressure, metabolic versatility and the ability to evade innate immune clearance mechanisms. These properties help to explain why the cystic fibrosis lung is infected by a relatively discrete group of organisms.

Implications for therapy

Whether cystic fibrosis is an innate or an acquired immune deficiency, the clinical benefits of targeting both the pathological immune response as well as the infecting organisms in this disease have been well documented in controlled clinical trials151, 152. Longitudinal studies of infants with cystic fibrosis in the first 2 years of life have documented loss of lung function, even in the absence of P. aeruginosa or S. aureus infection, both of which accelerate such a decline153. Antimicrobial therapy, in addition to the mechanical loosening of secretions and improved nutrition, have long been the mainstays of cystic fibrosis therapy, and longevity has paralleled the development of highly active antibiotics to P. aeruginosa (http://www.cff.org/). Recognition that young infants have airway inflammation that is out of proportion to the recovery of bacterial pathogens13, 22 led to clinical trials of anti-inflammatory agents such as systemic and inhaled steroids154, 155 and ibuprofen156, but each trial was limited by substantial toxicities. Azithromycin, a macrolide antibiotic that has a major anti-inflammatory activity, is effective in preventing loss of lung function in patients with cystic fibrosis157, even in those individuals who are not infected with susceptible bacteria, and is now a part of routine cystic fibrosis care at most centers. However, the accumulation of azithromycin within macrophages may also contribute to an increasing susceptibility to atypical mycobacterial infection in patients with cystic fibrosis who are treated with this drug158.

An alternative approach to cystic fibrosis therapy is based on targeting the dehydrated airway surface fluid present in individuals with the disease. Such strategies to counteract the physiological consequences of defective CFTR seem physiologically valid but have not been optimal thus far in controlled clinical trials. One approach to correct the defect in Cl− and water transport across the airway epithelium using a purinergic receptor P2Y purinoceptor 2 (P2Y2) receptor agonist to activate alternative Cl− channels was expected to result in hydrated airway surface fluid and improved mucociliary clearance. Despite promising initial results159, a large clinical trial using this drug failed to ameliorate lung disease in patients with a mild cystic fibrosis160. Whether these results were caused by a failure to target the innate immune defect, which is presumably CFTR dependent, is not clear. Similarly, the use of osmotic agents such as hypertonic saline or mannitol (Bronchitol) have had mixed results in clinical trials161–164.

The pharmacological potentiation of mutant CFTR function or the correction of mutant CFTR trafficking is a more desirable and realistic goal that, if achieved early enough in childhood, might prevent chronic infection and inflammation and irreversible airway damage. In vitro characterization of such compounds is currently in progress165. As clinical trials in human neonates are fraught with ethical and technical hazards, the newly developed _CFTR_−/− pig and _CFTR_−/− ferret models may be especially crucial in establishing whether pharmacologic therapy early in life can prevent cystic fibrosis lung disease (Box 2). Sequential bronchial biopsies could be done to ascertain whether endogenous upregulation of NF-κB activity is normalized, the amount of IL-8 and TNF in the airway is corrected and the numbers of recruited immune cells in the bronchoalveolar lavage fluid are normalized. Drugs to correct specific defects in CFTR biology, either by increasing the trafficking of mutant CFTR to the apical surface of the cell or by increasing Cl− transport through the mutant channel or both (by the use of two drugs in combination) are currently in clinical trials in patients with cystic fibrosis with specific genotypes166–168. However, there have not been any data published thus far that show that pharmacologic correction of channel function in vivo corrects abnormalities in innate immunity or prevents infection. This outcome should be a reasonable expectation, as in vitro correction of mutant CFTR decreases proinflammatory signaling in numerous model systems169, 170. Ideally, treatment of infants with cystic fibrosis with drugs that target specific CFTR genotypes would not only correct the defect in electrolyte transport in these infants but would also restore normal epithelial innate immune function, diminish endogenous proinflammatory signaling inflammation and prevent infection by correcting the many abnormalities detailed above. Such therapy has recently become available with the licensing of Kalydeco (ivacaftor) in January 2012, a CFTR potentiator that targets the CFTR G551D mutant, which has been found in approximately 1,200 people in the United States. Phase 3 clinical trials showed that the patients with cystic fibrosis carrying the G551D mutation (>6 years of age) who received this drug had improved pulmonary function. In the presence of established infection, as is the case in many patients with cystic fibrosis who are old enough to participate in clinical trials, a realistic goal of CFTR correctors and potentiators would be to stabilize lung function and decrease proinflammatory signaling. The availability of the porcine and, perhaps, the ferret models of cystic fibrosis should greatly expedite the evaluation of these new CFTR pharmacotherapeutics and facilitate the identification of optimal drug combinations, doses and methods of delivery.

BOX 2 Animal models of cystic fibrosis.

A tractable animal model of cystic fibrosis would be a tremendous tool in understanding how CFTR is linked to infection and for assessing the efficacy of therapy to correct CFTR dysfunction, particularly in young infants. The ideal animal model would accurately reflect the salient characteristics of human lung disease, namely excessive airway inflammation, spontaneous development of bacterial infection and progression to chronic infection with characteristic biofilm formation (Table 1). Although CFTR correction is typically monitored by the restoration of a cAMP-mediated Cl− current186, an animal model could be used to establish the clinically relevant outcome measurement, namely, the amount of CFTR correction that is necessary to prevent inflammation and infection. The much heralded _Cftr_-deficient mouse that was developed in 1992 (refs. 174, 187) unfortunately did not fulfill these goals; cystic fibrosis mice did not develop spontaneous lung disease, and an accurate reflection of cystic fibrosis pancreatic disease was dependent on the strain of mouse used175. Although these mice and, later, gut-corrected (_Cftrtm1Unc_-TgN(FABPCFTR) mice did show excessive proinflammatory signaling with increased chemokine (C-X-C motif) ligand 1 (CXCL1 or KC), TNF and PMN recruitment into the airways and increased weight loss during the course of infection, they spontaneously cleared even large inocula of typical cystic fibrosis pathogens. Thus, these mice have not been useful for testing the efficacy of antimicrobial regimens or of anti-inflammatory therapy or for monitoring effects of pharmacological correction of CFTR channel activity on inflammation or infection.

Recently developed _CFTR_-deficient171 and ΔF508/ΔF508 pigs188 spontaneously develop lung disease that is characterized by inflammation, mucus overproduction, airway obstruction and infection189. These features, which are typical of human disease, develop very early in life in these pigs. The _CFTR_-deficient pig fails to clear staphylococcal infection, a characteristic that may help to clarify the roles of CFTR expression in innate immunity. Physiological studies have suggested that the biology of porcine and human airway surface fluids are similar176. Notably, the antimicrobial activity of the airway surface fluid from cystic fibrosis pigs is not impaired172. _CFTR_-deficient pigs manifest the predicted defect in chloride and bicarbonate transport that is typical of human CFTR mutations, but they do not hyperabsorb sodium nor do they show diminished amounts of airway surface fluid. These results suggest that decreased hydration of airway surface fluid may not be central to the development of infection and inflammation190. Similarly, the cystic fibrosis ferret177 also develops lung infection very early in life, which is severe enough to require antibiotic treatment191. These models support the hypothesis that there is a direct role of CFTR in mucosal immunity beyond its contribution to the hydration of the airway surface fluids. These animal models may provide a useful model system to test the pharmacologic agents under development to correct specific CFTR mutations.

Concluding remarks

A tremendous amount of progress has been made in understanding the protean manifestations of cystic fibrosis: the genetic basis for cystic fibrosis and the influence of modifier genes in the disease has been established, the structure of the CFTR Cl− channel has been defined, and much of the complex physiology of CFTR and its central role in epithelial electrolyte transport has been explained. Cystic fibrosis pigs and ferrets are being developed that seem to reproduce human disease, providing model systems to explore the pathogenesis of cystic fibrosis and potential therapeutics, even in neonates. Clinical trials are ongoing to develop potent therapies based on this information to correct the multiple CFTR-associated abnormalities in epithelial function that together result in such a major defect in mucosal immune function. The involvement of CFTR in so many epithelial functions that are relevant to the host-pathogen interaction (constitutive activation of NF-κB, dysregulated TLR4 trafficking and signaling, associated defects in type I IFN regulation, impaired epithelial antioxidant activity and the failure of appropriate immune regulation) indicate that there is not a single ‘cause’ of the defective mucosal immunity associated with cystic fibrosis.

However, major questions remain unanswered. (i) Exactly how does CFTR participate in innate immune signaling? Is CFTR function involved in pathogen recognition, PAMP trafficking or the afferent limbs of signal transduction? (ii) Why do activated neutrophils in the cystic fibrosis airway fail to clear the initial P. aeruginosa infection? (iii) Does CFTR dysfunction affect dendritic cells, T cells or the development of adaptive immunity in the lung? (iv) What amount of CFTR ‘correction’ or ‘potentiation’ is necessary to restore normal innate immune function? (v) How early must therapy to correct CFTR function commence to prevent infection?

We anticipate that pharmacological correctors and potentiators of CFTR function, currently defined by restoration of chloride channel activity, will also ameliorate defects in mucosal immunity and diminish proinflamatory signaling and infection and the subsequent pulmonary pathology. Nonetheless, there are many unresolved issues that merit intensive investigation.

Table 1.

Characteristics of cystic fibrosis in human disease and animal models

Mouse
Human CFTR−/− andCFTRΔF508/ΔF505 ENaC overexpression Ferret Pig References
Spontaneousinfection P. aeruginosa; _S. aureus;_streptococci; anaerobes – – Streptococci;staphylococcienterococci Streptococci;staphylococci 116, 117 171, 172
Airwayinflammation PMNs; NF-κBactivation; IL-8;TNF-α Macrophage recruitment(strain dependent) PMNs; NF-κBactivation Uncharacterized PMNs 7–12, 173–176
Mucusaccumulationin airway Submucosal glandsecretion;mucus dehydration – Mucus plugs Uncharacterized lesionsin newborn lungs Mucus (hydrated);expanded submucosalglands 5, 6, 173, 172, 176 177

ACKNOWLEDGMENTS

Work in the laboratory is supported by US National Institutes of Health grants 5R21AI083491, 2R01HL079395 and 5R01HL073989 (A.P.) and a Parker B. Francis Fellowship (T.S.C.).

Footnotes

COMPETING FINANCIAL INTERESTS

The authors declare no competing financial interests.

References