Rational Drug Development for Mucous Dehydration and Mucous Metaplasia

Introduction

Mucociliary clearance (MCC) is a key component of the innate defenses of the airway, removing exogenous noxious particulates and endogenous pro-inflammatory substances that are enveloped in a mucous blanket and transported proximally by the rhythmic beating of cilia.^(1,2) Secretory dysfunction is a feature of most acute and chronic diseases of the airways due either to impairment of the ciliary function or to a change in the composition or quantity of secretions (Table 1). When MCC alone is not adequate to clear secretions, it is augmented by cough clearance. In severe illness, secretions may adhere to the airway walls, and it may become necessary to augment cough clearance with chest physiotherapy, for example, in patients with cystic fibrosis (CF).⁽³⁾

The well-recognized consequences of retained secretions include mechanical occlusion of airways leading to impaired gas exchange and the propensity of stagnant secretions to become infected. In addition, the chronic or recurrent acute lower respiratory tract infection and inflammation associated with secretory dysfunction can lead to lung remodeling, including mucous cell metaplasia, alterations in the quantity and composition of airway mucins, and bronchiectasis. In this review, the term “mucous metaplasia” is used in reference to alterations in the composition of airway mucins and the term “mucous cell metaplasia” describes the appearance of mucous cells (also known as goblet cells) in progressively more distal airways, and it is usually associated with hyperplasia/hypertrophy of the submucosal glands of the proximal airways.^(1,2)

The presence of metaplastic mucous cells in distal airways is of considerable pathophysiological interest because these airways have sparser cilia and weaker cough clearance than more proximal airways and, hence, the risk of chronic retention of secretions in this region is high. Further, the close proximity of these small airways to distal airspaces could increase the potential for retained inflammatory secretions to contribute to lung remodeling.⁽⁴⁾ It has been long agreed that a key aim of therapy in CF and other types of bronchiectasis is to interrupt the vicious cycle whereby retained infected secretions contaminate and injure previously healthy segments of the lung, which in turn serve as the nidus for further intrapulmonary contamination and injury.⁽³⁾

It is, however, becoming apparent that impaired MCC may amplify inflammation and promote remodeling to a more hypersecretory phenotype, even in the absence of established chronic infection or ongoing exposure to exogenous toxins.^(5–7) There is accumulating evidence that suboptimal hydration of airway mucins may be critical in initiating and sustaining the process of airway remodeling. For example, a mouse model that overexpresses the beta unit of the epithelial sodium channel (ENaC), which in turn produces dehydration of airway secretions due to an increase efflux of sodium from the airway lumen, is characterized by marked mucous cell metaplasia and emphysema that manifest within the first 2 weeks of life.^(6,7) These pathological changes develop in the absence of exposure to tobacco smoke or inoculation with pathogenic organisms. Thus, it appears that retention of secretions may, on its own, promote a hypersecretory phenotype.⁽⁷⁾

In other rodent models, administration of exogenous proteases and other pro-inflammatory cytokines have shown remodeling potential. For example in a rat model, instillation of neutrophil elastase (NE) led to a mucous metaplasia and was associated with increased production of a mucous glycoprotein, MUC5AC, without inoculation of bacteria.⁽⁸⁾ Data from a sheep model also suggest a mechanistic link between proteases and MCC.⁽⁹⁾ In that model, inhalation of exogenous NE or antigen associated with NE activation can lead to acute slowing of mucous clearance, reversible by specific inhibitors of NE and by amiloride, an inhibitor of epithelial sodium channels (ENaC). NE activates ENaC, which mediates the absorption of sodium from the airway lumen, and this promotes dehydration of secretions.^(5,6) Thus, NE may provide a link to both the acute slowing of MCC (i.e., ENaC) and long-term mucous hypersecretion (i.e., MUC5AC).

As will be discussed below, dehydration of airway mucins appears to be amenable to pharmacological intervention (Table 2) and, therefore, may have to potential to change the long-term clinical outcomes of a variety of airway diseases including COPD, asthma and bronchiectasis. First, however, a brief overview of the determinants of mucous hydration in health and disease is presented.

Mucous Hydration in Health and Disease

In health, airway mucous contains approximately 2% solids and 98% water.⁽¹⁾ In the solid phase of mucus, the secreted mucins designated as MUC5B and MUC5AC are, in terms of biophysical properties and mass, the most important constituents.⁽¹⁰⁾ These mucins are secreted by submucosal glands and goblet cells, predominantly in the more proximal ciliated airways. These free-floating mucins are located in an outer mucous layer that is separated from the surface of the respiratory epithelium by a periciliary layer (PCL) in which cilia beat rhythmically and move the mucous layer proximally, the process referred to above as mucociliary clearance. Recent evidence suggest that instead of a simple “Gel-on-Liquid” model of mucociliary clearance, that a ‘Gel-on-Brush’ model might better explain both the physiology and pathophysiology of mucociliary clearance in health and disease.⁽¹¹⁾

The Gel-on-Brush hypothesis is based in part, on the observation that, in addition to the secreted mucins, there are other mucins that are tethered to respiratory epithelium (MUC 1, 4, and 16), constituting a periciliary brush. The relative osmotic moduli of the free floating mucus layer and the periciliary brush control the depth of the periciliary layer. If the partial osmotic modulus of the mucus layer exceeds the osmotic modulus of the PCL, the depth of the PCL decreases, leading to a decrease in the rate of MCC. After a critical loss of depth, the mucus layer adheres to the collapsed PCL, and as a consequence, MCC will cease. The adherent mucus may occlude the airway lumen. Increases in partial osmotic modulus of the mucous layer may be due to an increase in solid content (e.g., increased mucin secretion from goblet cells) and/or a decrease in hydration (e.g., increased activation of ENaC leading to efflux of sodium ions from the airway lumen). Thus, maintenance of an adequate height of the PCL is critical to the integrity of MCC and variations in ion channel activity and secretion of mucins contribute to the different presentations of secretory dysfunction in CF, non-cystic fibrosis bronchiectasis (NCFB), chronic obstructive pulmonary disease (COPD), and asthma.⁽⁵⁾

Mucous Hydration and Mucins

The pathogenesis of the association between dehydration-induced impairment of MCC and mucous cell metaplasia is not well understood, including how impaired MCC could potentially reflect the relative abundance of the two free-floating mucins MUC5B and MUC5AC.⁽¹⁰⁾ Knockout and overexpression murine models suggest that these two mucins may have different roles in health and disease.⁽³³⁾ Constitutive MUC5B appears to be essential for maintenance of MCC and macrophage clearance of pathogens, whereas absence of MUC5AC was not associated with disease in the mouse model. Interpretation of studies on the levels of MUC5B and MUC5AC in inflammatory airway diseases is complicated by the fact that levels of the mucins have usually been measured using immunologic reagents. However in the presence of proteases, the levels of mucins appear to be decreased due to proteolysis of the targets for the antibodies.

In contrast, non-immune based biophysical assays show the mass and osmotic properties of mucins are much higher than measured by immune based assays.⁽³⁴⁾ Allowing for these caveats, there is evidence that MUC5AC increases more acutely than MUC5B in response to TH2/IL13 pathway activation, allergic airway hyper-responsiveness, smoke exposure, protease exposure, and infection.^(10,35) Thus, there is emerging evidence that MUC5B predominates in the healthy lung and that MUC5AC is upregulated in asthma.⁽³⁵⁾ It has also been suggested that MUC5AC is associated with goblet cells and MUC5B with submucosal glands, but this partitioning has been disputed.

In addition to soluble mucins, there are also tethered mucins (MUC 1,4,16),^(36,37) which play a role not only in regulating the height of the periciliary layer,¹¹ but may also be critical for regulating inflammation.⁽³⁷⁾ MUC 4 may play a role in resolving inflammation and mucous cell metaplasia after a proinflammatory stimulus has been removed. Exploring the link how mucous dehydration is associated with mucous cell metaplasia is an essential next step. The process is probably mediated through retention of mediators that would otherwise be cleared by MCC or cough. Being able to assess whether MUC5AC and MUC5B, and/or the tethered mucins, were regulated by different mediators or noxious stimuli would be a critical step as would better characterization of the regional distribution of mucins and an enhanced understanding of the role of tethered mucins in regulating inflammation.

Table 3 summarizes data on secreted and constitutive/tethered mucins.^(35–37) Some of the latter may become detached and contribute to the mass of free floating mucins. Specific tethered mucins are associated with cell types or glands. The tethered mucins may interact with inflammatory mediators, thus regulating inflammation.⁽³⁸⁾ As described above, they contribute to the osmotic gradient that preserves the integrity of airway surface liquid that is critical for maintenance of mucociliary clearance.

The role of impaired MCC (and by implication mucous hydration) in lung remodeling other than mucous cell metaplasia is more speculative but worthy of investigation. In COPD, the association of inflammatory mucous plugs in distal airways with loss of lung function suggests that impaired clearance of inflammatory exudates may amplify the process of remodeling in COPD (bronchiolitis).⁽⁴⁾ Even more intriguing is the consistent finding of an association of a single nucleotide polymorphism in the promoter region of an airway mucin gene, associated with overproduction of MUC5B, and an increased prevalence of idiopathic pulmonary fibrosis.⁽³⁹⁾ Whether this association reflects a relatively simple process such as the retention of inflammatory exudates in distal airspaces or a much more complex processes such as the endoplasmic reticulum stress in triggering mucous production, is not clear.⁽⁴⁰⁾

Prospects for Therapies Based on Mucous Hydration

There are several plausible pathways for drug development. They include topical osmolytes, drugs that act directly on epithelial ion channels, and drugs that target proteases that activate ENaC channels. While co-administration of mucolytics with hydrating agents is likely to be synergistic, a detailed review of mucolytic agents is beyond the scope of this review. Drug targets that focus primarily on the production and acute or chronic secretion of mucins are also beyond the scope.^(3,10)

Significant progress has been made in the development of medications that address the primary ion transport defect in CF.^(20,40) The impressive clinical efficacy of the potentiator ivacaftor in a subgroup of CF patients with gating mutations seems to validate the hypothesis that mucous hydration provides clinical benefit. Indeed, the impressive improvements in MCC with ivacaftor are consistent with this notion.⁽⁴¹⁾ There is, however, an important caveat to that assumed validation of the “hydration hypothesis” in CF. Correction of CFTR may have additional mechanisms beyond increasing the amount of intraluminal chloride. For example, restoration of bicarbonate secretion seems to be important in treatment of gastrointestinal mucous disorders in CF,⁽⁴²⁾ but the importance of bicarbonate in human lung secretions has not been as extensively investigated. In the neonatal pig model, however, acidification of secretions is associated with a decrease in clearance of secretions, that is correctable by bicarbonate therapy.^(14,15) It has also been suggested that CFTR may regulate epidermal growth factor receptor in a complex process involving modulation of neutrophil chemotaxis, protease activity, and mucus production.⁽⁴³⁾

In addition to the development of an effective potentiator for gating mutations of CFTR, there has been some progress in developing a corrector drugs for the most common mutation Delta 508. This is a more complex defect that involves failure of the protein to exit the endoplasmic reticulum. In this context, a combination of a potentiator (ivacaftor) and a corrector (lumicaftor) has been studied with evidence of efficacy in a phase 2 trial.⁽⁴⁴⁾ The results of large clinical studies demonstrated evidence of efficacy, and the combination is being is being evaluated by FDA (Vertex Press release: Vertex Submits Applications in the U.S. and Europe for Approval of Lumacaftor in Combination with Ivacaftor for People with Cystic Fibrosis Who Have Two Copies of the F508del Mutation November 5, 2014). However, the magnitude of clinical benefit was less than that observed with ivacaftor monotherapy for gating mutations.

There is evidence that cigarette smoking is associated with acquired defects in CFTR.^(24,45) Beta-sympathomimetic agonists can enhance CFTR function, but whether this of any clinical significance is not known.⁽⁴⁶⁾ To date, there is no clinical convincing evidence that CFTR specific correctors will provide benefit beyond CF.

CFTR is however not the only chloride channel that has been evaluated as a therapeutic target. However, clinical trials of denufusol with a P2Y2 agonist that stimulates the alternative were unsuccessful. Development of drugs for this pathway is complicated by the fact that these agents also stimulate mucin secretion.⁽⁴⁷⁾ Most likely the failure of denufusol reflects P2Y2 receptor desensitization following high dose denufusol administration.⁽⁴⁸⁾

Inhalation of exogenous osmolytes (sodium chloride and mannitol) can decrease exacerbations in CF and is tolerated by majority of the subjects, although a significant minority has to discontinue therapy due to cough and bronchospasm.⁽⁵⁾ In diseases other than CF, tolerability is likely to decrease because the increased prevalence of airway hyper-reactivity, especially in asthma. The delivery of a large mass of drug is time-consuming and discourages adherence. Unfortunately, decreasing the time of administration is likely to increase the density of the inhaled cloud and decrease tolerability. Intraluminal secretions increase the tolerability of osmolytes but exposure of mucosa to large osmotic loads can induce inflammation by altering intracellular osmotic gradients. Thus for highly motivated patients with marked mucous hypersecretion, these agents will continue to provide benefit, but a search for more effective and better tolerated alternatives will continue. Inhaled osmolytes may also have clinically relevant mucous modulating properties independent of their osmotic properties. For example, sodium chloride is a mucolytic that breaks ionic bonds, and mannitol may alter the surface tension of mucus.⁽⁴⁹⁾

The therapeutic role of ENaC in CF is an area of active clinical investigation. Clinical candidate molecules usually incorporate an amiloride pharmacophore.⁽⁵⁰⁾ Amiloride has been used orally as a potassium sparing diuretic for decades. It inhibits ENaC when delivered by aerosol to the airway epithelium. Topical delivery of potent ENaC inhibitors can, however, cause hyperkalemia, and the design of next generation molecules centers on achieving an increased therapeutic index.⁽⁵¹⁾ Another controversy centers on whether these agents can be combined with hypertonic saline to achieve a synergistic expansion of airway surface liquid. In a trial of patients with CF, treatment with hypertonic saline was associated with improvements in mucociliary clearance and lung function, but in contrast, a combination of hypertonic saline and amiloride was not effective.⁽⁵²⁾ The explanation for this unexpected result has been subject to speculation, including arguments for and against the potential role of amiloride as an inhibitor of aquaporins.^(49,52,53)

An alternative explanation was advanced by a recent study of human bronchial epithelial cells. Application of hypertonic saline to the apical surface led to increased intracellular sodium and cell shrinkage, provoking a feedback loop that led to a prolonged decrease in ENaC activity that persisted after the luminal hyperosmolarity resolved. This helps to explain why hypertonic saline provides a durable clinical response in CF. However if inhaled amiloride is co-administered with hypertonic saline, the immediate and short-lived inhibition of ENaC by amiloride decreases the magnitude of the sodium influx into epithelial cells, thus limiting the severity of the acute intracellular increases in sodium, and thus preventing a profound and prolonged inhibition of sodium transport. In other words, a possible explanation is that amiloride causes a short-lived inhibition of ENaC that prevents the intracellular increase in sodium from achieving a threshold severity that cause the more durable decrease in ENaC activity observed when hypertonic saline is administered alone.⁽⁵³⁾

Whether amiloride containing ENaC inhibitors should be used alone or in combination with hypertonic saline will need to be determined early in the development pathway to see if there are positive or negative synergies associated with combination. It can be hypothesized that a next generation amiloride pharmacophore should be able to provide sustained direct ENaC blockade, independent of intracellular sodium increases and cell shrinkage produced by hypertonic saline. Cell shrinkage is pro-inflammatory and theoretically it would be avoidable by the delivery of ENaC inhibitors in lower concentrations of saline over a longer delivery time to facilitate interstitial to luminal transport of water along a more tolerable osmotic gradient.⁽⁵⁴⁾

Development of inhibitors of the endogenous protease prostatin is a potential alternative approach to ENaC inhibition, but no clinical data have been published to date.⁽⁵⁵⁾

Macrolide therapy is associated with clinically meaningful benefit in CF, NCFB, COPD, post lung transplant obliterative bronchiolitis, and Japanese panbronchiolitis.^(3,56) However, it is not known with confidence if these benefits are related to the antimicrobial properties or to mucous modulation or anti-inflammatory properties. There is evidence that macrolides decrease secretion of mucus in response to inflammation by inhibiting extracellular signal regulated kinase pathways.⁽⁵⁷⁾ The recent findings in the sheep model that non-antimicrobial macrolides can reverse both antigen and NE-induced decreases in mucociliary clearance provide evidence that the benefits of macrolides may be due in part to improved mucous hydration. Macrolides do not appear to be direct inhibitors of ENaC or direct inhibitors of NE. Instead the authors postulated that the effect of macrolides was mediated through steric hindrance of the NE cleavage site on the γ-subunit of ENaC.⁽⁵⁸⁾

Inhibitors of soluble inflammatory proteases such as neutrophil elastase also may have potential applications in mucous cell metaplasia, due in part to their potential role as antagonists of protease-induced ENaC activation. Neutrophil elastase has been a target of drug development for years, but so far without success, except as replacement therapy in alpha-one antitrypsin deficiency.⁽⁵⁹⁾ However, systemically administered small molecule protease inhibitors are prone to toxicity due to the ubiquity of serine proteases and consequent potential for off target effects. Topical small molecule agents have been less extensively investigated. Large molecules such as secretory leukoprotease inhibitor (SLPI) and recombinant alpha-one antitrypsin have not been extensively studied. The clinical role of other non-serine proteases in regulating respiratory mucus has not been extensively elucidated, but it is plausible that these other proteases may activate ENaC and influence hydration.⁽⁶⁰⁾ Furthermore, dehydration of mucus induced by ENaC overexpression in a mouse model appears to trigger matrix metalloproteinase 12-dependent emphysema.⁽⁷⁾

Finally, the role of combination therapies should be considered. Mucolytic agents that decrease viscoelasticity by targeting disulfide bonds, or DNA, are likely to be synergistic with hydrating agents because hydration alone causes an exponential decrease in viscoelasticity.⁽³²⁾ Concomitant therapy with hydrating agents could enable lower and better tolerated dosing of mucolytics because agents that lyse disulfide bonds can potentially irritate respiratory mucosa.⁽⁶¹⁾

Conclusion

There is increasing evidence that dehydration of mucus decreases the clearance of pro-inflammatory mediators, including proteases, which can exacerbate the mucous dehydration leading to a positive feedback loop that may lead to mucous metaplasia and lung remodeling. Optimizing mucous hydration may provide a rational pathway for the development of novel agents that can alter the long-term outcome of common airway diseases.