It is now clear that NSPs (neutrophil serine proteases), including elastase, Pr3 (proteinase 3) and CatG (cathepsin G) are major pathogenic determinants in chronic inflammatory disorders of the lungs. Two unglycosylated natural protease inhibitors, SLPI (secretory leucocyte protease inhibitor) and elafin, and its precursor trappin-2 that are found in the lungs, have therapeutic potential for reducing the protease-induced inflammatory response. This review examines the multifaceted roles of SLPI and elafin/trappin-2 in the context of their possible use as inhaled drugs for treating chronic lung diseases such as CF (cystic fibrosis) and COPD (chronic obstructive pulmonary disease).
- inflammatory lung disease
- protease inhibitor
- secretory leucocyte protease inhibitor (SLPI)
- serine protease
Dysregulation of the protease/anti-protease balance is considered to be a major pathogenic determinant in a variety of acute and chronic inflammatory lung diseases, including adult respiratory distress syndrome, asthma, pulmonary fibrosis, COPD (chronic obstructive pulmonary disease) and CF (cystic fibrosis).
It was recognized early on that HNE (human neutrophil elastase) was probably involved in the development of pulmonary emphysema, by destroying lung elastin. It was found that a genetic deficiency of α1-PI (α1-proteinase inhibitor), the major physiological inhibitor of neutrophil elastase, was often associated with inherited emphysema [1,2]. HNE has many other biological activities that contribute to perpetuate inflammation, quite apart from its capacity to break down a wide variety of proteins. For example, it stimulates mucin production , activates MMPs (matrix metalloproteinases) , inactivates TIMPs (tissue inhibitors of MMPs)  and generates neutrophil-chemotactic elastin-derived fragments . Pr3 (proteinase 3) and CatG (cathepsin G), two elastase homologues released in massive quantities from neutrophils at inflammatory sites, have also recently been shown to have pro-inflammatory functions, acting through various mechanisms [7–9]. The lung epithelium of normal individuals is protected from the harmful effects of NSPs (neutrophil serine proteases) by a battery of anti-proteases. The most abundant is the plasma-derived serpin α1-PI that targets preferentially HNE in vivo and is also produced locally in the lungs by epithelial cells, monocytes and neutrophils. SLPI (secretory leucocyte protease inhibitor), an inhibitor of HNE and CatG but not of Pr3, is believed to control excess proteolysis in the upper airways, where it accounts for 80–90% of the total HNE inhibitory activity [10,11]. Elafin, the third NSP inhibitor, controls the activities of HNE and Pr3, and was first identified as a small 6 kDa acid-stable protein in human bronchial secretions . Elafin was later found to be derived from a biologically active 10 kDa precursor  called trappin-2  or pre-elafin. However, the term elafin is often used to denote either form of the inhibitors, so that it is not always clear whether published studies refer to elafin, trappin-2 or both molecules.
Both inhibitors have the same inhibitory potency towards their target proteases . They are essentially found in the lower respiratory tract, where they provide approximately 20% of the total HNE inhibitory activity of the lungs, as assessed from ELISA measurements on BALFs (bronchoalveolar lavage fluids) . Other NSP serpin inhibitors, including α1-anti-chymotrypsin and MNEI (monocyte/neutrophil elastase inhibitor) or serpinB1, are less abundant, but are also part of the lung anti-protease shield.
The huge quantities of NSPs released by neutrophils into the extracellular milieu in acute and chronic inflammatory conditions simply overwhelm the anti-proteases, resulting in excess proteolysis and its attendant deleterious effects. For example, we examined the relationships between active NSPs and their natural inhibitors in the sputa from patients with COPD using selective substrates for HNE, Pr3 and CatG. We detected nanomolar concentrations of all three NSPs (Figure 1). Although we also detected anti-proteases (α1-PI, SLPI and elafin/trappin-2), there was not enough of them to control NSP activities (molar ratio NSPs/inhibitors≫50). This protease/anti-protease imbalance was even more marked during exacerbation phases, when active protease concentrations increased up to approximately 20-fold and the inhibitory potential decreased (Figure 1). Thus anti-protease supplements should help to reduce the deleterious consequences of chronic inflammation induced by NSPs and it remains an attractive strategy for treating a number of inflammatory lung diseases. This review focuses on SLPI and elafin/trappin-2. Their therapeutic potential has generated great interest over the last decade when it was shown that they possess many other biological properties beyond their intrinsic ability to inhibit NSPs (Table 1).
The pleiotropic activities of SLPI and trappin-2/elafin anti-bacterial and anti-fungal activities
Both SLPI and trappin-2/elafin are anti-bacterial towards Gram-negative and Gram-positive bacteria, including bacteria with lung tropism (reviewed in [17,18]). They are therefore believed to be part of the mucosal host defence system, together with the defensin and cathelicidin families of anti-microbial peptides [18–20]. Trappin-2, like SLPI, also has fungicidal properties , especially against Aspergillus fumigatus and Candida albicans. These in vitro findings may have in vivo relevance, since a genetically driven increase in elafin in the lungs of mice protected these tissues against the inflammatory damage caused by Pseudomonas aeruginosa  and Staphylococcus aureus . We used a trappin-2 mutant that had no inhibitory activity (trappin-2 A62D/M63L) to show that the anti-bacterial activities of trappin-2 were independent of its inhibitory properties . The anti-microbial activities of both SLPI and trappin-2/elafin could be due to the cationic nature of these inhibitors, which could destabilize the anionic cell membranes of bacteria .
SLPI and trappin-2/elafin are substrates for transglutaminases
Trappin-2 contains five repeated sequence motifs having the general consensus sequence GQDPVK , which were recognized early on as motifs forming substrates for transglutaminase(s) . Tissue-type transglutaminase can cross-link trappin-2 to laminin in vitro  and links biotinylated variants of the consensus hexapeptide GQDPVK to various ECM (extracellular matrix) proteins  by forming intermolecular ϵ-(γ-glutamyl)lysine isopeptide bonds. We have shown that trappin-2, and, to a lesser extent, elafin, are readily covalently bound in vitro to ECM proteins including fibronectin, β-crystallin, collagen IV, fibrinogen and elastin by tissue-type transglutaminase or activated Factor XIIIa . We have also used several approaches, including enzyme-based assays and surface plasmon resonance, to demonstrate that fibronectin-bound trappin-2 or elafin can still inhibit their target proteases, HNE and Pr3 .
Although SLPI is structurally related to the elafin WAP (whey acidic protein) domain, it does not have the consensus GQDPVK transglutaminase substrate sequence in its two WAP domains. Hence it is not, a priori, a transglutaminase substrate. However, we have shown that SLPI can be readily cross-linked to fibronectin and elastin by both transglutaminase-2 and activated Factor XIII. Like trappin-2, cross-linked SLPI can still inhibit HNE and CatG. Interestingly, the transglutaminase-catalysed cross-linking of SLPI to fibronectin, which involves mainly lysine or glutamine residues of the N-terminal WAP domain, does not interfere with the cross-linking of trappin-2 to fibronectin. This unexpected capacity of SLPI to serve as a transglutaminase substrate may explain previous unexplained observations showing that SLPI was associated with elastin fibres in vivo [27,28].
These findings therefore suggest that the transglutaminase-mediated cross-linking of both SLPI and elafin/trappin-2 to soluble or insoluble structures in tissues provides local protection from excess proteolysis by NSPs. This covalent anchoring to ECM proteins may help prolong the actions of these inhibitors in the lungs of patients suffering from CF, where transglutaminase-2 is over-abundant .
SLPI and trappin-2/elafin as antiinflammatory proteins
SLPI and trappin-2 are not simply anti-proteases, but also have intrinsic anti-inflammatory properties that seem to be independent of their ability to complex proteases. Early experiments by Lentsch et al.  demonstrated that SLPI prevented the activation of NF-κB (nuclear factor-κB), which activates the synthesis of several pro-inflammatory mediators during lung inflammation. It has recently been demonstrated that SLPI is rapidly taken up by U937 monocytic cells and becomes located in the cytoplasm and nucleus . SLPI is thought to inhibit the breakdown of several key activation regulatory proteins [IκB (inhibitor of NF-κB) α, IκBβ and IRAK (interleukin-1-receptor-associated kinase)] by the ubiquitin-proteasome pathway. This blocks their release from the NF-κB complex, thereby preventing the entry of NF-κB into the nucleus and the activation of pro-inflammatory genes. SLPI also interferes with the NF-κB cascade in the nucleus by competing with p65 for binding to the NF-κB-binding sites in the promoter region of pro-inflammatory genes such as IL-8 (interleukin 8) and TNFα (tumour necrosis factor α) , thus inhibiting their transcription.
Elafin and trappin-2 have similar anti-inflammatory actions; they inhibit the activation of transcription factors AP-1 (activator protein 1) and NF-κB induced by LPS (lipopolysaccharide) in U937 cells by inhibiting the ubiquitin-proteasome pathway, as does SLPI [32–34]. In addition to interfering with the NF-κB signalling pathway, both SLPI  and elafin/trappin-2  bind bacterial LPS. This may inhibit the responses of macrophages to LPS, partly by blocking the transfer of LPS to soluble CD14 and the subsequent uptake of LPS–CD14 complexes by macrophages.
The benefits of aerosolized inhibitors
Inhaling inhibitors directly into the lungs to target extracellular proteases is very attractive since it limits treatment to the site of disease. It should provide high concentrations of active molecules where they are needed, so minimizing systemic side effects. Inhaled α1-PI significantly reduced the HNE-driven destruction of lung tissue in mice exposed to cigarette smoke , and in patients with CF [38,39]. The inhalation of α1-PI by patients with CF also restored the disabled ability of airway neutrophils to kill bacteria .
Aerosolized recombinant SLPI has been evaluated in sheep , in healthy patients and in patients with CF [41,42]. It was shown to reduce elastase activity  and the concentration of the potent neutrophil chemoattractant, IL-8 .
To our knowledge, inhaled forms of neither elafin nor trappin-2 have been evaluated, despite the fact that their anti-inflammatory properties have been demonstrated in several animal models of lung inflammation (reviewed in ). We have shown that trappin-2 can protect A549 lung epithelial cells from cell detachment induced by HNE (Figure 2) and Pr3 (results not shown). Preliminary studies indicate that trappin-2 A62L, a mutant that we designed to inhibit all three NSPs , also protects against the proteolytic damages caused by the three NSPs. We have recently initiated studies aiming to characterize trappin-2 aerosols delivered by various inhalation devices. We find that both trappin-2 and trappin-2 A62L retained their structural integrity and full inhibitory potency towards their respective target enzymes. We are now measuring the sizes of aerosol particles to select the most appropriate device for generating particles of the required size (typically 1–5 μm) that are compatible with pulmonary deposition.
Challenges for effective anti-protease aerosol therapy
The accessibility of pulmonary airways to aerosols of active drugs makes aerosol therapy with anti-proteases a potentially powerful method for treating acute and chronic lung diseases. However, depositing aerosols in the lungs requires not only particles of appropriate size but also that enough drug reaches the pulmonary territories to be treated, including poorly ventilated areas of the lungs. Treating patients with CF or emphysema with aerosols of recombinant SLPI proved to be effective only in well-ventilated areas of the lungs . This suggests that it may be more difficult to neutralize the NSP burden in poorly ventilated regions of the lungs.
As they are proteins, natural or engineered inhibitors may be degraded and/or inactivated in the inflammatory milieu, which is rich in oxidants and proteases. Although SLPI and trappin-2/elafin with their WAP domains are very stable, as are many small inhibitors, SLPI can be cleaved and inactivated by cysteine cathepsins . Elafin is also clipped at its N-terminal end by HNE. This does not affect its inhibitory potency, but Ps. aeruginosa proteases can inactivate it . Encapsulating these inhibitors within liposomes may circumvent unwanted proteolysis. Encapsulated SLPI was recently shown to be protected from degradation by cathepsin L . The aerosol delivery of liposome-encapsulated drugs is likely to become the preferential formulation because it has many other advantages including, for example, sustained release, increased biocompatibility and high-loading capacities.
The excess mucus in the airways of patients with most lung diseases not only affects the deposition of aerosols, but also acts as a barrier to the diffusion of drugs. In addition, the gel phase of the mucus contains NSPs that are partly resistant to inhibitors because access to protease active sites is limited. Treating mucus with DNase releases all three NSPs (S. Attucci A. Dubois and D. Bréa, personal communication), dramatically increasing the amount of harmful NSPs that must be inhibited. The situation is similar for NSPs bound to the chromatin fibres of NETs (neutrophil extracellular traps). Thus the mucolytic agents such as DNase that are currently used to treat CF may have unforeseen deleterious side effects due to their unmasking of NSP activities.
It has now become clear that all three NSPs, not just HNE, play a pivotal role in many inflammatory lung diseases. SLPI, trappin-2/elafin or their pan-inhibitory derivatives, such as trappin-2 A62L, may well be the long sought-after successful anti-inflammatory aerosol drugs. Their use could be of great benefit to patients suffering from chronic lung disorders. Future research will quite probably resort to nanotechnology to improve drug delivery and targeting to the lungs.
The work on proteases and protease inhibitors in our laboratory was supported by funds from the Institut National de la Santé et de la Recherche Médicale (INSERM) and grants from Vaincre la Mucoviscidose, ANR (NSP&COPD project), Région Centre (INFINHI project) and the European Community (Fonds Européen de Développement Régional, Région Centre).
The English text was edited before submission by Dr Owen Parkes. We also thank Professor François Lebargy (University of Reims, France) for providing the expectorations from patients with COPD.
Structure and Function of Whey Acidic Protein (WAP) Four-Disulfide Core (WFDC) Proteins: An Independent Meeting held at Robinson College, Cambridge, U.K., 12–14 April 2011. Organized and Edited by Colin Bingle (Sheffield, U.K.), Judith Hall (Newcastle, U.K.), Cliff Taggart (Queen's University Belfast, U.K.) and Annapurna Vyakarnam (King's College London, U.K.).
Abbreviations: α1-PI, α1-proteinase inhibitor; AP-1, activator protein 1; CatG, cathepsin G; CF, cystic fibrosis; COPD, chronic obstructive pulmonary disease; ECM, extracellular matrix; HNE, human neutrophil elastase; IκB, inhibitor of nuclear factor κB; IL-8, interleukin 8; LPS, lipopolysaccharide; NF-κB, nuclear factor-κB; NSP, neutrophil serine protease; Pr3, proteinase 3; SLPI, secretory leucocyte protease inhibitor; WAP, whey acidic protein
- © The Authors Journal compilation © 2011 Biochemical Society