Antiprotease therapy and cystic fibrosis

Antiprotease therapy in cystic fibrosis

Mike Henry. Jan, 2001. Antiprotease therapy in cystic fibrosis [online]. Leeds University Teaching Hospitals, Leeds, UK. Available from http://www.cysticfibrosismedicine.com

Introduction: Chronic progressive endobronchial infection in children and adults, associated with the development of bronchiectasis and a gradual loss of respiratory function is the predominant cause of morbidity and mortality in cystic fibrosis (CF) (Penketh 1987). Chronic infection, in particular by Pseudomonas aeruginosa, is associated with gradual tissue destruction and loss of pulmonary function. The inflammatory response to pulmonary infection is characterised by a sustained influx of neutrophils into the lung (Fick 1986). While this neutrophil influx is insufficient to overcome the bacterial load causing infection, there is considerable speculation that it may actually contribute to the perpetration of chronic infection and lung destruction (Berger 1991, Davies 1991).

An increase in the pro-inflammatory cytokine interleukin-8 (IL-8), associated with a concomitant increase in 'primed' neutrophils in pulmonary secretions of CF patients has been recognised not just in adults but also in the paediatric CF population (Khan 1993, Armstrong 1995, Nakamura1992). These 'primed' neutrophils in CF patients release many destructive oxidants and proteolytic enzymes including human neutrophil elastase (HNE) (Birrer 1993). The chronic release of HNE within the CF lung has the capacity to overwhelm the anti-protease shield, provided in the main by a1-proteinase inhibitor (a1-PI, a1-antitrypsin). It is thought that a substantial proportion of the pulmonary parenchymal tissue injury is mediated by unchecked activity of toxic proteases and oxygen metabolites (Birrer 1994).

In-vitro studies have shown that HNE can undermine the effectiveness of antibody-mediated phagocytosis by degrading opsonic proteins (including antibodies to P.aeruginosa) and C3b receptors on neutrophils (Fick 1984, Suter 1988, Berger 1989). The enzyme can also perpetuate lung inflammation and damage by degrading complement releasing C5a, a potent chemoattractant for neutrophils, and by increasing airway serous cell mucus production (Somerhoff 1990, Fahy 1992) and by undermining the immune response through release of IL-8 (Nakamura 1992). This chronic neutrophil mediated inflammatory response to continuous low-grade infection, punctuated by acute infective exacerbations, is associated with a waxing and waning of the HNE load within the lung and progressive pulmonary damage.

The protease-antiprotease imbalance hypothesis of chronic neutrophil mediated lung disease (Janoff 1997) suggests that HNE levels in the lower respiratory tract are a major determinant of pulmonary phenotype. There is indirect evidence in emphysema that severity of disease is directly related to pulmonary elastase load and inversely related to the anti-elastase shield (Fujita 1990). Likewise, in CF several studies have correlated tissue damage to elastase load and inversely to the a1-PI shield (Suter 1986, O'Connor 1993). Levels of active HNE in sputum increased with severity of disease. A positive correlation was found between sputum levels of active HNE and a1-PI and it has been assumed that this is an unsuccessful attempt by the lung to overcome the severe protease burden. An alternative interpretation might suggest that the increased a1-PI levels were independent markers of disease severity with unexplained pathological significance (Mahadeva 1998, Doring 1994, O'Connor - personal communication).

On a practical basis, HNE has several functions, which are detrimental to the lung. It degrades elastin, it's primary substrate, fibronectin and other structural proteins causing pulmonary extracellular matrix decay (Birrer 1993, Doring 1994). It cleaves cell surface fibronectin, thus enhancing P. aeruginosa adherence to the cells (Suter 1988) and it acts to impair ciliary function and thus reduce clearance of purulent pulmonary secretions.

Alpha1-Proteinase Inhibitor (a1-Antitrypsin): Fortunately, the normal lung is equipped with an impressive anti-elastase shield, which prevents HNE mediated destruction. Quantitative studies by Suter and colleagues (Suter 1986) have shown that a1-PI is the predominant elastase inhibitor in CF sputum hence, the investigation of the anti-elastase shield in the CF lung has focussed largely on this elastase inhibitor. Several studies have shown that the majority of endogenous a1-PI in the sputum of CF patients is in an inactive form. Similar work has demonstrated excess HNE activity in bronchoalveolar lavage fluid (BAL) representing lower airway activity (Goldstein 1986, Cantin 1989, Meyer 1991). Indeed the imbalance between HNE and the elastase shield is established by one year of age in the CF lung (Birrer 1994). This suggests the existence of a mechanism for inactivating a1-PI within the CF lung. It is thought that excess HNE and/or oxidants such as myeloperoxidase released from 'primed' neutrophils may be responsible for this inactivation of a1-PI.

Clearly, if such mechanisms are active in the CF lung, the potential efficacy of supplemental a1-PI is questionable, as the administered inhibitor may be inactivated in a manner similar to the endogenous protein. By contrast, if the major contributor to HNE excess is excessive production and release of HNE rather than inactivation of the inhibitor, then supplemental a1-PI may prove beneficial in CF. The work of O'Connor and co-workers (O'Connor 1993) has suggested that the latter may indeed be the case. Several potential mechanisms exist whereby, endogenous or supplemental a1-PI may be inactivated in CF. Structural and functional analysis of human a1-PI revealed that the antiprotease contains a critical methionine residue at position 358 in its reactive centre which determines its specificity towards HNE (Johnson 1979). This residue appears very sensitive to inactiviation by oxidants or matrix metalloproteinases (MMPs) released from 'primed' activated neutrophils (Ossanna 1986, Vissers 1988) such as may occur in CF. The triggered neutrophils may extend a zone of oxidising and proteolytic equivalents around the cell providing an environment where generated HNE would preferentially bind to host tissues before it collided with a native molecule of a1-PI. Once HNE is tissue bound, a1-PI cannot extricate or inhibit the active enzyme (Travis 1983, Janoff 1985). Even in non-infected sputum, up to 70% of a1-PI is inactive as an inhibitor (Morrison 1986, Stockley 1979). More recent work has suggested that activated neutrophils themselves may secrete small amounts of a1-PI (Pääkkö 1996). These tiny amounts may be disproportionally important in the active environment of HNE. Finally, there is evidence that most elastase -mediated damage occurs at the neutrophil / substrate interface. This is an area that a1-PI, predominantly released by the liver which diffuses into lung tissue, has difficulty accessing because of it's large size and negative charge (Campbell 1982). This highlights potential problems with intravenous or aerosolised a1-PI augmentation therapy. Reaching the neutrophil / elastin interface in areas of poor ventilation in the CF lung may prove very difficult for supplemental a1-PI for the reasons mentioned. The potential of several smaller serine proteinase inhibitors (serpins) in this respect will be highlighted later in this review. The theoretical argument surrounding supplemental a1-PI in CF was further complicated by the description of several authors of a polymorphism in the 3' flanking region of the a1-PI gene, which serves to upregulate a1-PI production during times of stress. Those with the Taq-1 polymorphism have been shown to have more severe disease in COPD and bronchiectasis. In-vitro studies have confirmed that these patients have an inability to increase a1-PI production in response to elastase load or IL-6 as occurs during acute infective exacerbations of CF. Those CF patients who carried either the Taq-1 polymorphism or the standard Pi (PiZZ) polymorphism leading to low baseline levels of serum a1-PI might then be expected to have a weaker antielastase shield. This has not proven to be the case however. Mahadeva and colleagues and other groups have described a trend towards better pulmonary function in both children and adult CF patients who were 'afflicted' with either of these mutations of the a1-PI gene (Mahadeva 1998, Doring 1994, Henry 1998).

Supplemental alpha 1-PI therapy: A commercial preparation of a1-PI is now available from pooled human plasma, called Prolastin (Bayer Corporation, Leverkeusen, Germany). Exogenous intravenous augmentation of the a1-PI shield via the infusion of pooled human a1-PI is currently recommended for those with severe hereditary (PiZZ / Null) deficiency of a1-PI (ATS Guidelines 1989). Infusion therapy seems safe and well tolerated (Dirksen 1999)

Dosage Schedules for a1-PI: Weekly (60 mg/kg), biweekly (120mg/kg) and monthly (250 mg/kg) infusions of a1-PI have been shown to be effective in maintaining a1-PI levels above plasma and ELF levels above the 'protective' threshold (80 mmol/L, 80 mg/dL) for the population in studies on patients with hereditary deficiency of a1-PI. Weekly infusions tend to be favoured because of better pharmacokinetics. (Gadek 1981, Wewers 1987, Moan 1983, Hubbard 1988). It is not reasonable to assume that these dosages of a1-PI would be enough to overcome the considerable elastase load present in the airways of CF patients.

When reconstituting the a1-PI protein (Prolastin) a concentration of 25 mg/ml should be used and an infusion maximal infusion rate of 50 mg/min (2 ml/min). Only 2% of the infused intravenous preparation will reach the lung (Rovner 1994) and not surprisingly researchers turned their attention to aerosolised preparations to improve efficacy, cut costs and save on the relatively scarce pooled human preparation. One of the main drawbacks of supplemental a1-PI in CF is the potential amount of a1-PI that would be needed to suppress high concentrations of HNE in the CF lung over a long period. The normal serum concentration of a1-PI in adult serum is 1.3 g/L which is adequate to overcome the HNE load. Levels up to 6 times normal may be produced during infection as part of the acute phase response. However, in CF lungs molar levels of immunoreactive HNE are 12 times higher that a1-PI in sputum in the stable patients without infective exacerbations (O'Connor 1993). The cost of such treatment would be a major factor given the lack of clinical efficacy displayed in CF and non-CF trials. At the dosages mentioned above, intravenous therapy was reckoned to cost between $25,000 and $35,000 in 1991! (Hay 1991)

Apart from the limited availability of Prolastin, there are also the inevitable concerns over the use of a human plasma product. Indeed, batches of Prolastin have been intermittently withdrawn over possible transmission of Creutzfeld-Jakob. It is recommended that in preparation for receiving prolastin, recipients be immunised against hepatitis B (ATS guidelines 1989).

Aerosolised a1-PI: This is an attractive alternative to intravenous infusion. Smaller amount of a1-PI are required.No intravenous access is required. Recombinant a1-PI has a short half life which precludes intravenous administration, yet maybe effective as an aerosol.Ease of administration of aerosol (Hubbard 1989, Hubbard 1990). Dosage schedules for the aerosol preparation tend to be 100 mg b.i.d. At this dosage plasma and ELF levels were significantly above the protective threshold and functionally intact (Hubbard 1990). These studies were in patients with hereditary deficiency of a1-PI and of course again may not mimic the situation of elastase excess in CF.

The first animal trials of supplemental a1-PI therapy in rat models of chronic P. aeruginosa lung infection showed that aerolised Prolastin, decreased elastase burden within the lung without having any bacteriacidal action. (Cantin 1989). Notwithstanding these contradictory arguments for and against supplemental a1-PI therapy in CF small phase I clinical trials of aerosolised a1-PI in CF have taken place (McElvaney 1991, Allen ED 1996) McElvaney and co-workers published the results of the first preliminary trial of a1-PI therapy in patients in 1991. This group administered aerosolised a1-PI (Prolastin) at dosages of 1.5-3.0 mg/kg b.i.d. for one week in 12 adult CF patients. Treatment suppressed neutrophil elastase in endolymphatic fluid (ELF) in the CF group. Furthermore, while pre-treatment a1-PI in ELF was elastase bound, inactive or cleaved, post-treatment ELF demonstrated not only suppressed elastase levels, but also fully active a1-PI (McElvaney 1991). Unfortunately, subsequent long-term phase II and phase III trials of supplemental a1-PI in CF have not been forthcoming. Large scale randomised controlled follow up trials of a1-PI treatment in individuals with hereditary deficiency of a1-PI and emphysema, however, have shown mixed results, though lower mortality rates and slower decline in FEV1 have been reported (Dirksen 1999, Hubbard 1989, The a1-antitrypsin deficiency registry study group 1998, Seersholm 1997). Vogelmeier and colleagues have demonstrated that anti-elastase activity is still present in airway secretions 36 hours after a single aerosolised dose of Prolastin (Vogelmeier 1997). Only one phase I study has followed on from these initial trials of supplemental a1-PI in CF. 22 adult CF patients were given supplemental nebulised Prolastin. While elastase levels were lowered, side effects related to the drug were described in 5 patients, one of whom had severe respiratory symptoms (Berger 1995). Until these safety issues and clinical efficacy of a1-PI can be demonstrated, the role of supplemental a1-PI in CF remains experimental.

Alternative strategies to pooled human a1-PI:

1. Recombinant a1-PI

2. Alternative antiproteases - SLPI / elafin

3. Synthetic inhaled anti-elastase drugs.

4. Gene therapy

Recombinant a1-PI: Given some of the clinical concerns surrounding human a1-PI, PP Therapeutics in Scotland have developed transgenic sheep which produce human a1-PI in their milk, which may overcome some of the safety and side effect concerns (Wright 1991). With the exception of several side chain sugars, the transgenic a1-PI (tg-a1-PI) is identical to the human protein. So far, human exposure to the transgenic protein in CF patients has proven safe with no documented allergic reactions (Balfour-Lynn 1999). Multicenter trials with tg-a1-PI are ongoing and results are awaited. Anti-proteases Secretory leucocyte proteinase inhibitor (SLPI) is a protein with a molecular weight of 11.7 kDa isolated from human parotid gland secretions (Thompson 1986). SLPI is a highly cationic non-glycosylated, single chain protein. It has been shown to be a potent inhibitor of HNE, human cathepsin G and human trypsin. It is produced in the central airways by serous glandular cells and in the lower respiratory by clara cells and goblet cells (Kramps 1981, de Water 1986). SLPI provides a significant component of the human anti-elastase shield within the lung. It has different reactivity from a1-PI and therefore the two anti-elastases may have complementary action. There are reasons why SLPI may have some advantages over exogenous a1-PI. SLPI is a hydrophobic cationic protein and, therefore, will readily bind to HNE and some of its substrates, for example the elastin / extracellular (ECM) complex. Unlike a1-PI, SLPI can inhibit elastin-bound HNE (Rice 1990). In addition, the interaction between SLPI and HNE is reversible, facilitating the transfer of HNE to a1-PI. Inactive a1-PI / HNE complexes are cleared by lymphatics and blood while SLPI is probably recycled to mop remaining uninhibited and active HNE (Gauthier 1982). SLPI is also far less susceptible to bacterial proteases than a1-PI (Spooner 1991). SLPI, along with a1-PI are regarded as the major defences against pulmonary HNE, providing anywhere between 10-50% of the total HNE inhibitory capacity throughout the lung depending on the disease state of the lung (Birrer 1992, Tetley 1987). Accordingly, SLPI has been suggested as a potential therapeutic agent in conditions of pulmonary neutrophil and HNE excess such as CF. Various protocols have been investigated for the administration of SLPI (GAST 1990, McElavaney 1993). Intravenous SLPI is unfeasible as it is a small molecule and quickly excreted in the urine. After encouraging in-vitro work, McElvaney and colleagues administered aersolised recombinant SLPI (rSLPI - Synergen, Boulder, CO) at a dosage of 100 mg b.i.d. for one week to 16 CF adult patients. After treatment with rSLPI levels of ELF HNE were significantly reduced and active SLPI levels were present in ELF (McElvaney 1993). As with a1-PI, phase II and phase III clinical trials are awaited. As with most other drugs the single greatest drawback is that aerosolised SLPI is not deposited particularly well in poorly ventilated areas of the lung that need treatment most. SLPI may also be susceptible to oxidation by activated neutrophils (Vogelmeier 1997), though oxidised SLPI may remain a potent elastase inhibitor, unlike a1-PI. The same applies to the degradation of SLPI by matrix metalloproteinases, a further group of potent proteases released by activated neutrophils which inactivate a1-PI but may not inactivate SLPI (Vissers 1988, Henry 1997).

Elafin, the elastase specific inhibitor, was first isolated from the skin of psoriatic patients (Wiedow 1991) and subsequently from sputum of normal subjects and those with farmer's lung (Sallenave 1992, Tremblay 1996). This 6 kDa peptide has many of the potential therapeutic attributes of SLPI. In-vivo, it is suggested that elafin contributes up to 20% of the anti-elastase capacity of the lung. Clinical trials of this novel peptide are outstanding.

Gene Therapy: Gene therapy represents the exogenous transfer of DNA which codes for normal a1-PI to deficient human cells allowing for ongoing endogenous production to augment deficient levels. cDNA can be packaged into the genome of viruses when the native genome has depletion of the normal regions which modulate viral replication. In-vitro work in fibroblasts and hepatocytes infected with reterovirus infused into host animals (Kay 1992) showed significantly increased a1-PI levels in plasma and ELF, but not enough to raise levels above the protective threshold. As the situation in the lungs of CF patients is one of relative a1-PI deficiency in the face of HNE excess, gene therapy would seem to be unlikely to be of potential value in this situation. Plasmid cationic liposome mediated a1-PI gene transfer to a CF bronchial epithelial cell line has been achieved (Canonico 1996), which has been shown in vitro to protect cells from elastase damage. No in-vivo work on a1-PI gene transfer has been yet reported.

Synthetic anti-elastases: Several synthetic oral agents are in development with the capacity to combat excess HNE activity. These agents have been shown to have action both within the azurophilic granules in which HNE is packaged within neutrophils prior to release from primed neutrophils and at the HNE / substrate interface (Vender 1998). One of these agents, L-658,758 has been shown to block over 90% of the elastolytic activity in-vitro in CF sputum (Rees 1997). Likewise, there are synthetic inhaled antielastase agents under investigation. Aerosolised FK706 was given twice daily to 16 adult CF patients for 10 days. It reduced sputum elastase activity and plasma inflammatory markers such as IL-8 (Jacobs 1998). Obviously the advantage of an oral agent lies with the potential to improve patient compliance. The results of phase II and phase III trials are awaited before early optimism and indeed continued pharmaceutical company development are justified.

Summary: Neutrophil mediated inflammation is a major component of pulmonary parenchymal damage in CF. Human neutrophil elastase is the main conduit of such damage. Despite early optimism that supplemental a1-PI administered either intravenously or by aerosol might have the potential to overcome this elastase burden in CF large scale clinical trials have not been forthcoming to support this early enthusiasm. Either plasma derived a1-PI or more likely it's recombinant form may overcome potential in vitro and in vivo hurdles to its efficacy. Recombinant SLPI or elafin used independently or in combination with supplemental a1-PI may prove more effective. The newer synthetic anti-elastase agents in development hold promise but require clinical data to justify early enthusiasm for their use.

References

Allen ED. Opportunities for the use of aerosolised a1- antitrypsin for the treatment of cystic fibrosis. Chest 1996; 110: 256S-260S.

American Thoracic Society. Guidelines for the approach to the patient with severe alpha-1-antitrypsin deficiency. Official statement of the ATS. Am Rev Respi Dis 1989; 140: 1494-1497.

Armstrong DS, Grimwood K, Carzino R, Carlin JB, Olinski A, Phelan PD. Lower respiratory infections and inflammation in infants with newly diagnosed cystic fibrosis. BMJ 1995; 310: 1571-1572.

Balfour-Lynn IM. The protease-antiprotease battle in the cystic fibrosis lung. J Royal Soc Med 1999; 37(suppl): 23-30.

Berger M, Soerensen RJ, Tosi MF, Dearborn DG, Doring G. Complement receptor expression on neutrophils at an inflammatory site, the pseudomonas infected lung in cystic fibrosis. J Clin Invest 1989; 84: 1302-1313.

Berger M. Inflammation in the lung in cystic fibrosis. A vicious cycle that does more harm than good ? Clin Rev Allergy 1991; 9: 119-142.

Berger M, Konstan M, Hilliard J and the CF Prolastin Study Group. Aerosolized Prolastin (a1-protease inhibitor) in CF. Pediatr Pulmonol 1995; 20: 421

Birrer P, McElvaney NG, Gillissen A, Hoyt RF, et al. Intravenous recombinant secretory leukoprotease inhibitor augments antineutrophil elastase defence. J Appl Physiol 1992; 73: 317-23.

Birrer P. Consequences of imbalanced protease in the lung: protease involvement in the destruction and local defence mechanisms of the lung. Agents Action Supplement 1993; 40: 3-12.

Birrer P. McElvaney NG, Rüdeberg A, Wirz Sommer C, Liechti-Gallati S, Kraemer R, Crystal RG. Protease-antiprotease imbalance in the lungs of children with cystic fibrosis. Am J Respir Crit Care Med 1994; 150: 207-213.

Campbell EJ, Senoir RM, McDonald JA, Cox DL. Proteolysis by neutrophils. Relative importance of cell substrate contact and oxidative inactive inactivation of proteinase inhibitors in vitro. J Clin Invest 1982; 70: 845-852.

Cantin A, Bilodeau G, Bégin R. Granulocyte elastase-mediated proteolysis of a1-antitrypsin in cystic fibrosis bronchopulmonary secretions. Pediatr Pulmonol 1989; 7: 12-17.

Canonico AE, Brigham KL, Carmichael LC et al. Plasmid-liposome transfer of the a1-antitrypsin gene to cystic fibrosis bronchial epithelial cells prevents elastase-induced cell detachment and cytokine release. Am J Respir Cell Mol Biol 1996; 14: 348-355.

Davis PB. Cystic fibrosis from bench to bedside. New Engl J Med 1991; 325: 575-577.

De Water R, Willems LN, Van Muijen GN, et al. Ultrastructural localisation of bronchial antileukoprotease in central and peripheral airways by a gold labelling technique using monoclonal antibodies. Am Rev Respir Dis 1986; 133: 882-890.

Dirksen A, Dijkman JH, Madsen F, et al. A randomised clinical trial of alpha-1-antitrypsin augmetation therapy. Am J Respir Crit Care Med 1999; 160:1468-1472.

Döring G, Krogh-Johansen H, Weidinger S, Høiby N. Allotypes of a1-antitrypsin in patients with cystic fibrosis, homozygous and heterozygous for DeltaF508. Paediatr Pulmonol 1994; 18:3-7.

Döring G. The role of neutrophil elastase in chronic inflammation. Am J Respir Crit Care Med 1994; 150: S114-117.

Gadek JE, Klein HG, Holland PV, Crystal RG. Replacement therapy of alpha 1-antitrypsin deficiency. J Clin Invest 1981; 68: 1158-1162.

Gauthier F, Frysmark U, Ohlsson K, Bieth JG. Kinetics of the inhibition of leucocyte elastase by the bronchial inhibitor. Biochem Biophys Acta 1982; 700: 178-183.

Hay JW, Robin ED. Cost-effectiveness of alpha 1-antitrypsin replacement of congenital chronic obstructive pulmonary disease. Am J Public Health 1991; 81: 427-430.

Henry M, Cave S, Morgan K, O'Connor CM, Kalsheker N, FitzGerald MX. a1PI mutations and severity of disease in cystic fibrosis. Eur Respir J 1998; 12(28): 234

Hubbard RC, Sellars S, Czerski D, et al. Biochemical efficacy and saftety of monthly augmentation therapy for alpha 1-antitrypsin deficiency. JAMA 1988; 260: 1259-1264.

Hubbard RC, Brantly ML, Sellers SE, et al. Anti-neutrophil elastase defenses of the lower respiratory tract in alpha 1-antitrypsin deficiency directly augmented with an aerosol of alpha 1-antitrypsin. Ann Intern Med 1989; 111: 206-211.

Hubbard RC, Crystal RG. Strategies for aerosol therapy of alpha 1-antitrypsin deficiency by the aerosol route. Lung 1990; 168: 565.

Khan TZ, Wagener JS, Best T, Martinez J, Accurso FJ, Riches DWH. Early pulmonary inflammation in infants with cystic fibrosis. Am J Respir Crit Care Med 1995; 151: 1075-1082.

Konstan MW, Hilliard KA, Norvell TM, Berger M. Bronchoalveolar lavage findings in cystic fibrosis patients with stable, clinically mild lung disease suggests ongoing infection and inflammation. Am J Respir Crit Care Med 1994; 150: 448-454.

Fahy JV, Schuster A, Ueki I, Boushey HA, Nadel JA. Mucus hypersecretion in bronchiestasis: the role of neutrophil proteases. Am Rev Respir Dis 1992; 146: 1430-1433.

Fick PA, Robins RA, Squier SU, Schoderbek WE, Russ WD. Complement activation in cystic fibrosis respirtatory fluids: in vivo and in vitro generation of C5a and chemotactic activity. Pediatr Res 1986; 20: 1258-1268.