Effects of Human Neutrophil Elastase and Pseudomonas aeruginosa Proteinases on Human Respiratory Epithelium Ryoichi Amitani, Robert Wilson, Andrew Rutman, Robert Read, Christopher Ward, David Burnett, Robert A. Stockley, and Peter J. Cole Host Defence Unit, Department of Thoracic Medicine, National Heart and Lung Institute, London, United Kingdom; Department of Infection and Inflammation, Chest Disease Research Institute, Kyoto University, Kyoto, Japan; and Lung Immunobiochemical Research Laboratory, The General Hospital, Birmingham, United Kingdom

It has been suggested that proteinase enzymes could play an important role in the pathogenesis of chronic bronchial infections including bronchiectasis and cystic fibrosis (CF). Because Pseudomonas aeruginosa frequently colonizes the respiratory tract in bronchiectasis and CF, we examined the in vitro effects of human neutrophil elastase (HNE) and proteinase enzymes produced by P. aeruginosa (elastase: PE; alka-

line proteinase: PAP) on the ciliary beat frequency (CBF) and ultrastructure of human nasal ciliated respiratory epithelium. HNE (500 J,tg/ml) progressively reduced CBF and caused marked epithelial disruption; lower concentrations (100 and 20 J,tg/ml) also caused epithelial disruption but without slowing CBF. The effects of HNE (500 J,tg/ml) were completely abolished by adding aI-antitrypsin (5 mg/ml). There was no synergy between HNE and pyocyanin, a product of P. aeruginosa which slows CBF. PE in phosphate-buffered saline also caused epithelial disruption without slowing CBF; however, PE in medium containing divalent metal ions caused CBF slowing as well as epithelial disruption at 100 J,tg/ml. PAP (500 J,tg/ml) had almost no effect on ciliated epithelium. The effects of HNE and PE on nasal and bronchial epithelium obtained from the same patient were similar. Light and transmission electron microscopy revealed that HNE and PE were cytotoxic and caused detachment of epithelial cells from neighboring cells and the basement membrane. There was cytoplasmic blebbing of the cell surface and mitochondrial damage; however, no increase of abnormalities in the ultrastructure of cilia on living cells was seen. These results support the hypothesis that HNE and PE contribute to the delayed mucociliary clearance and epithelial damage that is observed in patients with chronic bronchial infection.

It is widely accepted that human neutrophil elastase (HNE)

plays an important role in the pathogenesis of pulmonary emphysema (1, 2). During chronic respiratory infection, there is a large influx of polymorphonuclear leukocytes to the bronchial tree (3), and free proteinase enzyme activity is present in purulent sputum(4). For this reason, several investigators have suggested that HNE may also damage respiratory epithelium resulting in impairment of mucociliary transport, causing retention of infected secretions and escalation of the inflammatory response (5-8). This process (Received in original form December 16, 1989 and in revised form August 2,1990) Address correspondence to: Dr. R. Amitani, Department of Infection and Inflammation, Chest Disease Research Institute, Kyoto University, Sakyoku, Kyoto, 606, Japan. Abbreviations: ciliary beat frequency, CBF; cystic fibrosis, CF; human neutrophil elastase, HNE; Pseudmonas aeruginosa alkaline proteinase, PAP; phosphate-buffered saline, PBS; Pseudomonas aeruginosa elastase, PE. Am. J. Respir. Cell Mol. BioI. Vol. 4. pp. 26-32, 1991

could then contribute to a self-perpetuating "vicious circle" of host-mediated respiratory tract damage (9, 10) in diseases such as bronchiectasis and cystic fibrosis (CF). Previous studies have evaluated the in vitro effects of purulent sputum sol with elastolytic activity from bronchiectatic patients on human nasal cilia (6, 7), and the in vitro effects of purified HNE on rabbit tracheal cilia (5). However, there have been no studies of the in vitro effects of purified HNE on human respiratory epithelium. Consequently, the aim Ofthis study was to investigate the effects of purified HNE on human respiratory epithelium. A number of bacteria that colonize the respiratory tracts of patients with bronchiectasis also produce proteinase enzymes that have been implicated in their virulence (11) and other compounds that affect ciliary beating (10). The effects of HNE were therefore compared to two bacterial metalloproteinase enzymes produced by Pseudomonas aeruginosa, and the effect of HNE in the presence of pyocyanin (a product of P. aeruginosa which slows ciliary beating) was also studied.

Amitani, Wilson, Rutman et al.: HNE and P. aeruginosa Proteinases on Human Respiratory Epithelium

Materials and Methods Materials HNE was purified from an empyema. Sodium chloride solution was added to a final concentration of 1 mol/liter and Triton x-tOO to a concentration of 1% (vol/vol). The mixture was centrifuged, and the supernatant dialyzed versus 0.05 mol/liter Tris-HCl buffer, pH 7.5, containing 0.035 mol/liter MgClz. The dialyzed material (800 ml) was treated for 5 h (370 C) with 1 mg of proteinase-free bovine pancreatic DNAse (Sigma Chemical Company, Dorset, UK). The remaining purification procedure was based on that of Martodam and associates (12). Briefly, the solution was subjected to affinity chromatography using Sepharose-bound trasylol and separation of HNE from cathepsin G was achieved by chromatography on CM-Sephadex. The HNE was finally dialyzed versus phosphate-buffered saline (PBS) without calcium and magnesium. The HNE was shown to be pure by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and was shown to hydrolyze Ala-Ala-Pro-Val pNA and SueAla-Ala-Ala pNA but not a cathepsin G substrate (Suc-PheLeu-Phe pNA). Pure crystalline P. aeruginosa elastase (13) (PE) and P. aeruginosa alkaline proteinase (14) (PAP) were purchased from Nagase Biochemical (Fukuchiyama, Kyoto Prefecture, Japan), and aI-antitrypsin was purchased from Sigma. Pyocyanin was synthesized as previously described (15, 16). It was shown to be indistinguishable from biologic pyocyanin by ultraviolet spectrophotometry, high performance liquid chromatography running position, and mass spectrometry (17). Examination of Epithelium by Light Microscopy and Measurement of Ciliary Beat Frequency (CBF) Human nasal epithelium was obtained from the inferior turbinate of normal volunteers by a brushing technique using a cytology brush without local anesthesia (18). Human bronchial epithelium was obtained at bronchoscopy from the right or left main bronchus of two patients undergoing investigation for possible carcinoma. The area brushed was macroscopically normal. Informed written consent was obtained, and this procedure was approved by the Brompton Hospital Ethics Committee. Strips of ciliated epithelium were dislodged by brisk agitation of the cytology brush in Medium 199 cell culture fluid (Flow Laboratories). The suspension of ciliated epithelium was then divided into two, three, or four aliquots which were centrifuged at 200 X g for 5 min. One epithelial pellet was resuspended in PBS. The others were resuspended in 300 to 500 ttl of test solutions that consisted of different concentrations of proteinases (HNE, PE, and PAP) dissolved in PBS. Each sample was transferred to a sealed microscope slide-coverslip preparation, placed on an electronically controlled warm-stage (Microtec) at 3r C, and allowed to equilibrate for 15 min. The ciliated epithelium was viewed directly through a Leitz Dialux 20 phase-contrast microscope at magnification x320. Between six and to strips of ciliated epithelium each containing several hundred cells were identified on the slide, and their positions carefully recorded in order that serial readings ofCBF could be taken from the same strips through-

27

out the experiment. CBF was measured by a previously described photometric technique (18-21). Single cells and small groups of ciliated cells were disregarded, as they were considered inadequate for CBF measurement over a 6-h period. An assessment of the ciliary beating pattern and epithelial integrity was made in addition to measurement of CBF at each experimental timepoint. Ciliary dyskinesia was definedas disorganization of the normal ordered ciliary beating pattern; ciliostasis was defined as complete cessation of ciliary beating; epithelial disruption was defined as loss of the integrity of the usually smooth epithelial surface which became irregular. Epithelial disruption was assessed subjectively by the same investigator and scored from 0 to + + +: 0 was normal (see Figure 3a); + was undulation of the epithelial strip surface; ++ was undulation and extrusion of individual cells from the epithelial strip surface; +++ was total disorganization of the epithelial strip surface (see Figure 3b). Ten readings of CBF were taken from the epithelial strips selected at the beginning of the experiment and at l-h intervals for 6 h thereafter. Readings of CBF were taken from the same area on each epithelial strip at each timepoint. When more than one reading was taken from a single strip, the two

16

14

12

N 1:

.. 5

10

~

8

e : III

6

...".. c

It. ~

*

t

~

0

4

(-lffo )

* 2

o

2

3

4

5

6

Time (hours)

Figure 1. Mean (n = 10) ciliary beat frequency (CBF) values measured throughout experiments with phosphate-buffered saline (PBS) containing different concentrations of human neutrophil elastase (HNE). The mean CBF values are shown as follows: 20 jlg/ml HNE (closed circles), 100 jlg/ml HNE (open triangles), 500 jlg/ml HNE (closed triangles), and control (PBS) (open circles). Degree of disruption of ciliated epithelium is indicated in parentheses; +: irregular margin of ciliated epithelium; + +: undulation and extrusion of individual cells from the epithelium surface; + + + : disorganization of the epithelium surface. Closed star: ciliostasis first observed. Open star: dyskinesia first observed. Bars represent 1 SD. »P < 0.001 with reference to control value.

28

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 \99\

16

14

*

_.A-_-~(±)

12-

.., N

---0--------0-------1

-------0--------

**

10-

(± )

(-tt)

s:

>-



iii

8

::>

~

u,

10

co'"

(-tit )

6

t

~

U

4

2

V

I

I

0

2

3

4

5

6

Time (hours)

Figure 2. The inhibitory effect of ((t-antitrypsin on HNE. The mean (n = 10) CBF values are shown as follows: 500 Ilg/ml HNE (closed circles), 500 Ilglml HNE and 5 mglml ((t-antitrypsin (open triangles), 500 Ilglml HNE and 1 mg/ml ((I-antitrypsin (closed triangles), and control (PBS) (open circles). Degree of disruption of ciliated epithelium was indicated in parentheses as described in Figure 1. Bars represent 1 SD. -e < 0.001, **p < om with reference to CBF value of HNE (500 Ilg/ml) solution.

areas chosen were well separated and no more than two readings were taken from anyone strip. The slides were randomized at the beginning of each experiment so that the observer was unaware of the nature and concentration of the substances that were tested in each experiment. The mean CBF of control and test preparations were compared at each timepoint and the maximum percentage slowing of test CBF compared to control was identified (21). At this timepoint, the mean CBF (n = 10) of test and control preparations were compared using the unpaired Student's t test. Ultrastructural Study of the Effect of Proteinases on Human Respiratory Epithelium Human nasal turbinate tissue was obtained from patients un-

Figure 3. The effect of HNE on human nasal ciliated epithelium assessed by light microscopy. Nasal epithelium obtained by a cytology brush after 6-h experiments in PBS (a) and HNE (500 Ilg/ml) solution (b) (original magnification: 600X). Ciliated epithelium was extremely disrupted in HNE solution compared with control(PBS). Nasal turbinate tissue incubated in HNE (100 Ilg/ml) for 4 h is also shown (c). A thick toluidine blue-stained preparation (original magnification: 600x). The normal surface integrity of the epithelium has been disrupted, cells have become separated from each other and the basement membrane, and single cells are seen with cellular debris.

Amitani, Wilson, Rutman et al.: HNE and P. aeruginosa Proteinases on Human Respiratory Epithelium

29

TABLE I

Ciliary beat frequency of human nasal epithelium suspended in PBS or Ham's F-12 medium with different concentrations of Pseudomonas proteinases for 6 h * Concentration of Proteinase Enzymes

o PE in PBS PE in Ham's F-12 medium PAP in PBS PAP in Ham's F-12 medium

11.3 11.9 11.1 11.2

± ± ± ±

1.4 (+) 1.3 (+) 1.0(+) 1.6 (+ +)

100 jlg/ml

11.8±1.1(++) 11.3 ± 0.9 (+)

11.7 8.6 11.4 11.0

± ± ± ±

1.4 1.8 1.3 1.7

(+++) (+++)t (+) (+)*

11.8 ± 1.4 (+++) 11.7 ± 1.7 (+)

Definition of abbreviations: PAP = P. aeruginosa alkaline proteinase; PBS = phosphate-buffered saline; PE = P. aeruginosa elastase. * Data are mean values (Hz) ± I SO, with the degree of epithelial disruption in parentheses. t P < 0.01 compared to control (Ham's PI2 medium alone); in a concurrent experiment, CBP of epithelium from the same sample incubated with 100 jlg/ml PE in PBS was 11.7 ± 1.6 (++). No significant difference compared to control (Ham's PI2 medium alone); in a concurrent experiment, CBP of epithelium from the same sample incubated with 100 jlg/ml PAP in PBS was 12.4 ± 1.6 (+ +).

*

dergoing resection for nonallergic nasal obstruction (kindly supplied by Miss V. Lund, Royal Ear, Nose and Throat Hospital, Grays Inn, London, UK). Tissue was transported to the laboratory in Eagle's minimum essential medium (GIBCO, Middlesex, UK) supplemented with penicillin (50 Ulml), streptomycin (50 j.tg/ml) , and gentamicin (50 j.tg/ml). On arrival, it was inspected by light microscopy to ensure that the epithelial surface was normal and fully ciliated and that CBF was within the normal range (12 to 16 Hz). Adjacent 3-mm cross-sections of epithelium were carefully dissected for use in experiments. The mucosal sections were suspended in PBS alone (as control) and in PBS containing different concentrations of proteinases. The cell suspensions were incubated at 3r C for 4 h, then fixed in cacodylatebuffered 2.5% glutaraldehyde, followed by postfixation in 1% osmium tetroxide, and processed for transmission electron microscopy (22). Sections were examined using a morphometric technique by an observer unaware of their origin. An ultra-thin section through the central portion of each specimen was examined. Sections consisted of 150 to 200 cells, and each cell was scored using the following criteria: extrusion of the cell from the epithelial surface, loss of cilia from ciliated cells, cytoplasmic blebbing of the epithelial surface, and mitochondrial damage. The amount of the epithelium that was ciliated was assessed by counting the number of ciliated and unciliated cells and by making a subjective score of the number of cilia present on ciliated cells. Changes in ciliary ultrastructure were scored as abnormalities of microtubules (disarrangement, addition or loss of microtubuIes, abnormalities in the central pair, compound cilia) and of dynein arms (loss of either inner, outer, or both dynein arms).

Results HNE had no effecton the beat frequency of human nasal cilia during the first hour. There was, however, subsequently progressive slowing with marked epithelial disruption at a concentration of 500 j.tg/ml (Figure 1). Ciliary dyskinesia occurred after 2 h when the ciliary beating was beginning to slow, and areas of epithelium with complete ciliostasis occurred after 5 h. At lower concentrations (20 and 100 j.tg/ml), whe~e there was no ciliary slowing over 6 h, epithelial disruption was still observed although the degree was de-

pendent upon the dose used (Figure 1). Figure 2 shows that human at-antitrypsin (5 mg/ml) completely prevented CBF slowing and epithelial disruption caused by HNE (500 j.tg/ml), whereas lower concentrations of at-antitrypsin (1 mg/ml) only partially inhibited the effects. at-antitrypsin (l to 5 mg/ml) itself had no effecton CBF. On light microscopy, marked epithelial disruption was demonstrated in the strips of epithelium obtained by nasal brushing after 6-h incubation at 37° C in HNE solution (500 j.tg/ml), although the ciliated epithelium showed only minor disruption after the 6-h incubation in control experiments with PBS (Figures 3a and 3b). Incubation of human nasal turbinate tissue with HNE (100 j.tg/ml) showed that there was detachment of epithelial cells from neighboring cells and basement membrane (Figure 3c). Table 1 shows results of experiments' with PE and PAP. When both proteinases were diluted in PBS without divalent metal ions, they had no effect on CBF, even at 500 j.tg/m1. However, PE caused epithelial disruption in a dose-dependent manner between 20 and 500 j.tg/ml, whereas PAP had no effecton epithelial structure at concentrations between 20 and 500 j.tg/ml. When PBS was replaced by Ham's F-12 tissue culture medium (Flow Laboratories, containing Ca2+ [12.0 mg/liter], Mg2+ [14.6 mg/liter], Zn2+ [0.196 mg/liter], Fe2+ [0.168 mg/liter], but not containing Co2+), PE (100 j.tg/ ml) caused a modest fall in CBF after 6 h which was associated with marked epithelial disruption, whereas PAP (100 j.tg/ml) still had no effect on CBF and epithelial structure. The effects of HNE (100 j.tg/ml) and PE (100 j.tg/mI) on human nasal and bronchial epithelium were similar. With epithelium from the first patient after 4 h of incubation with HNE, there was 7% (not significant) slowing of nasal CBF and epithelial disruption graded +++ (test mean CBF, 11.5 Hz; control mean CBF, 12.4 Hz) and no slowing of bronchial CBF and disruption graded +++ (test mean CBF, 12.5 Hz; control mean CBF, 12.4 Hz). With epithelium from the second patient after 4 h of incubation with PE, there was 5 % (not significant) slowing of nasal CBF and epithelial disruption graded ++ to +++ (test mean CBF, 11.7 Hz; control mean CBF, 12.3 Hz) and 4% (not significant) slowing of bronchial CBF and epithelial disruption graded ++ to +++ (test mean CBF, 10.7Hz; control mean CBF, 11.1 Hz). A Pseudomonas product with known cilio-inhibitory

30

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 41991

a

and HNE (100 Itg/ml) in their effects on CBF was demonstrated. Ultrastructural study of ciliated epithelium revealed epithelial damage in elastase solution but not PAP, compared with that in control solution (PBS) (Figures 4a through 4c and Table 2). The characteristic findings on transmission electron microscopy were extrusion and desquamation of epithelial cells, cytoplasmic blebbing of the cell surface, and mitochondrial damage. The amount of the epithelial surface that was ciliated was identical in control and test experiments. Although abnormal ultrastructure was seen in cilia arising from dead epithelial cells, when counts were made from cilia arising from living cells only small numbers of abnormalities were observed (Table 2). Therefore, there was probably no direct effect of proteinase enzymes on the cilia themselves.

Discussion

Figure 4. Transmission electron microscopy of human nasal turbinate tissue incubated in PBS (a) and HNE (100 p.g/ml) for 4 h (b, c). Normal architecture of the ciliated epithelium was maintained in control (PBS) solution through the experiment (a), whereas a cell is seen separated from its neighbors and extruding from the epithelial surface (b) and cytoplasmic blebbing of the epithelial surface is shown in (c) (original magnification: 7,500X [a, b) and 12,000x [cl).

properties, the phenazine pigment pyocyanin (21, 23) was tested either with or without HNE (100 Itg/ml) at a concentration (10 JLgiml) not affecting CBF, but causing epithelial disruption. There was no potentiation of the slowing of CBF in the preparations containing pyocyanin and HNE, compared to those with pyocyanin alone; pyocyanin alone (at oh and after 6 h): 13.1 Hz, 2.6 Hz; pyocyanin + HNE: 12.2 Hz, 3.4 Hz. The gradual reduction in CBF was identical in the two experiments, and no synergism between pyocyanin

Since the first report (1) in 1963 of an association between the early onset of pulmonary emphysema and an inheritable deficiency of serine proteinase inhibitor a -antitrypsin, many studies of the role of HNE in the pathogenesis of pulmonary emphysema have been carried out, and the hypothesis of proteinase-antiproteinase imbalance in the pathogenesis of emphysema has gained wide acceptance (2). For the past decade, several investigators have suggested that HNE also caused damage to the respiratory epithelium and played an important role in the pathogenesis and the progression of chronic diseases of the bronchi such as bronchiectasis. Tegner and colleagues (5) first reported that HNE reduced CBF and caused superficial tissue destruction of the trachea, but used rabbit trachea as the target organ and cautioned that corresponding experiments on human mucosa were necessary before any extrapolation could be made to human disease. This disparity between results obtained with animal and human tissue was confirmed by Rutland and coworkers (24), who showed that cystic fibrosis serum reduced beat frequency of rabbit tracheal cilia but not human nasal CBF in vitro. Thus it is clear that species differences do exist in response to factors that affect cilia. It is therefore important that the effect of HNE on human respiratory cilia be examined. Smallman and associates (6) reported that elastase-positive sputum sols from patients with purulent bronchiectasis reduced beat frequency of human nasal cilia significantly in vitro and that this effect was abolished by prior addition of al-antiproteinase to the sputum sol. These workers suggested that the ciliotoxic effect was related to a serine proteinase, probably neutrophil elastase, on ciliary beating. Nevertheless; there is the possibility that sputum sols from bronchiectatic patients contain a variety of cilioinhibitory factors other than HNE and that interactions between HNE and other factors cause reduction of CBF (7, 21, 23). Sykes and colleagues (7) demonstrated at least two factors in purulent sputum sol from patients with chronic bronchial sepsis that impaired ciliary beating; one a serine proteinase probably released by the host's phagocytic defenses, the other probably a bacterial product. In that study, there was a poor correlation between sputum sol's elastolytic activity and its potency for slowing CBE Moreover, there was a lack of ciliary beat slowing in four patients whose sputum

31

Amitani, Wilson, Rutman et al.: HNE and P. aeruginosa Proteinases on Human Respiratory Epithelium

TABLE 2

Ultrastructure of human nasal ciliated epithelium incubated in proteinase enzymes for 4 h Mitochondrial Swelling

Percentage of Cilia with Abnormalitities (Number of Cilia Counted)

Number of Cells Counted

Extruding Cells

Epithelium Surface Blebbing

(%)

(%)

(%)

Microtubules

Dynein Anus

Control (PBS) HNE (100 /Lg/mJ)

153 130

15 48

8 28

5 29

0.7 (293) 1.7 (363)

2.2 (90) 2.7 (151)

Control (PBS) PE (100 /Lg/mJ)

186 146

4 47

3 36

3 43

2.3 (129) 4.4 (270)

0(48) 3 (100)

Control (PBS) PAP (100 /Lg/mJ)

184 178

13 19

7 22

5 14

0.7 (272) 1.2 (248)

1.2 (82) 0.9 (112)

Definition of abbreviations: HNE = human neutrophil elastase; for other abbreviations. see Table I.

sol displayed considerable elastolytic activity (7). Consequently, it is necessary to examine the effectof purified HNE alone on human ciliated epithelium in order to clarify the role of HNE in the pathogenesis of chronic bronchial infection. In the present study, we examined the effects of purified HNE on human nasal ciliated epithelium in vitro. Stockley and coworkers (4) have shown free elastase activity equivalent to more than 50 p,g porcine pancreatic elastase activity/ ml sputum sol in 29 out of 62 purulent sputum samples obtained from patients with bronchiectasis. In the present study, HNE (20 to 500 p,g HNE activity/ml sputum sol) affected ciliated epithelium. At high concentration of HNE (500 p,g/ml), progressive slowing of ciliary beating occurred, but not during the first hour of the experiment. When CBF slowing occurred, it was associated with marked epithelial disruption. Epithelial disruption was noted to a lesser extent at lower concentrations (100 and 20 p,g/ml), but CBF remained normal. Ql-antiproteinase inhibited the effects of HNE in a dose-dependent manner, showing that the effect of HNE related to the enzymes' activity. Bronchial mucus inhibitor is present in proportionally greater amounts than QI-antiproteinase in the lower respiratory tract, and as it is a reversible inhibitor of HNE, its effecton the interaction between HNE and epithelium may be different. Transmission electron microscopy showed separation between cells and extrusion of cells from the epithelial surface. There were cytotoxic changes in the epithelial cells and dead cells were seen, but the number and ultrastructure of the cilia remained normal on living cells. Although a small number of microtubular and dynein arm abnormalities were observed, similar changes are also seen in normal specimens (25). The effects of HNE were therefore predominantly on the cells themselves and their attachment to each other and to the basement membrane, rather than on ciliary structure and function. At high concentration (500 p,g/ml), the reduction in CBF probably reflects the cytotoxic effects of HNE. The concentration of active HNE in purulent sputum is much lower than this concentration. However, the concentration of HNE in the microenvironment of the neutrophil close to the respiratory mucosa is likely to be higher than that in the sputum sol phase. Thus the results presented here do not completely exclude an effect of HNE on ciliary beating in vivo. In patients with chronic bronchial sepsis including CF

and bronchiectasis, P. aeruginosa frequently colonizes the respiratory tract (9). It can release at least two kinds of metallo-proteinases: elastase (13) and alkaline proteinase (14). In our in vitro experiments, these bacterial proteinase enzymes were less potent than HNE in their effecton respiratory epithelium. In PBS, neither enzyme affected CBF, although PE caused epithelial disruption. When tested in Ham's F-12 medium (containing divalent cations), the effect of PE on epithelium was increased, while PAP still failed to alter CBF or cause epithelial disruption. These in vitro findings are compatible with the reports that the elastolytic activity of PE could be enhanced by Zn2+ whereas PAP requires C02+ (26, 27). However, the relevance of these in vitro results to the in vivo environment of the bronchial tree is unknown, as they imply that extra cations are needed for optimal metallo-proteinase activity. Because of the easier availability of tissue, we have used human nasal epithelium in the majority of our experiments. However, there may be considerable longitudinal heterogenicity in the structure, function, and response to injury of the respiratory tract epithelium. We therefore compared the effect of HNE and PE on samples taken simultaneously from the nose and bronchial tree of two patients. The effects were broadly similar, suggesting that our observations using upper respiratory tract epithelium also have relevance to lower respiratory tract disease. Proteinases are cytotoxic and cause detachment of ciliated cells from their neighbors and from the basement membrane, possibly by degrading extracellular (or intercellular) matrix. This causes epithelial disruption which, if it occurs in vivo, would impair mucociliary transport independent of ciliary beating through loss of ciliated cells and disturbance of the critical relationship between mucus and cilia which is in turn dependent upon the depth of the periciliary fluid layer. Wilson and associates reported that a variety of products of P. aeruginosa, besides the proteinases, have potential to affect ciliated epithelium (10, 20, 21, 23). Further studies are necessary to determine the effect of proteinase enzymes on ciliated epithelium in the presence of other bacterial products. However, in this study we could not demonstrate any synergy between the effects of HNE and the Pseudomonas phenazine pigment pyocyanin, using concentrations of both compounds similar to those measured in vivo (4, 23). We conclude that HNE and PE damage epithelium struc-

32

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 41991

ture in vitro, and they therefore have the potential to damage human respiratory epithelium in vivo and could play important roles in the pathogenesis of airways damage during chronic bronchial sepsis. Their effectis primarily on the epithelial cells themselves and not on ciliary function or structure. Acknowledgments: We would like to thank Dr. Graham Taylor (Department of Clinical Pharmacology, Royal Postgraduate Medical School, Hammersmith Hospital, London, UK) for providing pyocyanin. Dr. Amitani was a recipient of a grant from the Kyoto University 70th Anniversary Memorial Foundation, Inc. Dr. Read was supported by the National Fund for Research into Crippling Diseases (Action Research for the Crippled Child). Drs. Ward, Burnett, and Stockley were supported by the British Lung Foundation.

References I. LaurelI, C. B., and S. Erikson. 1963. The electrophoretic alpha-I-globulin pattern of serum in alpha-l-antitrypsin deficiency. Scand. J. Clin. Lab. Invest. 15:132-140. 2. Janoff, A. 1985. Elastase and emphysema: current assessment of the protease-antiprotease hypothesis. Am. Rev. Respir. Dis. 132:417-433. 3. Currie, D. C., S. H. Saverymuttu, A. M. Peters et al. 1987. Indium-l l lIabelIed granulocyte accumulation in respiratory tract of patients with bronchiectasis. Lancet i:1335-1339. 4. Stockley, R. A., S. L. Hill, H. M. Morrison, and C. M. Starkie. 1984. Elastolytic activity of sputum and its relation to purulence and to lung function in patients with bronchiectasis. Thorax 39:408-413. 5. Tegner, H., K. Ohlsson, N. G. Toremalm, and C. von Mecklenburg. 1979. Effectof human leukocyte enzymes on tracheal mucosa and its mucociliary activity. Rhinology 17:199-206. 6. SmalIman, L. A., S. L. Hill, and R. A. Stockley. 1984. Reduction of ciliary beat frequency in vitro by sputum from patients with bronchiectasis: a serine proteinase effect. Thorax 39:663-667. 7. Sykes, D. A., R. Wilson, M. Greenstone, D. C. Currie, C. Steinfort, and P. J. Cole. 1987. Deleterious effects of purulent sputum solon human ciliary function in vitro: at least two factors identified. Thorax 42: 256-261. 8. LungarelIa, G., and L. Fonzi. 1980. Ultrastructural evidence of mucociliary function impairment induced by elastase. Bull. Eur. Physiopathol. Respir. 16(Suppl.):167-172. 9. Cole, P. J. 1984. A new look at the pathogenesis and management of persistent bronchial sepsis: a "vicious circle" hypothesis and its logical therapeutic connotations. In: Strategies for the Management of Chronic Bronchial Sepsis. R. 1. Davis, editor. The Medicine Publishing Foundation, Oxford. 1-20. 10. Cole, P. 1., and R. Wilson. 1989. Host-microbial interrelationships in re-

spiratory infection. Chest 95:217S-22IS. 11. Morihara, K., and J. Y. Homma. 1985. Pseudomonas proteases. In Bacterial Enzymes and Virulence. I. A. Holder, editor. CRC Press, Boca Raton, FL. 41-79. 12. Martodam, R. R., R. J. Baugh, D. Y. Twumasi, and I. E. Liener. 1979. A rapid procedure for the large scale purification of elastase and cathepsin G from human sputum. Prep. Biochem. 9:15-31. 13. Morihara, K., H. Tsuzuki, T. Oka, and M. Ebata. 1965. Pseudomonas aeruginosa elastase. Isolation, crystallization and preliminary characterization. J. Biol. Chern. 240:3295-3304. 14. Morihara, K. 1963. Pseudomonas aeruginosa proteinase. I. Purification and general properties. Biochim. Biophys. Acta 73:113-124. 15. Flood, M. E., R. B. Herbert, and F. G. HolIiman. 1972. Pigments ofPseudomonas species. V. Biosynthesis of pyocyanin and the pigments of Pseudomonas aeruginosa. J. Chern. Soc. (Perkin I) 4:622-626. 16. Knight, M., P. E. Hartman, Z. Hartman, and V. M. Young. 1979. A new method of preparation of pyocyanin and demonstration of an unusual bacterial sensitivity. Ann. Biochem. 95:19-23. 17. Watson, D., J. MacDermot, R. Wilson, P. J. Cole, and G. W. Taylor. 1986. Purification and structural analysis of pyocyanin and 1hydroxyphenazine. Eur. J. Biochem. 159:309-313. 18. Rutland, J., and P. Cole. 1980. Non-invasive sampling of nasal cilia for measurement of beat frequency and study of ultrastructure. Lancet ii:564-565. 19. Greenstone, M., R. Logan-Sinclair, and P. J. Cole. 1984. An automated method of recording ciliary beat frequency. IRCS Medical Science 12:715-716. 20. Wilson, R., D. Roberts, and P. Cole. 1985. Effect of bacterial products on human ciliary function in vitro. Thorax 40: 125-13 I. 21. Wilson, R., T. Pitt, G. Taylor et al. 1987. Pyocyanin and 1hydroxyphenazine produced by Pseudomonas aeruginosa inhibit the beating of human respiratory cilia in vitro. J. Clin. Invest. 79:221-229. 22. Rutland, J., A. Dewar, T. Cox, and P. Cole. 1982. Nasal brushing for the study of ciliary ultrastructure. J. Clin. Pathol. 35:357-359. 23. Wilson, R., D. A. Sykes, D. Watson, A. Rutman, G. W. Taylor, and P. J. Cole. 1988. Measurement of Pseudomonas aeruginosa phenazine pigments in sputum and assessment of their contribution to sputum sol toxicity for respiratory epithelium. Infect. Immun. 56:2515-2517. 24. Rutland, 1., A. Penketh, W. M. Griffin, M. E. Hodson, J. C. Batten, and P. J. Cole. 1983. Cystic fibrosis serum does not inhibit human ciliary beat frequency. Am. Rev. Respir. Dis. 128:1030-1034. 25. Greenstone, M., A. Rutman, A. Dewar, I. Mackay, and P. J. Cole. 1988. Primary ciliary dyskinesia: cytological and clinical features. Q. J. Med. 67:405-430. 26. Morihara, K., and H. Tsuzuki. 1975. Pseudmonas aeruginosa elastase: affinity chromatography and some properties as a metalIo-neutral proteinase. Agric. Bioi. Chem. Tokyo 39:1123-1128. 27. Morihara, K., and H. Tsuzuki. 1974. Effect of cobalt ion on the enzymatic activity of Pseudomonas aeruginosa alkaline proteinase. Agric. Bioi. Chern. Tokyo 38:621-626.

Effects of human neutrophil elastase and Pseudomonas aeruginosa proteinases on human respiratory epithelium.

It has been suggested that proteinase enzymes could play an important role in the pathogenesis of chronic bronchial infections including bronchiectasi...
779KB Sizes 0 Downloads 0 Views