Eur Respir J 1993, 6, 27-34

Reactive pleural inflammation caused by intratracheal instillation of killed microbes

X.Y. Li*, G.M. Brown**, D. Lambt, K. Donaldson**

Reactive pleural inflammation caused by intratracheal instillation of killed microbes. X.Y. Li, G.M. Brown, D. Lamb, K. Donaldson.

*Unit of Respiratory Medicine. Dept of Medicine (RlE), The University of Edinburgh, Edinburgh, UK. **Institute of Occupational Medicine, Edinburgh, UK. ' University of Edinburgh, Dept of Pathology, Edinburgh, UK.

ABSTRACT: To investigate the pleural leucocyte response to severe alveolar inflammation, heat-killed Corynebacterium parvum were instilled intratracheally into the lungs of PVG rats and pleural lavage was performed.

Polymorphonuclear neutrophils are not normally resident in the pleural space but were found transiently after intratracheal instillation of C. parvum. Macro­phages increased gradually in the pleural space following instillation, reaching a peak at day 5. The activity of plasminogen activator inhibitor in the pleural leucocyte supernatants was increased at day 1, but returned to control levels by day 5. The activities of interleukin-1 and tumour necrosis factor secreted by pleural leucocytes were decreased compared with control pleural leucocytes at day 1 and were further reduced at day 5. The analysis of particle translo­cation showed that intratracheally instilled C. parvum and fluorescent beads with or without C. parvum did not reach the pleural space.

Correspondence: K. Donaldson rnstitute of Occupational Medicine 8 Roxburgh Place Edinburgh EH8 9SU, UK

Keywords: Inflammation leucocytes monokines pleura rat

We hypothesize tbat pleural inflammation resulting from C. parvum-induced inflammation in the lung is the result of transfer of a diffusible factor from the adjacent parenchyma.

Received: January 21 , 1992 Accepted after revision July 13, 1992

This study was supported by the British Council and the Chinese Government. Eur Respir J. , 1993, 6, 27-34.

Lung inflammation has been studied extensively, however, very few experimental studies have been car­ried out on pleural inflammation [1, 2]. The pleura is a structure adapted to form a lubricated surface that allows the lungs to expand and recoil without resist­ance, and to resist the accumulation of fluid despite negative intrapleural pressure. Pleural pathological changes, such as pleural effusions, pleural fibrosis and pleural mesothelioma may occur as a result of oral drug treatment, and also inhalation exposure to asbestos and other dusts (3, 4]. In a ral:>bit study, for instance, intravenous Ethchlorvynol caused a patchy, subpleural, haemorrhagic polymorphonuclear neutrophilic in­flammatory response [5]. A reactive pleuritis may also occur when there is adjacent pneumonia in the under­lying lung parenchyma. For example, a para-pneu­monic effusion is a common complication of bacterial pneumonia [6]. Approximately 60% of patients with pneumococcal pneumonia and 40% of all bacterial pneumonias have associated pleural effusion (7].

Interleukin-1 (IL-l) [8] and tumour necrosis factor (TNF) [9] are cytokines which have effects on recruit­ment and activation of polymorphonuclear leucocytes (PMNs), on endothelial cell activity, fibroblast prolif­eration, macrophage activation and immune response. Plasminogen activator (PA) converts the abundant extracellular zymogen plasminogen into plasmin, an active protease that promotes degradation of all com­ponents of the extracellular matrix. Plasminogen

activator inhibitor (PAl) [10] can influence cellular function and tissue remodelling by finely regulating the location and extent of PA-induced plasminogen activation. We have reported modulation of the pro­duction of these inflammatory mediators by pleural leucocytes following deposition of fibrous dusts in the bronchoalveolar space [11]. We previously described the inflammatory response of the lung parenchyma to a range of agents [12] and noted, in preliminary stud­ies, that intratracheal Corynebacterium parvum caused a transient pleuritis.

In the present study, in an attempt to study reactive pleural inflammation caused by C. parvum pneumoni­tis, we measured the translocation of particles to the pleura following alveolar deposition. We also assessed the kinetics of the production of the cytokines, TNF and IL-l by pleural leucocytes and determined sec­retion of P AI by pleural leucocytes during pleural inflammation.

Materials and Methods

Animals

Syngeneic, male and female PVG rats, aged 12 weeks or more, obtained from the Institute of Occu­pational Medicine Animal Unit, were used throughout.

28 X.Y. L1 ET AL.

intratracheal instillation of C. parvum and microspheres

Anaesthetized rats were instilled intratracheally with 1.4 mg of heat-killed C. parvum (Wellcome, Beckenham, UK) in a volume of 0.2 ml. Bronchoalveolar and pleural lavage were carried out at intervals of 1-5 days thereafter. To study the possible transfer of particles from the alveolar to the pleural space, animals were intratracheally instilled with 0.4 ml of phosphate­buffered saline (PBS) containing 1 :20 diluted mono­dispersed fluorescent carboxylated microspheres which have a geometric diameter of 0.770 f..Lm (Polysciences Inc., Warrington, UK), since they are easily seen when viewed by UV illumination. The microspheres were injected together with C. parvum when required.

Bronchoalveolar and pleuraL Lavage

Rats were sacrificed by intraperitoneal InJection of Nembutal (Ceva Ltd, Watford, UK). The rat trachea was cannulated with a blunt, 18-gauge needle. The lungs were removed from the thorax and broncho­alveolar lavage was performed by instilling 8 ml of PBS at 37·c into the lungs via the cannula. PBS was then withdrawn after gentle massage and the entire pro­cedure repeated 4 times.

Pleural lavage was carried out as follows: the rat peritoneal cavity was opened and the skin covering the right chest removed. A round-ended 8-gauge cannula was inserted through the chest, between the lower two ribs, into the pleural space, for injection of 5 ml PBS at 37•c into the space, monitored via the opened peritoneal cavity. The PBS was then withdrawn after massaging the rib cage and this procedure was repea­ted on four occasions.

The cell suspensions were spun at 1,000 rpm for 5 min and resuspended in F10 medium (Gibco, Paisley, UK) containing 2% bovine serum albumin (BSA). (Sigma, Poole, UK). Cell viability, as judged by trypan blue exclusion, was normally more than 95%. For dif­ferential counts, cell suspensions were prepared as cytoccntrifuge smears and then fixed and stained with May Grtinwald-Giemsa stain.

MesotheliaL cell characterization

To determine whether there were significant numbers of mesothelial cells present in the pleural leucocyte populations, a monoclonal antibody w rat leucocyte common antigen (CD45) (Antibody MRC OX-I, MCA 43, Serotec, Oxford, UK) was used. Pleural lavage cells obtained from control rats and rats at 16 h and 5 days after intratracheal instillation of C. parvum were stained with the antibody. The second antibody was a fluorescein-conjugated F(ab') 2 fragment of rabbit immunoglobulins raised against mouse irnmunoglobu­Hn (F313, Dako Ltd, High Wycombe, Bucks, UK). The fluorescence was determined in a Coulter Flow

Cytometer (Epics-Profile 11, CouJter Electronic Ltd, Luton, Beds, UK).

Observation of beads and C. parvum in bronchoalveo­lar and pleural lavage leucocytes

Following instillation of fluorescent beads, lOO PMNs and macrophages of cytospin preparation of broncho­alveolar and pleural leucocytes were observed on each smear by light rnicroscopy under epi-fluorescent illu­mination. The macrophages and PMNs from both lav­aged .populations were divided into two groups: with or without beads. In the group with beads, the cells were further divided into three subgroups: cells con­taining <5 beads, cells containing 6-10 beads and cells containing >10 beads. During bead counting, a dif­ferential count was also carried out.

In addition, following C. parvum exposure alone, the lavaged leucocytes were also examined for intracellular organisms. Both macrophages and PMNs were divided into one of two groups: with or without intracellular organisms.

Preparation of supernatants

Pleural leucocytes were adjusted to a concentration of Ixl06·ml 1 in FlO medium + 2% BSA and cultured for 24 h at 37"C in a 5% C0

2 atmosphere. For leuco­

cyte stimulation, lipopolysaccharide (LPS), (Sigma, Poole, UK), was added to the culture at 100 ng·mt·•. The supematants were collected and centrifuged at 3,000 rpm for 10 min to remove cell debris, then stored at -7o·c prior to assay.

Assay for TNF activity

The assay was performed according to the method of WARNER and LIBBY (13). In brief, murine L929 cells grown in minimal essential medium (MEM) (Gibco, Paisley, UK), containing 5% foetal calf serum (FCS) were trypsinized and suspended in assay medium (MEM + 5% FCS, no antibiotics), 100 f..Ll aliquots containing 3x l 05 cells were added to each well of microtitre plate and incubated at 3TC for 24 h. The supematant was then discarded, and 100 f..Ll of assay medium contain­ing 1 J..Lg·ml·1 actinomycin D (Sigma, Poole, UK), 50 f..LI of pleural leucocyte supematants and 50 fll of assay medium with 2 J..Lg·mi·• of actinomycin D was added to experimental wells; then double diluted. Follow­ing further 18-20 h incubation, the wells were emp­tied, stained with crystal violet methanol solution, and then washed with water and allowed to dry. Optical density at 540 nm was determined with a MR650 plate reader (Dynatech Laboratories Inc., USA). TNF activ­ity in the supematant was determined by comparison with a TNF-alpha standard (a gift kindly supplied by J. Symonds, Dept of Rheumatology, Northern General Hospital, Edinburgh, UK) dilution curve.

C. PARVUM-lNDUCED PLEURAL INFLAMMATION 29

Assay for IL-l activity

IL-l activity in the supernatant was quantified using the C3H/HeJ mouse thymocyte assay [14]. Briefly, thymus glands of mice killed by ether were removed and single cell suspensions of thymocytes were prepared by homogenization. The cells were washed and suspended at 6xl06·ml·1 in RPMI-1640 medium (Gibco, Paisley, UK) with 10% FCS and 20 j.tM 2-mecaptoethanol (Sigma, Poole, UK). One hundred J.!l of the cell suspension was added to each well, together with 50 1.11 of RPMJ medium contain­ing 20 j.!g·mt·• of phytohaemagglutinin (PHA) (Sigma, Poole, UK) and 50 J.!l of dilutions of test pleural leu­cocyte supematants (PLS). The cells were then incu­bated at 37°C in 5% C0

2 for 48 h and pulsed with

0.25 j.!Ci 3H-methyl-thymidine per well (Amersham International, Amersham, UK). After one more day of culture, the cells were harvested onto glass filters, and 3H incorporation was quantitated by scintillation spectrophotometry.

Assay for plasminogen activator inhibitor activity

PAl activity was assayed in 1251-fibrin coated tissue culture wells as described previously [15]. Briefly, dissolved fibrinogen (Sigma, Poole, UK) was labelled with Na1251 (Amersham International, Amersham, UK). Free 125! was removed by chromatography on a Sephadex G25 column (Pharmacia, Milton Keynes, UK). 1251-labelled fibrinogen was diluted with PBS and added to microtitre plates so that each well contained lxl05

cpm; the plates were then dried at 45·c for 3 days. The fibrinogen was converted to fibrin by treatment with FIO medium containing 2.5% freshly thawed FCS. The plates were washed twice with PBS prior to assay. To determine the PAl activity in the super­natants, the concentrations of urokinase (kindly pro­vided by I. McGregor, Blood Transfusion Centre, Edinburgh, UK) which caused 50% of total 1251-fibrin lysis when cultured with an excess of plasminogen (Kabi, Stockholm, Sweden and a gift from I. McGregor),

40

<D 30 0

x ~ 20 E ~ c::

~ 10

Bronchoalveolar leucocytes

0~--------~--------~---------

Control 3 5 Day after instillation

were used. The activity of PAl in the PLS was deter­mined by titration against urokinase and plasminogen. Solubilized 1251-fibrin degraded by urokinase in the cul­ture was measured after incubation for 24 h at 3TC, 5% C0

2• The PAl activity was expressed as percent­

age inhibition of plasminogen-dependent fibrinolysis according to the following equation:

Percentage inhibition = (Fib without PLS - Blank)-(Fib with PLS -Biank)x iOO

(Fib without PLS - Blank)

Where units of fibrinolysis = cpm; and Fib = fibrino­lysis.

Statistical analysis

All experiments were repeated at least twice and details of experiments are given in legends to figures and tables. Results of repeated experiments were sub­jected to analysis of variance. When there was a sig­nificant F value for the effect of treatment, individual means were compared for significance using t-test [16].

Results

Alveolar and pleural leucocyte populations after intra­tracheal instillation of C. parvum

Normal rat alveolar leucocyte populations contained more than 95% macrophages with a total number of less than 5xl 06• One day after intratracheal instilla­tion of C. parvum, large number of PMN were recruited to the bronchoalveolar space and this increased the total cell number to around 30xl06, although the mac­rophage number remained at the same level (fig. 1). Three days after instillation, the number of PMNs had decreased dramatically and was minimal by day 5. The number of macrophages was markedly increased by day 3 and remained elevated at day 5.

... ------o ... ,.: ____ __ o

5

Fig. I. - Bronchoalveolar and pleural leucocyte populations at various times after intratracheal instillation of C. parvum. Each point shows means of 2- 13 separate experiments. For clarity, SEM which varies from 8 to 30% of mean, is not presented in the graph. Aster­isks denote a significant difference from control: *: p<0.05; **: p<O.Ol; ***: p<O.OOI. ~= total cell number; -B-: macrophage;~: PMN;-A-: lymphocyte;-.-: mast cell;~: eosinophil.

30 X.Y. LI ET AL.

There were no pleural PMNs detectable in normal pleural leucocyte populations. One day after intra­tracheal instillation of C. parvum there was a slight increase in number of macrophages but a marked increase in number of PMNs in the pleural space (fig. I). The peak of PMNs at day 1 disappeared quickly and no PMNs were found by day 5 after C. parvum instillation. However, number of macrophages increased gradually from day 3 and this was sustained up to day 5. When the components of pleural leucocytes were compared with those of alveolar leucocytes, the pat­tern of change in the macrophage and PMN popula­tions was similar.

Analysis of beads and C. parvum organisms in alveo­lar and pleural leucocytes

When beads were instilled intratracheally alone, they produced a moderate alveolitis characterized by 24% PMNs in the bronchoalveolar lavage (table 1 ).

A high percentage (72%) of alveolar macrophages engulfed beads and most macrophages (>70%) were found to contain more than 10 beads (table 2). One day after intratracheal C. parvum plus beads, the total number of alveolar leucocytes increased up to eight times compared with bead instillation alone. A severe inflammatory response was observed with PMNs accounting for more than 80% of the total alveolar population (table 1). The distribution of beads in the macrophages changed with most macro­phages (>80%) containing less than 10 beads. The numbers of PMN with beads reduced from 33 to 4% (table 2).

In the pleural space, little effect on cell componems was seen following alveolar deposition of beads alone and beads were negligible in the pleural leucocytes. Intratracheal instillation of C. parvum and beads together caused recruitment of PMNs into the pleural space (29% of total population) and the total cell number increased nearly threefold (table 1). In spite of such severe pleural inflammation resulting from

Table 1. - Alveolar and pleural leucocyte populations after intratracheal instillation of beads <!nd/or C. parvum

Treatment Total cell no. Differential percentage

x106 Mac PMN Lymph Mast Eos

Alveolar Jcucocytcs B alone 7.36 (1.26) 75 (3.5) 24 (2.8) 1 (0.7) 0 0 B+Cp (day 1) 58.97 (2.95) 16(2.1) 84 (2.8) 0 0 0 B+Cp (day 5) 13.87 (1.76) 81 (9.5) 18 (9.5) 1 (0) 0 0 Pleural leucocytes B alone 4.10 (0.70) 76 (6.4) 0 1 (0) 6 (3.5) 7 (2.8) B+Cp (day I) 11.80 (3.38) 60 (0.7) 29 (1 1) 0 6 (5.7) 5 (5.7) B+Cp (day 5) 9.47 (0.55) 86 (0) 3 (1.5) 6 ( 1) 4 (1.7) 1 (0.6)

Results are presented as mean (so) of two separate experiments. B: beads; Cp: C . parvum; Mac: mac-rophages; PMN: polymorphonuclear leucocytcs; Lymph: lymphocytes; Mast: mast cells; Eos: eosinophils.

Table 2. - Analysis of beads in the alveolar and pleural leucocytes, following intratracheal instillation of beads and/or C. parvum

Treatment % of cells

with beads

Alveola r macropbagcs B alone 72(7.1) B+Cp (day 1) 37 (6.4) B+Cp (day 5) 15 (1.5) Alveolar PMNs B alone 33 (18.4) B+Cp (day 1) 4 (0.7) B+Cp (day 5) 0 (0) Pleural macrophages B alone <I B+Cp (day I) <I B+Cp (day 5) 0 Pleural PMNs B alone 0 B+Cp (day 1) 0 B+Cp (day 5) 0

Distributjon of beads in cells - ---1-5 5- 10 >11

12 (8.5) 17(4.2) 71 (12.7) 57 (7.1) 26 (8.5) 17 (1.4) 52 (8.7) 25 (6.2) 23 (2.6)

90 (2.8) 9 (1.4) I (1.4) 100 (0) 0 0

0 0 0

100 0 0 100 0 0

0 0 0

0 0 0 0 0 0 0 0 0

Results are presented as mean (so) of two separate experiments. For abbrevia-tions see legend to table I.

C. PARVUM-lNDUCED PLEURAL INFLAMMATION 31

C. parvum in the lung, negligible numbers of beads could be observed in the pleural leucocytes. We found that the beads seen very occasionally in cytospins of pleural lavage leucocytes were extracellular. In another study (data not shown), we demonstrated that even one month after intratracheal instillation of beads alone, or together with mineral dusts, there was no detectable translocation of beads from airspace to pleu­ral space.

Our results suggest that there is no direct commu­nication of particles between the alveolar space and the pleural space, even during acute inflammation. To confirm this hypothesis, we extended our observations to assess numbers of cells containing C. parvum organisms in both spaces. Table 3 shows that one day after intratracheal instillation of C. parvum, numbers of macrophages and PMNs containing organisms intracellularly in the alveolar space, were 26 and 34%, respectively. By day 5, the cells with organisms were negligible in the alveolar space and the percentages of macrophages and PMNs containing identifiable organ­isms had dropped to 1 and 0%. No organisms could be found in either macrophages or PMNs in the pleu­ral space at any time.

Table 3. - Analysis of C. parvum organisms in the alveolar and pleural leucocytes after intratracheal instil­lation of C. parvum

Cells

Alveolar macrophages Day I Day 5 Alveolar PMNs Day 1 Day 5 Pleural macrophages Day I Day 5 Pleur al PMNs Day 1 Day 5

% of cells with organisms

26(3.1) I (0.6)

34 (5.6) 0

0 0

0 0

Results are presented as mean (so) of 3 to 5 separate experiments. PMNs: polymorphonuclear leucocytes.

Characterization of pleural leucocytes

In all treatments the number of leucocyte common antigen positive cells was around 95% (mean (so) of two experiments: Control: 96.6 (1.2)%; 16 h after intratracheal instillation of C. parvum: 94.9 (3.5)%; and 5 days after C. parvum instillation: 95.9 (0.9)%). The remainder of the cells (around 5%) was assumed to be exfoliated mesothelial cells. Any difference between the pleural populations was very likely, therefore, to be a result of change in leucocytes and their functions and not the population of mesothelial cells.

Secretion of cytokines by pleural leucocytes in vitro: effect of intratracheal instillation of C. parvum

Normal pleural leucocytes produced high levels of IL- l and TNF in culture and this production was dra­matically enhanced whilst the leucocytes were stimu­lated with LPS (fig. 2). One day after intratracheal instillation of C. parvum, the activities of both IL-l and TNF released by pleural leucocytes decreased compared to control levels. Five days after instilla­tion, this reduction had become even more marked (fig. 3).

4 50

•• *

40 v 3 C>

-;= 30 ::::>

x E a_

-~ > u 20 <1:1

LL

0

:;::;. 2 ·:;; t5

<1:1 z t-~

10

0 Without LPS With LPS

Fig. 2. - IL-l and TNF production by normal pleural leucocytes with and without stimulation with LPS. Each bar represents the mean and SEM of 4-10 separate experiments performed in tripli­cate. The dilution of pleural leucocyte supematant is I :8 for IL­l and 1:32 for TNF. ***: p<O.OO!. .<:iP,nificantly different from control. IZ1 : interleukin- 1 (IL-l);~·: tumour necrosis fac10r (TNF); LPS: lipopolysaccharide.

3

.... C>

x 2 E a_ 0

~ ·:;; ·-g

<1:1 ~

:::':!

0 Control 5

Day after instillation

30

20 E ::, :;::;. :2: u

<1:1

10 ~ t-

Fig. 3. - IL-l and TNF production by pleural leucocytes at vari­ous times after intratracheal inst illation of C. parvum. Each bar represents the mean and SEM of 4-9 separate experiments per­fanned in triplicate. The dilution of pleural leucocyte supematant is 1:8 for IL- l and 1:32 for TNF. **: p<O.Ol, significantly different from control; ***: p<O.OOJ. ~ : intcrleukin- 1 (lL- 1); ~ : tumour necrosis factor {TNF).

32 X.Y. LT ET AL.

PAl produced by pleural leucocytes in vitro after intratracheal instillation of C. parvum

Figure 4 shows that the production of PAl by pleural leucocytes lavaged from rats one day after C. parvum instillation had increased significantly. However, five days after instillation of C. parvum, PAl activity pro­duced by pleural leucocytes had returned to normal levels.

100

en ·u; 80 >-0 c:: ... i3 60 ;;;::

0 c:: 0 40 ~ :2 .s

20 ~ 0

0 Control 5

Day after instillation

Fig. 4. - Fibrinolytic inhibition caused by pleural Jeucoeytes supernatants at various times after intratracheal instillation of C. parvum. Results are presented as mean and SBM of 3-8 sepa­rate experiments performed in triplicate. The dilution of pleural leucocyte supematant is 1:32. ***: p<O.OOl, significantly differ­ent from control.

Discussion

This study has shown that intratracheal instillation of C. parvum induces severe alveolar inflammation, characterized by transient recruitment of large numbers of PMN. The C. parvum-elicited macrophage alveo­litis was a relatively long-lived process. The same inflammatory pattern, but of a much lesser magnitude, was found in the pleural space after intratracheal instillation of C. parvum. The very similar inflam­matory patterns in both alveolar and pleural spaces indicated that there may be some communication of inflammatory messages between the lung and the pleu­ral space. To address this question, we investigated the possibility that particles might be translocated from the lung to the pleural space.

The present study on the translocation of particles (i.e. beads and C. parvum organisms) indicated that, following deposition in the lung, those particles were rapidly engulfed by alveolar macrophages and PMNs. Meanwhile, in the pleural space, no macrophages or PMNs were found to contain beads. Analysis of mac­rophages and PMNs containing C. parvum organisms also confjrmed that no microbial particles reach the pleural space. These findings are consistent with the study of LEHNERT et al. [17] which showed that, after instillation of particles with diameter of 2 jl into the lung, there was no detectable translocation of particles to the pleural spaces.

Although there is no translocation of particles to the pleural space, this region is, nevertheless, influenced by inflammation in the lung as shown here by the fact that one day after intratracheal in stillation of C. parvum, there was an increase in the number of pleural PMNs. The visceral pleura is thick and com­posed of five different layers, including mesothelial layer, submesothelial layer, external elastic layer, interstitial layer and internal elastic layer [4]. This represents a substantial barrier to passive movement of particles from alveolar to pleural spaces and could explain absence of particle transfer. Since particles do not cross the pleura, we presume that pleuritis is due to the transfer of a diffusible sigoal(s) to the pleural space from the parenchyma.

Asbestos and other fibres have been found at vis­ceral and parietal sites of the pleura, and in pleural fluid after deposition in airspaces [18, 19], but failure to find fibres has also been reported [20).

MILLER et al. l5] found that intravenous injection of Ethchlorvynol (ECV) produced a PMN-predominant exudative alveolitis and pleural effusion in rabbits, and they proposed that ECV -induced lung injury results in migration of PMN into the alveolar and interstitial space. Furthermore, damaged lung tissue was also suggested to stimulate PMN infiltration into and across the adjacent visceral pleura. FLENLEY [6] also sug­gested that pneumonia, in a subpleural location with an abundance of PMNs, causes increased pulmonary and pleural capillary permeability resulting in a leak of protein-rich fluid into the pleural space. Once the pulmonary endothelium becomes more permeable, the pleura is likely to offer less of a barrier to fluid, protein, and cellular movement, promoted by a pulmonary­pleural pressure gradient that is enhanced by the lung leak [5).

IL-l levels normally increase during inflammation [21]. Our present study indicated that IL-l activity released by pleural leucocytes one day after C. parvum instillation was decreased and the reduction was even more dramatic by day 5 compared with control lev­els. Presumably, the initial reduction in IL-l activity may be due to the accumulation of PMNs in the pleu­ral space, since IL-l inhibitor is constitutively present in normal PMNs [22). During the later stages after C. parvum instillarjon, when PMN numbers decreased, the further reduction of lL-1 activity might result from the additional secretion of IL-l inhibitor by newly recruited macrophages [23]. The same explanation may be extended to the reduction in TNF activity [24]. Our findings, which indicate the existence of IL-l and TNF inhibitor activity in the normal pleural lavage fluid (data not presented), favour this hypothesis.

RICH et al. [25] found that in vitro cultivation of blood monocytes resulted in decreased expression of TNF in response to LPS, which suggests that matura­tion of mononuclear phagocytes leads to a phenotype capable of producing higher levels of TNF. From this point of view, it is likely that the decrease in TNF production by the pleural leucocytes may partly result from the observed recruitment of blood monocytes to

C. PARVUM-lNDUCED PLEURAL INFLAMMATION 33

the pleural space. However, IL-l elaboration does not show this pattern, and in fact decreases with mono­cyte maturation [25, 26].

Decreased fibrinolytic activity has been reported in oleic acid-induced lung injury [27] and in quinacrine pleurodesis treatment of malignant pleural effusion [281. We also demonstrated an increased PAI pro­duction by rat pleural leucocytes after intratracheal instillation of crocidolite asbestos [llj. The newly recruited PMNs may be activated and release products, such as oxidants, bioactive lipids and proteases, which influence the function of macrophagcs [29] including production of PAL The similarity of changes of pleu­ral PMN numbers and PAl production during the study period favours this hypothesis.

An altered balance of coagulation and fibrinolysis, caused by increased PAl, may establish local condi­tions which could promote acute fibrin deposition in the pleura. This is supported by the findings of WroSTROM et al. [30], that fibrin deposition is associated with increased formation of thrombin in pleural fluid after intrapleural BCG. Fibrin is not a normal protein con­stituent of either the pulmonary interstitium or the alveolar space. Fibrin deposition commonly accom­panies acute lung injury as well as influencing the sub­sequent repair processes [31], including the formation of a scaffold for fibroblast proliferation leading to fibrosis. Intrapleural urokinase [32) inhibited the development of pleural fibrosis in rabbits treated with high-dose tetracycline, implying that enhanced pleural fluid clotting and inhibition of fibrinolysis are impor­tant in drug-induced pleural fibrosis. The rapid return of the P AI levels to normal in the present study sup­ports the observation that pleural fibrosis rarely results from acute, resolving pneumonia.

Our findings suggest that inflammatory pleural mac­rophages may produce cytokine inhibitors. In addition, their production of P AI, which may cause fibrin depo­sition and promote an inflammatory reaction in local connective tissue, is also reduced to normal levels as soon as PMN recruitment passes. These findings may represent an important protective role exerted by pleu­ral macrophages. The phenomenon of a mild, long­standing pleural macrophage accumulation in our study may represent a beneficial reaction to the PMN recruit­ment engendered in the pleural space by inflammation in the adjacent lung parenchyma. By recruiting mono­cytes, tissue damage may be restricted through their ability to control the production of a series of inflam­matory regulators.

In conclusion, after intratracheal instillation of C. parvum or beads, there was no penetration of parti­cles from the airspaces to the pleural space. The PMN­macrophage inflammation occurring in the pleura, characterized by egress of PMNs into the pleural cav­ity, followed by an increase in macrophage numbers, is therefore presumed to result from transfer of a signal(s) from the parenchyma. The effects on the pleural leucocytes are a generalized decrease in cyto­kine activities and temporary enhancement of PAl activity. The role that these changes play in pleural

effusion is not clear but this rat model of reactive pleural inflammation caused by intratracheal instillation of C. parvum may be useful in explaining some pleu­ral events. resulting from infection in the airspaces or deposition of agents such as asbestos.

Acknowledgements: The authors would like to thank W. MacNee for his helpful comments on this manuscript.

Reference

I. Sahn SA, Good JT. - The effect of common sclera­sing agents on the rabbit pleural space. Am Rev Respir Dis, 1981; 124: 65- 67. 2. Sahn SA, Antony VB. - Pathogenesis of pleural plaques. Relationship of early cellular response and pathol­ogy. Am Rev Respir Dis, 1984; 130: 884-887. 3. Seaton A, Seaton D, Leitch AG. - Drug-induced lung diseases. In: Crofton and Douglas's Respiratory Diseases, Seaton A, Seaton D, Leitch AG. eds Oxford, Blackwell Scientific Publications, 1989; p. 857. 4. Herbert A. Pathogenesis of pleuritis, pleural fibrosis and mesothelial proliferation. Thorax, 1986; 41: 176-189. 5. Miller KS, Harley RH, Sahn SA. - Pleural effusions associated with ethchlorvynol lung injury result from vis­ceral pleural leak. Am Res Respir Dis, 1989; 140: 764-770. 6. Flenley DC. - Diseases of the pleura. In: Flenley DC, ed., Respiratory Medicine, London, Bailliere Tindall, 1981; pp. 320-338. 7. Light RW, Girard WM, Jenkinson SG, George RB. -Parapneumonic effusions. Am J Med, 1980; 69: 985- 986. 8. Dinarello CA. - Interleukin-1 and its related cyto­kines. In: Sorg C, ed. Macrophage-derived Cell Regula­tory Factors. Cytokines. Base!, Karger, Vol. I , 1989; pp. 105-145. 9. Beutler B, Cerami A. - Tumor necrosis, cachexia, shock, and inflammation: a common mediator. Ann Rev Boichem, 1988; 57: 505-518. 10. Hert DA, Rehemtulla A. - Plasminogen activators and their inhibitors: regulators of extracellular proteolysis and cell function. Biochem Physiol, 1988; 90: 691- 708. 11. Li XY, Brown GM, Lamb D, Donaldson K. - Increased production of plasminogen activator inhibitor in vitro by pleural leukocytes from rats intratracheally instilled with crocidolite asbestos. Environ Res, !991; 55: 135- 144. 12. Donaldson K, Brown GM, Brown OM, et al. - Con­trasting bronchoalveolar leukocyte responses in rats inhal­ing coal mine dust, quartz or titanium dioxide: effects of coal rank, airborne mass concentration and cessation of exposure. Environ Res, 1990; 52: 62-76. 13. Warner SC. Libby P. - Human vascular smooth mus­cle cells, target for and source of tumor necrosis factor. J lmmunol, 1989; 142: 100-109. 14. Mizel S. - Production and quantitation of lymphocyte activating factor (interleukin 1 ). In: Herscowitz HB , Holden HT, Bellanti JA, Ghaffar A, eds. Manual of Macro­phage methodology, New York, Marcel Dekker, 1981; pp. 329-336. 15. Li XY, Brown GM, Lamb D, Donaldson K. - Secre­tion of plasminogen activator inhibitor by nonnal rat pleural leukocytes in culture. Lung, 1990; 168: 309- 322. 16. Ryan BF, Joiner BL, Ryan TA. In: Minitab hand-book. Second Edn. Boston, Duxbury Press, 1985.

34 X.Y. LJ ET AL.

17. Lehnert BE, Baldez YE, Bomalaski SH. - Anal~ses of particles in the lung free cell, tracheobronchial lymph nodal and pleural space compartments following their depo­sition in the lung as related to lung clearance mechanisms. Ann Occup Hyg, 1988; 32: 125-140. 18. Sebastien P, Janson X, Bonnaud G, er al. - Translo­cation of asbestos fibres through respiratory tract and gastrointestinal tract according to fibre type and size. /n: Lemen R, Dement JM eds. Dust and Disease. Proceed­ings of the Conference on Occupational Exposures to Fibrous and Particulate Dust and Their Extension into the Environment. Park Forest South, Illinois, Pathotox Pub­lisher lnc; 1979; pp. 65-76. 19. Viallat J, Raybuad F, Passarel M, Boutin C. - Pleural migration of chrysotile fibres after intratracheal injection in rats. Arch Environ Health, 1986; 41: 282-286. 20. Dodson RF, Ford JO. - Early response of the vis­ceral pleura following asbestos exposure: an ultrastructural study. J Toxicol Environ Health, 1985; 15: 673-686. 21. Lasalle P, Gossct P, Aerts C, et al. Abnormal secretion of interleukin-1 and tumor necrosis factor alpha by alveolar macrophages in coal worker's pneumoconiosis: comparison between simple pneumoconiosis and progressive massive fibrosis. Exp Lung Res. 1990; 16: 73-80. 22. Tiku K, Tiku ML, Liu S, Skosey JL. Normal human neutrophils are a source of a specific interleukin-1 inhibitor. J lmmunol, 1986; 136: 3686- 3692. 23. Arcnd WP, Joslin GG, Massoni RJ. Effects of immune complexes on production by human monocytes of interleukin- I or an interleukin-1 inhibitor. J lmmunol, 1985; 134: 3868- 3875. 24. Seckinger P, Isaaz S, Dayer J-M. - Purification and

biologic characterization of a specific tumor necrosis fac­tor alpha inhibitor. J Bioi Chem, 1989; 264: 11966-11973. 25. Rich EA, Panuska JR, Wallis RS, et al. - Discoor­dinate expression of tumor necrosis factor-alpha by human blood monocytes and alveolar macrophages. Am Rev Respir Dis, 1989, 139: 1010- 1016. 26. Wewers MD, Rennard SI, Hance AJ, Bitterman PB, Crystal RG. - Normal human alveolar macrophages ob­tained by bronchoalveolar lavage have a limited capacity to release interleukin-1 . J Clin Invest, 1984, 74: 2208-2218. 27. Idell S, Peterson BT, Gonzales KK, et al. - Local abnormalities of coagulation and fibrinolysis and alveolar fibrin deposition in sheep with oleic acid-induced lung injury. Am Rev Respir Dis, 1988; 138: 1282-1294. 28. Agrenius V, Chmielewska J, Widstrom 0, Blomback M. - Pleural fibrinolytic activity is decreased in inflam­mation as demonstrated in quinacrine pleurodesis treatment of malignant pleural effusion. Am Rev Respir Dis, 1989; 140: 1381-1385. 29. Sibille Y, Reynolds HY. - Macrophages and poly­morphonuclear neutrophils in lung defense and injury. Am Rev Respir Dis, 1990; 141: 471- 501. 30. Widstrom 0, Chmielewska J, Blomback M. - Meas­urements of enzymatic activities in pleural exudates with chromogenic substrates. Thromb Res, 1986; 42: 859-864. 31. Chapman NA, Stone OL. - A fibrinolytic inhibitor of human alveolar macrophages. Induction with endotoxin. Am Rev Respir Dis, 1985; 132: 569-575. 32. Good JT, Hyers TM, Maulitz RM, Kaplan RL, Sahn SA. - An hypothesis for tetracycline induced pleural sym­physis: impairment of the fibrinolytic system. Am Rev Respir Dis, 1979; I 19: 118.