Regulatory Effect of Prostaglandin E2 on Fibronectin Release from Human Alveolar Macrophages1 •2

TOSHIO OZAKI, HIROKI MORIGUCHI, YOUICHI NAKAMURA, TOSHIHIKO KAMEl, SUSUMU YASUOKA, and TAKESHI OGURA

Introduction

Alveolar macrophages (AM) are thought to playa central role in inflammatory reactions in the lower respiratory tract by releasing a variety of mediators such as fibronectin (Fn), prostaglandin (PG), and interleukin 1 (IL-l) (1-8). But the mechanisms controlling their release are unknown. Fn, which is a glycoprotein, has several biological activities related to inflammatory reactions (9). And AM obtained from patients with idiopathic pulmonary fibrosis (IPF) release large amounts of Fn (3, 4), suggesting that increased release of Fn from these cells may be closely related to the pathophysiology of IPF. Thus, information on the mechanisms regulating Fn release from AM may be important for understanding the pathophysiology of IPF. On the other hand, PGE 2 is known to be released from AM and affect the release of other biologically active materials (915), but the role of PGE 2 on regulating Fn production has not been clarified yet. In the present study, to test the possible role of PGE 2 in regulating release of Fn from AM, we examined the in vitro release of Fn and PGE 2 from AM of normal volunteers, control patients, and patients with IPF and the effect of various stimuli on the release of Fn and PGE 2 from AM. Methods Study Population Patients with IPF. This group consisted of five patients in whom diagnoses were based on clinical symptoms, roentgenograms, respiratory function tests, and histologic findings that showed interstitial fibrosis without granulomas in more than two specimens obtained by repeated transbronchiallung biopsy. None of the patients had evidence of collagenvascular diseases, hypersensitivity pneumonia, sarcoidosis, left ventricular dysfunction, or significant occupational exposures to organic or inorganic dusts. None of the patients was treated with corticosteroid. The average age of this group was 58.6 ± 5.1 yr (all data

SUMMARY Flbronectin (Fn), which is released from several kinds of cells including alveolar maerophages (AM), is important in inflammatory reactions in the certain lung diseases such as idiopathic pulmonary fibrosis (IPF). Therefore, information on the mechanisms regulating Fn release from AM may be useful for elucidating the pathogenesis of these diseases and developing therapeutic modalities. We supposed that prostaglandin E, (PGE,), which is known to modulate cellular functions, might be involved in regulation of Fn release, and, accordingly, we measured the release of Fn and PGE, from AM from normal volunteers (NY), control patients (CP), and patients with IPF. AM from patients with IPF were found to release more Fn than AM from NV (IPF: 250 ± 58.8110· cells'24 h, NV: 53.0 ± 7.3 ngl10· cells'24 h) and to release less PGE, than the latter (IPF: 0.48 ± 0.12 ng/10· cells'24 h, NV: 1.35 ± 0.24 ngl10· cells'24 h). A negative correlation was found between the contents of Fn and PGE, in the culture media of AM from NV,CPo and patients with IPF.Lipopolysaccharide, phorbol myristate acetate, and zymosan suppressed Fn release from AM but stimulated their PGE, release, and these effects were reversed by indomethacin. Exogenous PGE, (> 1 x 10-· M) suppressed Fn release. The albumin-antialbumin complex stimulated Fn release but did not affect PGE, release. These results indicate that Fn release from AM changed in response to various stimuli, and that PGE, is important in suppressing Fn release from AM, suggesting a negative feedback mechanism of PGE, in releasing Fn. AM REV RESPIR DIS 1990; 141:965-969

are presented as mean ± SEM), and four patients were men. Control patients (CP). This group consisted of seven patients with lung cancer localized in one lung as judged by roentgenographic examination. Their average age was 54.5 ± 4.0 yr, and five were men. Normal volunteers (NY). This group consisted of eight healthy volunteers who had no clinical symptoms and who had normal roentgenograms and respiratory function tests. Their average age was 23.2 ± 0.4 yr; all were men. Before the study, informed consent was obtained from all patients and volunteers after a full explanation of the procedures involved.

Bronchoalveolar Lavage BAL was performed using a flexible fiberoptic bronchoscope (Model BF-lT; Olympus Corp. of America, New Hyde Park, NY) as described previously (16). Briefly, the tip of the bronchoscope was wedged into the segmental bronchus (B4 or B5) of the right middle lobe in the NY and IPF groups, and into the segmental bronchus (B4 or B5) of the healthy lobe in the CP group. A 50-ml volume of sterilized saline was instilled through the bronchoscope, and the lavage fluid was aspirated gently by a syringe after a deep inspiration. This procedure was repeated three times, and all the lavage fluids werecombined and filtered through three layersof gauze, and

the filtrate was centrifuged at 250 x g for 10 min to collect cells. The total cell number was counted in a hemocytometer, and the differential cell count was made by microscopy after staining the cellswith May-Giemsa, Papanicolaou, and nonspecific esterase stain. Cell viability was measured by the trypan blue dye exclusion test.

Culture of AM Cells obtained by BAL were suspended in RPMI 1640medium at 1 x 106 AM/ml and plated in culture dishes (No. 3047; Falcon Plastics, Oxnard, CA). After incubation for 3 h under 50/0 CO 2 in air at 370 C, nonadherent cells were removed by washing the plates twice with phosphate-buffered saline (PBS). The adherent cells (about 98% of them were AM) were cultured for 4 or 24 h with or without stimuli. Then culture media wereharvest-

(Received in original form April27, 1989 and in revised form September 7, 1989) 1 From the Third Department of Internal Medicine, School of Medicine, Tokushima University, Tokushima City, Japan. , Correspondence and requests for reprints should be addressed to Dr. T. Ozaki, The Third Department of Internal Medicine, School of Medicine, Tokushima University, Kuramoto-cho, Tokushima City, 770 Japan.

965

966

OZAKI, MORIGUCHI, NAKAMURA, KAMEl, YASUOKA, AND OGURA TABLE 1 RECOVERY OF FLUID AND DIFFERENTIAL CELL COUNTS IN THE BAL FLUID'

Group

n

Recovery of BAL Fluid (%)

Normal volunteers Control patients Patients with IPF

8 7 5

65.6 ± 3.3 50.4 ± 4.1 54.0 ± 7.4

Total Cells (x 10'/m/)

AM (%)

Lym (%)

Neut (%)

Eos (%)

Others (%)

0.18 ± 0.02 0.31 ± 0.08 0.32 ± 0.10

92.1 ± 1.9 92.6 ± 1.3 72.2 ± 5.0

6.8 ± 1.9 5.6 ± 1.4 14.3 ± 4.9

0.7 ± 0.2 1.1 ± 0.3 8.7 ± 5.4

0.1 ± 0.2 0.5 ± 0.2 3.5 ± 1.9

0.3 ± 0.1 0.2 ± 0.1 1.6 ± 0.6

Definition of abbreviations: AM = alveolar macrophages; SAL = bronchoalveolar lavage; Eos = eosinophils; IPF = idiopathic pulmonary fibrosis; Lym = lymphocytes; Neut = neutrophils . • Values are mean ± SEM.

ed and centrifuged at 2,500 x g for 10 min, and the supernatants werestored at - 70° C. Their Fn and PGE 2 contents were measured within 4 wk.

Stimuli Solutions of lipopolysaccharide (LPS) (l ug/ ml), phorbol myristate acetate (PMA) (1 x 10-6 M), indomethacin (1 x 10-6 M), and PGE 2 (1 x 10-6 M to 1 X 10-9 M) (all from Sigma Chemical, St. Louis, MO) were prepared before use. Zymosan (Sigma, 200 ug/ ml) was opsonized with fresh serum as described by Gadek and coworkers (17). The albumin-antialbumin immune complex was prepared by incubating 1 ml of bovine serum albumin (BSA) (50 ug/ml; Miles Laboratories, Kankakee, IL) with 200 IIIof rabbit antiBSA antiserum (Organon Teknica Co., Durham, NC) diluted 1:400with saline for 30min at 37° C. Measurement of Fn Fn in the culture media of AM wasmeasured by ELISA as described by Rennard and coworkers (18). Briefly, volumesof 110 III ofthe culture mediumserially diluted with PBS contained 0.0250J0 Tween 20(PBS Tween 20)were incubated with 100 III of goat antihuman Fn antiserum (Calbiochem Co., San Diego, CA) diluted 1:1,000 with PBS Tween 20 for 30min at room temperature. The specificity of the goat antihuman Fn antiserum to human Fn was determined by the double diffusion test. After incubation, 200 III of the solution were transferred to another 96-well plate that had been coated overnight with 400 ng/ml of human Fn (Behringer Diagnostics, Somerville, NJ) at 4° C. The plate was stored for 2 h to allow free anti-Fn antibody to bind to Fn on the plate, and it was then washed three times with PBS Tween 20. It was then treated with peroxidase-conjugated rabbit antigoat IgG (Miles) diluted 1:2,000 with PBS Tween 20 for 2 h, washed, and treated with the peroxidase substrate o-phenylenediamine (Sigma) dissolved in methanol and diluted with distilled watercontainingH 20 2 • The absorbance of the product of the enzyme reaction was measured at 490 nm with a microplate photometer (Colona Electric,Tokyo, Japan). The Fn concentrationwasdetermined from a standard curve. Measurement of PGE2 PGE 2 in culture media of AM was measured

with a PGE 2 [ 125 I]RIA Kit (NewEngland Nuclear, Boston, MA). Results

Cell Number and Differential Cell Count in BAL Fluid The recoveries of BAL fluid in the NV, CP, and IPF groups and the differential cell counts in these fluids are shown in table 1. There was no significant difference in the cell numbers per unit of BAL fluid in the three groups (p > 0.1). The percentage of AM in the BAL fluid was significantly less in the IPF group than in the other two groups (p < 0.01). The percentages of lymphocytes, neutrophils, and eosinophils in the BAL fluid tended to be larger in the IPF group than in the NV and CP groups (p < 0.1).

Releases of Fn and POE2 from AM The Fn content of the culture medium of AM was significantly higher in the IPF group (4-h culture, 57.0 ± 6.3 ng/l06 cells; 24-h culture, 250.0 ± 58.8 ng/If)" cells) than in the NV group (4-h culture, 33.0 ± 5.1 ng/t06 cells; 24-h culture, 53.0 ± 7.3 ng/If)" cells), whereas the POE 2 content in the medium was lower in the

. i rD(J

c

The effects of various stimuli on the release of Fn from AM from NV, CP, and patients with IPF was shown in table 2. The release of Fn from AM of the NV, CP, and IPF groups was suppressed significantly by LPS, PMA, and zymosan. In contrast, the albumin-antialbumin immune complex stimulated the release of Fn from AM of the CP and IPF groups, but not of the NV group. In this experiment, LPS, PMA, and zymosan increased POE 2 levels (LPS, 4tO ± 340/0; PMA, 393 ± 33%; zymosan, 348 ± 30%) markedly, whereas albumin-antialbumin immune complex did not affect the POE 2 level (figure 3).

50

=u

Ii

;:! 1.5

.c

t;,

.:

~ 40

N

UJ

200

e '6

¥ o .l;

Effects of Various Stimuli on the Release of Fn and POE2

i

**

300

;:! e

ng/l0 6 cells) than in the NV group (4-h culture, 0.31 ± 0.01 ng/lf)" cells; 24-h culture, 1.35 ± 0.24 ng/t06 cells) (figure 1). The contents of Fn in the culture medium of AM from subjects in the NV, CP, and IPF groups were inversely correlated with their POE 2 contents (y = - 0.5, P < 0.05) (figure 2).

;; 2.0

400

g. -

IPF group (4-h culture, 0.07 ± 0.01

ng/Iu" cells; 24-h culture, 0.48 ± 0.12

. E

•

1.0

'i

~

..

c

100

u:

0.5

i!l

30

c

- 20

It 4 hr

24 hr

4 hr

24 hr

culture

culture

culture

culture

Fig. 1. Releases of fibronectin and prostaglandin E2 from alveolar macrophages from eight normal volunteers (open columns) and from five patients with IPF (solid columns). Alveolar macrophages (1 x 10'/ml) were cultured in RPMI 1640 medium for 4 and for 24 h. Fibronectin and prostaglandin E, in culture supernatant were measured as described in METHODS. Values are shown as mean ± SEM. Single asterisk indicates p < 0.05 versus value for macrophages from normal volunteers; double asterisks indicate p < 0.01 versus value for macrophages from normal volunteers.

oSu )10

;;:

.

o •

..

0 .....- - -.......- - - - - ' - - - - - " 1 2 3 Prostaglandin Ez (ng/10 6cena. 24hrs)

Fig. 2. Relationship between releases of fibronectin and prostaglandin E2 from alveolar macrophages from normal volunteers (closed circles), control patients (open circles), and patients with IPF (triangles). There is a significant correlation between these values (y = 0.5, P < 0.05).

967

PROSTAGLANDIN E, REGULATES FIBRONECTIN RELEASE

TABLE 2 EFFECT OF LPS, PMA, ZYMOSAN AND ALBUMIN-ANT/ALBUMIN IMMUNE COMPLEX ON RELEASE OF FIBRONECTIN FROM ALVEOLAR MACROPHAGES' Fibronectin Release (% of control)

Control LPS, 1 ~g/ml PMA, 10-' M Zymosan, 200 ~g/ml Albumin-antialbumin immune complex, 50

NV (n = 5)

CP (n = 6)

100

100

59 ±

at

70 ± 12:1: 73 ± 12:1: 105 ± 12

~g/ml

P

To elucidate the relationship between PGE z and Fn release, we examined the Fn release by stimulated AM in the presence and absence of indomethacin. As shown in figure 4, Fn release from AM of the NY and IPF groups cultured with LPS, PMA, or zymosan tended to be increased by 1 x 10-6 M indomethacin (p < 0.1). In this experiment, indomethacin reduced PGE z content in the culture media of AM from NY and patients with IPF to less than 20070 of the original level.

Effect of PGEz on Release of Fn To substantiate the regulating role of PGE z, we studied the effect of exogenous PGE z on Fn production from AM. Fn release from AM of the NY and IPF

500

..

400

~

300

c 0

37 54 51 210

se ± lot 77 ± at 154 ± llt

± 9t ± 6t ± at ± 25t

= normal volunteers; CP = con-

0.05.

Effects of Indomethacin on Release of Fn

ii::

100

56 ± 9t

Definition of abbreviations: LPS = lipopolysaccharide; PMA = phorbol myristate acetate; NV trol patients; IPF = patients with idiopathic pulmonary tibrosis. .. Data are shown as mean ± SEM. t p < O.Ol.

*