de- grees in diseases such as idiopathic pulmonary fibrosis, sarcoidosis, siicosis, and asbestosis, among others. [8]. An increase in collagen metabolism directly.
Journal
of Leukocyte
Biology
42:51 -60
(1987)
Alveolar Macrophage/Peripheral Blood Monocyte-Derived Factors Modulate Proliferation of Primary Lines of Human Lung Fibroblasts Manel Department
Jordana, of Pathology, St.
Michael McMaster
Joseph’s
Hospital
T. Newhouse, University
(M.J.,
(M.T.N.),
Hamilton,
and Jack Gauldie
J.G.)
and
Department
Ontario,
of Medicine,
Canada
Pulmonary fibrosis Is characterized by an alteration in lung collagen synthesis and deposition, as well as by increased tibroblast proliferation. It is also characterized by an Intermittent influx of immune and inflammatory cells in the lung. To investigate the nature of the target cell in this dIsorder, we established a series of prImary lines of human adult lung fibroblasts and studied the effect of mediators released from activated normal human alveolar macrophages (AM) and peripheral blood monocytes (PBM) on the proliferation of both normal lung fibroblasts and fibroblasts established from lung tissue of patients with active fibrosis. Our data show that monocyte supernatants containing a 15-18 kD monokine from either AM or PBM inhibits growth of logarithmic phase proliferating lung fibroblasts in a dose-dependent manner. This effect can be entirely abrogated by treating the fibroblasts with indomethacin and is reconstituted by adding exogenous PGE2. A study of the kinetics of this Interaction shows that exposure to monocyte supernatant for 30 mm to 1 hr is sufficient to cause significant inhibition of fibroblast proliferation and that this effect can be halted, but not reversed, at any stage by incubation with indomethacin. We also show that fibroblasts derived from patients with pulmonary fibrosis are affected more quickly by exposure to the mediators, although the final extent of inhibition seen at each concentration of mediators is similar in normal and “fibrotic” fibroblasts. These studies Indicate that activated AM or PBM release cytokines (including IL-i) which Inhibit the growth of proliferating normal and fibrotic fibroblasts through activation of the intrinsic arachidonic acid pathway of this cell and also that this effect
requires
a continuous
activation
of this pathway
to be fully
expressed. Key words:
fibroblast, pulmonary
alveolar fibrosis
macrophage,
INTRODUCTION Pulmonary fibrosis is a disorder histologically characterized by an alteration in the synthesis and deposition of collagen within the lung structures [14,26,30]. It is a very common pathologic finding present to different degrees in diseases such as idiopathic pulmonary fibrosis, sarcoidosis, siicosis, and asbestosis, among others [8]. An increase in collagen metabolism directly implicates the lung fibroblast in the basic pathogenesis of this disorder since this cell is the main collagen producer [3]. The precise nature of the mechanisms regulating fibroblast behaviour are not fully understood. However, there is evidence indicating that collagen synthesis and fibroblast proliferation are independently regulated mechanisms [4] and that both play a role in the final alteration of lung collagen content [14,26]. During the last few years, evidence has emerged indicating that a number of © 1987 Alan
R. Liss,
Inc.
peripheral
blood
monocyte,
lnterleukin-1,
mediators such as interleukin-l (IL-i) released by immune and inflammatory lung cells have the ability to interact with the lung fibroblast as a target cell and thereby modulate its functional behaviour [1,12,17,35]. It is clear that the alveolar macrophage, in addition to being a major lung scavenger, has the ability to release effector molecules (cytokines), some of which interact with the lung fibroblast. In addition, peripheral blood monocytes can be activated to release the same/similar series of cytokines which also act on fibroblasts [11,25]. We have established cultures of primary human adult fibroblast cell lines derived from normal lung and from specimens obtained from patients with pulmonary fibroReceived
July
Reprint
requests:
University,
1, 1986;
Hamilton,
Jack
accepted Gauldie, Ontario,
December
I,
Department Canada
1986. of Pathology,
L8N
325.
McMaster
52
Jordana,
Newhouse,
and Gauldie
sis and have studied the effect of exposure of these cells in vitro to both normal alveolar macrophage (AM) and peripheral blood monocyte (PBM) products. In this report, we show that when these phagocytic cells are activated, they release nondialysable cytokines which inhibit the proliferation of log-phase growing fibroblasts in a dose-dependent fashion. This effect can be abrogated by treating the fibroblasts with indomethacin and is reconstituted by adding exogenous prostaglandin F, suggesting that the arachidonic acid pathway of the fibroblast plays a role in modulating the response of the fibroblast to cytokines. In addition, a study of the kinetics of interaction between fibroblasts and cytokines reveals that only a very short exposure, 30 mm to 1 hr, to the active material is sufficient to cause a substantial effect and that fibrotic fibroblasts are more quickly responsive to these cytokines compared to normal lung fibroblasts. Finally, our data show that the active molecule in the monokine preparations responsible for fibroblast inhibition co-chromatographs with interleukin1 (lymphocyte activating factor) or gel filtration and that purified human IL- 1$ causes similar effects in the lung fibroblasts.
MATERIALS Primary
AND METHODS
Lung
Fibroblast
Cell Lines
Cultures of normal lung fibroblasts were established from histologically normal areas of surgical lung specimens from patients undergoing resective surgery. Cultures of fibrotic fibroblasts were established from grossly abnormal areas of lung obtained at the time of open lung biopsy and were part of the originally excised specimen submitted for histology and diagnosed as having the histologic characteristics of pulmonary fibrosis associated to silicosis. A more detailed description of the source and characterization of these cell lines is the subject of a separate publication [19]. Briefly, these specimens were chopped in pieces < 1 mm3 and washed once with phosphate buffered saline (PBS) and twice with regular growth medium (RGM) (Dulbecco’s modified Eagle medium; Grand Island Biological Co., Grand Island, NY) containing 10% fetal bovine serum (FBS) with an antibiotic solution consisting of penicillin 100 U/mI, streptomycin 100 tg/m1, and fungizone 25 pg/ml). Three pieces were then plated per 60-mm dish (Corning Glassworks, Corning, NY) and covered with a slide stuck to the dish with sterile vaseline. Five ml of RGM were added and the tissue was incubated at 37#{176}Cand 5% CO2. The medium was changed twice weekly. When the bottom of each dish was coated with a layer of fibroblasts at near confluency, usually 5 to 6 wk later, the pieces of tissue were removed, the cells trypsinized for 10 mm, resuspended in RGM, and replated in 100-mm culture dishes with
10 ml
of RGM.
Subsequently,
the
cells
were
split
1:2 at confluency, usually weekly. Fibroblast aliquots at different passages were frozen and stored in liquid nitrogen. Using this procedure, we have been able to establish ten normal and six fibrotic fibroblast cell lines. Macrophage/Monocyte Preparation
Cell Supernatant
Bronchoalveolar lavage (BAL) of normal volunteers was performed as previously described by Weinberger, Kelma, and Elson [36]. After centrifugation of the BAL fluid, the cell pellet was resuspended in RPMI containing 10% FBS and 2 x 106 cells in 1 ml were plated in each well of a T24 NUNC plate. After 2 hr of incubation at 37#{176}Cand 5% CO2 to ensure adherence of the alveolar macrophages, the nonadherent cell population was carefully washed off. Fresh medium (1 ml) was then added and the adherent cells (95-99% alveolar macrophages) were incubated for a further 24 hr with (stimulated) or without (unstimulated) lipopolysaccharide 10 jig/mI (LPS, TCA precipitated; Sigma, St. Louis, Missouri). This level of LPS is not toxic to the cells and results in activation of macrophage/monocytes and the maximum generation of cytokines from cells such as the alveolar macrophage [24]. The supernatant was collected, filtered, and extensively dialysed against PBS using dialysis tubing with a molecular weight cutoff of 7-8,000 daltons. After dialysis, the samples were sterilized by micropore filtration (0.22 ), aliquoted, and frozen at -20#{176}Cbefore being used. No samples thawed more than twice were used in the fibroblast proliferation assay. Peripheral blood monocyte supematant was generated from normal volunteers in an identical manner after buffy-coat preparation and Ficoll-Hypaque centrifugation to prepare a mononuclear cell suspension. The cell suspension was further purified by adherence (RPMI, 10% FBS, 37#{176}C, 5% C02, 2 hr) at a cell density of 2 x lO6cells/ml in T24 NUNC plates. After removal of the nonadherent cells by washing with PBS, the adherent cells were further incubated with 1 ml of medium with or without LPS for 24 hr. The supernatant was removed and treated as for AM supernatant. Because the supematants
blast
from AM and PBM showed similar effects on fibroproliferation, experiments are described using either
or both
sources.
Fibroblast
Proliferation
Assay
To study the response of human adult lung fibroblasts to either AM or PBM supernatants as well as to exogenous PGE2 (Pharmacia, Piscataway, NJ), 10 fibroblasts, removed by trypsinization from confluent cultures, were seeded in flat-bottomed microtiter wells (96 wells/plate, NUNC, InterMed, Denmark) in 100 ILl of RGM. The cells were incubated for 24 hr at 37#{176}C and 5% CO2. The medium was then removed, and several concentrations
Cytokines of the test samples were added. We ensured that regardless of the dilution used, the final concentration of FBS was 10% in all cases. The cells were incubated for an additional 24 hr before being trypsinized and harvested. Tritiated thymidine, 1 ILCi/well (20 Ci/mmol New England Nuclear Corp., Boston, MA) was added during the last 12 hr of the incubation. To examine the kinetics of the effect of monocyte supernatants on fibroblast proliferation, i#{248} fibroblasts per well were seeded and incubated for 24 hr. The medium was then removed and monocyte supernatant was added to the cells for various periods of time ranging from 30 mm to 24 hr. After each exposure, the medium was removed and the cells washed with fresh medium. Fresh medium (RGM) was added, and the incubation continued to a total of 24 hr. In all cases, the fibroblasts were incubated for a total period of 48 hr before being trypsinized and harvested, with thymidine being added for the last 12 hr (Fig. 6). Results are expressed as mean cpm ± SD of thymidine incorporated (n = 6 to 8). In experiments to confirm thymidine incorporation as representing fibroblast proliferation, 3.5 x l0 cells were seeded in 35-mm dishes (the same density as in the thymidine assay), the supernatants added as described above, and the cultures terminated by trypsinization. Cells were then counted using trypan-blue exclusion and a hemocytometer. Levels of significance between different values of cell counts of 3H-thymidine incorporation were established by the paired Student’s t-test. Lymphocyte-Activating
Factor
(LAF)
Assay
Assay of LAF activity was carried out essentially as described by Simon and Willoughby [31]. Thymocytes from C3H/HeJ mice (1.5 x 10 cells) in 100 IL’ of medium (RPMI with 10% FBS) were cultured for 30 mm at 37#{176}C, 5% CO2 in the presence of 1 g of PHA-P (Difco Laboratories, St. Louis, MO). One hundred IL’ of medium, or AM/PBM culture supernatant, or dilutions of such were added, and the thymocytes were cultured for a further 48 hr. Thymocyte proliferation was detected after cell harvest and counting as the uptake of 3Hthymidine (20 ILCi/rnl) (New England Nuclear) added for the last 16 hr of culture. Results are expressed as mean cpm ± SEM of 3 to 5 wells.
RESULTS The techniques described for growing out fibroblasts from tissue explants result in heterogeneous polyclonal populations of diploid fibroblasts which proliferate as lines in culture in fetal calf-containing media. In the experiments described below, we have compared the characteristics of two normal and two fibrotic cell lines. The results are typical of the cells as a group, and all
Modulate
Lung
Fibroblast
Growth
53
cells were used at a passage earlier than the tenth. The human lung fibroblast lines had normal fibroblast morphology and characteristics (Fig. lA). There were no obvious differences between normal and fibrotic fibroblasts and there was no evidence of myofibroblast characteristics by electron microscopy (Fig. lB). All cell lines had a normal karyotype preparation.
Effect of Macrophage/Monocyte on Fibroblast Proliferation
Supernatants
When the fibroblasts were cultured under normal growth conditions, the rate of incorporation of 3H-thymidine over the first 60 hr of culture was directly related to cell proliferation (Fig. 2), and this time period was used to examine the effect of monocyte supernatants on fibroblast proliferation in subsequent experiments by 3Hthymidine uptake alone. As shown in Figure 3, unstimulated AM supernatant or medium alone containing lipopolysaccharide (LPS) did not significantly affect fibroblast proliferation. In contrast, LPS-stimulated AM supernatant inhibited fibroblast proliferation in a dose-dependent manner, from up to 70% inhibition at ‘A dilution to approximately 40% at 1/64 dilution. On further dilution, normal proliferation was observed. LPS-stimulated PBM supernatants caused similar inhibition, and some preparations had higher titres than AM supernatants (Fig. 4). At no time did we see any stimulation of fibroblast proliferation by AM or PBM supernatants in our assay. These experiments were performed on several normal fibroblast lines with similar results. The inhibitory effect of AM or PBM supernatants on fibroblast proliferation was not due to a cytotoxic effect as shown by trypan blue exclusion assessment of viability (> 90% at 96 hr). Studies on the Mechanism of Fibroblast Inhibition by AM or PBM Products
Growth
Figure 5 shows the effects of activated monocyte supernatant, indomethacin, and exogenous PGE2 alone and in combination on fibroblast proliferation. The inhibition caused by monocyte supernatant could be duplicated by adding exogenous PGE2 alone (500 ng/ml added twice at 12 hr intervals). Indomethacin at 1 g/m1 caused no detectable modulation of fibroblast proliferation by itself, but could reverse the inhibition caused by monocyte supernatants. However, indomethacin could not reverse the inhibition caused by the addition of exogenous PGE2, implicating the arachidonate pathway of the fibroblast as being involved in growth modulation of that cell by cytokines. Dexamethasone (1 ILM) in the culture system had a very similar effect to indomethacin (data not shown).
54
Jordana,
Newhouse,
and Gauldie
z 0
C
I.-
0
x
16
a.-. Ox
Ca
-j-J 12
Ui
400 8
Ui
Q >-
=
0 U.
0
300
4
I-
S
z
0
0
12
24
36
INCUBATION
48
60 TIME
72
84
96
(hours)
Fig. 2. Growth characteristics of normal human adult lung fibroblast cell lines. Over each 24 hr interval, cell counts (U-U) by trypan blue exclusion and 3H-thymidmne uptake (#{149}-#{149}) were determined. Data for 3H-thymidmne is expressed as mean cpm ± SD of six replicates.
FIg.
1.
A) Fibrobiast
morphology.
Normal
human
lung
fibro-
blast monolayers stained with Alcian blue showing a typical fibroblast cytoskeleton and nuclear/cytoplasm distribution. B) Electron microscopic examination of normal (N) and fibrotic (P9 fibroblasts.
Kinetics of Cytokine-Mediated Fibroblast Proliferation we
In order to examine cultured fibroblasts
Modulation
of
the rate of induction of inhibition, with monocyte supernatant for
increasing periods, as outlined in Figure 6. Significant inhibition was seen with as little as 30-mm exposure to cytokmnes. When we compared normal and fibrotic cell lines by this method, the fibrotic fibroblasts showed a greater inhibition at each time of exposure (at 30 mm, 47% vs 23% inhibition; 52% vs 25%; 4 hr. 49% vs 36%; 8 hr 64% vs 41%) up to 12 hr, but the maximum degree of inhibition over 24 hr was indistinguishable between the two cell lines. Figure 7 shows the typical results obtained for such an experimental protocol. In several experiments involving two normal and two fibrotic cell lines, similar results were obtained.
Cytokines z 0
Modulate
Lung Fibroblast
Growth
55
RGM+LPS
40
medium+LPS
C4
1-LPS.Stimulated
r__Un5t1mt_i
0 a. 0 0
1/4
0
1/64
1/16
1/4
-
1/64
1/16
z
Ui
z a
.4t)
S I CUi 0
z 4 I 0
-80
Fig. 3. Effect of alveolar macrophage supernatants (AMQ) on in vitro fibroblast proliferation. Data are expressed as percentage change from a control culture of the same fibroblasts (incubation in RGM with 10% FBS alone). Each data point represents mean ± SD of four replicates.
l0’c.II,
z
pl.d
in ROM
10%
FBS
PBM
ADDED
ADDED
THYMIDINE
HARVEST
10
0 C-
1/4
1/8
1/16
1/32
0
a. 0
of
-10
24
C.)
INCUBATION
z
w
z
f.osn
J
TIME
- 30
-20
-Cc,, -
0
4.’, RI,’
-30
)
x
)12n’
I
. 24I1I
IUi
*pBM
0
REPLACED
BY FRESH
RGM
10%
FBS
z
.(
Fig. 6. ExperImental design of study of the kinetics action between peripheral blood monocyte supematants and fibroblasts.
I C)
of Inter(PBM)
-
Fig. 4. Effect of LPS-stlmulated peripheral blood supernatants on in vitro fibroblast proliferation. Data expressed as in Figure 3. Each data point represents mean ± SD of four replicates. PBM
z 0 I..
TIME
OF
EXPOSURE
z
0 IND
PBM
IND
0 a.
z Ui
I4
0 a. 0 C) 1 Ui
z
a
z a
S
S
= Ui 0
z 4
I C,
=
-50
I-
Ui 0
z 4 =
-60 -70
0
S Fig. 5. Effect of LPS-stlmulated peripheral blood monocyte supematants (PBM), indomethacin (1 1og/ml), and prostaglandmn E2 (500 ng/ml added twice), alone and in combination on in vitro fibrobiast proliferation. Data expressed as in Figure 3. Each data point represents mean ± SD of six replicates.
-40
Fig.
-80
7.
Kinetic
study
of the effect
of peripheral
blood
mono-
cyte (PBM) supernatants on proliferation of both normal (0) and fibrotic (#{149}) fibroblasts. Data expressed as in Figure 3. Each data point represents mean ± SD of six replicates. p < 0.05, normal vs fibrotic at each time point.
56
Jordana,
Newhouse,
and Gauldie PGE2
z
PeN
4h-..Indo,,,
Ih-..PBM
DOSE
(nglml)
19h
Ui
0
Ih-.*.PBM
Is,,
z
1050
500
50
a
4
-z
0 a. 0
0
I-
z
20
zo
4.
z
03.
0
40
5.
I
xz z-
4.
I.,
z
60
4 I
0
80
Fig. 8. Effect of indomethacin (1 pg/mi) for lh added at various intervals during the incubation period on fibroblast proliferation. Data expressed as in Figure 3. Each data point represents mean ± SD of eight replicates.
TABLE 1. Effect
of Exogenous
PGE2 on Lung
Fig. 9. Effect of exogenous PGE2 on in vitro proliferation of lung fibroblasts. Results are expressed as mean ± SO of two normal (empty bars) and two fibrotic (dashed bars) cell lines. The value assigned to each cell type was the mean response (% change compared to control) to any given concentration of PGE2 on repeated testing (Table 1). Control value was obtained by incubating fibroblasts in RGM with 10% FBS alone. * p < 0.05, normal vs fibrotic.
Fibroblast
Proilferationa Co ntrol
fibroblast lines
Fibrotic
fibroblast lines
PGE2 dose (ng/ml)
NI
Fl
F2
1.000
-58
-66
+3
-38
-62
-64
-20
-37
500
-55 -52
-59
+9
-21
-57
-52
-17
-25
N2
-24
-52 50
-7
-40
-45
+12
-15
-25
-40
-5
-18
-30 5
aResults
are
expressed
-5
-20
-17
+11
-12
-9 -3
-10
-11
+6
-2 as percentage
change
compared
to control
(cells incubated in RGM with 10% FBS). Three separate experiments were carried out with Nl and Fl and two with N2 and F2. Each experiment, either control or exposure to any given concentration of POE2, was performed in 12 replicate wells, and the standard deviation of the mean was always less than 15%.
adding indomethof fibroblast proliferation, which was marginally less than that caused by a full 24 hr exposure. Indomethacin had no protective effect if added after fibroblasts had been exposed for 12 hr to
blasts acm
PBM
to PBM
resulted
to 8 hr before
products
in 52%
supernatants
inhibition
(Fig.
Effect of Exogenous Proliferation
8).
PGE2
on Fibroblast
The effect of exogenous PGE2 addition on fibroblast thymidine incorporation is shown in Table 1. Two cell lines, one normal and one fibrotic, were tested three times, and another two on two separate occasions. To allow comparisons, the results are expressed as the percentage change in thymidine incorporation compared to background exogenous
(cells PGE2
dose-dependent
grown in RGM alone). The to the fibroblast cultures inhibition
and fibrotic fibroblasts. fibrotic fibroblasts was PGE2
at each
dose
tested.
of proliferation
However, significantly The
addition induced
on both
normal
the proliferation less depressed
differential
of a
response
of by of
When we added indomethacin (1 g/ml) 1 hr either at normal and fibrotic fibroblasts to PGE2 is illustrated in Figure 9, which depicts the combined response of two the start or variously throughout the culture with monocyte supernatant, we were able to totally or partially control and two fibrotic lines. The value assigned to each reverse the inhibition. Pretreatment of fibroblasts for 1 cell type was its average response to PGE2 on repeated hr with indomethacin followed by PBM supernatants for testing (Table I). 23 hr resulted in complete abrogation of the previously of the Active Macrophage/ shown PBM-induced inhibition of fibroblast growth. Ex- Characterization Monocyte Product posure of fibroblasts for 4 hr to PBM supernatants, folIn order to characterize the monocyte product(s) active lowed by indomethacin for 1 hr and PBM supernatants for a further 19 hr, resulted in 30% inhibition of fibroon lung fibroblasts, crude dialyzed supernatants from blast growth, substantially less in comparison to the ef- both activated AM and PBM were chromatographed in PBS on a G-100 Sephadex column (3.0 x 110 cm) as fect induced by exposing fibroblasts for 24 hr to the PBM previously described [21], and the fractions were micromaterial, 64% (Fig. 8). Increasing the exposure of fibro-
Cytokines
Modulate
44kD
18000
Lung
Fibroblast
Growth
57
preparations
of
15kD
14000
10000
LAF ASSAVI 6000
2000 14
16
18
24
26
28
30
31
,
32
33
34
36
38
40
42
45
FRACTION NUMBER medIum
10000
E
a. U
6000
I FIBROBLAST
ASSAY
I
2000
Fig. 10. Effect of fractionated (Sephadex G-100 column, PBS) LPS-stimulated alveolar macrophage supematant on the fibroblast and lymphocyte activating factor (LAF) assays. Each data point represents mean of three replicates. SD was less than 15% in all cases in the fibrobiast assay and 20% in the LAF assay. The molecular weight markers used to calibrate the column were ovalbumin, 44 kD and cytochrome C, 13.5 kD.
TABLE 2. Effect of Purified Fibroblast Units
Human
IL-ill
on Normal
Inhibition
ILll3a
12.5 25 50 100 aIL..lll
53.5 74.1 82.6 82.2 activity
is expressed
as anti-viral
cultures us 10% FBS. blnhibition is calculated incorporation (n
=
Lung
Proliferation
compared
as
percent
to fibroblasts
units [34] added inhibition cultured
of in
± ± ± ±
17.6 8.9 2.9 7.6
3H-thymidine FBS
alone
6).
porefiltered and tested directly in both the fibroblast and LAF assays. Figure 10 shows that fractions 30 to 32 eluting at approximately 15 kD were active in both the fibroblast and LAF assays, and the activities appeared coincident.
The
inhibitory
effect
of these
active
subsequently IL-l with
tested other similar results.
(%)b
to fibroblast
10%
acm. We have purified human
fractions
on fibroblasts could be abrogated by incubating the fibroblasts with indomethacin (data not shown). As a direct assessment of IL-i being the active component in the supernatant, we tested the effect of highly purified human IL-113, kindly supplied by Dr. J. Van Damme [34]. Table 2 shows that pure IL- 13 causes a similar inhibition of proliferation, as does the purified factor from alveolar macrophage and peripheral blood monocytes. This inhibition could also be reversed by the addition of indometh-
DISCUSSION The group
interstitial of
disorders
lung
diseases sharing,
(ILD) in
are a heterogeneous
different
degrees,
two
main features: inflammation within the lung structures, usually referred to as alveolitis, and fibrosis, which is characterized by an alteration in collagen metabolism and deposition in the lung [8,15]. It has become clear in recent years that the alveolar macrophage, the most prevalent inflammatory cell in the lung, has not only phagocytic activity but also a relevant effector role [10,33]. Indeed, these cells can release a number of molecules which can attract other cells to the lung or interact with a variety of cells and thereby regulate their function. Given that the fibroblast is the main target cell making collagen in the lung, and given that evidence exists regarding heterogeneity in a number of functional aspects in fibroblasts derived from different tissue sites and even within the same tissue [7,20,29], we established a number of primary cultures of human adult lung fibroblasts and examined the in vitro proliferative behaviour of these cells when exposed to mediators derived from activated peripheral blood monocytes and alveolar macrophages. We recognize that these primary lines are polyclonal
58
Jordana,
outgrowths ulation of ential in fibroblast currently [19]. We first
Newhouse,
and Gauldie
and therefore represent a heterogeneous popfibroblasts. However, the lines showed differvitro behaviour as a group. Clonally derived lines from these heterogeneous populations are being derived and examined in our laboratory established
that
these
polyclonal
lines
had the
morphology and cellular characteristics typical of diploid fibroblasts (Fig. lA,B) and had a normal karyotype. When we compared the proliferation as detected by cell counts or by uptake of 3H-thymidine, we saw that for incubation periods up to 60 hr, 3H-thymidine incorporation was directly related to cell proliferation and could be used more efficiently to assess the modulating effects of exogenous mediators. Our data show that in vitro LPS-activated alveolar macrophages or peripheral blood monocytes release a nondialysable factor (Mr approx. 15 kD) which inhibits in vitro fibroblast proliferation in a dose-dependent manner. To our knowledge there are only two reports in the literature studying the interaction of human AM products and human adult lung fibroblasts. Bitterman et al [1] showed that activated AM release a molecule, alveolar macrophage derived growth factor (AMDGF, Mr 18,000) which stimulates proliferation of fibronectin-primed fibroblasts in an assay system using quiescent lung fibroblasts. Elias et al [12] have recently shown, in an assay system very similar to ours, that activated AM supernatants inhibit lung fibroblast proliferation. Our data confirm and expand the findings of Elias et al and show that the activity present in alveolar macrophage and peripheral blood monocyte supernatants is consistent with the presence of IL-i, since the two activities (fibroblast growth modulation and lymphocyte activating factor) co-elute on molecular sieve chromatography, and highly purified IL-1j3 causes similar changes in lung fibroblast behaviour (Table 2). Figure 5 shows that indomethacin, which by itself did not have a significant effect on fibroblast proliferation, entirely reversed the inhibition induced by monocyte! macrophage products. Initially, we exposed fibroblasts to indomethacin for 24 hr, concurrently with PBM supernatants, but it was equally effective to simply pre-expose the target cell for 1 hr to indomethacin before adding the PBM supernatants (data not shown). The fact that the addition of exogenous PGE2 reconstituted the growth inhibition even when indomethacin was present (Fig. 5) suggests that the response of fibroblasts to AM or PBM cytokines is mediated through the prostaglandin pathway of the fibroblast. We have not measured PGE2 in our fibroblast supernatants after AM- or PBM-mediated inhibition, but Elias and coworkers found that AM products stimulated acm, at a dose
fibroblast PGE2 release, and similar to that in our studies,
indomethmarkedly
prevented the release of PGE2 [12]. Mizel et al [27] showed that a partially purified IL-i preparation stimulated PGE2 production by rheumatoid synovial cells, and recent reports using rDNA-derived IL-i confirm that the fibroblast is a target cell for this cytokine and responds by activation of the arachidonic acid pathway and generation of PGE2 [9,28]. Our assay and that of Elias et al [121, using rapidly proliferating (10% FBS), newly established lung fibroblast cell lines, result in an inhibition of proliferation with monocyte/macrophage supernatants. Other assays using quiescent fibroblasts (0.4% FBS) and fetal lung fibroblasts result in stimulation [27,28]. This apparent anomaly has been previously noted [51 and suggests that heterogeneity of the fibroblast lines used and assay conditions influence the outcome of treating these cells with cytokine preparations. A recent review by Freundlich et al summarized the differences in outcomes of fibroblast proliferation assays [13]. Furthermore, different activities, such as alveolar macrophagederived growth factor, platelet-derived growth factor, IL1, or tumor necrosis factor, all of which are present in activated monocyte/macrophage supernatants, can be detected in different fibroblast assays. Thus, our results showing purified IL-i cause inhibition of lung fibroblast proliferation are not inconsistent with those of Schmidt et al [28], who found stimulation of proliferation using a different cell line and assay system. The fact that we demonstrate inhibition in an in vitro assay of proliferation cannot be directly related to outcomes in the lung in vivo. However, whether the cells respond to cytokines by inhibition or stimulation of proliferation, the implications remain that mediators released by the alveolar macrophage and peripheral blood monocyte can modulate the behaviour of the lung fibroblast. An examination of Figure 7 shows that a short period of exposure to monocyte-derived cytokine causes significant inhibition of fibroblast proliferation and that the cytokine must continually be present in order that the full inhibitory effect be manifest. In addition, for the two fibrotic cell lines studied, these cells appear to be more rapidly inhibited by the cytokine. However, both the normal and fibrotic lines reach the same level of inhibition after prolonged exposure to the same dose of cytokmne. This latter finding is consistent with the work of Jimenez et al [18] showing that normal and scleroderma skin fibroblast lines display similar inhibition of collagen synthesis when exposed to mononuclear cell products. The different kinetic response seen here in the fibrotic fibroblasts suggests an alteration in the pattern of activation of the intrinsic prostaglandin pathway of the fibrotic cell. This requires further investigation. The results seen with various exposures to monocyte supernatant and indomethacin (Fig. 8) indicate that the cytokines rapidly
Cytokines activate the arachidonate pathway of the fibroblast and that although the inhibition can be halted at any stage, it cannot be reversed. The implications of these results to the development of pulmonary fibrosis remain at the moment speculative. Clearly, the lung
an increase parenchyma
in the number of occurs in human
fibroblasts within [6] as well as in
experimentally induced pulmonary fibrosis [32]. This could be the result of lung effector cell cytokines continually signalling fibroblasts to proliferate. An increase in the number of fibroblasts could also result from the emergence, in sustained inflammmed tissue, of fibrobiast subpopulations with distinct proliferative features. The evidence demonstrating that apparently homogeneous mass fibroblast cultures are heterogeneous in their proliferative and synthetic behaviour as well as in their response to inflammatory mediators [2,7,16,20,22,23] is the first requirement to support this hypothesis. A number of inflammatory mediators could potentially promote the preferential expression of fibroblast populations with distinct phenotypic characteristics. The data we report here showing that the overall response of both control and fibrotic fibroblasts to dialyzed supernatants from in vitro activated monocytes is very similar, suggest that the responsiveness of fibroblasts to these supernatants is not a directly selecting variable. However, monocyte cytokines could participate in the selection of certain fibroblast populations via activation of a second modulatory signal. Ko, Page, and Narayan have shown that incubation of normal skin fibroblasts in PGE2-contamning medium for 2 weeks results in the emergence of a PGE2insensitive cell population [20]. In addition, Korn, Torres, and Dowme have shown that normal skin fibroblasts exposed to mononuclear cell supernatants for several weeks show a greater PGE2 synthetic response upon reexposure to those supernatants [22]. As pointed out [20,22], this is the opposite result of what might be expected unless those fibroblasts with greater PGE2 synthetic response were also the most resistant to PGE2mediated antiproliferative effects. This situation would be entirely consistent with the results we have reported showing that fibrotic fibroblasts, which presumably have been chronically exposed in vivo to lung effector cell mediators, are hyporesponsive to PGE2 compared to control fibroblasts. The emergence of fibroblast populations with such characteristics would further limit the proliferation of PGE2-sensitive fibroblast populations and eventually result in an alteration of the clonal composition of fibroblasts in the lung. In summary, it is very likely that mediators released by alveolar and peripheral blood macrophages modulate the behaviour of the lung fibroblast. This modulation may result in enhanced proliferation and/or collagen production or may allow the preferential emergence of fibro-
Modulate
Lung
Fibroblast
Growth
59
blast populations with phenotypic characteristics such as promotion of disease expression. These areas are being actively pursued in our laboratory in experiments using clonally-derived human adult lung fibroblasts.
ACKNOWLEDGMENTS The authors thank the editorial Butera and the technical assistance Jane-Ann Schroeder, and Margaret debted to Drs. J.E.M. Young and kind collaboration in supplying the Dr.
J. Van
Damme
kindly
assistance of Janice of Gloria Jordana, Kardish. We are inW. Bennet for their surgical specimens.
supplied
the
purified
IL-1f3,
initially described as a 22K interferon inducing factor from human monocyte [34]. This work was supported in part by the MRC of Canada, Ontario Thoracic Society, the Council for Tobacco Research Inc., USA, and the St. Joseph’s Hospital Foundation. M.J. is a Fellow of the Canadian Lung Association.
REFERENCES I. Bitterman, P.B., Rennard, SI., Hunninghake, G.W., and Crystal, R.G. Human alveolar macrophage growth factor for fibroblasts: Regulation and partial characterization. J. Clin. Invest. 70, 806, 1982. 2. Botstein, G.R., Sherer, O.K., and LeRoy, E.C. Fibroblast selection in scleroderma: An alternative model of fibrosis. Arthr. Rheum. 27, 174, 1984. 3. Bradley, K.H., Kawanami, 0., Ferrans, V.J., and Crystal, R.G. The fibroblast of human lung alveolar structures: A differentiated cell with a major role in lung structure and function. Meth. Cell. Biol. 21A, 57, 1980. 4. Breul, S.D., Bradley, K.H., Hance, A.J., Schafter, M.P., Berg, R.A., and Crystal, R.G. Control of collagen production by human diploid lung fibroblasts. J. Biol. Chem. 255, 5250, 1979. 5. Clark, J., Kostal, K.M., and Marino, B.A. Bleomycin induced pulmonary fibrosis in hamsters: An alveolar macrophage product that increases fibroblast prostaglandin E2 and cyclic adenosine monophosphate and suppresses fibroblast proliferation and collagen production. J. Clin. Invest. 72, 2082, 1983. 6. Coalson, J.J. The ultrastructure of human fibrosing alveolitis. Virchows
Archiv
(Pathol
Anat)
395,
181,
1982.
7. Connor, N.S., Aubin, i.E., and Sodek, J. Independent expression of type I collage and fibronectin by normal fibroblast-like cells. J. Cell Sci. 63, 233, 1983. 8. Crystal. R.G., Bitterman, P.B.. Rennard. SI.. Hance. A.J.. and Keogh, B.A. Interstitial lung diseases of unknown cause (first of two parts). N. EngI. J. Med. 310, 154, 1984. 9. Dinarello, C.A. An update on human interleukin-l: From molecular biology to clinical relevance. J. Clin. Immunol. 5, 287, 1985. 10.
DuBois,
R.M.
The
alveolar
macrophage.
Thorax
40,
32,
1985.
11. Elias, J.A., Rossman, M.D., and Daniele, R.P. Inhibition of human lung fibroblast growth by mononuclear cells. Am. Rev. Respir. Dis. 125, 701, 1982. 12. Elias, iA., Rossman, M.D., Zurier, RB., and Daniele, R.P. Human alveolar macrophage inhibition of lung fibroblast growth:
60
Jordana,
Newhouse,
A prostaglandin-dependent 94, l3.
and GauIdie
process.
Am.
Rev.
Respir.
Dis.
131,
25.
1985.
Freundlich,
B.,
Regulation
Bomalski,
of
iS.,
fibroblast
Neilson,
proliferation
E., and
and
Jimenez,
collagen
cytokines. Immunol. Today 14. Fulmer, iD., Bienkowski, Bradley,
K.M.,
Collagen
7, 303, 1986. R.S., Cowan, M.J., Breul, Vi., Roberts, W.C., and Crystal,
Ferrans,
concentration
monary
fibrosis.
and
Am.
rates
Rev.
of
Respir.
synthesis
Dis.
S.A.
synthesis
by 26.
S.D., R.G.
in idiopathic
122, 289,
pul-
21.
Ko,
S.D.,
Page,
R.C.,
neity
and
prostaglandin
Acad
Sci
USA
Koj, G.D. tem
A.,
Narayan,
regulations
74, 3429,
Gauldie,
The
and
J.,
acute-phase
Fibroblast
of subpopulations.
Regoeczi,
Proc
E.,
Sauder,
and the effect
D.N.,
and
of human
cytokines.
23.
the
Arth. 24.
fibroblast
Rheum.
response
to
mononuclear
cell
derived
Gauldie,
S.M.,
synovial
32.
in cultured
Cell Res. 108, I, 1977. Seyer, J.M., Hutcheson, morphism in idiopathic vest. 57, 1498, 1976. Simon,
P.L.,
and
and
cell
of two
produced by 1534, 1981. Snider, G.L., tial
El., chronic
Mergenhagen,
collagenase
distinct
Hayes,
diploid
and Kang, pulmonary W.F.
immune
rabbbit
pulmonary
human
Willoughby,
in pulmonary
terization
S.E.
and
prosta-
activating 78, 2474,
J.A.,
Korthy,
produced
Tissue Exp.
A.H. Collagen polyfibrosis. J. Clin. Inrole
of
subcellular
Physicochemical
species of lymphocyte alveolar macrophages.
fibrosis
S.S.
fibroblasts.
The
function:
I. Interleui. Immu-
AL., in
charac-
activating
factors
J. Immunol.
et al. Chronic
hamsters
by
26,
intersti-
endotracheal
bleomycin. Am. Rev. Respir. Dis. 117, 1099, 1978. Unanue, E.R. Symbiotic relationships between macrophages and lymphocytes. In Macrophages and Natural Killer Cells (Normann, Si., and Sorkin, E., Eds.) New York: Plenum, p. 49, 1981.
34.
35.
Van Damme, J., De Ley, M., Opdenakker, DeSomer, P. Homogeneous interferon-inducing lated to endogenous pyrogen and interleukin1985. WahI,
S.M.,
diated
activation
tion. 36.
J., Stadnyk,
Krane,
by partially purified lymphocyte Proc. NatI. Acad. Sci. USA
differences
factors
mediators.
27, 174, 1984. L.R.,
31.
Sys-
A.W., Richards, C.D., and Jenkins, E. In vivo initiation of unstimulated in vitro interleukin-I release by alveolar macrophages. Am. Rev. Respir. Dis. Lamontagne,
30.
Biochem.
1984. Korn, J.H., Torres, D., and Downie, E. Fibroblast prostaglandin E2 synthesis. Persistence of an abnormal phenotype after short term exposure to monoculear cell products. J. Clin. Invest. 71, 1240, 1983. Korn, J.H., Torres, D., and Downie, E. Clonal heterogeneity in
J.M.,
rheumatoid
Schmidt, J.A., Mizel, SB., Cohen, D., and Green, kin-i, a potential regulator of fibroblast proliferation. nol. 128. 2177, 1982. Schneider, EL., Mitsui, Y., Au, KS., and Shorr, specific
NatI
J. 224, 505,
22.
29.
Sweeney,
rat hepatocytes.
Dayer, of
1981.
28.
33. of cultured
S.B.,
glandin production factor (interleukin-l).
heteroge-
1977.
response
characterization
A.S.
Mizel, Stimulation
15. Fulmer, J.D., Roberts, W.C., Von Gal, E.R., and Crystal, R.G. Morphologic-physiologic correlations of the severity of fibrosis and degree of cellularity in idiopathic pulmonary fibrosis. J. Clin. Invest. 63, 665, 1979. 16. Harper, R.A., and Grove, G. Human skin fibroblasts derived from papillary and reticular dermis: Differences in growth potential in vitro. Science 204, 525, 1979. 17. Jimenez, S.A., McArthur, W., and Rosenbloom, J. Inhibition of collagen synthesis by mononuclear cell supernatants. J. Exp. Med. 150, 1421, 1979. 18. Jimenez, S.A., McArthur, W.M., Bashey, R.I., and Rosenbloom, J. Selective inhibition of excessive scleroderma fibroblast collagen production by lymphokines from normal human mononuclear cells. Arth. Rheum. 28, 502, 1985. 19. Jordana, M., Schulman, J., McSharry, C., Jordana, G., and Gauldie, J. Proliferative characteristics of human adult lung fibroblast lines and clonally derived fibroblasts from control and fibrotic tissue (submitted for publication). 20.
27.
1980.
131, 326, 1985. Leslie, C.C., Musson, R.A., and Henson, P.M. Production of growth factor activity for fibroblasts by human monocyte-derived macrophages. J. Leuk. Biol. 36, 143, 1984. Madri, J.A., and Furthmayr, H. Collagen polymorphism in the lung: An immunochemical study of pulmonary fibrosis. Human Pathol. II, 353. 1980.
459,
lavage 1978.
L.M.,
of fibroblast
J. Immunol.
Weinberger,
veolar
Wahl,
121, 942,
SE.,
Kelman,
in interstitial
and
McCarthy, proliferation
G., Billian, 2K
I. Nature J.B. and
A.,
factor
and
is re-
314, 266,
Lymphocyte collagen
meproduc-
1978. iA.,
and
lung disease.
Elson,
Ann.
N.A.
Intern.
Bronchoal-
Med.
89,