Cytokine signaling in lung: transforming growth factor-p secretion by lung fibroblasts JASON KELLEY, JAMES P. FABISIAK, KAREN HAWES, AND MARLENE ABSHER Department of Medicine, University of Vermont College of Medicine, Burlington, Vermont 05405; Department of Medicine, Loyola University Medical Center, Maywood 60153; Medicine Service, Hines Veterans Administration Hospital, Hines, Illinois 60141; and Departments of Medicine and Pharmacology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 KELLEY,JASON,JAMES P. FABISIAK, KAREN HAWES, AND MARLENE ABSHER. Cytokine signaling in lung: transforming growth factor-p secretion by lung fibroblasts. Am. J. Physiol. 260 (Lung Cell. Mol. Physiol. 4): L123-L128, 1991.-Control of growth and phenotypic expression of interstitial fibroblasts is a critical determinant of lung architecture and physiology during processes of growth and remodeling. We examined the ability of lung fibroblasts to produce transforming growth factor-@ (TGF-P), a cytokine that is known to modulate proliferation and phenotypic expression of mesenchymal cells. Cultures of fibroblasts isolated from rat lungs spontaneously secrete TGF-6 as measured in the standard bioassay of anchorageindependent growth of normal rat kidney (NRK) cells in soft agar. Rat lung fibroblasts secrete TGF-P in an inactive precursor form. Fibroblasts cultured from adult and fetal rat lungs produced comparable amounts of TGF-P. The ability of lung fibroblast supernatant fluids to induce colony formation in soft agar could be completely neutralized by preincubation of samples with anti-TGF-P immunoglobulin (Ig). Anti-platelet-derived growth factor IgG had no effect on anchorage-independent growth of NRK cells driven by rat fibroblast culture supernatant samples. These results indicate that TGF-P does not require the presence of and interaction with secondary cytokines for its activity. In contrast to the results obtained with rat cells, neither human fetal nor adult lung fibroblasts secreted detectable amount of active TGF-P or its inactive precursor. This was not due to the presence of TGF-P inhibitors in fibroblast culture media, because the addition of purified porcine TGF-P to conditioned medium from human lung fibroblast cultures yielded the expected increase in NRK cell growth in soft agar. These results point to differing cytokine control patterns in the lungs of the two species. cell proliferation; rat
lung injury;
platelet-derived
growth
factor;
RECOGNITION of cytokines (also termed peptide growth factors or biological response modulators) as important mediators of cellular phenotype represents a major advance in the scientific basis of cellular physiology. Most initial examples of cytokine action on cellular phenotype involved cytokines released by immune effector cells such as alveolar macrophages. However, there is growing recognition that structural cells found in parenchyma are capable of producing cytokines that can locally modulate the behavior of proximate cells. This observation implies that structural cells cannot be viewed simply as passive targets modulated by cytokines reTHE
1040-0605/91 $1.50 Copyright
leased by immmune effector cells (15). A number of cytokines have now been shown to be produced in abundance by fibroblasts and other mesenchymal cells. These include among others platelet-derived growth factor (PDGF, 24), fibroblast growth factor (FGF, 8), interleukin (IL) 6(4), IL-8 (30), insulin-like growth factor (IGF) I, and IGF-II (1). Of the hematological growth factors, fibroblasts produce granulocyte-macrophage colonystimulating factor and granulocyte colony-stimulating factor (13) but not IL-3 (21). We have recently observed that rat lung fibroblasts secrete a competence-type growth factor that resembles PDGF and express the gene for PDGF-A as determined by Northern blot analysis (Fabisiak, Absher, and Kelley; unpublished data). We report here that fibroblasts cultured from the lungs of fetal and adult rats are also capable of producing and secreting transforming growth factor-p (TGF-P). TGF-P is a multipotent cytokine capable of modulating bidirectionally the proliferative capacity and phenotypic expression of a variety of cells (28). The observation that structural cells obtained from the pulmonary interstitium are capable of secreting TGF-P, PDGF, and other cytokines locally has potentially important implications for the pathogenesis of interstitial lung diseases. METHODS
Materials. PDGF-AB purified by high-pressure liquid chromatography from human platelets, anti-PDGF polyclonal antibody, IGF-I, and epidermal growth factor (EGF) came from Collaborative Research (Lexington, MA). We obtained TGF-P1 extracted from porcine platelets and polyclonal anti-TGF-P antibody from R&D Systems (Minneapolis, MN). Cell cultures. Cell cultures obtained from American Type Culture Collection (Rockville, MD) included normal rat kidney (NRK, 49F) cells, BALB/c fetal mouse embryo 3T3 (A3l) cells, and rat fetal lung (RFL-6) fibroblasts isolated from a germ-free 18-day gestation Sprague-Dawley rat. IMR-90 human fetal lung fibroblasts came from the Coriell Institute for Medical Research (Camden, NJ) as did human adult lung fibroblasts (AG02603). We isolated fibroblasts from adult Fischer 344 rat lungs. Tissue was obtained in a sterile fashion, and fibroblasts were isolated by enzymatic dissociation of
0 1991 the American
Physiological
Society
1,123
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
LlZ4
TRANSFORMING
GROWTH
FACTOR-8
IN
LIJNG
lung tissue using 0.1% (wt/vol) trypsin and 0.1% colla- isolated from rat expresses both the genes for the two genase. Cultures below population doubling level 2 were peptides of PDGF (6). We were therefore concerned that frozen and stored in liquid nitrogen. These cells were the presence of PDGF in conditioned media from fibroshown to have a characteristic fibroblast morphology, blasts might be detected in the assay for anchoragegrowth cycle, and aging profile, and stained positively independent growth, yielding a false-positive signal for for actin and vimentin but not keratin, and were able to TGF-P. Therefore, we examined how much authentic produce both types I and III collagen. The four fibroblast PDGF could be detected in the anchorage-independent cultures tested are referred to as HAL (human adult agar assay for TGF-P (Fig. 1). We found that a level of lung, AG02603), HFL (human fetal lung, IMR-90), RAL PDGF (3 U/ml) that yielded a near-maximal (>90%) (rat adult lung, RL-87), RFL (rat fetal lung, RFL-6). All proliferative response in the BALB/c 3T3 (A3l) cell cell cultures employed were shown to be free of myco- bioassay induced ~15% as many NRK(49F) colonies in plasma contamination using a DNA fluorochrome test soft agar as produced by maximal doses of authentic kit (Bionique Laboratories, Saranac Lake, NY). TGF-P (3 rig/ml). Hence a maximal dose of PDGF mimTo harvest conditioned medium from cultures of lung icked ~80 pg/ml of TGF-P. fibroblasts, we seeded 7.5 x 10’ human lung fibroblasts PDGF assay. We used the standard PDGF bioassay (population doubling level 25) or 3 x 10’ rat lung fibrodescribed by Pledger and colleagues (23) with minor blasts (population doubling level 7) in each of two T-75 modifications. This assay requires quiescent densityflasks and allowed them to grow to confluence in a inhibited BALB/c 3T3 (A31) cells to undergo DNA synminimum essential medium (MEM) containing 10% fetal thesis and cell division in response to added PDGF. bovine serum, penicillin (100 U/ml), and streptomycin Samples were assayed in quadruplicate in eight-well slide (100 pg/ml). After 6 days of culture in an incubator at chambers (Scientific Products, McGaw Park, IL). Con37°C under an atmosphere of 95% air-5% carbon dioxide ditions of assay are described elsewhere (23). For imthe cell layer was washed with phosphate-buffered saline munohistochemical quantitation of mitotic activity, we and fresh medium composed of MEM containing 1% added 5-bromo-2’-deoxyuridine (BrdU) and 5-fluoro-2’platelet-poor plasma was added. Platelet-poor plasma deoxyuridine (FrdU), (10 and 1 PM, respectively) to be was used because it contains low background levels of incorporated into DNA of dividing cells (9) using a PDGF (7). After 48 h of further culture, the conditioned commercially available kit (Amersham, Arlington medium was harvested, filtered (pore size 45 pm), ali- Heights, IL). Nuclei staining positively for BrdU/FrdU quoted, and stored at -20°C until used. Cell counts at DNA complex were counted using the image analyzer. the end of the experiment varied between 5.0 x 10’ and Approximately 800 cells were counted in each well. 9.1 X 10” cells per flask. Data analysis. Conditions for both bioassays were deTGF-P bioassay. We used a modification of the stand- termined to be rate limiting for the measured cytokine. ard soft agar assay as standardized by Rizzino (26) with This was effected by carrying out titration curves for NRK(49F) as target cells. The assay was carried out in each basal component (secondary cytokines and serum) 24-well tissue culture plates (Falcon 3047; Becton Dickbefore carrying out critical experiments. Results are inson, Lincoln Park, NJ) containing a lower layer of 0.5 based on at least three experiments replicating the obml of 0.5% noble agar (Difco, Detroit, MI), an upper servations. Antibodies used in these studies were shown layer of 0.5 ml of 0.3% noble agar, and 0.5 ml of culture to be active at the concentrations employed in inhibiting medium containing the sample or standard to be assayed. purified cytokines at concentrations fivefold above their Medium consisted of Dulbecco’s modified Eagle’s me- 50% inhibitory concentration. dium (DMEM)/F12 (1:l vol/vol) to which we added 50 Na,HCO, 2.4 g/l, 5% calf serum (previously heat-treated r at 56°C for 1 h), antibiotics as listed above, NazSeOz325 ,o TGF-8 nM, and 0.5 rig/ml EGF. 7 0 Samples of cell culture supernatants were assayed with or without prior acidification to activate TGF-P. Acid 1 I treatment was carried out by the addition of 4 M HCl to I E 10 a final concentration of 125 mM at room temperature 0. ,A PDGF / for 2 h, followed by extensive dialysis against 4 mM HCl, 0-0 b=&---= IGF-I z 0 centrifugation, and lyophilization of the supernatant. Samples were then solubilized in 2-3 ml DMEM/F12, 5 0 1 2 3 4 filtered (0.22 pm) to sterilize, frozen, and aliquoted for Log [Cytokine] (pg/mL) later assay. FIG. 1. Effect of purified cytokines on anchorage-independent We performed the soft agar assay for TGF-P otherwise curves are shown for induction of as described by Roberts et al. (27). Colony formation in growth. Concentration-response colony formation by normal rat kidney (NRK,49F) cells in soft agar. 0.3%-0.5% soft agar was assayed after 12-14 days of Response to transforming growth factor-@ (TGF-(-), circles) is compagrowth. The number and size of colonies formed in agar rable to response expected based on previously published values (halfmaximal stimulation at 200 pg/ml). Purified human platelet-derived were determined using a Quantex QX-7 image analyzer growth factor (PDGF) alone (triangles) induces only 14% as much system (Sunnyvale, CA) connected to a Zeiss Axiovert colony formation as maximal response to TGF-@. This amount of 35 microscope at x10 magnification. The image analyzer PDGF is adequate to cause maximal stimulation of BALB/c 3T3 cells was gated to count only colonies larger than 200 pm’. in classic competence factor assay (23). Insulin-like growth factor (IGFWe have previously shown that whole lung tissue I, squares) has no effect when tested alone in this assay.
l0
t
I
L
I
1
I
1
,
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
TRANSFORMING
GROWTH
Results of bioassays were transformed from colony number or cell counts to units of cytokine based on a standard curve generated in the same experiment with varying doses of authentic purified cytokine. All dilutions of reagents given as percentages indicate volume-volume measure unless otherwise noted. Unknown samples were tested at concentrations that were on the midportion of the dose-response curve. Bioassay results of unknown samples were not considered significant unless there was a semilogarithmic relationship between the amount tested and the response over a 5 to IO-fold range of concentrations. RESULTS
TGF-P induced anchorage-independent growth of lung fibroblasts in agar, as demonstrated by others. As a negative control, conditioned medium from BALB/c 3T3 cells contained no such activity. We found that conditioned media from rat lung fibroblasts contain titratable transforming activity as measured in the standard bioassay of colony formation by NRK(49F) cells in soft agar (Fig. 2). Conditioned media of fibroblasts cultured from adult and fetal rat lungs contained comparable amounts of TGF-P (Table 1). The requirement for treatment with acid to activate this activity strongly suggested that TGF-P was the responsible cytokine. More specific confirmation of the 30
c
0
RFL
-
-g b n .-ii t n
s
2om
RAL
10, -
0
c>
HFL HAL
....#
t
I I
1
0
Volume
1
I
100
200
a
I
300
of Sample
400
1. TGF-P secretion by lung fibroblasts Fibroblast Source
Rat fetal lung Rat adult lung Human fetal lung Human adult lung
TGF-p ng. ml-‘.
IN
LUNG
L125
identity of this activity as TGF-0 was obtained using an antibody inhibition protocol. The ability of conditioned media from cultures of rat lung fibroblasts to induce NRK colony formation in soft agar could be completely neutralized by preincubation of the samples for 1 h with an immunoglobulin (Ig) G antibody to TGF-P at 50 ,ug/ ml (Fig. 3A ). By comparison, incubation of conditioned medium samples with an anti-PDGF IgG had no effect on their ability to induce the anchorage-independent growth of NRK(49F) cells, indicating that the presence of this cytokine was not responsible for NRK cell growth in the agar assay. In contrast to the pattern of cytokine release by lung fibroblasts from rats, conditioned media from cultures of fibroblasts derived from human lungs contained no detectable amounts of TGF activity (Fig. 2, Table 1). This was true for fibroblasts cultured from both fetal and adult human lungs. Acid treatment of inactive samples of culture medium from human lung fibroblasts failed to activate any latent TGF-P activity (Fig. 3B). In contrast, acid-treatment of rat samples elicited a 40- to 5O-fold enhancement of TGF-P activity. This observation rules out the possibility that human cells release TGF-P in a latent form. Free TGF-P can be inactivated by binding to proteins found in serum, most notably az-macroglobulin (20). As an explanation for the absence of TGF-P activity in medium samples conditioned by human lung fibroblasts, we questioned whether these cells might be producing and secreting a specific inhibitor of TGF-P into their medium. To test for the presence of such an inhibitor in the conditioned medium samples, we added purified active porcine platelet TGF-P to samples of conditioned media from human fibroblast cultures. The amount of anchorage-independent growth in agar in response to exogenously added TGF-P was close to the expected value regardless of whether fibroblast-conditioned media harvested from cultures of human lung fibroblasts was present (Fig. 4). This finding speaks against the presence of an inhibitor to TGF-P in human fibroblast media as the cause of their apparent lack of TGF-P activity.
(FL)
2. Concentration-response curves for induction of anchorageindependent growth of NRK(49F) cells in soft agar by samples of conditioned media from cultures of lung fibroblasts. Rat fetal (RFL) and adult (RAL) lung fibroblast cultures produce TGF-@. Neither fetal (HFL) nor adult (HAL) human lung fibroblasts produce titratable TGF-6. Sham-conditioned medium composed of MEM-1% plateletpoor plasma incubated in absence of cells yielded no detectable colony formation. All samples of conditioned media were treated with acid and then neutralized before assay (see METHODS). FIG.
TABLE
FACTOR-$
24 h-’
13o-e24 109*21 co. 1 CO.1
Values are means + SD. Data are derived response curves of 4 conditioned media samples lated from 4 sources (Fig. 2). TGF-fi, transforming
ng. 10’ cells-‘.
24 h-*
18.4t3.6 16.3t3.2 20 genes related to connective tissue metabolism. If the results we report in cultured lung fibroblasts apply to intact tissue, then interstitial lung fibroblasts may be directing local cellular phenotypic changes within the interstitium. Lawrence and colleagues (17) found TGF-P to be secreted by embryonic fibroblasts from chick, mouse, and human tissues in a latent form. Danielpour and colleagues (3) have reported that the human lung fibroblast line WI38 secretes some detectable TGF-P that is composed almost entirely of the TGF -& isoform . However, neither of these reports indicated the actual amount of TGF-P produced. It also deserves pointing out that we have recently cultured fibroblasts from several adult human biopsy and autopsy specimens using a protocol identical to the one used to isolate rat lung fibroblasts. In preliminary studies, these cultures of human lung fibroblasts do not release detectable amounts of TGF-P, in accord with the other lines tested. This rules against the possibility that observed differences between human
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
TRANSFORMING
GROWTH
and rat lines are an artifact of the isolation method. An important corollary drawn from our observations on the differences between various fibroblast lines with respect to spontaneous release of TGF-P is that control of cell proliferation and extracellular matrix protein production may differ between rat and human. This would suggest, for example, that studies of cytokine action in animal models of pulmonary fibrosis and remodeling may not reflect the situation found in the spontaneous human disorder. Although there have been few previous demonstrations of the production of TGF-P by lung cells, Hoyt and Lazo (12) have shown that mRNA levels for TGF-/? rise during lung injury induced by constant infusion of bleomycin in the mouse. Yamauchi et al. (31) found constitutive levels of TGF-P in human lung epithelial lining fluid. If our observation that unstimulated human fibroblasts fail to produce TGF-p in vitro can be applied to similar cells in the intact lung, it suggests that cells other than lung fibroblasts are responsible for the levels of TGF-P detected in lavage fluid by Yamauchi and colleagues (31). In this regard, in other studies we have found that rat alveolar macrophages, like murine wound macrophages (25), express the gene for TGF-@ (unpublished observation), and also secrete TGF-/3 as do monocytes (11). Although TGF-P is a potent inhibitor of proliferation of a number of cell types, it is also capable of acting as a mitogen for certain cells. Leof and colleagues (18) have described a novel mechanism by which it can serve as a mitogen. The ability of certain cells to undergo mitosis in response to TGF-0 is a consequence of their ability to produce and respond to PDGF after exposure to TGF-P. Under circumstances in which TGF-P induces cell replication, as for instance in NRK(49F) cells, it has been shown to work by inducing autocrine release of PDGF. NRK(49F) cells are also capable of responding to the PDGF by virtue of having PDGF receptors on their surfaces (18). Because TGF-0 acts as a mitogen only indirectly, sufficient time must elapse for PDGF synthesis and secretion by the target cells. Other growth factors such as EGF, FGF, or PDGF do not appear to act via such an autocrine response to endogenously produced PDGF (29). TGF-p treatment of human foreskin fibroblasts induces the expression of the gene for the PDGF-A chain and not the B chain. At the same time, TGF-P modulates the responsiveness of certain cells to the different forms of PDGF (10). After exposure of Swiss mouse 3T3 cells to TGF-/3, binding of PDGF-AA and -AB is completely lost, whereas the binding of PDGF-BB is reduced by only 40%. The loss of binding sites for PDGF-AA is accompanied by a decreased mitogenic response of these cells to PDGF-AA but not to PDGF-AB or PDGF-BB. Other cytokines such as IL-l (24) and even PDGF-B (22) also appear to act as mitogens via mechanisms involving PDGF-A gene modulation.
FACTOR+ Address 317, IJniv. Received
for reprint of Vermont 27 ,January
L127
LIJNG requests: J. Kelley, College of Medicine, 1990; accepted
in final
Pulmonary Burlington, form
6 July
Unit, Given VT 05405.
C-
1990.
REFERENCES 1. ADAMS,
O., S. P. NISSLEY, S. HANDWERGER, ANI) M. M. Developmental patterns of insulin-like growth factor-I and -11 synthesis and regulation in rat fibroblasts. Nature Land. 302: 150-152, 1983. DANIELPOUR, D., L. L. DART, K. C. FLANDERS, A. B. ROBERTS, AND M. B. SPORN. Immunodetection and quantitation of the two forms of transforming growth factor-beta (TGF-beta 1 and TGFbeta 2) secreted by cells in culture. J. Cell. Physiol. 138: 79-96, 1989. ELIAS, J. A., G. TRINCHIERI, J. M. BECK, P. L. SIMON, P. B. SEHGAI,, ANII J. A. KERN. A synergistic interaction of IL-6 and IL-l mediates the thymocyte-stimulating activity produced by recombinant IL-l -stimulated fibroblasts. J. Immunol. 142: 509-514, 1989. FABISIAK, J. F., M. P. ABSHER, AND J. KEI~LEY. Production of platelet-derived growth factor (PDGF)-like cytokines by rat lung fibroblasts in vitro. (Abstract) Am. Rev. Respir. Dis. 141: A915, 1990. FABISIAK, J. P., J. N. EVANS, AND ,J. KELLEY. Increased expression of PDGF-@ (c-sis) mRNA in rat lung precedes DNA synthesis and tissue repair during chronic hyperoxia. Am. J. Respir. Cell. Mol. Biol. 1: 181-189, 1989. GLENN, K. C., AND R. ROSS. Human monocyte-derived growt,h factor(s) for mesenchymal cells: activation of secretion by endotoxin and concanavalin A. Cell 25: 603-615, 1981. GOSPODAROWICZ, D., G. NEUFELD, AND L. SCHWEIGERER. Fibroblast growth factor. Mol. Cell Endocrinol. 46: 187-204, 1986. GRATZNER, H. G. Monoclonal antibody to 5-bromoand Ij-iododeoxyuridine: a new reagent for detection of DNA replication. Science Wash. DC 218: 474-475, 1982. GRONWALD, R. G. K., R. SEIFERT, AND D. F. BOWEN-POPE. Differential regulation of expression of two platelet-derived growth factor receptor subunits by transforming growth factor-p. J. Hiol. Chem. 264: 8120-8125, 1989. GROTENDORST, G. R., G. SMALE, AND D. PENCEV. Production of transforming growth factor beta by human peripheral blood monocytes and neutrophils. J. Cell. Physiol. 140: 396-402, 1989. HOYT, D. G., AND J. S. LAZO. Alterations in pulmonary mRNA encoding procollagens, fibronectin and transforming growth factor@ precede bleomycin-induced pulmonary fibrosis in mice. J. Pharmacol. Exp. Ther. 246: 765-771, 1988. KAUSHANSKY, K., N. LIN, AND J. W. ANDERSON. Interleukin 1 stimulates fibroblasts to synthesize granulocyte-macrophage and granulocyte colony-stimulating factors. J. Clin. Invest. 81: 92-97, 1988. KEHRL, J. H., L. M. WAKEFIELII, A. B. ROBERTS, S. JAKOWI,EW, M. ALVAREZ-M• N, R. DERYNCK, M. B. SPORN, AND A. S. FA~JCI. Production of transforming growth factor /3 by human T lymphocytes and its potential role in the regulation of T cell growth. J. Exp. Med. 163: 1037-1050, 1986. KELLEY, J. State of the art: cytokines of the lung. Am. Rev. Respir. Dis. 141: 765-788, 1990. KONDAIAH, P., E. VAN OBBERCHEN-SCHIUINC., R. L. LUDWIG;, R. DHAR, M. B. SPORN, AND A. B. ROBERTS. cDNA cloning of porcine transforming growth factor-@, mRNAs. Evidence for alternate splicing and polyadenylation. J. Hiol. Chem. 263: 18313-18:317, 1988. LAWRENCE, D. A., R. PIRCHER, C. KRYCI~VE-MARTINERIE, AND P. JULLIEN. Normal embryo fibroblasts release transforming growth factors in a latent form. J. Cell. Physiol. 121: 184-188, 1984. LEOF, E. B., J. A. PROPER, A. S. GOIJSTIN, G. D. SHIPLEY, P. E. DICORLETO, AND H. L. MOSES. Induction of c-sis mRNA and activity similar to platelet-derived growth factor by transforming growth factor /3: a proposed model for indirect mitogenesis involving autocrine activity. Proc. Natl. Acad. Sci. IJSA 83: 2453-2457, 1986. LYONS, R. M., J. KESKI-OJA, AND H. L. MOSES. Proteolytic activation of latent transforming growth factor-p from fibroblastconditioned medium. J. Cell. Biol. 106: 1659-1665, 1988. MCCAFFREY, T. A., D. J. FALCONE, C. F. BRAYTON, L. A. AGARWAL, F. G. P. WELT, AND B. B. WEKSI,ER. Transforming growth S.
RECHLER.
3.
4.
5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15. 16.
17.
18.
19. This work was supported by Pulmonary Specialized Center of Research Grant (HL-14212) awarded to the University of Vermont College of Medicine, and Veterans Administration Research Funds. J. P. Fabisiak is a Parker B. Francis Fellow in Pulmonary Research.
IN
20.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
L128
21.
22.
23.
24. ‘>r LitI.
26.
TRANSFORMING
GROWTH
factor-@ activity is potentiated by heparin via dissociation of the transf’orming growth factor-e/ tu?-macroglobulin inactive complex. J. Cell. Riol. 109: 441-448, 1989. NIEMEYER, C. M., C. A. SIEFF, B. MATHEY-PREVOT, J. Z. WIMPERIS, B. E. BIERER, S. C. CLARK, AND D. G. NATHAN. Expression of human interleukin-3 (multi-CSF) is restricted to human lymphocytes and T-cell lines. Blood 73: 945-951, 1989. PAUI,SSON, Y., A. HAMMACHER, C-H. HELDIN, AND B. WESTERMARK. Possible positive autocrine feedback in the prereplicative phase of human fibroblasts. Nature Lond. 328: 715-717, 1987. PI,EIXER, W. J., C. D. STILES, H. N. ANTONIADES, AND D. &HER. An ordered sequence of events is required before BALB/c-3T3 cells become committed to DNA synthesis. Proc. Natl. Acad. Sci. USA 74: 4481-4490, 1977. RAINES, E. W., S. K. DOWER, AND R. ROSS. Interleukin-1 mitogenie activity for fibroblasts and smooth muscle cells is due to PDGF-AA. Science Wash. DC 243: 393-396, 1989. RAPPOI,EE, D. A., D. MARK, M. ?J. BANDA, AND Z. WERB. Wound macrophages express TGF-cu and other growth factors in vivo: analysis by mRNA phenotyping. Science Wash. IX 241: 708-712, 1988. RIZZINO, A. Soft agar growth assays for transforming growth
FACTOR-P factors 1987.
IN and
LUNG
mitogenic
peptides.
Methods
Enzymol.
146: 341-:352,
27. ROBERTS, A. B., M. A. ANZANO, L. C. LAMB, J. M. SMITH, AND M. B. SPORN. New class of transforming growth factors potentiated by epidermal growth fact,or: isolation from non-neoplastic tissues. Proc. Natl. Acad. Sci. USA 78: 5339-5343, 1981. A. B., AND M. B. SPORN. The transforming growth 28. ROBERTS, factor-&. In: Handbook of Experimental Pharmacology. Peptide Growth Factors and Their Receptors, edited by M. B. Sporn and A. B. Roberts. Heidelberg: Springer-Verlag, 1990, vol. 95, p. 419-473. 29. SOMA, Y., AND G. R. GROTENDORST. TGF-IJ stimulates primary human skin fibroblast DNA synthesis via an autocrine production of PDGF-related peptides. J. Cell. Physiol. 140: 246-253, 1989. 30. STRIETER, R. M., S. H. PHAN, H. J. SHOWELL, D. G. REMICK, J. P. LYNCH, M. GENORD, R. M. MARKS, AND S. L. KUNKEL. Monokine-induced neutrophil chemotactic factor gene expression in human fibroblasts. J. Riol. Chem. 264: 10621-10626, 1989. 31. YAMAUCHI, K., Y. MARTINET, P. BASSET, G. A. FELLS, AND R. G. CRYSTAL. High levels of transforming growth factor-p are present in the epithelial lining fluid of the normal lower respiratory tract. Am. Rev. Respir. Dis. 137: 1360-1363, 1988.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
lung injury;
platelet-derived
growth
factor;
RECOGNITION of cytokines (also termed peptide growth factors or biological response modulators) as important mediators of cellular phenotype represents a major advance in the scientific basis of cellular physiology. Most initial examples of cytokine action on cellular phenotype involved cytokines released by immune effector cells such as alveolar macrophages. However, there is growing recognition that structural cells found in parenchyma are capable of producing cytokines that can locally modulate the behavior of proximate cells. This observation implies that structural cells cannot be viewed simply as passive targets modulated by cytokines reTHE
1040-0605/91 $1.50 Copyright
leased by immmune effector cells (15). A number of cytokines have now been shown to be produced in abundance by fibroblasts and other mesenchymal cells. These include among others platelet-derived growth factor (PDGF, 24), fibroblast growth factor (FGF, 8), interleukin (IL) 6(4), IL-8 (30), insulin-like growth factor (IGF) I, and IGF-II (1). Of the hematological growth factors, fibroblasts produce granulocyte-macrophage colonystimulating factor and granulocyte colony-stimulating factor (13) but not IL-3 (21). We have recently observed that rat lung fibroblasts secrete a competence-type growth factor that resembles PDGF and express the gene for PDGF-A as determined by Northern blot analysis (Fabisiak, Absher, and Kelley; unpublished data). We report here that fibroblasts cultured from the lungs of fetal and adult rats are also capable of producing and secreting transforming growth factor-p (TGF-P). TGF-P is a multipotent cytokine capable of modulating bidirectionally the proliferative capacity and phenotypic expression of a variety of cells (28). The observation that structural cells obtained from the pulmonary interstitium are capable of secreting TGF-P, PDGF, and other cytokines locally has potentially important implications for the pathogenesis of interstitial lung diseases. METHODS
Materials. PDGF-AB purified by high-pressure liquid chromatography from human platelets, anti-PDGF polyclonal antibody, IGF-I, and epidermal growth factor (EGF) came from Collaborative Research (Lexington, MA). We obtained TGF-P1 extracted from porcine platelets and polyclonal anti-TGF-P antibody from R&D Systems (Minneapolis, MN). Cell cultures. Cell cultures obtained from American Type Culture Collection (Rockville, MD) included normal rat kidney (NRK, 49F) cells, BALB/c fetal mouse embryo 3T3 (A3l) cells, and rat fetal lung (RFL-6) fibroblasts isolated from a germ-free 18-day gestation Sprague-Dawley rat. IMR-90 human fetal lung fibroblasts came from the Coriell Institute for Medical Research (Camden, NJ) as did human adult lung fibroblasts (AG02603). We isolated fibroblasts from adult Fischer 344 rat lungs. Tissue was obtained in a sterile fashion, and fibroblasts were isolated by enzymatic dissociation of
0 1991 the American
Physiological
Society
1,123
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
LlZ4
TRANSFORMING
GROWTH
FACTOR-8
IN
LIJNG
lung tissue using 0.1% (wt/vol) trypsin and 0.1% colla- isolated from rat expresses both the genes for the two genase. Cultures below population doubling level 2 were peptides of PDGF (6). We were therefore concerned that frozen and stored in liquid nitrogen. These cells were the presence of PDGF in conditioned media from fibroshown to have a characteristic fibroblast morphology, blasts might be detected in the assay for anchoragegrowth cycle, and aging profile, and stained positively independent growth, yielding a false-positive signal for for actin and vimentin but not keratin, and were able to TGF-P. Therefore, we examined how much authentic produce both types I and III collagen. The four fibroblast PDGF could be detected in the anchorage-independent cultures tested are referred to as HAL (human adult agar assay for TGF-P (Fig. 1). We found that a level of lung, AG02603), HFL (human fetal lung, IMR-90), RAL PDGF (3 U/ml) that yielded a near-maximal (>90%) (rat adult lung, RL-87), RFL (rat fetal lung, RFL-6). All proliferative response in the BALB/c 3T3 (A3l) cell cell cultures employed were shown to be free of myco- bioassay induced ~15% as many NRK(49F) colonies in plasma contamination using a DNA fluorochrome test soft agar as produced by maximal doses of authentic kit (Bionique Laboratories, Saranac Lake, NY). TGF-P (3 rig/ml). Hence a maximal dose of PDGF mimTo harvest conditioned medium from cultures of lung icked ~80 pg/ml of TGF-P. fibroblasts, we seeded 7.5 x 10’ human lung fibroblasts PDGF assay. We used the standard PDGF bioassay (population doubling level 25) or 3 x 10’ rat lung fibrodescribed by Pledger and colleagues (23) with minor blasts (population doubling level 7) in each of two T-75 modifications. This assay requires quiescent densityflasks and allowed them to grow to confluence in a inhibited BALB/c 3T3 (A31) cells to undergo DNA synminimum essential medium (MEM) containing 10% fetal thesis and cell division in response to added PDGF. bovine serum, penicillin (100 U/ml), and streptomycin Samples were assayed in quadruplicate in eight-well slide (100 pg/ml). After 6 days of culture in an incubator at chambers (Scientific Products, McGaw Park, IL). Con37°C under an atmosphere of 95% air-5% carbon dioxide ditions of assay are described elsewhere (23). For imthe cell layer was washed with phosphate-buffered saline munohistochemical quantitation of mitotic activity, we and fresh medium composed of MEM containing 1% added 5-bromo-2’-deoxyuridine (BrdU) and 5-fluoro-2’platelet-poor plasma was added. Platelet-poor plasma deoxyuridine (FrdU), (10 and 1 PM, respectively) to be was used because it contains low background levels of incorporated into DNA of dividing cells (9) using a PDGF (7). After 48 h of further culture, the conditioned commercially available kit (Amersham, Arlington medium was harvested, filtered (pore size 45 pm), ali- Heights, IL). Nuclei staining positively for BrdU/FrdU quoted, and stored at -20°C until used. Cell counts at DNA complex were counted using the image analyzer. the end of the experiment varied between 5.0 x 10’ and Approximately 800 cells were counted in each well. 9.1 X 10” cells per flask. Data analysis. Conditions for both bioassays were deTGF-P bioassay. We used a modification of the stand- termined to be rate limiting for the measured cytokine. ard soft agar assay as standardized by Rizzino (26) with This was effected by carrying out titration curves for NRK(49F) as target cells. The assay was carried out in each basal component (secondary cytokines and serum) 24-well tissue culture plates (Falcon 3047; Becton Dickbefore carrying out critical experiments. Results are inson, Lincoln Park, NJ) containing a lower layer of 0.5 based on at least three experiments replicating the obml of 0.5% noble agar (Difco, Detroit, MI), an upper servations. Antibodies used in these studies were shown layer of 0.5 ml of 0.3% noble agar, and 0.5 ml of culture to be active at the concentrations employed in inhibiting medium containing the sample or standard to be assayed. purified cytokines at concentrations fivefold above their Medium consisted of Dulbecco’s modified Eagle’s me- 50% inhibitory concentration. dium (DMEM)/F12 (1:l vol/vol) to which we added 50 Na,HCO, 2.4 g/l, 5% calf serum (previously heat-treated r at 56°C for 1 h), antibiotics as listed above, NazSeOz325 ,o TGF-8 nM, and 0.5 rig/ml EGF. 7 0 Samples of cell culture supernatants were assayed with or without prior acidification to activate TGF-P. Acid 1 I treatment was carried out by the addition of 4 M HCl to I E 10 a final concentration of 125 mM at room temperature 0. ,A PDGF / for 2 h, followed by extensive dialysis against 4 mM HCl, 0-0 b=&---= IGF-I z 0 centrifugation, and lyophilization of the supernatant. Samples were then solubilized in 2-3 ml DMEM/F12, 5 0 1 2 3 4 filtered (0.22 pm) to sterilize, frozen, and aliquoted for Log [Cytokine] (pg/mL) later assay. FIG. 1. Effect of purified cytokines on anchorage-independent We performed the soft agar assay for TGF-P otherwise curves are shown for induction of as described by Roberts et al. (27). Colony formation in growth. Concentration-response colony formation by normal rat kidney (NRK,49F) cells in soft agar. 0.3%-0.5% soft agar was assayed after 12-14 days of Response to transforming growth factor-@ (TGF-(-), circles) is compagrowth. The number and size of colonies formed in agar rable to response expected based on previously published values (halfmaximal stimulation at 200 pg/ml). Purified human platelet-derived were determined using a Quantex QX-7 image analyzer growth factor (PDGF) alone (triangles) induces only 14% as much system (Sunnyvale, CA) connected to a Zeiss Axiovert colony formation as maximal response to TGF-@. This amount of 35 microscope at x10 magnification. The image analyzer PDGF is adequate to cause maximal stimulation of BALB/c 3T3 cells was gated to count only colonies larger than 200 pm’. in classic competence factor assay (23). Insulin-like growth factor (IGFWe have previously shown that whole lung tissue I, squares) has no effect when tested alone in this assay.
l0
t
I
L
I
1
I
1
,
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
TRANSFORMING
GROWTH
Results of bioassays were transformed from colony number or cell counts to units of cytokine based on a standard curve generated in the same experiment with varying doses of authentic purified cytokine. All dilutions of reagents given as percentages indicate volume-volume measure unless otherwise noted. Unknown samples were tested at concentrations that were on the midportion of the dose-response curve. Bioassay results of unknown samples were not considered significant unless there was a semilogarithmic relationship between the amount tested and the response over a 5 to IO-fold range of concentrations. RESULTS
TGF-P induced anchorage-independent growth of lung fibroblasts in agar, as demonstrated by others. As a negative control, conditioned medium from BALB/c 3T3 cells contained no such activity. We found that conditioned media from rat lung fibroblasts contain titratable transforming activity as measured in the standard bioassay of colony formation by NRK(49F) cells in soft agar (Fig. 2). Conditioned media of fibroblasts cultured from adult and fetal rat lungs contained comparable amounts of TGF-P (Table 1). The requirement for treatment with acid to activate this activity strongly suggested that TGF-P was the responsible cytokine. More specific confirmation of the 30
c
0
RFL
-
-g b n .-ii t n
s
2om
RAL
10, -
0
c>
HFL HAL
....#
t
I I
1
0
Volume
1
I
100
200
a
I
300
of Sample
400
1. TGF-P secretion by lung fibroblasts Fibroblast Source
Rat fetal lung Rat adult lung Human fetal lung Human adult lung
TGF-p ng. ml-‘.
IN
LUNG
L125
identity of this activity as TGF-0 was obtained using an antibody inhibition protocol. The ability of conditioned media from cultures of rat lung fibroblasts to induce NRK colony formation in soft agar could be completely neutralized by preincubation of the samples for 1 h with an immunoglobulin (Ig) G antibody to TGF-P at 50 ,ug/ ml (Fig. 3A ). By comparison, incubation of conditioned medium samples with an anti-PDGF IgG had no effect on their ability to induce the anchorage-independent growth of NRK(49F) cells, indicating that the presence of this cytokine was not responsible for NRK cell growth in the agar assay. In contrast to the pattern of cytokine release by lung fibroblasts from rats, conditioned media from cultures of fibroblasts derived from human lungs contained no detectable amounts of TGF activity (Fig. 2, Table 1). This was true for fibroblasts cultured from both fetal and adult human lungs. Acid treatment of inactive samples of culture medium from human lung fibroblasts failed to activate any latent TGF-P activity (Fig. 3B). In contrast, acid-treatment of rat samples elicited a 40- to 5O-fold enhancement of TGF-P activity. This observation rules out the possibility that human cells release TGF-P in a latent form. Free TGF-P can be inactivated by binding to proteins found in serum, most notably az-macroglobulin (20). As an explanation for the absence of TGF-P activity in medium samples conditioned by human lung fibroblasts, we questioned whether these cells might be producing and secreting a specific inhibitor of TGF-P into their medium. To test for the presence of such an inhibitor in the conditioned medium samples, we added purified active porcine platelet TGF-P to samples of conditioned media from human fibroblast cultures. The amount of anchorage-independent growth in agar in response to exogenously added TGF-P was close to the expected value regardless of whether fibroblast-conditioned media harvested from cultures of human lung fibroblasts was present (Fig. 4). This finding speaks against the presence of an inhibitor to TGF-P in human fibroblast media as the cause of their apparent lack of TGF-P activity.
(FL)
2. Concentration-response curves for induction of anchorageindependent growth of NRK(49F) cells in soft agar by samples of conditioned media from cultures of lung fibroblasts. Rat fetal (RFL) and adult (RAL) lung fibroblast cultures produce TGF-@. Neither fetal (HFL) nor adult (HAL) human lung fibroblasts produce titratable TGF-6. Sham-conditioned medium composed of MEM-1% plateletpoor plasma incubated in absence of cells yielded no detectable colony formation. All samples of conditioned media were treated with acid and then neutralized before assay (see METHODS). FIG.
TABLE
FACTOR-$
24 h-’
13o-e24 109*21 co. 1 CO.1
Values are means + SD. Data are derived response curves of 4 conditioned media samples lated from 4 sources (Fig. 2). TGF-fi, transforming
ng. 10’ cells-‘.
24 h-*
18.4t3.6 16.3t3.2 20 genes related to connective tissue metabolism. If the results we report in cultured lung fibroblasts apply to intact tissue, then interstitial lung fibroblasts may be directing local cellular phenotypic changes within the interstitium. Lawrence and colleagues (17) found TGF-P to be secreted by embryonic fibroblasts from chick, mouse, and human tissues in a latent form. Danielpour and colleagues (3) have reported that the human lung fibroblast line WI38 secretes some detectable TGF-P that is composed almost entirely of the TGF -& isoform . However, neither of these reports indicated the actual amount of TGF-P produced. It also deserves pointing out that we have recently cultured fibroblasts from several adult human biopsy and autopsy specimens using a protocol identical to the one used to isolate rat lung fibroblasts. In preliminary studies, these cultures of human lung fibroblasts do not release detectable amounts of TGF-P, in accord with the other lines tested. This rules against the possibility that observed differences between human
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
TRANSFORMING
GROWTH
and rat lines are an artifact of the isolation method. An important corollary drawn from our observations on the differences between various fibroblast lines with respect to spontaneous release of TGF-P is that control of cell proliferation and extracellular matrix protein production may differ between rat and human. This would suggest, for example, that studies of cytokine action in animal models of pulmonary fibrosis and remodeling may not reflect the situation found in the spontaneous human disorder. Although there have been few previous demonstrations of the production of TGF-P by lung cells, Hoyt and Lazo (12) have shown that mRNA levels for TGF-/? rise during lung injury induced by constant infusion of bleomycin in the mouse. Yamauchi et al. (31) found constitutive levels of TGF-P in human lung epithelial lining fluid. If our observation that unstimulated human fibroblasts fail to produce TGF-p in vitro can be applied to similar cells in the intact lung, it suggests that cells other than lung fibroblasts are responsible for the levels of TGF-P detected in lavage fluid by Yamauchi and colleagues (31). In this regard, in other studies we have found that rat alveolar macrophages, like murine wound macrophages (25), express the gene for TGF-@ (unpublished observation), and also secrete TGF-/3 as do monocytes (11). Although TGF-P is a potent inhibitor of proliferation of a number of cell types, it is also capable of acting as a mitogen for certain cells. Leof and colleagues (18) have described a novel mechanism by which it can serve as a mitogen. The ability of certain cells to undergo mitosis in response to TGF-0 is a consequence of their ability to produce and respond to PDGF after exposure to TGF-P. Under circumstances in which TGF-P induces cell replication, as for instance in NRK(49F) cells, it has been shown to work by inducing autocrine release of PDGF. NRK(49F) cells are also capable of responding to the PDGF by virtue of having PDGF receptors on their surfaces (18). Because TGF-0 acts as a mitogen only indirectly, sufficient time must elapse for PDGF synthesis and secretion by the target cells. Other growth factors such as EGF, FGF, or PDGF do not appear to act via such an autocrine response to endogenously produced PDGF (29). TGF-p treatment of human foreskin fibroblasts induces the expression of the gene for the PDGF-A chain and not the B chain. At the same time, TGF-P modulates the responsiveness of certain cells to the different forms of PDGF (10). After exposure of Swiss mouse 3T3 cells to TGF-/3, binding of PDGF-AA and -AB is completely lost, whereas the binding of PDGF-BB is reduced by only 40%. The loss of binding sites for PDGF-AA is accompanied by a decreased mitogenic response of these cells to PDGF-AA but not to PDGF-AB or PDGF-BB. Other cytokines such as IL-l (24) and even PDGF-B (22) also appear to act as mitogens via mechanisms involving PDGF-A gene modulation.
FACTOR+ Address 317, IJniv. Received
for reprint of Vermont 27 ,January
L127
LIJNG requests: J. Kelley, College of Medicine, 1990; accepted
in final
Pulmonary Burlington, form
6 July
Unit, Given VT 05405.
C-
1990.
REFERENCES 1. ADAMS,
O., S. P. NISSLEY, S. HANDWERGER, ANI) M. M. Developmental patterns of insulin-like growth factor-I and -11 synthesis and regulation in rat fibroblasts. Nature Land. 302: 150-152, 1983. DANIELPOUR, D., L. L. DART, K. C. FLANDERS, A. B. ROBERTS, AND M. B. SPORN. Immunodetection and quantitation of the two forms of transforming growth factor-beta (TGF-beta 1 and TGFbeta 2) secreted by cells in culture. J. Cell. Physiol. 138: 79-96, 1989. ELIAS, J. A., G. TRINCHIERI, J. M. BECK, P. L. SIMON, P. B. SEHGAI,, ANII J. A. KERN. A synergistic interaction of IL-6 and IL-l mediates the thymocyte-stimulating activity produced by recombinant IL-l -stimulated fibroblasts. J. Immunol. 142: 509-514, 1989. FABISIAK, J. F., M. P. ABSHER, AND J. KEI~LEY. Production of platelet-derived growth factor (PDGF)-like cytokines by rat lung fibroblasts in vitro. (Abstract) Am. Rev. Respir. Dis. 141: A915, 1990. FABISIAK, J. P., J. N. EVANS, AND ,J. KELLEY. Increased expression of PDGF-@ (c-sis) mRNA in rat lung precedes DNA synthesis and tissue repair during chronic hyperoxia. Am. J. Respir. Cell. Mol. Biol. 1: 181-189, 1989. GLENN, K. C., AND R. ROSS. Human monocyte-derived growt,h factor(s) for mesenchymal cells: activation of secretion by endotoxin and concanavalin A. Cell 25: 603-615, 1981. GOSPODAROWICZ, D., G. NEUFELD, AND L. SCHWEIGERER. Fibroblast growth factor. Mol. Cell Endocrinol. 46: 187-204, 1986. GRATZNER, H. G. Monoclonal antibody to 5-bromoand Ij-iododeoxyuridine: a new reagent for detection of DNA replication. Science Wash. DC 218: 474-475, 1982. GRONWALD, R. G. K., R. SEIFERT, AND D. F. BOWEN-POPE. Differential regulation of expression of two platelet-derived growth factor receptor subunits by transforming growth factor-p. J. Hiol. Chem. 264: 8120-8125, 1989. GROTENDORST, G. R., G. SMALE, AND D. PENCEV. Production of transforming growth factor beta by human peripheral blood monocytes and neutrophils. J. Cell. Physiol. 140: 396-402, 1989. HOYT, D. G., AND J. S. LAZO. Alterations in pulmonary mRNA encoding procollagens, fibronectin and transforming growth factor@ precede bleomycin-induced pulmonary fibrosis in mice. J. Pharmacol. Exp. Ther. 246: 765-771, 1988. KAUSHANSKY, K., N. LIN, AND J. W. ANDERSON. Interleukin 1 stimulates fibroblasts to synthesize granulocyte-macrophage and granulocyte colony-stimulating factors. J. Clin. Invest. 81: 92-97, 1988. KEHRL, J. H., L. M. WAKEFIELII, A. B. ROBERTS, S. JAKOWI,EW, M. ALVAREZ-M• N, R. DERYNCK, M. B. SPORN, AND A. S. FA~JCI. Production of transforming growth factor /3 by human T lymphocytes and its potential role in the regulation of T cell growth. J. Exp. Med. 163: 1037-1050, 1986. KELLEY, J. State of the art: cytokines of the lung. Am. Rev. Respir. Dis. 141: 765-788, 1990. KONDAIAH, P., E. VAN OBBERCHEN-SCHIUINC., R. L. LUDWIG;, R. DHAR, M. B. SPORN, AND A. B. ROBERTS. cDNA cloning of porcine transforming growth factor-@, mRNAs. Evidence for alternate splicing and polyadenylation. J. Hiol. Chem. 263: 18313-18:317, 1988. LAWRENCE, D. A., R. PIRCHER, C. KRYCI~VE-MARTINERIE, AND P. JULLIEN. Normal embryo fibroblasts release transforming growth factors in a latent form. J. Cell. Physiol. 121: 184-188, 1984. LEOF, E. B., J. A. PROPER, A. S. GOIJSTIN, G. D. SHIPLEY, P. E. DICORLETO, AND H. L. MOSES. Induction of c-sis mRNA and activity similar to platelet-derived growth factor by transforming growth factor /3: a proposed model for indirect mitogenesis involving autocrine activity. Proc. Natl. Acad. Sci. IJSA 83: 2453-2457, 1986. LYONS, R. M., J. KESKI-OJA, AND H. L. MOSES. Proteolytic activation of latent transforming growth factor-p from fibroblastconditioned medium. J. Cell. Biol. 106: 1659-1665, 1988. MCCAFFREY, T. A., D. J. FALCONE, C. F. BRAYTON, L. A. AGARWAL, F. G. P. WELT, AND B. B. WEKSI,ER. Transforming growth S.
RECHLER.
3.
4.
5.
6.
7.
8. 9.
10.
11.
12.
13.
14.
15. 16.
17.
18.
19. This work was supported by Pulmonary Specialized Center of Research Grant (HL-14212) awarded to the University of Vermont College of Medicine, and Veterans Administration Research Funds. J. P. Fabisiak is a Parker B. Francis Fellow in Pulmonary Research.
IN
20.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
L128
21.
22.
23.
24. ‘>r LitI.
26.
TRANSFORMING
GROWTH
factor-@ activity is potentiated by heparin via dissociation of the transf’orming growth factor-e/ tu?-macroglobulin inactive complex. J. Cell. Riol. 109: 441-448, 1989. NIEMEYER, C. M., C. A. SIEFF, B. MATHEY-PREVOT, J. Z. WIMPERIS, B. E. BIERER, S. C. CLARK, AND D. G. NATHAN. Expression of human interleukin-3 (multi-CSF) is restricted to human lymphocytes and T-cell lines. Blood 73: 945-951, 1989. PAUI,SSON, Y., A. HAMMACHER, C-H. HELDIN, AND B. WESTERMARK. Possible positive autocrine feedback in the prereplicative phase of human fibroblasts. Nature Lond. 328: 715-717, 1987. PI,EIXER, W. J., C. D. STILES, H. N. ANTONIADES, AND D. &HER. An ordered sequence of events is required before BALB/c-3T3 cells become committed to DNA synthesis. Proc. Natl. Acad. Sci. USA 74: 4481-4490, 1977. RAINES, E. W., S. K. DOWER, AND R. ROSS. Interleukin-1 mitogenie activity for fibroblasts and smooth muscle cells is due to PDGF-AA. Science Wash. DC 243: 393-396, 1989. RAPPOI,EE, D. A., D. MARK, M. ?J. BANDA, AND Z. WERB. Wound macrophages express TGF-cu and other growth factors in vivo: analysis by mRNA phenotyping. Science Wash. IX 241: 708-712, 1988. RIZZINO, A. Soft agar growth assays for transforming growth
FACTOR-P factors 1987.
IN and
LUNG
mitogenic
peptides.
Methods
Enzymol.
146: 341-:352,
27. ROBERTS, A. B., M. A. ANZANO, L. C. LAMB, J. M. SMITH, AND M. B. SPORN. New class of transforming growth factors potentiated by epidermal growth fact,or: isolation from non-neoplastic tissues. Proc. Natl. Acad. Sci. USA 78: 5339-5343, 1981. A. B., AND M. B. SPORN. The transforming growth 28. ROBERTS, factor-&. In: Handbook of Experimental Pharmacology. Peptide Growth Factors and Their Receptors, edited by M. B. Sporn and A. B. Roberts. Heidelberg: Springer-Verlag, 1990, vol. 95, p. 419-473. 29. SOMA, Y., AND G. R. GROTENDORST. TGF-IJ stimulates primary human skin fibroblast DNA synthesis via an autocrine production of PDGF-related peptides. J. Cell. Physiol. 140: 246-253, 1989. 30. STRIETER, R. M., S. H. PHAN, H. J. SHOWELL, D. G. REMICK, J. P. LYNCH, M. GENORD, R. M. MARKS, AND S. L. KUNKEL. Monokine-induced neutrophil chemotactic factor gene expression in human fibroblasts. J. Riol. Chem. 264: 10621-10626, 1989. 31. YAMAUCHI, K., Y. MARTINET, P. BASSET, G. A. FELLS, AND R. G. CRYSTAL. High levels of transforming growth factor-p are present in the epithelial lining fluid of the normal lower respiratory tract. Am. Rev. Respir. Dis. 137: 1360-1363, 1988.
Downloaded from www.physiology.org/journal/ajplung by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 27, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
