Int. 1. Exp. Path. (1992) 73, 793-800
Role of lymphocytes in silicosis: regulation of secretion of macrophage-derived mitogenic activity for fibroblasts Wei Li, Rakesh K. Kumar, Roslynn O'Grady and Gary M. Velan School of Pathology, University of New South Wales, Sydney, Australia
Received for publication 9 April 1992 Accepted for publication 28 July 1992
Summary. We investigated the role of pulmonary lymphocytes in regulating the secretion by alveolar macrophages (AM) of mitogenic activity for lung fibroblasts, in an experimental model of the initial stages of silicotic inflammation and fibrosis. Following intratracheal instillation of silica, pulmonary parenchymal lymphocytes produced a lymphokine(s) that caused modest stimulation of the secretion of mitogenic activity by normal AM. Co-culture of small numbers of lymphocytes from silica-injected animals with AM induced enhanced secretion of fibroblast growth factor activity which was comparable to the maximal response elicited by recombinant interferon-y. Lymphocytes from animals given non-fibrogenic titanium dioxide exhibited no such effects. The stimulatory effect of lymphocytes from silica-treated animals in co-culture with macrophages was abrogated when the cells were separated by a microporous membrane. Our findings demonstrate that lymphocytes participating in the response to pulmonary deposition of silica are able to induce the secretion of a growth factor(s) for fibroblasts by pulmonary macrophages, possibly via lymphokines expressed on the cell surface or secreted at sites of cell-to-cell contact.
Keywords: pulmonary fibrosis, lymphokines, growth factors Inhalation of inorganic dusts such as silica and asbestos can induce pulmonary inflammation and the subsequent development of alveolar or interstitial fibrosis. It is widely believed that the pathogenesis of fibrotic responses evoked by particulates involves the generation of macrophage-derived growth factor(s) which stimulate proliferation and collagen synthesis by fibroblastic cells in the lung (Davis 1986; Rom et al. 1991). However, regardless of their fibrogenic potential, particulates exhibit comparable ability to stimulate the production of
alveolar macrophage-derived mitogens in vitro (Kumar et al. 1992). We have therefore proposed that the secretory activity of pulmonary macrophages may be regulated in vivo by other inflammatory cells such as lymphocytes. Consistent with this, we have demonstrated that a marked influx of Tlymphocytes is a significant component of the pulmonary inflammatory response induced by silica, which is not elicited by non-fibrogenic titanium dioxide particles (Kumar 1989). The lymphocytes participating in silicotic inflammation are activated, as
Correspondence: R.K. Kumar, School of Pathology, The University of New South Wales, P0 Box 1, Kensington, NSW, Australia 2033. 793
Wei Li et al. 794 assessed by increased expression of inter- Isolation of pulmonary lymphocytes and leukin-2 receptors and enhanced sponta- macrophages neous DNA synthesis (Kumar et al. 1990). AM were obtained by bronchoalveoActivated lung lymphocytes might secrete Normal lar lavage, Dulbecco's phosphatelymphokines which could in turn induce the buffered salineusing (DPBS) containing 100 U/ml secretion of pulmonary macrophage-derived of penicillin and 100 jug/ml of streptomycin growth factors for fibroblasts. To test this (Cytosystems, Sydney) (Kumar al. 1992). hypothesis, we obtained a population of cells These cells were > 95% pure asetassessed by enriched for parenchymal lymphocytes from non-specific esterase staining and their viabithe lungs of animals following intratracheal lity was > 98% by trypan blue exclusion. injection of silica or titanium dioxide partiA population of cells enriched for pulmoncles, and evaluated the effects of these cells ary parenchymal lymphocytes was obtained and their products on the secretion by noras described (Kumar et al. 1990). previously mal AM of mitogenic activity for fibroblasts. Briefly, animals were exsanguinated under Our findings establish that lymphocytes anaesthesia and leucocytes in the pulmonisolated from the lung tissue of animals removed by perfusion were ary capillaries administered intratracheal silica are able saline via the with 0.9% right ventricle. The to stimulate the secretion of macrophageand disaggregated was minced tissue lung derived growth factor activity and that this with collagenase (Boehringer Mannheim, process is facilitated by cell-to-cell contact of cell debris and removal After Sydney). between lymphocytes and macrophages. adherent cells were erythrocytes, depleted by This novel pathway of cellular interaction a column of passing through G-10 Sephadex may contribute to the pathogenesis of silico(Pharmacia, Sydney). The recovered cells sis. contained >85% lymphocytes, as assessed by immunocytochemistry and enzyme histoMethods chemistry. Cells staining positive for nonspecific esterase usually accounted for 2-4% Animals of the separated population and a further 3Specific pathogen-free female BALB/c mice 5% were morphologically recognizable as aged 8-10 weeks were supplied by the SPF neutrophils, while 1-2% were epithelial Animal Facility at the University of New cells. These percentages were similar for cells South Wales. Maintenance and experi- obtained from both silica and titanium dioxmental procedures were in accordance with ide-treated animals. However, the proporthe regulations of the University's Animal tion of T-lymphocytes varied from approximately one-third in the cells obtained from Care and Ethics Committee (Ref. No. ACE 90/ titanium dioxide-treated mice to over two107). thirds in cells from silica-injected animals (Kumar et al. 1990). Viability of the populaIntratracheal injection of particles tion was >90% by trypan blue exclusion. Mice were anaesthetized with a mixture of For convenience, these cell preparations are ketamine and xylazine and 2 mg of either hereafter simply referred to as pulmonary silica (Min-U-Sil 5, US Silica, Berkeley lymphocytes. Springs, WV) or titanium dioxide (Sigma, St Louis, MO) was directly injected into the surgically exposed trachea in 0.05 ml of Serum-free culture 0.9% saline (Kumar 1989). The animals were killed 2 weeks after administration of Lavaged cells were resuspended in serumfree basal medium MCDB 201 (Sigma) and particles.
Lymphocytes in silicosis dispensed at 5 x 105 cells/well of a 24-well culture plate (Nunc, Roskilde, Denmark). AM were allowed to adhere at 3 7°C for 1 hour in a humidified atmosphere of 2% C02 in air and non-adherent cells were removed by washing the wells three times with DPBS. Cultures then received 0.6 ml/well of MCDB 201 medium supplemented with 0.5% fatty acid-free bovine serum albumin, 3 jug/ml cholesterol, 1 jug/ml sphingomyelin, 0.2 ug/ ml vitamin E acetate (all from Sigma), 30 Mug/ ml partially saturated human transferrin (Boehringer Mannheim), 50 U/ml penicillin and 50 jug/ml streptomycin. Culture supernatants were collected after 24 hours. The same medium was used for culture of pulmonary lymphocytes. To detect lymphokine secretion, these cells were incubated in tissue culture tubes (Greiner Labortechnik, Frickenhausen, Germany) at either 106/ml or 5 x 106/ml for 24 hours. Lymphocyte culture supernatants were added to AM at a final concentration of 50% and AM culture supernatants were subsequently collected to assess any effect upon the production of mitogenic activity for fibroblasts. Pulmonary lymphocytes were also tested for their ability to regulate the secretion of fibroblast growth factor activity when directly placed into co-culture with normal AM (Mohler & Butler 1989). Preliminary experiments using lymphocytes from silicainjected animals established that the addition of approximately 3 x 105 lymphocytes per well containing AM caused maximal stimulation of secretion. To facilitate assessment of differences between the effects of coculture of lymphocytes from silica or titanium dioxide-injected mice with AM, a lymphocyte: macrophage ratio that produced submaximal stimulation was chosen. Accordingly, 105 lymphocytes were added to each well. Because it was technically impracticable to perform pulmonary lymphocyte isolation from more than four or five tissue samples while simultaneously collecting AM by bronchoalveolar lavage, these experiments usually comprised two or three sam-
795
ples of lymphocytes from each group of particle-injected animals. To evaluate whether stimulatory effects were dependent upon cellular contact, lymphocytes isolated from silica-injected animals were also placed in co-cultures in well inserts which separated them from the AM by a microporous membrane (pore diameter 0.45 gm; Millicell-HA, Millipore, Sydney). For comparison with a known macrophage-activating factor, each experiment included AM stimulated with 1 ng/ml of recombinant mouse interferon-y (IFN-y) (from E. coli; 107 U/mg) generously supplied by Boehringer Ingelheim, Sydney. Before testing in the bioassay, residual IFN-y and other acid-labile fibroblast inhibitors were inactivated by dialysis against glycine-HCl buffer (pH 2) overnight, in a continuous flow microdialysis unit (Bethesda Research Laboratories, Gaithersburg, MD) employing a membrane with a molecular weight cut-off of 6000-8000. Subsequently, samples were dialysed against DPBS for 24 hours and further dialysed against basal MCDB 201 for 4 hours before being sterilized by filtration.
Bioassayforfibroblast DNA synthesis Adult mouse lung fibroblasts (MLF) were established and maintained in serum-free culture as described in detail elsewhere (Kumar et al. 1991). Mitogenic activity in AM supernatants was assessed by their ability to enhance DNA synthesis by earlypassage non-cycling MLF in a bioassay. Briefly, non-cycling cells were obtained by incubation in the absence of epidermal growth factor (EGF) (the major mitogen in the culture medium) for 96 hours. The medium was then replaced with the AM supernatants to be tested, at a final concentration of 33% in EGF-deficient medium. DNA synthesis was estimated by quantifying the incorporation of methyl-3H-thymidine (Dupont-NEN, Sydney) as previously described (Kumar et al. 1992). As a control, each experiment included assessment of the
Wei Li et al. the ability of lymphocytes to stimulate AM in response to EGF at concentrations from 0 to co-culture in the presence or absence of a 1 5 ng/ml. separating membrane was by a paired t-test. Statistical analysis Results Isotope incorporation data are presented as mean counts per minute (c.p.m.) ± standard Stimulation of AM by supernatants of lung error of replicate samples. Significant differlymphocytes ences between the mitogenic activity of coculture supernatants and that of normal AM We examined whether mediators produced by lung parenchymal lymphocytes could supernatant were assessed by Dunnett's test (Zar 1984). To permit comparative evalu- regulate the secretion of fibroblast mitogenic activity by normal AM. In initial experiation of separate experiments, data were expressed as a percentage of the tritiated ments, we found that supernatants from thymidine incorporation by MLF in response lymphocytes cultured at 106/ml did not stimulate any significant increase in the to normal AM supernatant from the same ability of AM supernatants to enhance DNA experiment and significant differences assessed by an unpaired t-test. Comparison of synthesis by MLF. However, when the cells
796
Table 1. Effect of co-culture with pulmonary lymphocytes on mitogenic activity of AM supernatants for MLF AM co-cultured with lymphocytes from animals injected with
Normal AM Experiment A 7035±983 Experiment B 12122 ± 538
Silica
Titanium dioxide
13517±492** 10316±1300*
7276±499 8324±569
19359±1675** 18526±2172* 15981±1151*
12721 ± 1107 11763±1207
Tritiated thymidine incorporation by MLF exposed to supernatants of AM is shown as mean c.p.m. ± standard error of quadruplicate cultures. In order to obtain a sufficient number of lymphocytes for study, each sample from animals that received intratracheal titanium dioxide comprised pooled cells from the lungs of two animals. Samples from silica-injected mice were cells recovered from individual animals. In Experiment A, isotope incorporation by unstimulated MLF was 2124 ± 480 c.p.m. and the maximal response to EGF was 62200 ± 2463 c.p.m., while the corresponding values for Experiment B were 3418±386 c.p.m. and 32326±914 c.p.m. Significant differences between individual samples and normal AM supernatant from the same experiment are denoted as *P
Role of lymphocytes in silicosis: regulation of secretion of macrophage-derived mitogenic activity for fibroblasts Wei Li, Rakesh K. Kumar, Roslynn O'Grady and Gary M. Velan School of Pathology, University of New South Wales, Sydney, Australia
Received for publication 9 April 1992 Accepted for publication 28 July 1992
Summary. We investigated the role of pulmonary lymphocytes in regulating the secretion by alveolar macrophages (AM) of mitogenic activity for lung fibroblasts, in an experimental model of the initial stages of silicotic inflammation and fibrosis. Following intratracheal instillation of silica, pulmonary parenchymal lymphocytes produced a lymphokine(s) that caused modest stimulation of the secretion of mitogenic activity by normal AM. Co-culture of small numbers of lymphocytes from silica-injected animals with AM induced enhanced secretion of fibroblast growth factor activity which was comparable to the maximal response elicited by recombinant interferon-y. Lymphocytes from animals given non-fibrogenic titanium dioxide exhibited no such effects. The stimulatory effect of lymphocytes from silica-treated animals in co-culture with macrophages was abrogated when the cells were separated by a microporous membrane. Our findings demonstrate that lymphocytes participating in the response to pulmonary deposition of silica are able to induce the secretion of a growth factor(s) for fibroblasts by pulmonary macrophages, possibly via lymphokines expressed on the cell surface or secreted at sites of cell-to-cell contact.
Keywords: pulmonary fibrosis, lymphokines, growth factors Inhalation of inorganic dusts such as silica and asbestos can induce pulmonary inflammation and the subsequent development of alveolar or interstitial fibrosis. It is widely believed that the pathogenesis of fibrotic responses evoked by particulates involves the generation of macrophage-derived growth factor(s) which stimulate proliferation and collagen synthesis by fibroblastic cells in the lung (Davis 1986; Rom et al. 1991). However, regardless of their fibrogenic potential, particulates exhibit comparable ability to stimulate the production of
alveolar macrophage-derived mitogens in vitro (Kumar et al. 1992). We have therefore proposed that the secretory activity of pulmonary macrophages may be regulated in vivo by other inflammatory cells such as lymphocytes. Consistent with this, we have demonstrated that a marked influx of Tlymphocytes is a significant component of the pulmonary inflammatory response induced by silica, which is not elicited by non-fibrogenic titanium dioxide particles (Kumar 1989). The lymphocytes participating in silicotic inflammation are activated, as
Correspondence: R.K. Kumar, School of Pathology, The University of New South Wales, P0 Box 1, Kensington, NSW, Australia 2033. 793
Wei Li et al. 794 assessed by increased expression of inter- Isolation of pulmonary lymphocytes and leukin-2 receptors and enhanced sponta- macrophages neous DNA synthesis (Kumar et al. 1990). AM were obtained by bronchoalveoActivated lung lymphocytes might secrete Normal lar lavage, Dulbecco's phosphatelymphokines which could in turn induce the buffered salineusing (DPBS) containing 100 U/ml secretion of pulmonary macrophage-derived of penicillin and 100 jug/ml of streptomycin growth factors for fibroblasts. To test this (Cytosystems, Sydney) (Kumar al. 1992). hypothesis, we obtained a population of cells These cells were > 95% pure asetassessed by enriched for parenchymal lymphocytes from non-specific esterase staining and their viabithe lungs of animals following intratracheal lity was > 98% by trypan blue exclusion. injection of silica or titanium dioxide partiA population of cells enriched for pulmoncles, and evaluated the effects of these cells ary parenchymal lymphocytes was obtained and their products on the secretion by noras described (Kumar et al. 1990). previously mal AM of mitogenic activity for fibroblasts. Briefly, animals were exsanguinated under Our findings establish that lymphocytes anaesthesia and leucocytes in the pulmonisolated from the lung tissue of animals removed by perfusion were ary capillaries administered intratracheal silica are able saline via the with 0.9% right ventricle. The to stimulate the secretion of macrophageand disaggregated was minced tissue lung derived growth factor activity and that this with collagenase (Boehringer Mannheim, process is facilitated by cell-to-cell contact of cell debris and removal After Sydney). between lymphocytes and macrophages. adherent cells were erythrocytes, depleted by This novel pathway of cellular interaction a column of passing through G-10 Sephadex may contribute to the pathogenesis of silico(Pharmacia, Sydney). The recovered cells sis. contained >85% lymphocytes, as assessed by immunocytochemistry and enzyme histoMethods chemistry. Cells staining positive for nonspecific esterase usually accounted for 2-4% Animals of the separated population and a further 3Specific pathogen-free female BALB/c mice 5% were morphologically recognizable as aged 8-10 weeks were supplied by the SPF neutrophils, while 1-2% were epithelial Animal Facility at the University of New cells. These percentages were similar for cells South Wales. Maintenance and experi- obtained from both silica and titanium dioxmental procedures were in accordance with ide-treated animals. However, the proporthe regulations of the University's Animal tion of T-lymphocytes varied from approximately one-third in the cells obtained from Care and Ethics Committee (Ref. No. ACE 90/ titanium dioxide-treated mice to over two107). thirds in cells from silica-injected animals (Kumar et al. 1990). Viability of the populaIntratracheal injection of particles tion was >90% by trypan blue exclusion. Mice were anaesthetized with a mixture of For convenience, these cell preparations are ketamine and xylazine and 2 mg of either hereafter simply referred to as pulmonary silica (Min-U-Sil 5, US Silica, Berkeley lymphocytes. Springs, WV) or titanium dioxide (Sigma, St Louis, MO) was directly injected into the surgically exposed trachea in 0.05 ml of Serum-free culture 0.9% saline (Kumar 1989). The animals were killed 2 weeks after administration of Lavaged cells were resuspended in serumfree basal medium MCDB 201 (Sigma) and particles.
Lymphocytes in silicosis dispensed at 5 x 105 cells/well of a 24-well culture plate (Nunc, Roskilde, Denmark). AM were allowed to adhere at 3 7°C for 1 hour in a humidified atmosphere of 2% C02 in air and non-adherent cells were removed by washing the wells three times with DPBS. Cultures then received 0.6 ml/well of MCDB 201 medium supplemented with 0.5% fatty acid-free bovine serum albumin, 3 jug/ml cholesterol, 1 jug/ml sphingomyelin, 0.2 ug/ ml vitamin E acetate (all from Sigma), 30 Mug/ ml partially saturated human transferrin (Boehringer Mannheim), 50 U/ml penicillin and 50 jug/ml streptomycin. Culture supernatants were collected after 24 hours. The same medium was used for culture of pulmonary lymphocytes. To detect lymphokine secretion, these cells were incubated in tissue culture tubes (Greiner Labortechnik, Frickenhausen, Germany) at either 106/ml or 5 x 106/ml for 24 hours. Lymphocyte culture supernatants were added to AM at a final concentration of 50% and AM culture supernatants were subsequently collected to assess any effect upon the production of mitogenic activity for fibroblasts. Pulmonary lymphocytes were also tested for their ability to regulate the secretion of fibroblast growth factor activity when directly placed into co-culture with normal AM (Mohler & Butler 1989). Preliminary experiments using lymphocytes from silicainjected animals established that the addition of approximately 3 x 105 lymphocytes per well containing AM caused maximal stimulation of secretion. To facilitate assessment of differences between the effects of coculture of lymphocytes from silica or titanium dioxide-injected mice with AM, a lymphocyte: macrophage ratio that produced submaximal stimulation was chosen. Accordingly, 105 lymphocytes were added to each well. Because it was technically impracticable to perform pulmonary lymphocyte isolation from more than four or five tissue samples while simultaneously collecting AM by bronchoalveolar lavage, these experiments usually comprised two or three sam-
795
ples of lymphocytes from each group of particle-injected animals. To evaluate whether stimulatory effects were dependent upon cellular contact, lymphocytes isolated from silica-injected animals were also placed in co-cultures in well inserts which separated them from the AM by a microporous membrane (pore diameter 0.45 gm; Millicell-HA, Millipore, Sydney). For comparison with a known macrophage-activating factor, each experiment included AM stimulated with 1 ng/ml of recombinant mouse interferon-y (IFN-y) (from E. coli; 107 U/mg) generously supplied by Boehringer Ingelheim, Sydney. Before testing in the bioassay, residual IFN-y and other acid-labile fibroblast inhibitors were inactivated by dialysis against glycine-HCl buffer (pH 2) overnight, in a continuous flow microdialysis unit (Bethesda Research Laboratories, Gaithersburg, MD) employing a membrane with a molecular weight cut-off of 6000-8000. Subsequently, samples were dialysed against DPBS for 24 hours and further dialysed against basal MCDB 201 for 4 hours before being sterilized by filtration.
Bioassayforfibroblast DNA synthesis Adult mouse lung fibroblasts (MLF) were established and maintained in serum-free culture as described in detail elsewhere (Kumar et al. 1991). Mitogenic activity in AM supernatants was assessed by their ability to enhance DNA synthesis by earlypassage non-cycling MLF in a bioassay. Briefly, non-cycling cells were obtained by incubation in the absence of epidermal growth factor (EGF) (the major mitogen in the culture medium) for 96 hours. The medium was then replaced with the AM supernatants to be tested, at a final concentration of 33% in EGF-deficient medium. DNA synthesis was estimated by quantifying the incorporation of methyl-3H-thymidine (Dupont-NEN, Sydney) as previously described (Kumar et al. 1992). As a control, each experiment included assessment of the
Wei Li et al. the ability of lymphocytes to stimulate AM in response to EGF at concentrations from 0 to co-culture in the presence or absence of a 1 5 ng/ml. separating membrane was by a paired t-test. Statistical analysis Results Isotope incorporation data are presented as mean counts per minute (c.p.m.) ± standard Stimulation of AM by supernatants of lung error of replicate samples. Significant differlymphocytes ences between the mitogenic activity of coculture supernatants and that of normal AM We examined whether mediators produced by lung parenchymal lymphocytes could supernatant were assessed by Dunnett's test (Zar 1984). To permit comparative evalu- regulate the secretion of fibroblast mitogenic activity by normal AM. In initial experiation of separate experiments, data were expressed as a percentage of the tritiated ments, we found that supernatants from thymidine incorporation by MLF in response lymphocytes cultured at 106/ml did not stimulate any significant increase in the to normal AM supernatant from the same ability of AM supernatants to enhance DNA experiment and significant differences assessed by an unpaired t-test. Comparison of synthesis by MLF. However, when the cells
796
Table 1. Effect of co-culture with pulmonary lymphocytes on mitogenic activity of AM supernatants for MLF AM co-cultured with lymphocytes from animals injected with
Normal AM Experiment A 7035±983 Experiment B 12122 ± 538
Silica
Titanium dioxide
13517±492** 10316±1300*
7276±499 8324±569
19359±1675** 18526±2172* 15981±1151*
12721 ± 1107 11763±1207
Tritiated thymidine incorporation by MLF exposed to supernatants of AM is shown as mean c.p.m. ± standard error of quadruplicate cultures. In order to obtain a sufficient number of lymphocytes for study, each sample from animals that received intratracheal titanium dioxide comprised pooled cells from the lungs of two animals. Samples from silica-injected mice were cells recovered from individual animals. In Experiment A, isotope incorporation by unstimulated MLF was 2124 ± 480 c.p.m. and the maximal response to EGF was 62200 ± 2463 c.p.m., while the corresponding values for Experiment B were 3418±386 c.p.m. and 32326±914 c.p.m. Significant differences between individual samples and normal AM supernatant from the same experiment are denoted as *P
