Document NEzXg2Y85BNOd5Qb2j5Y4pyn8

AmericanJournal ofPathology, Vol. 134, No. 2, February 1989 Copyright American Association o fPathologists Enhanced M acrophage-Fibroblast Interactions in the Pulmonary Interstitium Increases Fibrosis After Silica Injection to MonocyteDepleted Mice I. Y. R. Adamson, H. L. Letourneau, and D. H. Bowden From the Departm ent o f Pathology, University o f Manitoba, Winnipeg, Manitoba, Canada The role o f interstitial vs. alveolar macrophages in the generation o f pulmonary fibrosis after silica was examined. Using whole body irradiation to de lay the inflammatory response and so retardpartic ulate clearance, many more instilled silica parti cles reached the interstitial macrophages in thefirst 2 weeks than after silica alone. This was follow ed by greatly increased fibroblast proliferation and deposition o f collagen in the irradiation pius silica group, which developed large interstitial granulo mas at th sites o f silica retention. Although alveo lar macrophages containing silica were seen in both silica groups, more interstitial particles were observed after combined irradiation and silica, sig nificantly more silica was recovered in a residue from the Ipngs at ! 6 weeks, and pulmonaryfibrosis at 8 - /6 weeks was greater than in all other groups. The results indicate that increased fibroblast growth and collagen synthesis in vivo are associ ated with phagocytosis o f silica by interstitial mac rophages rather than byfree alveolar macrophages It is suggested that transfer o f a macrophage-de rived growth factor to fibroblasts is more efficient when it occurs within the pulmonary interstitium. (A m f Pathpl 1989, 134:411-418) Deposition of silica in the lung is associated with the de velopment of pulmonary fibrosis. In this and other exam ples of environmental lung disease, it is believed that the interaction of particulates with the alveolar macrophage (AM) is responsible for the generation of growth factors for fibroblasts 12 The demonstration of macrophage-de rived factors has been made after in vitro exposure of normal AM to particles34 and using supernatants of AM lavaged from fibrotic lungs.56 Pulmonary fibrosis is, however, an interstitial lung dis ease and it is not clear whether factors secreted by mac rophages in the alveoli can reach and act on fibroblasts in the interstitium, or whether secretory activity of the inter stitial macrophage plays the critical role in fibroblast stimu lation. We have previously observed that, in mice, short particles of crocidolite asbestos are ingested by AM with no subsequent fibrosis7 whereas the same dose of long fibers, many of which reach the interstitial macrophages, induce pulmonary fibrosis.8 These results lead to the hy pothesis that the interaction between particles and mac rophages within the interstitium may be more important than that occurring in the alveolar space in stimulating the fibrotic process. The close contact between macro phages and fibroblasts within the interstitium may permit more efficient transfer of macrophage-derived growth fac tors for fibroblasts than would secretion by macrophages into the alveolar space. To evaluate the relative roles of alveolar versus intersti tial locations of the same particle in stimulating fibrosis, we chose to administer a single dose of silica to mice. In this group of animals, most particles were phagocytized by the AM. In a second group of mice, we created a situa tion in which reduced phagocytosis and clearance of sil ica from the alveoli allowed a greater number of particles to reach the interstitium. This was done using whole-body irradiation to reduce the leukocyte number, thereby limit ing the inflammatory response and the clearance of sub sequently instilled silica. This allowed increased transport of silica to the pulmonary interstitium for phagocytosis by interstitial macrophages. The cellular responses seen in the alveoli, the cytokinetics of various pulmonary cell types, and the development of fibrosis were correlated Supported by the Medical Research Council of Canada. Accepted for publication October 3. 1988 Address reprint requests to Dr. I Adamson, Pathology Department, University of Manitoba. 770 Bannatyne Avenue. Winnipeg, Manitoba, Can ada R3E 0W3. 411 412 Adamson, Letoumeau, and Bowden AJP February I9H9. Voi 1.14, No 2 &control xR adn. o S i # S i + Radn. Figure 1. White blood cell (WBC) counts in the various groups o f mice. Irradiation was given to 2 groups 3 days before time zero. *P < 0 .0 / compared with controls at the same time. X, irradiation only, O, silica only. , silica after irradiation. A, no treatment with the location and retention of silica in the lung, and compared in the various groups. Materials and Methods A group of 50 Swiss-Webster mice (25 g males) were placed in separate sections of a plexiglas box and re ceived 650 rads of whole-body irradiation.9 Three days later the mice received an intratracheal instillation of 1 mg silica (crystalline quartz, Dowson and Dobson, South Af rica) in 0.1 ml stenle water while under mild nembutal an esthesia.'0 Other groups of mice received silica only, irra diation only, or no treatment. The animals, in groups of 4, were killed at days 0 and 3 then at weeks 1,2, 4, 8, 12, and 16 with time counted from the point of silica injection. Each mouse received 2 iCi/g tritiated thymidine 1 hour before death. The heart was punctured and 0.3 ml blood withdrawn and the total white blood cell count (WBC) was deter mined using a Coulter counter. A tracheotomy was then performed, and the lungs were washed four times with 1 0 ml saline. The lavage fluid was pooled for each animal, and the total number of cells was counted on a hemocytometer The cell suspension was centrifuged, and a smear was made and stained; differential counts of poly morphonuclear leukocytes (PMN) and AM were made on 500 cells per slide. The total number of cells of each type was calculated for each time studied, and the mean SE of the total cells per lung in the four mice per group was plotted against time. The supernatant of the lavage fluid from each mouse was used to determine total protein 1' After lavage, the bronchus leading to the right lung was clamped; this lung was removed, weighed, and fro zen for biochemical analysis. The left lung was inflated with 0.5 ml 2% buffered glutaraldehyde and removed; most of the tissue was processed for embedding in glycol methacrylate Sections (0.75 ^ thick) from three random blocks per animal were prepared for autoradiography with the use of Kodak NTB2 emulsion. We determined the per centages of 3H-thymidine-labeled cells at each time point by counting 3000 lung cells per animal. The means and standard error were calculated for each group. These sec tions were thin enough to allow identification of pulmonary cell types, and differential counts of labeled cells were performed on 300 labeled cells per animal. The product of the differential percentage and the total labeling per centage gave the labeling index for each cell type, The index for epithelial, interstitial, and endothelial cells was calculated at each time studied. The right lung of each mouse was homogenized in wa ter, and biochemical assays were performed on duplicate samples. Determinations of DNA and total protein were performed by conventional methods. As an index of colla gen content, hydroxyproline levels were determined after hydrolysis with hydrochloric acid.'2 At 16 weeks, 4 extra mice from each group were killed and the lungs removed immediately. These were chopped up and incubated in 40% KOH overnight in a 80 C waterbath. When the tissue was completely digested, the solution was centrifuged at 1500 rpm for 15 minutes and a residue was obtained. This was washed twice in distilled water, resuspended in water, and 1 drop placed on a copper grid for examination by electron microscopy. The remainder was transferred to a weighed tube, dried down, and the weight of the residue was determined. Results Blood The numbers of WBC recovered from the various groups are shown in Figure 1 The values for controls and silicainjected animals were equal and rose slightly over the 16week period. The other two groups were irradiated 3 days before the time zero on the graph. Irradiated mice re ceived silica when the WBC count was at 1 X 103. The WBC fell to low values and did not rise until after a further week. The counts were in the normal range at 4 weeks. Bronchoalveolar Lavage In the lavage from groups of normal and from irradiated mice, very few if any PMN were recovered at any time Interstitial Macrophages in Pulmonary Fibrosis 413 AJP February 1989, Vol 134, No. 2 Figure 2. The numbers o f PMN recovered in HAL flu id up to 10 weeks The valuesfrom control atid irradiated groups uere at or near zero so that all values shown are significantly greater than control (Figure 2). In contrast, mice that received silica alone showed a rapid eflux of PMN, which peaked at 5 X 10s in 3 days. The number of these cells dropped rapidly but never reached zero up to 16 weeks later. When irradiated mice received silica, there was a small increase in PMN in the BAL fluid over the first week then a sharp increase to a peak at 2 weeks (Figure 2). Although the number of PMN fell subsequently, it was always higher than that found in all other groups to 16 weeks The AM response to silica in irradiated mice was also delayed. The AM numbers in normal and irradiated ani mals were constant at about 2 X 10s(Figure 3). After silica alone, there was a rapid large increase in AM that declined over the next few weeks; however, the number recovered remained above normal up to 12 weeks after silica. When irradiated mice were instilled with particles, little effect on AM numbers was seen initially, then a fourfold rise oc curred (Figure 3). Although the values fell somewhat, the number of AM recovered remained above normal over the remainder of the 16 week expenment. The protein level in the bronchoalveolar lavage (BAL) supernatants was measured as an indicator of lung injury The 650 rads of irradiation alone did not result in a signifi cant increase in protein over that lavaged from control mice (Figure 4). Silica administration alone resulted in a fourfold increase in BAL protein at 3 days and the level rapidly returned to control values The combined irradia tion plus silica group showed a much higher peak in BAL protein and the increase was maintained for several weeks. The protein level was still significantly higher than in controls at 16 weeks, indicating ongoing lung injury (Figure 4) Morphology Animals that received irradiation only showed morpho logic changes similar to these described earlier.9 There Figure 3. The number o f AM recovered in BAL flu id up to 16 weeks P < 0.01 compared with controls. was some evidence of endothelial injury and interstitial edema in the first 2 weeks whereas the epithelium was normal. From 8 weeks on there were focal areas of linear fibrosis resulting in a thickened air-blood barrier, espe cially at perivascular locations. After silica alone, there was an initial inflammatory re sponse with many AM and PMN seen in alveolar spaces. After 1 week the silica appeared confined to these cells with small amounts only in the pulmonary interstitium Fo cal areas of type 2 epithelial cell proliferation were also seen and the interstitium showed a mixed population of macrophages and fibroblasts. From 4 weeks on, small focal interstitial granulomas were observed (Figure 5). These became less cellular and more fibrotic with time and resembled those described earlier in the pulmonary response to silica.13 Free alveolar macrophages contain ing silica were seen in the cytospin preparations of la vaged cells and in lung sections up to 16 weeks. In the peripheral alveolar regions of the lung, there was no evi- a control x Radn. 0 S1 Si + Radn Figure 4. Protein levels in HAL flu id over 16 weeks in the ran ousgroups P < (LUI compared with controls. 414 Adamson. Letourneau, and Bowden AJF February 1989, Vol 1 $4, No J dence of fibrosis in the alveolar walls, and the air-blood barrier was thin as in controls, even at areas where these appeared to be trapped AM containing silica. When irradiated mice received silica, the normal in flammatory response was impaired and fewer AM and PMN were seen. Free silica particles were found up to 2 weeks in the alveoli and many particles reached the interstitium, where they were phagocytized by macrophages. Many large interstitial granulomas composed of macro phages and fibroblasts formed as early as 2-4 weeks af ter silica injection (Figure 6). In these areas many of the macrophages contained silica crystals that could be iden tified using polarized light. These large granulomas be Figure 5. Section o f lung 8 weeks after sil ica alone. A small interstitial granuloma composed of macrophages a nd fibroblasts is seen Alveolar walls appear thin as usual (arrows. X850) came more fibrotic with time although they still retained a high silica content. In these lungs, fibrosis seemed limited to the interstitial granulomas and little or no change in thickness was seen in alveolar walls even in regions of high AM content. Autoradiography The percentage of thymidine-labeled cells in normal mouse lung was approximately 0.3% (Figure 7), After irra diation alone, labeling increased in a biphasic manner over a 4-week period. From the radiographic indices, it Figure 6. Lung 4 weeks after silica with prior irradiation Interstitial granuloma is very large and cellular Many macro phages containing silica (arrows) are ob served (X850). Interstitial Macrophages in Pulmonary Fibrosis 415 AJP February 1989, Voi 134, No 2 B IPfT M fLIA L CCLLt fNOOTHCLIAL CCLLt Q M T IM S T rr tAL C1LLS Acontrol x Radn oSi Si+Radn Figure 7. Percentages o f fH thymidine labeled nuclei in lung sections. *P < 0 01 compared with controls; a, P < 0 01 com pared with all other groups was found that the initial peak was due to an increase in endothelial cell labeling subsequent to the endothelial damage, whereas the later phase reflected a period of increased fibroblast growth (Figure 8). After silica alone, there was an immediate increase in labeling to 2.5% of lung cells This value dropped steadily to normal after 8 weeks (Figure 7). Differential counts showed that there was an early phase of type 2 epithelial cell proliferation that probably followed type 1 cell necrosis.'0 Interstitial cell labeling was much more prominent, however (Figure 8). Up to 2 weeks after irradiation most of these labeled cells resembled mononuclear phagocytes, whereas after 2 weeks the majority of labeled interstitial cells resembled fibroblasts The greatest increase in cell proliferation in the lung was seen in the combined irradiation-silica group. The ini tial peak was over 3% and a large increase over normal was seen throughout; from 4 weeks DNA synthesis was significantly higher than in all other groups (Figure 7). These mice showed periods of endothelial and epithelial regeneration equivalent to that seen after irradiation or sil ica alone, but the large increase in labeling was due to thymidine uptake by interstitial cells (Figure 8). Initially the labeled nuclei were a mixture of mononuclear cells resem bling macrophages and fibroblasts (Figure 9) After 2 weeks, however, it appeared that most of the interstitial labeling was due to fibroblast proliferation (Figure 10) Figure 8. Radiographic indices fo r epithelial, endothelial and interstitial cells in the various treatment groups *P < 0.01 compared with controls: a, P < 0.01 compared with all other groups at the same time HYP beginning at 4 weeks and giving a greater than 50% increase at 16 weeks. The combination of irradiation fol lowed by silica injection produced an increase in HYP by 2 weeks and the level continued to rise and remained 100% greater than that found in controls. After 8 weeks the level was significantly higher than all other groups. A similar increase in fibrosis was seen when the results were ex pressed as HYP/protein (Figure 12). Biochemistry To quantitate pulmonary fibrosis, total hydroxyproline (HYP) content of the right lung was measured (Figure 11). Irradiation or silica exposure alone resulted in increased Figure 9, Autoradiograph o f lung I week after silica ( 10 days after irradiation ). Thymidine uptake occurred in interstitial mononuclear cells, many resembling macrophages (arrows. X 10(H)). 416 Adamson, Letourneau, and Bowden AJP February 1989, Vol. 1J4, No. J a control xRadn O Si S i+ R a d n Figure 12. The ratio o f HYP toprotein in right lungs o f the vari ous groups. *P < 0,01 greater than controls, a, P < 0.01 greater than all other groups. Figure 10. Autoradiograph o f lung 4 weeks after silica and irradiation. Labeled cells are mostly fibroblasts (arrows. XIOOO). Tissue Residue The weights of residues recovered after tissue digestion are shown in Table 1. A small increase in residue was found at 16 weeks in control and irradiated mice, probably due to insoluble elastin. The residue was amorphous and always weighed less than 0.2 mg for controls and for irra diated mice that showed some fibrosis. The residue from silica-treated lungs was white in appearance and was sig nificantly heavier. After silica plus irradiation, a white pow- der was obtained after lung digestion. This residue was almost all particulate, as seen in the electron microscope, and was twice the weight of the residue from the silica only group and was four times greater than the weight of the control material. Discussion In general, small amounts of inhaled particulates can be phagocytized and cleared from the lung by alveolar mac rophages with no pathologic sequelae. Larger loads re quire an adaptive increase in the number of AM involving migration of blood monocytes to the alveoli supple mented by division and migration of mononuclear phago cytes from the interstitium. These cells, plus the initial influx of PMN, constitute the normal inflammatory re sponse and usually clear the airways. Some particles, however, especially at higher dose levels, escape phago cytosis and cross the type 1 epithelial cell to reach the interstitial macrophage.1013 Although macrophagic up take of particles such as silica is associated with the pro duction and secretion of various fibroblast growth factors (FGF), the relative importance of macrophage-particle in- Figure 11. Hydroxyprotine (HYP) content o f the right lung to 16 weeks *P < 0 01 greater than controls, a, P < 0 01 greater than all other groups. Table 1. Weight oj Residue After Total Lung Digestion Group (N = 4) Dry weight (mg SE) Control, week 0 Control, week 16 Irradiation Silica Silica and irradiation 0.05 0.02 0.12 + 0.04 0.17 + 0.05 0.40 0 .07 * 0.78 + 0.09 "t * Experimental greater than control, P < 0.01 t Value greater than all other groups. P < 0.01 Interstitial Macrophages in Pulmonary Fibrosis 417 AJP February 1989. Vol. 134, No. 2 teractions in the alveolus and in the interstitium to fibro blast stimulation in vivo is not known. Macrophage-derived growth factors have been dem onstrated in pure in vitro systems15 and in studies using supernatants of AM lavaged from silicotic lungs.2-5 This may be a general property of macrophages because FGF are associated with various types of pulmonary fibrosis6 and may be common to a variety of chronic inflammatory states.16 In the lung, most studies have involved the readily accessible AM and its secretion under serum-free test conditions. However, whether a secretion by free al veolar macrophages in vivo can reach and affect intersti tial fibroblasts is not known and the possibility of direct transfer of FGF from interstitial macrophages to neigh bouring fibroblasts has not been investigated. The model used in the present experiments allows comparison of the fibrotic response to silica in vivo when particles are either largely confined to AM or are located predominantly within interstitial macrophages. This experiment is based on our earlier finding that reduction of the alveolar inflammatory response to intratracheal carbon significantly increases the transfer of free particles to the interstitium 17 After the instillation of silica to the lung, we observed a delayed inflammatory response in irradiated mice, with few cells in the alveoli to 2 weeks and a large increase in the translocation of silica to the interstitium. Many parti cles were found in interstitial macrophages over the 16week period, in association with fibroblasts within colla gen forming granulomas. Even in these lungs, regions where silica was confined to free AM within the air sacs did not show thickening of alveolar walls. When the lung tissue at 16 weeks was digested, twice as much residue, mainly silica, was recovered from the irradiation plus silica group. The prolonged retention of silica mostly in intersti tial macrophages over 16 weeks was associated with a significant increase in fibroblastic proliferation and colla gen deposition. Thus, the enhanced translocation of silica to interstitial macrophages results in greater pulmonary fibrosis than when the same dose of silica is predomi nantly handled within the air sacs by the AM Although the possibility that the enhanced fibrotic re sponse results from synergistic action of silica and irradia tion on lung cell injury cannot be ailed out, there are sev eral arguments against this possibility. The injury induced by irradiation alone is small, as shown by unchanged alve olar protein levels, and is confined to the endothelium, which is rapidly repaired (Figure 8). After irradiation fol lowed by silica, the same degree and timing of endothelial repair is seen, indicating an equivalent amount of endo thelial injury to the irradiation alone group. In addition, epi thelial damage associated with this level of silica adminis tration is also repaired within 2 weeks in the silica plus radiation group, so that the subsequent progressive fi brosis is not likely the result of continuing injury to the cells of the air-blood barrier. The type of granulomatous fibrosis seen after combined irradiation plus silica is clearly an exaggeration of the pulmonary response to sil ica, and more in keeping with the observed increase in interstitial silica. The predominant dividing cell in the lung after 2 weeks is the fibroblast and the continued high level of fibroblast proliferation and collagen deposition in the lung in the silica plus irradiation group is more likely to reflect ongoing secretion of a FGF by particle laden interstitial macrophages trapped within the gran ulomas. These results demonstrate that limiting the initial in flammatory response is a useful method of increasing transepithelial passage of particles and their subsequent long-term retention in the lung. This provides a useful model for exaggerating the fibrotic response to particles such as silica, and suggests that, in vivo, the interstitial macrophage may play a more important role than the al veolar macrophage in fibrogenesis. In examining the properties of pulmonary macrophages, almost all studies have been carried out in vitro, usually on the alveolar mac rophage; only recently has the interstitial macrophage re ceived much attention. It is more difficult to isolate a pure population of these cells and it appears that the prolifera tive capacity and some functional characteristics are different from those of the AM.1819 Further in vitro experi ments will be required on pure populations to determine whether interstitial macrophages are more potent secretors of FGF than alveolar macrophages. It is also possible that secretion by these two cell types is equal, but that in vivo, growth factors secreted into the alveolar space are more likely to be inactivated by serum factors for exam ple, whereas close contact between activated macro phages and fibroblasts in the pulmonary interstitium may allow more efficient transfer of FGF. In either case, the results of the present in vivo study suggest that the key events in the generation of severe pulmonary fibrosis are the translocation of particulates across the epithelium and the stimulation of the interstitial macrophage. References 1. Davis GS. Pathogenesis of silicosis: Current concepts and hypotheses Lung 1986, 164:139-154 2. Goldstein RH, Fine A: Fibrotic reactions in the lung: The acti vation of the lung fibroblast Exp Lung Res 1986, 11:245 261 3. Bateman ED, Emerson RJ, Cole PJ: A study of macrophage mediated initiation of fibrosis by asbestos and silica using a diffusion chamber technique Br J Exp Pathol 1982. 63:414 425 418 Adamson. Letourneau, and Bowden AJP F e b ru a ry 1< W 9 . Vol. I.i4 , No. 2 4. Heppleston AG: Pulmonary toxicology ot silica, coal and as bestos. Environ Health Perspect 1984, 55:111-127 5. Lugano EM, Dauber JH, Elias JA, Bosley Rl. Jimenez SA, Daniele RP: The regulation of lung fibroblast proliferation by alveolar macrophages in expenmental silicosis. Am Rev RespirDis 1984, 129:767-771 6. Bitterman PB, Adelberg S, Crystal RG: Mechanisms of pul monary fibrosis: Spontaneous release of the alveolar macro phage-derived growth factor in the interstitial lung disorders. J Clin Invest 1983, 72:1801-1813 7 Adamson IYR, Bowden DH: Response of mouse lung to crocidolite asbestos: 1. Minimal fibrotic reaction to short fibers. J Pathol 1987, 152:99-107 8. Adamson IYR, Bowden DH: Response of mouse lung to cro cidolite asbestos: 2. Pulmonary fibrosis after long fibers J Pathol 1987, 152:109-117 9. Adamson IYR, Bowden DH: Endothelial injury and repair in radiation-induced pulmonary fibrosis Am J Pathol 1983, 112:224-230 10 Adamson IYR, Bowden DH: Role of polymorphonuclear leu kocytes in silica-induced pulmonary fibrosis Am J Pathol 1984, 117:37-43 11. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ: Protein measurement with the Folin phenol reagent. J Biol Chem 151, 193:265-269 12 Woessner JF: Determination of hydroxyproline in connective tissue, The Methodology of Connective Tissue Research. Edited by DA Hall. Oxford, Joynson and Bruwres, 1976, pp 227-233 . 13. Bowden DH, Adamson IYR: The role of cell injury and the continuing inflammatory response in generation of silicotic pulmonary fibrosis. J Pathol 1984,144:149-161 14. Bowden DH, Adamson IYR: Adaptive responses of the pul monary macrophage system to carbon: 1 Kinetic studies. Lab Invest 1978, 38:422-429 15. Aalto M, Potila M, Kulonen E: The effect of silica treated mac rophages on the synthesis of collagen and other proteins in vitro Exp Cell Res 1976, 97:193-202 16. Wahl SM: Inflammatory cell regulation of connective tissue metabolism. Rheumatol Int 1986. 10:404-429 17 Adamson IYR, Bowden DH: Effects of irradiation in macro- phagic response and transport of particles across the alveo lar epithelium. Am J Pathol 1982. 106:40-46 18. Dethloff LA. Lehnert BE: Pulmonary interstitial macrophages: Isolation and flow cytometric comparisons with alveolar macrophages and blood monocytes. J Leuk Biol 1988. 43: 80-90 19. Brain JD Lung macrophages How many kinds are there? What do they do7 Am Rev Respir Dis 1988. 137:507-509