Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The genetic and immunologic basis of pulmonary fibrosis is poorly understood. In rodents, intratracheal bleomycin administration induces progressive inflammation and fibrosis with histologic changes that model pulmonary fibrosis in human diseases such as idiopathic pulmonary fibrosis (1, 2). Within 5 to 7 d after bleomycin exposure, an intense interstitial pneumonitis develops characterized by a mononuclear cell alveolitis and interstitial infiltration with epithelial cell necrosis and subsequent interstitial fibrosis (3). Previous studies have demonstrated a genetic susceptibility to bleomycin-induced pulmonary toxicity based on the close association between mouse strain and the fibrotic outcome (4). For example, C57BL/6J and C3H/HeN mice are considered to be fibrosis-prone, and BALB/c and C3H/fKam mice are relatively fibrosis-resistant (3). The mechanisms behind this genetic susceptibility are incompletely understood but may include differences in bleomycin pharmacokinetics, susceptibility to oxidative stress, ability to repair DNA damage, or immune system responses to lung injury (6).

Several studies have suggested that T cells play a role in the development of bleomycin-induced pulmonary toxicity in the rodent models. These studies include reports that mice which have undergone T-cell depletion by anti-CD3 antibody treatment or cyclosporine A treatment have an attenuated fibrotic response to bleomycin in susceptible mouse strains (7). Other studies using T cell-deficient nude/athymic mice or SCID mice question whether T cells play an essential role in bleomycin-induced pulmonary fibrosis because they report either slightly reduced or no alteration in bleomycin susceptibility (11). The mechanisms by which T cells might contribute to the fibrotic response have not been determined but likely involve the production of T-cell cytokines and/or direct T-cell cytotoxic effects.

The pattern of cytokines produced during an inflammatory response to many infectious agents tracks closely with mouse strain; for example, C57BL/6J mice tend to express T helper (Th)1 cytokines (interferon [IFN]-, interleukin [IL]-2), and BALB/c mice tend to express Th2 cytokines (IL-4, IL-5) in response to certain infectious agents such as Leishmania major (14). Interestingly, several mouse strains that are susceptible to bleomycin-induced pulmonary toxicity, such as the C57BL/6J and C3H/HeN strains, are those that tend to produce Th1 cytokines in response to soluble antigens or infectious agents, whereas the "fibrosis-resistant" BALB/c strain tends to produce Th2 cytokines in response to these stimuli. Based on these observations, we hypothesized that the pulmonary inflammatory and fibrotic response to bleomycin is mediated, in part, by local production of the major Th1 cytokine IFN-. To address this question, we analyzed expression of Th1 and Th2 cytokines in susceptible and resistant mice and directly evaluated a role for IFN- in bleomycin-induced pulmonary toxicity using genetically altered IFN--deficient (knock-out) mice.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Lyophilized bleomycin sulfate (Bristol-Myers Squibb, New York, NY) was diluted to a stock concentration of 5 U/ml in sterile normal saline and kept at 20°C until used.

Animal Procedures

Wild-type A/J, C57BL/6J, BALB/c, and C57BL6/J IFN--deficient (IFN-[/]) mice were purchased from Jackson Laboratories (Bar Harbor, ME) (15). The presence of the disrupted IFN- gene in the IFN-(/) mice was confirmed by the absence of native IFN- messenger RNA (mRNA) by reverse transcription polymerase chain reaction (data not shown). All animals were 6- to 8-wk-old female mice and were housed in the same animal facility. Mice were randomly assigned to receive 30 to 40 µl of either bleomycin (5, 1.5, or 0.5 U/kg) or sterile saline. Mice were anesthetized with intraperitoneal ketamine and xylazine, and then had intratracheal administration of either bleomycin or saline solution via a Pipetman (16). Mice were killed under phenobarbital anesthesia by exsanguination, consistent with the recommendations of the Panel on Euthanasia of the American Veterinary Medical Association. Single lung bronchoalveolar lavage (BAL) was performed with a single aliquot of 400 µl of sterile saline that was flushed five times into the airways using a Teflon 20-guage Jelco intravenous catheter (Johnson & Johnson, New Brusnwick, NJ). The lungs were excised for histology and RNA extraction. All animal handling and procedures were performed according to protocols approved by the Johns Hopkins University Animal Care and Use Committee.

Histologic Analysis

Lungs removed for histology were inflated with 250 to 350 µl of 10% neutral formalin and then immersed in this fixative solution and imbedded in paraffin. Lung sections were taken from at least three levels (apical, midlung, and basal) of the left lobe and were stained with hematoxylin and eosin (H&E) or Sirius Red III for collagen with fast green FCF counterstaining.

Alveolar and interstitial cellular infiltration was graded by a blinded reviewer according to the following scale modified from Raisfeld (3): ratings of 0, 1, 2, 3, 4, and 5 correspond, respectively, to 0, < 10, 10 to 25, 25 to 50, 50 to 75 and > 75% of lung area involved with mononuclear cell infiltration, interstitial thickening, distortion of native lung architecture, or abnormal collagen deposition.

Cytokine Analysis of BAL Fluid and Whole Lung

Concentrations of IFN- and IL-4 protein in BAL fluid from individual mice were determined by enzyme-linked immunosorbent assay (ELISA) using OptEIA kits (PharMingen, San Diego, CA). Concentrations of IFN- in whole lung tissue were determined by ELISA after manual homogenization of lung samples in saline. Samples were measured photometrically by an automated plate reader (Microplate Reader Model 550; Bio-Rad, Hercules, CA). All assays were performed in duplicate.

Analysis of Cytokine mRNA Expression

Lungs removed for RNA analysis were placed immediately in 1 ml of TRIzol (GIBCO/Life Technologies, Inc., Rockville, MD), snap-frozen in liquid nitrogen, and stored at 80°C. Samples were later thawed on ice, homogenized using a Polytron apparatus (Brinkman Instruments, Westbury, NY), and RNA extracted by the modified single-step method as recommended by the manufacturer (17).

Steady-state mRNA levels of IL-12p40, IL-4, IL-5, and transforming growth factor (TGF)-1 were determined by RNase protection assay (RiboQuant; PharMingen) according to the manufacturer's recommended protocol and as previously reported (18). Briefly, total lung mRNA was hybridized overnight with RNA probes radiolabeled with {-³²P}uridine triphosphate, digested with RNase A, and the protected fragments resolved by 8% polyacrylamide gel electrophoresis. Autoradiographs were analyzed by laser densitometry. To control for differences in the amount of mRNA loaded per lane, cytokine mRNA was normalized to the amount of ribosomal L32 mRNA in the same lane.

Analysis of Collagen Content

Single lung total collagen was quantified by analysis of hydroxyproline, an amino acid unique to this protein (19). Briefly, lung tissue from the right lung was manually homogenized in saline. An aliquot of lung homogenate was hydrolyzed in 4N NaOH at 120°C for 10 min in an autoclave. The mixture was reacted with chloramine-T and Ehrlich's reagent to produce a hydroxyproline-chromophore that was quantified by 550 nm spectrophotometry. A second aliquot of the original lung homogenate was analyzed by the bicinchroninic acid method for colorimetric detection and quantification for total protein content (BCA Protein Assay; Pierce, Rockford, IL). All protein assays were performed in duplicate or triplicate.

Statistical Analysis

Data are reported as mean ± standard error (SE). After testing for normality, statistical analyses of parametric group data were performed by analysis of variance (ANOVA) with post-hoc comparisons by the method of Scheffé. Histologic data were evaluated by group with the nonparametric Kruskal-Wallis test followed by post-hoc comparisons by the nonparametric Mann-Whitney U test. Survival data were evaluated by chi square analysis. A probability value of P < 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Morphologic Changes and Cytokine Expression in Response to Intratracheal Bleomycin Exposure

Consistent with prior reports, C57BL/6J mice demonstrated interstitial inflammation and fibrosis 10 d after intratracheal bleomycin administration but not saline control exposures (Figure 1) (4). A/J mice also demonstrated a marked inflammatory response after intratracheal bleomycin administration, whereas saline control lungs showed normal lung histology (Figure 1). Histologic changes seen after bleomycin exposure in both of these mouse strains included interstitial pneumonitis with increased interstitial wall thickness, interstitial mononuclear cell infiltrates, fibroblasts, and interstitial collagen deposition associated with architectural distortion of lung tissue. Histologic analysis with Sirius Red III staining for collagen confirmed the presence of collagen deposition in areas of inflammation in bleomycin-exposed C57BL/6J and A/J mice (data not shown). BALB/c mice demonstrated minimal degrees of interstitial and alveolar inflammation or collagen deposition 10 d after intratracheal bleomycin administration. Consistent with histologic changes demonstrating marked increases in the number of inflammatory cells in bleomycin-susceptible mice, greater amounts of total RNA were recovered from the lungs 7 d after bleomycin exposure in C57BL/6J and A/J mice compared with their own saline controls or bleomycin-exposed BALB/c mice (Figure 2).

View larger version (129K): [in this window] [in a new window]   View larger version (150K): [in this window] [in a new window]   View larger version (129K): [in this window] [in a new window]   View larger version (156K): [in this window] [in a new window]   View larger version (130K): [in this window] [in a new window]   View larger version (120K): [in this window] [in a new window]   Figure 1.   Lung histologic changes in mice after intratracheal instillation of saline or bleomycin. Shown are representative photomicrographs of H&E-stained sections of lung tissue from C57BL/6J (A and B), A/J (C and D), and BALB/c (E and F ) mice 10 d after intratracheal instillation of either saline (A, C, and E) or 5 U/kg bleomycin (B, D, and F ). Extensive interstitial mononuclear cell infiltration, interstitial thickening, and architectural distortion are seen in both C57BL/6J and A/J mice but not in BALB/c mice after bleomycin exposure.

View larger version (42K): [in this window] [in a new window]   Figure 2.   Total lung RNA content after intratracheal saline or bleomycin administration. C57BL/6J (open bars), A/J (shaded bars), and BALB/c (solid bars) mice were exposed to either saline or 5 U/kg bleomycin via intratracheal instillation. Total RNA was isolated from lung tissue collected at baseline, 3 d (3d saline and 3d bleo), or 7 d (7d saline and 7d bleo) after exposure. Results are shown as the mean ± SE of six mice per time point. Comparisons are denoted by brackets: *P < 0.0001; **P < 0.05; P > 0.05; P = 0.12, not significant.

To determine whether Th1 or Th2 cytokine expression correlated with the histologic changes in these mouse strains, concentrations of IFN- and IL-4 protein were measured in BAL fluid after intratracheal bleomycin or saline administration. In both bleomycin-sensitive strains (C57BL/ 6J and A/J), the concentration of IFN- protein in BAL fluid was significantly higher 12 and/or 24 h after bleomycin administration compared with their respective saline controls (Figure 3). In contrast, lower concentrations of IFN- protein were seen in bleomycin-exposed BALB/c mice compared with saline controls at all time points with the differences achieving statistical significance at 3 d and 7 d. Levels of IFN- protein in whole lung tissue were also significantly higher at 24 h after bleomycin exposure compared with saline exposure in C57BL/6J mice (bleomycin versus saline, 540 ± 56 versus 278 ± 33 pg/ml; P  0.005) and A/J mice (480 ± 49 pg/ml versus 249 ± 34 pg/ml; P < 0.05), and this same regulation was not seen in the bleomycin-resistant BALB/c mice (351 ± 25 versus 195 ± 11 pg/ ml; P = 0.33, not significant). The Th2 cytokine IL-4 was either not detected or detected at minimal levels in BAL fluid from mice in all groups (Figure 3). Together, these results suggest that bleomycin-induced pulmonary toxicity is associated with enhanced production of IFN- within 24 h after bleomycin exposure in the bleomycin-susceptible C57BL/6J and A/J mice.

View larger version (41K): [in this window] [in a new window]   Figure 3.   Expression of IFN- and IL-4 protein in BAL fluid after intratracheal saline or bleomycin administration. A/J (A and B), C57BL/6J (C and D), and BALB/c (E and F ) mice were exposed to either saline (open bars) or 5 U/kg bleomycin (solid bars) via intratracheal instillation. BAL fluid was obtained at baseline, 12 h, 24 h, 3 d, or 7 d later. Concentrations of IFN- (A, C, and E) and IL-4 (B, D, and F ) were determined by ELISA. Each time point represents the mean ± SE of six mice per group. Data were analyzed using pair-wise t tests. Comparisons are denoted by brackets: *P < 0.05, **P < 0.005.

To evaluate cytokines that are known to regulate IFN- production, we measured mRNA levels of IL-12, IL-4, and TGF- after bleomycin exposure in these mice by RNase protection assay. We also measured levels of the Th2 cytokine IL-5 that has been previously implicated in bleomycin-induced pulmonary toxicity (20). IL-12 upregulates production of IFN- and, conversely, IFN- is a potent costimulator of IL-12 expression, resulting in a positive feedback loop that promotes upregulation of both cytokines (21). Consistent with an upregulation of IFN- expression, steady-state mRNA expression of the highly regulated (and IFN--inducible) IL-12p40 subunit gene was significantly higher in bleomycin-exposed A/J and C57BL/6J mice but not BALB/c mice at 24 h compared with saline-exposed animals (Figure 4). Expression of the less regulated IL-12p35 subunit mRNA did not demonstrate significant differences in expression in bleomycin exposure compared with saline exposure at any time point (data not shown). IL-4 mRNA was not detectable in any of the mouse lungs at any time point, suggesting the relative absence of IL-4 BAL protein levels was reflective of minimal levels of IL-4 transcription in the lung in response to intratracheal bleomycin administration. Expression of IL-5 was greater at 7 d in bleomycin-exposed animals compared with saline controls, though this difference was statistically significant only in A/J mice (Figure 4). Consistent with prior studies, expression of TGF-1 was higher in bleomycin-exposed animals at 12 h in all strains (ratio of TGF-1:L32 mRNA densitometry units: A/J saline, 0.92 ± 0.03 versus bleomycin, 4.9 ± 0.3, P < 0.0001; C57BL/6J saline, 0.79 ± 0.06 versus bleomycin, 4.6 ± 0.3, P < 0.0001; BALB/c saline, 0.85 ± 0.04 versus bleomycin, 4.3 ± 0.2, P < 0.0001). No strain-dependent differences were noted at other time points as well. Because TGF-1 is a potent inhibitor of IL-12 production and IFN- expression, the lack of strain differences in the expression of TGF-1 suggests that differential production of this cytokine does not play a dominant role in determining IFN- production or susceptibility to bleomycin pulmonary toxicity in this mouse model.

View larger version (36K): [in this window] [in a new window]   Figure 4.   Expression of IL-12p40 and IL-5 mRNA in whole mouse lung after intratracheal saline or bleomycin administration. A/J (A and B), C57BL/6J (C and D), and BALB/c (E and F ) mice were exposed to either intratracheal saline (open bars) or 5 U/kg bleomycin (solid bars) via intratracheal instillation. Concentrations of IL-12p40 (A, C, and E) and IL-5 (B, D, and F ) mRNA from lung tissue obtained at 12 h, 24 h, 3d, or 7 d were determined by RNase protection assay. Each time point represents the mean ± SE of six mice per group. Comparisons are denoted by brackets: *P < 0.005, **P < 0.05.

Effect of Intratracheal Bleomycin Exposure in IFN--Deficient Mice

To directly evaluate a role for IFN- in mediating the pulmonary inflammatory response to bleomycin, we analyzed the response of C57BL/6J mice with a targeted disruption of the IFN- gene (IFN-[/] mice) to intratracheal bleomycin exposure. In response to intratracheal saline administration, no significant differences were noted in either the degree of weight change or the mortality rate at 10 d between IFN-(/) mice and their wild-type controls (Table 1). When exposed to 5 U/kg intratracheal bleomycin, IFN-(/) mice had less weight loss compared with their wild-type controls, though this difference did not achieve statistical significance (P = 0.48). Importantly, IFN-(/) mice experienced significantly reduced mortality at 10 d compared with wild-type control mice (P < 0.0001). Deaths in the saline-exposed mice occurred within 24 h from lack of recovery from anesthesia in one set of mice. To correlate the reduction in weight loss and mortality in IFN-(/) mice with the pathologic response of the lung to intratracheal bleomycin exposure, we examined lung tissue for evidence of inflammation and fibrosis after intratracheal bleomycin administration in these mouse strains (Table 1, Figure 5). Similar to wild-type C57BL/6J mice, IFN-(/) mice demonstrated minimal or no evidence of lung inflammation in response to intratracheal saline administration. Ten days after exposure to 5 U/kg intratracheal bleomycin, IFN-(/) mice demonstrated patchy interstitial and alveolar inflammation, and architectural distortion typically involving approximately 50% of cross-sectional lung area, significantly less than the more confluent inflammatory changes that were present in C57BL/6J wild-type control mice (P < 0.05). Consistent with the greater degree of interstitial pneumonitis observed in the wild-type C57BL/6J mice, a greater amount of total lung RNA was isolated from wild-type C57BL/6J mice compared with IFN-(/) mice after bleomycin exposure (P < 0.005) (Figure 6).

View larger version (124K): [in this window] [in a new window]   View larger version (134K): [in this window] [in a new window]   View larger version (111K): [in this window] [in a new window]   View larger version (121K): [in this window] [in a new window]   Figure 5.   Effect of intratracheal bleomycin administration on lung inflammation in IFN--deficient and wild-type control mice. Shown are photomicrographs of H&E-stained sections of lung tissue. Wild-type C57BL/6J (A and B) and IFN--deficient (C and D) mice were exposed to saline (A and C) or 5 U/kg bleomycin (B and D) of bleomycin by intratracheal instillation and analyzed 10 d later for lung inflammation by histologic techniques. Representative photomicrographs from each experimental group (six to 12 mice/group) are shown. Lower degrees of interstitial pneumonitis and architectural distortion were present in IFN-(/) mice compared with their wild-type counterparts.

View larger version (31K): [in this window] [in a new window]   Figure 6.   Total lung mRNA content of C57BL/6J or IFN-(/) mice after intratracheal saline or bleomycin administration. Mice were exposed to intratracheal saline (open bars) or 5 U/kg bleomycin (solid bars). Total RNA was isolated from one lung collected 10 d after exposure. Results are shown as mean ± SE of six to 12 mice per group. Comparisons are denoted by brackets: *P < 0.0001, **P < 0.005.

Effect of Intratracheal Bleomycin Exposure on Lung Fibrosis

Increased amounts of collagen fibers were visualized by histologic staining with Sirius Red III in both IFN-(/) mice and wild-type control mice 10 d after intratracheal bleomycin exposure compared with saline control mice (Figure 7). To quantitate total collagen content, single lung hydroxyproline (HP) was measured as a surrogate marker of fibrosis. Both IFN-(/) and wild-type C57BL/6J mice demonstrated increased lung HP content compared with their saline control mice after 5 U/kg intratracheal bleomycin (data not shown). Despite observed differences in the pulmonary inflammatory response, there was no significant difference in HP content between IFN-(/) and C57BL/6J control mice after 5 U/kg bleomycin at 10 d (C57BL/6J, 35.1 ± 2.3 µg/lung; IFN-(/), 37.7 ± 1.6 µg/ lung; P = 0.85).

View larger version (158K): [in this window] [in a new window]   View larger version (142K): [in this window] [in a new window]   View larger version (143K): [in this window] [in a new window]   View larger version (144K): [in this window] [in a new window]   Figure 7.   Effect of intratracheal bleomycin administration on histologic fibrosis in IFN--deficient and wild-type control mice. Shown are photomicrographs of lung tissue stained with Sirius Red III for visualization of collagen with fast green FCF counter-staining. Wild-type C57BL/6J (A and B) and IFN-(/) (C and D) mice were exposed to either saline (A and C) or to 5 U/kg bleomycin (B and D) via intratracheal route. Lung tissue was obtained 10 d after treatment, sectioned, and stained for histologic analysis. Representative photomicrographs from each experimental group (six to 12 mice/group) are shown. Red staining material is collagen against a blue-green background. Fibrosis was seen in areas of inflammation in both IFN-(/) mice and C57BL/6J control mice.

Because the lack of difference in collagen content at 10 d could reflect insufficient time for any potential differences in the fibrotic processes to be manifest, we evaluated IFN- (/) and wild-type C57BL/6J mice at 3 wk after exposure to two lower doses of bleomycin, 1.5 and 0.5 U/kg. These doses resulted in longer term survival of the bleomycin- sensitive mice. IFN-(/) mice experienced significantly less mortality and greater weight gain than wild-type control mice at 3 wk in response to the 1.5 U/kg bleomycin dose (Table 2). As with the higher dose of bleomycin, almost all deaths occurred during the height of the inflammatory response phase at 7 to 14 d after bleomycin administration when there is not yet extensive pulmonary fibrosis. Importantly, IFN-(/) mice demonstrated lower degrees of pulmonary inflammation by graded histology compared with C57BL/6J mice receiving 1.5 U/kg bleomycin, consistent with data from the earlier 10-d time point in response to 5 U/kg bleomycin (P < 0.05). Consistent with a more intense inflammatory response in the wild-type mice, single lung total protein was higher in wild-type C57BL/6J mice 3 wk after 1.5 U/kg bleomycin exposure compared with IFN- (/) mice, though this comparison barely failed to achieve statistical significance (C57BL/6J, 7.37 ± 0.84 mg/lung; IFN- (/) 5.08 ± 0.33 mg/lung; P = 0.055). When collagen content was analyzed at 3 wk after 1.5 U/kg bleomycin, we found significantly less single lung HP in IFN-(/) mice than in wild-type C57BL/6J mice (Figure 8). Together, these results indicate that IFN-(/) mice experience lower degrees of fibrosis and lower degrees of lung inflammation after a dose of intratracheal bleomycin that results in significant interstitial pneumonitis yet longer term survival for at least 3 wk. Differences in single lung HP between wild-type C57BL/6J and IFN-(/) mice were not apparent after the lower 0.5 U/kg dose, likely due to the limited inflammatory response seen in both strains of mice in response to this lower dose of intratracheal bleomycin (Table 2, Figure 8).

View larger version (34K): [in this window] [in a new window]   Figure 8.   Single lung hydroxyproline (HP) content 3 wk after intratracheal bleomycin administration. Wild-type C57BL/6J control mice (open bars) and IFN--deficient mice (solid bars) were exposed to either saline, 0.5 U/kg bleomycin, or 1.5 U/kg bleomycin. Single lung HP was determined from whole right lung homogenate from individual animals 3 wk after exposure. Results shown represent mean ± SE of HP isolated from one lung from each of three to eight mice per group. Comparisons are denoted by brackets: *P < 0.0001; **P < 0.005; ***P < 0.05; P > 0.05, not significant.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study provides evidence that IFN- plays a role in mediating bleomycin-induced pulmonary inflammation and fibrosis. A role for IFN- was demonstrated by the presence of elevated IFN- levels in BAL within 24 h after intratracheal bleomycin administration in susceptible strains of mice, as well as the reduced inflammatory response to intratracheal bleomycin exposure in IFN-(/) mice as measured by histologic analysis, total RNA, and total protein content. The attenuated initial inflammatory response to bleomycin in IFN-(/) mice was also associated with lower degrees of weight loss and mortality when compared with wild-type C57BL/6J mice. Together, these results suggest that the inflammatory response to intratracheal bleomycin exposure is amplified by the local expression of IFN- in the lung.

A role for IFN- in bleomycin-induced pulmonary toxicity is consistent with studies that provide evidence that this inflammatory response is T-cell dependent (7). IFN- is the major effector cytokine produced by type 1 CD4⁺ and CD8⁺ T cells and natural killer (NK) cells. Early release of IFN- from type 1 CD4⁺ and CD8⁺ T cells in the lung could theoretically occur after either local activation or systemic activation followed by migration of these preactivated T cells into the lung. Consistent with an upregulation of IFN- expression, IL-12 was found to be expressed in the lungs of the two bleomycin-susceptible mouse strains investigated in this study. Because production of IFN- is potently enhanced by IL-12, the finding that the highly regulated IL-12p40 mRNA is upregulated in susceptible mice in response to bleomycin suggests that IL-12 may play an important regulatory role by enhancing IFN- production in this inflammatory response. T cells may not be the only or major source of IFN- in these murine models given a recent study that found that C57BL/6J SCID mice (which lack both B and T cells) developed pulmonary toxicity comparable to wild-type C57BL/6J mice after bleomycin exposure (13). Interestingly, this study found the bleomycin-susceptible phenotype to be linked with increased lung IFN- mRNA expression in bleomycin-susceptible F1 offspring, suggesting that bleomycin toxicity, although not T-cell dependent, may be dependent on IFN-. These results, together with the results from this present study, suggest that a non-T cell source of IFN-, such as NK cells, may be an important source of IFN- in bleomycin-susceptible mice. NK cells could theoretically be stimulated by damaged lung epithelium or endothelium with subsequent release of IFN-. Alternatively, bleomycin pulmonary toxicity in some T cell-deficient mice may be mediated through IFN--independent mechanisms. Whether the major source of IFN- is NK cells or T cells in bleomycin-susceptible mice, a role for IFN- in amplifying the inflammatory response to bleomycin is clearly demonstrated by the use of genetically altered, IFN--deficient mice in this present study.

This study presents data that is also consistent with a role for IFN- in mediating not only pulmonary inflammation but the degree of pulmonary fibrosis after bleomycin exposure. The difference in fibrotic response between IFN--deficient mice and their wild-type counterparts was dose and time dependent with attenuation of fibrosis readily apparent when using a dose of bleomycin that led to the survival of most mice for at least 3 wk, yet produced significant interstitial pneumonitis. Ten days after exposure to 5 U/kg bleomycin, differences in HP content were not apparent, despite lower degrees of inflammation in IFN-(/) mice, possibly because 10 d was too early for differences in the fibrotic processes to be manifest. Significantly lower amounts of HP were found in IFN-(/) mice 3 wk after 1.5 U/kg bleomycin compared with wild-type mice, consistent with the reduced amounts of inflammation found in these genetically altered mice. Differences in HP content were not seen in response to 0.5 U/kg, likely due to the very limited inflammatory response in both IFN-(/) and wild-type control mice with this exposure dose. Together, these results support the hypothesis that the fibrotic response in bleomycin-susceptible mice is dependent on the degree and persistence of pulmonary inflammation and injury.

The mechanisms through which IFN- modulates the inflammatory and fibrotic response to bleomycin are likely to be related to its ability to enhance the production of other proinflammatory mediators with subsequent lung injury. For example, IFN- upregulates and synergizes with the proinflammatory cytokine tumor necrosis factor (TNF)- (22). The importance of TNF- in bleomycin- induced pulmonary fibrosis has been established by early seminal studies of the protective effects of TNF- neutralization and, more recently, by the attenuated response to bleomycin in TNF--receptor deficient mice (23, 24). Other studies have clearly demonstrated that the inflammatory and fibrotic response to bleomycin exposure is related to oxidative stress by showing, for example, that the administration of antioxidant agents or dietary supplements attenuates bleomycin-induced lung injury and fibrosis (25, 26). Because inducible nitric oxide synthase and other mediators of oxidative stress are upregulated in bleomycin-susceptible mice, the general ability of IFN- and TNF- to synergistically upregulate oxidation-induced tissue injury may be central to the pathogenesis of bleomycin inflammation and fibrosis (27, 28). We hypothesize that the early upregulation of IFN- after bleomycin exposure, at a time when there is concomitant upregulation of other proinflammatory cytokines such as TNF- and IL-6, results in an important amplification of the inflammatory response and subsequent lung injury from this agent. Because many studies including the present one have found a close correlation between the degree of inflammation and fibrosis, an absence of IFN- production soon after the injury stimulus might be expected to result not only in a reduced inflammatory response but also subsequently a reduced fibrotic response. Similar to this hypothesized role for IFN- in bleomycin-induced pulmonary fibrosis, the fibrotic response to intratracheal silica (another lung injury-inducing agent) exposure in rodents has also been shown to be dependent on IFN-, IL-12, and TNF- (29).

The current study stands in contrast, at least superficially, to the well-documented antifibrotic effects of IFN-. IFN- has been shown to directly suppress fibroblast collagen production in vitro and can suppress TGF- expression, a cytokine that promotes new collagen synthesis and deposition (30, 31). In the Th2-dependent Schistosoma egg antigen (SEA) mouse model of granulomatous inflammation and fibrosis, administration of IFN- or IL-12 suppresses the fibrotic response, suggesting that Th1 cytokines have counter-regulatory and antifibrotic effects in this model system (32). Importantly, previous reports have found that in bleomycin-susceptible mouse strains, repeated administration of IFN- or poly-ICLC (polyinosinic-polycytidylic acid complexed with poly-L-lysine), an inducer of IFN- expression, actually retards bleomycin-induced pulmonary fibrosis (35, 36). All of these experimental models involved the chronic, repeated upregulation of IFN- from either repeated systemic administration of IFN- itself or via the administration of poly-ICLC or IL-12, both of which induce IFN- production. In these models, IFN- is present in abnormally high concentrations at multiple time points distant from the initial bleomycin administration, time points that are not characterized by a similar concomitant upregulation of other proinflammatory mediators. These reports indicate that under conditions of chronic administration or pharmacologic induction of IFN-, the net direct antifibrotic effects of IFN- outweigh its potential to increase fibrosis from enhancing lung injury after bleomycin administration. The systemic administration of IFN- or its inducers may also serve to decrease its local toxic proinflammatory effects and allow direct antifibrotic effects to dominate in these models.

In the current study, a role for IFN- in mediating bleomycin-induced pulmonary inflammation and fibrosis was inferred by the attenuation of these responses in IFN-(/) bleomycin-susecptible strains. Under conditions of the current study, IFN- was upregulated within 24 h in response to intratracheal bleomycin exposure in bleomycin-sensitive mice at a time when there was significant upregulation of other proinflammatory mediators. In support of an important role for early IFN- production in bleomycin- induced pulmonary interstital pneumonitis, we have preliminary data that intratracheal IFN- administered 24 h before 1.5 U intratracheal bleomycin significantly enhances the inflammatory lung response in genetically "resistant" BALB/c mice as measured by graded histology (unpublished data). Whether IFN- pretreatment results in significant fibrosis and the dose/timing dependency of the altered inflammatory response is under investigation.

In contrast to IFN-, we found little evidence that IL-4 is upregulated in the inflammatory response to bleomycin in our mouse model. Other models, such as the SEA mouse model of granulomatous inflammation and fibrosis, have demonstrated a role for Th2 cytokines (IL-4 and IL-5) in the pulmonary fibrotic response to this parasite (32). This observation is supported by previous studies that have found that IL-4 mRNA expression is not upregulated in bleomycin-susceptible mice (37). Consistent with these findings, a recent preliminary report provides strong evidence that bleomycin-induced pulmonary fibrosis is not dependent on IL-4 by demonstrating that IL-4-deficient (knock-out) mice are not less susceptible to bleomycin than wild-type mice, and that IL-4-overexpressing (transgenic) mice are not more susceptible to the effects of bleomycin (38). In contrast to IL-4, we found that IL-5 mRNA expression was increased in susceptible strains 7 d after bleomycin exposure, supporting previous studies that found that this cytokine was upregulated in bleomycin-induced pulmonary fibrosis (20). Increased numbers of eosinophils have been observed in BAL and lung tissue sections after bleomycin exposure in mice, supporting a proposed role for IL-5 in this response. These studies are consistent with a role for IL-5 but not IL-4 in bleomycin-induced pulmonary fibrosis. How IFN- interacts with IL-5 in bleomycin-induced pulmonary fibrosis requires further study.

Our findings that IFN- may play a profibrotic role under certain conditions in experimental bleomycin-induced pulmonary toxicity have possible implications for human disease. First, these observations suggest that IFN- could play a profibrotic role in the lung fibrosis that occurs in diseases associated with enhanced chronic production of IFN- such as sarcoidosis, chronic beryllium disease, silicosis, and hypersensitivity pneumonitis (29, 39). In these diseases, chronic lung injury secondary to the proinflammatory effects of IFN- could theoretically result in progressive lung fibrosis (43). Second, although IFN- may have antifibrotic effects through direct suppression of fibroblast type I and III collagen synthesis and through the suppression of TGF- expression, the clinical use of IFN- in pulmonary fibrotic disorders may be a double-edged sword, with the potential for toxicity related to enhancing proinflammatory processes (44). A recent study found that repeated administration of IFN- together with a low dose of prednisolone was superior to prednisolone alone in preserving lung function in patients with mild to moderate idiopathic pulmonary fibrosis (45). The potentially negative, proinflammatory effects of IFN- may have been mitigated by the concurrent administration of low-dose prednisolone in this human trial, allowing the antifibrotic effects of pharmacologically administered IFN- to dominate. Although the long-term administration of IFN- may favor a net antifibrotic result in diseases where fibrotic processes dominate over inflammatory processes, more extensive clinical trials are required to assess the appropriate role of this therapeutic modality in the general treatment of lung fibrosis.