Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Acute lung injury is characterized by endothelial and epithelial cell injury that subsequently leads to pulmonary edema, alveolar collapse, and insufficient gas exchange (1). In humans, acute lung injury can be initiated by a diverse array of factors, including infection, aspiration, trauma, and inhaled irritants. In animal models, acute lung injury is also initiated by lipopolysaccharide (LPS), hyperoxia, embolism, oleic acid, and ozone (2, 3). Particulate nickel sulfate found in cigarette smoke, occupational environments, and ambient particulate matter (4, 5) also produces acute lung injury in mice (6, 7). Inbred mouse strains were found to vary in their sensitivity to acute nickel-induced lung injury with A/J (A) mice being sensitive and C57BL/6J (B6) being resistant (7). Analysis of lung gene expression using 8,734 sequence-verified murine cDNAs revealed clusters of coregulated genes, including decreased surfactant protein (SP)-B following nickel exposure (8). A subsequent genome-wide analysis with 307 nickel-exposed backcross mice, generated from resistant (B6xA) F₁ and susceptible A mice, revealed linkage to a quantitative trait locus (QTL) on chromosome 6 (9). Candidate genes in the QTL interval include SP-B and transforming growth factor- (TGF-).

TGF- is a member of the epidermal growth factor (EGF) peptide family, which are ligands for the EGF receptor (EGF-R) and share many similar biologic properties with EGF (10). Found in several epithelial and mesenchymal tissues, TGF- is synthesized as a 160-amino acid precursor polypeptide with a mature 50-amino acid TGF- peptide that is released at sites of injury through proteolytic cleavage by specific elastase-like enzymes (10). Although the precise physiologic role of TGF- and other EGF-R ligands in tissue repair are not completely understood, topical administration of either recombinant TGF- or EGF accelerated epidermal regeneration in partial thickness skin burns in pigs or humans (13), and accelerated healing of corneal ulcers in humans (16). In rodents, parenteral pretreatment with TGF- or EGF decreased irritant-induced gastric mucosal damage (17) and TGF- knockout mice develop increased dextran sulfate-induced colitis compared with wild-type mice (20).

TGF- also may play a role in repair of an injured lung. TGF- significantly accelerated re-epithelialization of denuded areas of confluent monolayers of primary rat type II cells and the addition of a monoclonal antibody to TGF- with serum decreased the rate of re-epithelialization (21). To examine the effects of TGF- during acute lung injury in vivo, we utilized transgenic mice overexpressing human TGF- in the distal lung epithelium under control of the human SP-C promoter (22). TGF- transgenic lines, exposed to ultrafine polytetrafluoroethylene (PTFE) particles, survived significantly longer with less inflammation and pulmonary edema than nontransgenic mice, indicating protection of the lung from inhalant exposure (23). Because the QTL studies identified TGF- as a candidate gene in nickel resistance, we used the TGF- transgenic models to test the role of TGF- in protection against nickel-induced acute lung injury. Enhanced survival of the transgenic mice and reduced inflammation and pulmonary edema, with preservation of SP-B levels in the highest expressing line, supports the hypothesis that localized production of TGF- in the lungs protects against acute lung injury.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Experimental Design

Lung injury in mice was induced by exposure to submicrometer nickel aerosol generated as previously described (7). To determine the effect of increased TGF- levels on nickel-induced lung injury, four transgenic lines expressing different amounts of human TGF- (hTGF-) in the lung and strain-matched nontransgenic mice were exposed continuously to nickel for up to 14 d and survival time recorded. To characterize nickel-induced lung injury, pulmonary histology, lung wet-to-dry ratios, bronchoalveolar lavage (BAL) fluid cell differential and protein, and lung homogenate inflammatory cytokine, SP-B and SP-D levels were compared between TGF- transgenic lines and age- and strain-matched nontransgenic mice before exposure or up to 72 h of continuous nickel exposure.

Transgenic Mice and Southern Blot Analysis

Methods for generating transgenic mice expressing hTGF- under control of the human SP-C 3.7 kb promoter enhancer sequence were previously described (22). Transgenic mice were identified by a diagnostic 1.4-kb band on genomic Southern blots of Pst1 digested genomic tail DNA. Transgenic mice and nontransgenic control mice were age- and strain (FVB/N)-matched. Levels of hTGF- for each line are based on historic controls (24). Line 6,108 (388 pg/ml hTGF- or 1.3-fold increase normalized to nontransgenic controls to correct for endogenous mouse TGF-) express the lowest level of hTGF-, and demonstrate mild alveolar emphysema with no fibrosis. Line 4 (682 pg/ml) express 2.3 times nontransgenic mice and are phenotypically identical to line 6,108 mice with mild alveolar emphysema and no fibrosis. Line 2 (899 pg/ml) express 3.1 times nontransgenic mice and develop significantly greater alveolar emphysema than lines 6,108 and 4 with minimal to no pleural fibrosis. Line 28 mice (1,247 pg/ml) express 4.3 times nontransgenic mice and consistently develop emphysema greater than all the transgenic lines, along with pleural fibrosis. Mice were handled in accordance with protocols approved by the Institutional Animal Care and Use Committee of the Children's Hospital Research Foundation and the University of Cincinnati Medical Center (Cincinnati, OH).

Aerosol Generation

Mice were exposed to 70 µg Ni/m³ nickel in stainless steel cages placed inside of a 0.32 m³ stainless steel inhalation chamber. Nickel aerosol (mass median aerodynamic diameter = 0.22 µm, geometric standard deviation [_g] = 1.85) was generated from a solution of nickel sulfate hexahydrate (NiSO₄ × 6H₂O; Sigma, St. Louis, MO) using a modified Collison three-jet nebulizer (3.5 liters/ min) (BGI Incorporated, Waltham, MA) placed inside a narrow wall glass tube (inner diameter, 24 mm). The particle number concentration and particle size were determined using a differential mobility analyzer consisting of an electrostatic classifier (Model 3071A; Thermo-Systems, Inc [TSI], St. Paul, MN), a condensation nucleus counter (Model 3022A; TSI), and scanning mobility particle sizer fast-scanning software (TSI). The chamber nickel concentration was determined using the dimethylglyoxime method (25). Samples of the chamber atmosphere were collected with two midget impingers (Ace Glass, Inc., Vineland, NJ) in series, each containing 10 ml distilled water (flowrate: 11.3 l/min). Collected samples were mixed with a solution containing 1 M HCl, 0.2 M Br, 12 M NH₄OH (Fisher, Fair Lawn, NJ), 1% dimethylglyoxime (Sigma), and 95% ethanol. Absorbance was measured at 445 nm (Model DU-64; Beckman, Fullerton, CA). During nickel exposures, mice had unlimited access to food and water.

Pulmonary Histology

Nickel-exposed mice dedicated for histological analysis were removed from the chamber at selected times, injected with pentobarbital sodium (50 mg/kg; Nembutal; Abbott Laboratories, N. Chicago, IL), and exsanguinated. A cannula (inner diameter, 0.58 mm) was inserted into the trachea and the lungs were instilled in situ (1 min, 25 cm H₂O) with phosphate-buffered formaldehyde (pH = 7.1), removed, and immersed in fixative (24 h). The lung was washed with phosphate-buffered saline (PBS), dehydrated through a series of graded ethanol solutions (30-70%), and processed into paraffin blocks (Hypercenter XP; Shandon Scientific, Pittsburgh, PA). Paraffin-embedded tissues were sagittally sectioned (5 µm) and placed on Polysine glass slides (Erie Scientific, Portsmouth, NH). Each specimen was placed in xylene, rehydrated with graded alcohol washes and PBS, and stained with hematoxylin and eosin.

Lung Wet-to-Dry Weight Ratios and Lung Nickel Retention

At selected times mice were removed from exposure and anesthetized, the chest cavity was opened, and the heart-lung block was removed. The lungs were rinsed with PBS, trimmed (heart, connective tissue, esophagus), blotted dry with gauze, and the wet weight was recorded. Lungs were then dried in a plastic desiccator containing silica gel (Sargent-Welch, Skokie, IL) for up to 2 wk before recording dry weights.

To compare pulmonary nickel retention between nontransgenic and line 28 mice, lungs were obtained before or 72 h after exposure to 70 µg Ni/m³ NiSO₄, wet weight determined and dried as above. Each sample was placed in a 1.25-ml vial and sent to the University of Wisconsin Nuclear Reactor Laboratory (UWNRL) for neutron activation. Vials with known amounts of nickel standards or lung samples from exposed mice were passed through the nuclear reactor and irradiated for 4 h (using an irradiation position designated as Whale 8). Beginning 375 h later, each vial was counted for 2 h at zero cm (vertical axis) with a 170 cm³ Intrinsic Germanium Detector (GEM-40190; Ortec Products, Perkin Elmer Instruments, Wellesley, MA) coupled to a multichannel analyzer (PCAII PC-based). Peak areas were computed using a UWNRL compiled basic program, NAACALC. Fluxes were calculated based on nickel content in known standards and samples had counter dead times of < 0.20%.

BAL Fluid Cell Differential and Protein

After mice were killed, the trachea was cannulated and the lungs lavaged three times with 1 ml Ca²⁺, Mg²⁺-free Hanks' balanced salt solution (HBSS: 137 mM NaCl, 5.4 mM KCl, 0.44 mM KH₂PO₄, 0.34 mM Na₂HPO₄, 4.2 mM NaHCO₃, and 5.6 mM glucose). BAL fluid was pooled and immediately cooled to 4°C. Cells from a 250-µl sample of BAL fluid were placed on microscopic slides by centrifugation (150 × g for 4 min, Cytospin 3; Shandon Scientific) and differential cell counts were performed following Diff-Quik staining (Baxter Diagnostics, McGraw Park, IL) with 200 cells/slide counted. The rest of the lavage fluid was centrifuged (1,300 rpm for 10 min, 4°C) and the supernatant was decanted. The cell pellet from each lavage was resuspended in 1 ml HBSS. Total cell counts were determined with a hemocytometer. Total BAL protein was measured from the BAL supernatant as previously described using the bicinchoninic acid method with bovine serum albumin for standards (26).

Cytokine and SP-B Levels

At selected times mice were removed from nickel exposure, anesthetized, and the lungs were removed and homogenized in 2 ml PBS, pH 7.2. The homogenate was centrifuged at 1,500 × g and the supernatant removed and stored at 70°C. Interleukin (IL)-1, IL-6, and macrophage inflammatory protein-2 (MIP-2) were determined using quantitative murine sandwich enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN) according to the manufacturer's directions. All plates were read on a microplate reader (Molecular Devices, Menlo Park, CA) and analyzed with the use of a computer-assisted analysis program (Softmax; Molecular Devices). Only assays having standard curves with a calculated regression line value r² > 0.95 were accepted for analysis.

Total SP- B concentration was determined with ELISA using a rabbit polyclonal antibody directed against mature bovine SP-B (27). For SP-D concentration, plates were coated with a polyclonal immunoglobulin G (IgG) fraction of rabbit anti-mouse SP-D (28). The second antibody is a guinea pig polyclonal IgG fraction of guinea pig anti-mouse SP-D. This antibody is previously absorbed with solid phase lung from SP-D-deficient mice to block nonspecific binding. The reporter antibody is a peroxidase-conjugated anti-guinea pig IgG. Each assay plate included a standard curve generated with purified bovine SP-B or purified mouse SP-D obtained using column fractionation and gel filtration as previously described (27, 28).

Statistics

Statistics for survival were performed using the LIFETEST procedure in SAS for Windows version 8.2. Multiple pairwise comparisons among groups were performed separately for each pair of groups, with the level of significance set to account for the multiple comparisons. To assess nickel-induced changes in lung wet-to-dry weight ratios, protein, polymorphonuclear leukocytes in lavage fluid, cytokines, and SP-B and SP-D, statistical analysis was performed using a two-way analysis of variance followed by a Student-Newman-Keul test of significance. The factors for each analysis were transgenic line (typically nontransgenic versus line 28) and exposure (exposed versus unexposed). Statistical significance for all comparisons of means was accepted at P < 0.05 and values are presented as the means ± SE.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Survival

Survival varied among the different transgenic-mouse lines, with survival correlating with increasing levels of hTGF- expression (Figure 1). The survival times of nontransgenic (FVB/N strain matched) mice (68 ± 3 h) and line 6,108 mice (76 ± 2 h) were not significantly different. In contrast, all line 4 mice (87 ± 1 h) survived longer than nontransgenic and line 6,108 mice (P < 0.05). Line 2 and line 28 mice survived the longest, with 50% of line 2 mice and 60% of line 28 mice surviving the entire 14-d exposure period, and both groups survived significantly longer than nontransgenic and lines 6,108 and 4 (P < 0.01).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Survival time of nontransgenic and TGF- mice exposed to 14 d of continuous nickel aerosols (70 µgNi/m3). There were no significant differences between nontransgenic (n = 10) and line 6,108 mice (n = 10). Line 4 mice (n = 10) survived significantly longer than nontransgenic and line 6,108 mice (P < 0.05). Lines 2 (n = 8) and 28 (n = 10) survived significantly longer than nontransgenic and lines 6,108 and 4 (P < 0.01).

Histology

Within 24 h of exposure, mild perivascular edema began to develop in the lungs of nontransgenic mice, although the lungs of line 28 mice had minimal changes (arrowheads, Figure 2). By 48 h, perivascular widening consistent with edema became more pronounced in nontransgenic mice with evidence of inflammatory cell infiltration. In contrast, line 28 mice did not demonstrate evidence of edema or inflammation. By 72 h, extensive widening of the perivascular space was detected in nontransgenic mice with apparent extension into the adjacent alveoli (arrow, Figure 2). Again, line 28 mice remained unaffected at 72 h with little or no evidence of pulmonary edema or inflammation. Lungs from line 4 mice demonstrated edema and inflammatory cell infiltration at 72 h similar to nontransgenic mice, whereas lungs from line 2 were much less effected, with focal areas of mild perivascular edema and inflammatory cell infiltration.

View larger version (110K): [in this window] [in a new window]   Figure 2.   Lung histology after 70 µgNi/m3 nickel exposure. Lungs from transgenic (line 28 TGF-) and age- and strain-matched nontransgenic mice were inflation-fixed and stained with hematoxylin and eosin. Sections are representative of four individual mice sacrificed at each time point. Arrowheads point to widening of perivascular space in nontransgenic mice and unaffected vessels in line 28 mice. Arrow points to extensive perivascular widening and septal thickening in nontransgenic mice 72 h into nickel exposure. Original magnification was approximately ×25 and ×50 for the 72 h nontransgenic panel.

Lung Wet-to-Dry Weight Ratios and Nickel Retention

Before exposure, lung wet-to-dry weight ratios of nontransgenic mice did not differ from those of line 28 mice. With nickel exposure, wet-to-dry weight ratios of lungs from nontransgenic mice increased above control and were significantly increased by 72 h. This effect was attenuated in line 28 mice compared with nontransgenic mice (Figure 3).

View larger version (14K): [in this window] [in a new window]   Figure 3.   Lung wet-to-dry weight ratios after 70 µgNi/m3 nickel exposure. Open bars, nontransgenic mice; solid bars, line 28 transgenic mice. Values are means ± SE; n = 4-5 mice at each time. *P < 0.05 compared with line 28 mice at the 72 h time period.

The amount of nickel present in the lungs before exposure was below the limit of detection (0.4 µg Ni/gm tissue) for either nontransgenic or line 28 mice; however, the amount retained at 72 h increased to 0.7 ± 0.2 µg Ni/gm tissue in nontransgenic mice compared with 2.1 ± 0.3 µg Ni/gm tissue in line 28. Paradoxically, despite the lack of injury, the increased nickel lung burden in Line 28 following exposure was significantly greater than nontransgenic controls (P < 0.001).

BAL Fluid Cell Counts, Differential, and Protein

Before exposure, few neutrophils (< 1%) were detected in the BAL of nontransgenic or all TGF- transgenic lines, and there were no differences in the total cell number. At 72 h, total cell counts increased significantly from time zero in all groups of mice, with no differences noted between nontransgenic and TGF- transgenic lines. At 72 h the neutrophil percentage increased in all groups of mice. However, neutrophils were decreased in the higher-expressing transgenic lines 2 and 28 as compared with nontransgenic controls (Figure 4). Similarly, at 72 h BAL fluid protein levels were elevated in all groups of mice compared with line-matched control values, yet decreased BAL protein was noted in the higher expressing lines (lines 2 and 28) as compared with nontransgenic control (Figure 5).

View larger version (25K): [in this window] [in a new window]   Figure 4.   Percentage of neutrophils in bronchoalveolar lavage fluid after 72 h of 70 µgNi/m3 nickel. Values are means ± SE. (n = 4-10 mice/line.) *P < 0.05 compared with nontransgenic mice.

View larger version (27K): [in this window] [in a new window]   Figure 5.   Bronchoalveolar lavage fluid protein levels after 70 µgNi/m3 nickel.exposure. Values are means ± SE. (n = 4-8 mice/line). *P < 0.05 compared with nontransgenic mice.

Cytokine Levels

Before exposure IL-1, IL-6, and MIP-2 levels in lung homogenates from nontransgenic and line 28 mice were similar (Figure 6). After 24 and 48 h, small but insignificant increases in cytokine levels for both nontransgenic and line 28 mice were seen. However, by 72 h all three cytokine levels increased markedly in nontransgenic mice, whereas cytokines remained near control levels in line 28 mice. There was a trend toward decreased cytokine levels in the intermediate lines compared with nontransgenic mice 72 h after nickel exposure, although differences did not reach statistical significance. Similar values for IL-1, IL-6, and MIP-2 were noted between line 6,108 and nontransgenics. For line 4 mice IL-1 was 18% less than nontransgenic values, 30% less for IL-6 values, and 14% less for MIP-2. For line 2 mice IL-1 was 41% less than nontransgenic values, 21% less for IL-6 values, and 24% less for MIP-2.

View larger version (14K): [in this window] [in a new window]   Figure 6.   Lung homogenate cytokine levels after 24, 48, and 72 h of 70 µgNi/m3 nickel exposure. (A) Interleukin-1; (B) interleukin-6; (C) macrophage inflammatory protein-2. Open bars, nontransgenic mice; solid bars, line 28 transgenic mice. Values are means ± SE. (n = 5 mice/line). *P = 0.01 compared with line 28 mice at the 72 h time period.

Surfactant Proteins

Before exposure, levels of SP-B in line 2 and 28 mice were almost half the amount (54%) found in the nontransgenic FVB/N strain (Table 1). Lines 6,108 and 4 had slightly lower, but equivalent levels of SP-B compared with nontransgenic mice. After 72 h of NiSO₄ exposure, SP-B levels decreased from time zero in all lines except line 28, where values were slightly increased from time zero.

In contrast to SP-B, line 28 and nontransgenic mice have similar levels of SP-D at time zero (480 ± 31 ng/ml for nontransgenic; 529 ± 55 for line 28). Following 72 h of exposure, SP-D decreased significantly from baseline in both groups, falling to 350 ± 40 ng/ml in nontransgenic mice (27% from time zero), and to 300 ± 26 ng/ml in line 28 mice (43% from time zero).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study demonstrates differences between TGF- transgenic and nontransgenic mice in the development of acute lung injury following NiSO₄ exposure. Mice from the three highest-expressing TGF- transgenic mouse lines survived significantly longer than the nontransgenic strain-matched mice. Characterization of mice at selected time points demonstrated reduced evidence of pulmonary edema, inflammatory cytokine production, cellular infiltration, and preservation of SP-B levels in the highest-expressing transgenic line (line 28). These findings support the hypothesis that human TGF- expression in transgenic mice protects the lung from injury by attenuating pulmonary edema and inflammation.

Survival time, among transgenic mice continuously exposed to submicrometer NiSO₄, increased with increasing levels of transgene-derived TGF-. Survival of the lowest expressing line, line 6,108 (388 pg hTGF-/ml), was slightly longer than nontransgenic mice but did not differ statistically. However, line 4 mice (682 pg hTGF-/ml) survived significantly longer with all line 4 mice surviving longer than the longest surviving nontransgenic or line 6,108 mouse. The next highest-expressing transgenic mice, line 2 (899 pg hTGF-/ml), survived longer than lower-expressing transgenic lines, and the highest-expressing line, line 28 (1,247 pg hTGF-/ml), had the best survival outcome. In both line 2 and line 28, 50% or more mice survived the entire exposure period. Thus, TGF- levels between 900 and 1,250 pg hTGF-/ml in lung homogenates, as produced by lines 2 and 28 mice, respectively, markedly increased the length of survival from inhaled nickel injury. Similar increased survival was detected among the transgenic lines exposed to PTFE inhalation, which caused an acute lung injury within 8-12 h of exposure in FVB/N control mice (23). Although the two studies used different types of particulates, the present study indicates that TGF- expression in transgenic mice extends protection into a prolonged period of inhalation injury.

As both PTFE and nickel studies demonstrate increased survival in higher-expressing TGF- transgenic mice, we further characterized indices of pulmonary edema, inflammation, and SP-B in nontransgenic and transgenic lines to begin exploring underlying mechanisms explaining this survival advantage. Line 28 mice demonstrated reduced pulmonary edema as assessed by histology, lung wet-to-dry weight ratios, and BAL protein levels, with attenuated edema noted on lung histology and BAL protein for line 2 mice. This attenuation of pulmonary edema following nickel exposure in TGF- transgenic mice is similar to findings in line 28 mice exposed to ultrafine PTFE (23). There is experimental evidence that EGF-R ligands increase fluid clearance across the alveolar epithelial membrane. Folkesson and coworkers (29) reported that instillation of TGF- rapidly (within 1-4 h) increased alveolar fluid clearance in anesthetized ventilated rats through a mechanism involving sodium channel activation in alveolar Type II cells. In rat alveolar epithelial cells, EGF increased active Na⁺ absorption via direct effects on Na⁺ pump subunit mRNA expression and protein synthesis which led to increased numbers of functional Na⁺ pumps in the basolateral cell membranes (30). The tested transgenic lines in our studies express TGF- in Type II cells, and pulmonary remodeling is mediated through type II cell signaling (31). Taken together, it is possible that reduced edema in the higher-expressing transgenic lines occurs through TGF- signaling in type II cells, resulting in increased alveolar fluid clearance.

The extent of lung inflammation was attenuated in line 28 as indicated by the reduced BAL neutrophils and lung homogenate cytokines 72 h into nickel exposure. Before exposure, the lung homogenate levels of IL-1, IL-6, and MIP-2 were nearly identical between nontransgenic and line 28 mice and not at proinflammatory levels. During nickel exposure, nontransgenic mice had increased neutrophil and cytokine levels by 72 h, differing significantly from the line 28 transgenic mice. Significantly reduced BAL neutrophils were also noted in line 2 mice, with an attenuated response in proinflammatory cytokine levels in lines 4 and 2. These findings suggest that higher levels of TGF- expression in the lung attenuate the induction of neutrophil influx and proinflammatory cytokine response in the lung following nickel exposure. Although the mechanism has not been reported, TGF- preconditioning downregulates bradykinin-induced acute prostaglandin production in human colonic epithelial lines, suggesting that TGF- has a role in limiting the acute inflammatory responses induced by bradykinin (32). The present study suggests an additional role of TGF- in modification of selective components of the inflammatory cascade.

Expression of TGF- may also protect against nickel-induced lung injury through maintenance of SP-B levels, although this preservation of SP-B was observed only in line 28 mice and not in other transgenic lines. Before exposure, both lines 2 and 28 mice demonstrate significantly reduced levels of SP-B compared with nontransgenic mice. This reduced SP-B in lines 2 and 28 may reflect a decrease in the total numbers of type II cells in these mice as a result of the alveolar emphysema (24). Following nickel exposure SP-B levels fell from time zero in all mice except line 28 mice. The decrease in SP-B in nontransgenic mice supports previous studies demonstrating reduced SP-A, -B, and -C mRNA levels in mice after nickel exposure (8). Produced only in the lung, SP-B is a small hydrophobic peptide that enhances surfactant absorption and spreading and is necessary for the surface tension reducing properties of pulmonary surfactant. Gene targeted mice, lacking SP-B, succumb shortly after birth to respiratory failure, whereas heterozygous SP-B gene-targeted mice, containing 50% of wild-type SP-B levels, survive, suggesting that a loss of 50% SP-B can be tolerated in normoxic conditions (33). However, heterozygous SP-B mice developed increased lung permeability and BAL protein leakage, as well as histological evidence of pulmonary edema, hemorrhage, and inflammation compared with wild type mice following 72 h of hyperoxia (34). The administration of surfactant with the active SP-B peptide to heterozygous SP-B mice prevented bronchoalveolar protein increases and histologic abnormalities caused by oxygen-induced injury (35). Although heterozygous gene-targeted SP-B mice have not been studied during nickel exposure, the preservation of SP-B levels in line 28 mice following nickel exposure may contribute to the increased survival and reduced lung injury. However, this mechanism does not completely explain the increased survival and attenuated lung injury noted in lines 4 and 2, where SP-B levels 72 h into exposure fell significantly from baseline.

In contrast to SP-B, SP-D levels in nontransgenic and line 28 mice were nearly equivalent at time zero. SP-D fell significantly from time zero at the 72-h point for both groups of mice but there were no differences between either group. SP-D, along with SP-A, are members of a family of host defense lectins, called collectins, and are important components of innate immunity against microbial pathogens in the lung (36). The similar decrease of SP-D in the TGF- transgenic compared with nontransgenic mice suggests that SP-D levels do not play a role in protection of the lung from inhaled nickel injury.

We have previously reported that line 28 TGF- transgenic mouse lungs are remodeled with emphysema and fibrosis (22, 24). We cannot discount that the remodeled lungs alone may contribute to the reduced lung injury observed in our transgenic lines. The remodeling in lines 2 and 28 may have altered the susceptibility of subcell populations in the alveoli and airways through mechanisms not yet understood. However, three lines of evidence suggest that TGF- attenuates lung injury independent of lung remodeling. Although nickel particle deposition and clearance were not directly measured, lung retention of nontransgenic and line 28 mice suggests that both groups inhaled similar amounts of nickel, with line 28 mice retaining more nickel than nontransgenic mice. Thus, a decrease in nickel retention does not explain the observed difference in susceptibility. Second, the higher-expressing line 4 mice survived significantly longer than the lower-expressing 6,108 mice, despite a lack of detectable histologic, morphologic, or physiologic differences between these two lines. The increased survival of line 4 mice would thus suggest that TGF- provides a survival benefit independent of any lung remodeling. Third, previous studies in animal models of emphysema have demonstrated that emphysematous lungs are no more or less susceptible to injury induced by irritants than normal lungs. Emphysema experimentally induced in rats, hamsters, and guinea pigs demonstrate no changes in survival, inflammation, or pulmonary function testing, compared with normal animals when exposed to pulmonary insults including hyperoxia, ammonium sulfate, or olefin-ozone-sulfur dioxide (37).

Our data suggests a role for TGF- in protecting the lung from acute lung injury and is supported by other experimental and clinical studies. Increased TGF- was identified in alveolar epithelial and septal cells of rat lungs following bleomycin injury (40). Naphthalene injury induced increased EGF and TGF- in proliferating bronchiolar epithelial and interstitial cells (41). Asbestos fibers increased TGF- mRNA and protein in bronchiolar-alveolar duct cells in known sites of asbestos-induced injury (42). Although these studies associate increased release of TGF- following injury but do not assign a specific role, other work suggests TGF- is active in repairing the injured lung. In rat alveolar type II cells, mechanically denuded regions of confluent cell monolayers re-epithelialized more rapidly in the presence of exogenously supplied TGF-. Moreover, a neutralizing antibody to TGF-, in the absence of exogenous peptide, reduced the rate of repair (21). Enhanced EGF-R was detected in injured regions of bronchial epithelium of biopsies from patients with asthma. In the same study, inhibitors of EGF-R signaling blocked the repair of wounds in monolayers of bronchial epithelial cells (43). Administration of EGF to rats following pneumonectomy enhanced postpneumonectomy lung growth compared with untreated rats (44). These studies provide strong evidence for the involvement of EGF family, including TGF-, in the repair of pulmonary injury.

In clinical studies TGF- was identified in epithelial and interstitial cells of patients with interstitial fibrosis or cystic fibrosis (45, 46). Soluble TGF- was increased in pulmonary edema fluid from patients with acute lung injury one day following lung injury, but not increased in fluid from patients with hydrostatic edema (47). TGF- levels were also significantly increased in BAL fluid of patients with adult respiratory distress syndrome and idiopathic pulmonary fibrosis compared with normal volunteers (48). The latter two studies linked increased TGF- with diseases causing pulmonary permeability, such as occurred with particulate injury. Taken together, the experimental and clinical data support the hypothesis that TGF- is released following lung injury. While the precise role of TGF- following injury remains to be elucidated, our data and the results from other laboratories suggest TGF- may be involved in attenuating acute lung injury.

In summary, the present study demonstrates that TGF- transgenic mice, constitutively expressing human TGF-, have significantly greater survival in a dose-dependent manner when exposed to nickel particulate. Lung edema, inflammation, and levels of proinflammatory cytokine expression are significantly reduced 72 h into nickel exposure in line 28 mice, the highest-expressing transgenic line, whereas SP-B levels were preserved compared with nontransgenic mice. The survival patterns and reduced lung injury in transgenic mice exposed to nickel is similar to that observed in TGF- transgenic mice exposed to PTFE inhalation and suggest a role for TGF- in protecting the lung from acute lung injury.