INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Idiopathic pneumonia syndrome (IPS) remains a highly fatal complication in recipients of allogeneic bone marrow transplantation (BMT) (1). Although risk factors associated with IPS generation in humans include the severity of graft-versus-host disease (GVHD) (2), and intensity of conditioning regimens (3), the pathophysiological mechanisms of IPS are unknown. In a murine model of IPS that closely simulates the human condition (4), we reported the generation of large amounts of proinflammatory cytokines and reactive oxygen and nitrogen species during the course of lung dysfunction associated with infusion of allogeneic T cells, cyclophosphamide (Cy) conditioning, and irradiation (5). The most severe form of lung dysfunction was observed in irradiated mice injected with both Cy and allogeneic T cells and was associated with the production of peroxynitrite (ONOO), a potent oxidant and nitrating species (6). ONOO generation most likely resulted from the simultaneous production of nitric oxide (·NO) by T cell-activated macrophages and superoxide (O₂·) generated in the lungs following exposure to Cy (7).

The preinsult intratracheal and systemic administration of keratinocyte growth factor (KGF), a specific mediator of epithelial cell proliferation and differentiation, is protective in experimental models of lung injury including hyperoxia (8), bleomycin, and -irradiation (9), but the protective mechanisms are incompletely understood. In our murine BMT model, we reported that the systemic pre-BMT administration of recombinant human KGF prevented permeability edema, and suppressed the allogeneic T cell-mediated production of inflammatory mediators (tumor necrosis factor [TNF]-, interferon [IFN]-, and ·NO) without modifying the number or type of cells infiltrating the lungs (10). The same regimen of KGF given to mice injected with T cells plus Cy failed to prevent the high levels of inflammatory mediators, or ameliorate the severe permeability edema associated with the presence of nitrative/ oxidative stress (10). The mechanisms by which KGF inhibits the T cell-dependent inflammatory events and the reasons for impaired KGF responses in mice receiving both T cells and Cy remain unknown.

Because the antiinflammatory effects of KGF were observed 10 d after cessation of KGF treatment, and because predominantly epithelial cells are known to respond to KGF (11), we reasoned that an epithelial cell-derived, oxidant-sensitive mediator is responsible for the antiinflammatory effects of KGF. Transtracheal injection of KGF enhances surfactant protein A (SP-A) production in vivo (12). However, the effects of systemic KGF on SP-A production have not been elucidated. Emerging evidence indicates that in addition to its host defense against microbial agents, SP-A is a potent immunoregulator of activated T cell responses (13, 14). Therefore, we hypothesized that systemic KGF up-regulates SP-A production during intense allogeneic T cell-induced inflammation (Day 7 post-BMT). Because our previous studies indicate that strong oxidants inhibit surfactant synthesis (15) and injure SP-A function (16), we further hypothesized that the ability of KGF to up-regulate SP-A is impaired in mice injected with allogeneic T cells and Cy. Our data show increased Day 7 post-BMT SP-A expression in lethally irradiated mice that received allogeneic T cells at time of BMT, but not in mice that also received Cy as part of the conditioning regimen. A potential role for SP-A in mediating the antiinflammatory effects of KGF after allogeneic transplantation is suggested.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mice

Female B10.BR (H2^K) and C57BL/6 (H2^b) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Mice were housed in microisolator cages in the specific pathogen-free (SPF) facility of the University of Minnesota and cared for according to the Research Animal Resources guidelines of our institution. For BMT, donors were 8-12 wk of age and recipients were used at 8-10 wk of age. Sentinel mice were found to be negative for 15 known murine viruses including those that contribute to pneumonitis (e.g., cytomegalovirus, pneumonia virus of mice, K-virus) by our animal facility during repeated extensive evaluations over the study period.

Bone Marrow Transplant

BMT was performed as previously described (4). B10.BR mice received phosphate-buffered saline (PBS) or KGF (5 mg/kg/d subcutaneously; Amgen, Thousand Oaks, CA) on Days 6, 5, and 4 pre-BMT. Mice were then segregated into those receiving either PBS or cyclophosphamide (Cytoxan; Bristol Myers Squibb, Seattle, WA) 120 mg/kg/d as a conditioning regimen on Days 3 and 2. All mice were lethally total body irradiated (7.5 Gy TBI by X-ray at a dose rate of 0.41 Gy/min) on the day before BMT. Donor C57BL/6 bone marrow (BM) was T cell depleted with a monoclonal anti-Thy 1.2 antibody (clone 30-H-12, rat immunoglobulin G [IgG_2b], kindly provided by Dr. David Sachs, Charlestown, MA) plus complement (Neiffenegger Co., Woodland, CA). Recipient mice were transplanted via caudal vein with 20 × 10⁶ C57BL/6 marrow cells supplemented with or without 15 × 10⁶ natural killer (NK) cell-depleted (PK136, anti-NK 1.1 plus complement) spleen cells (BMS) as a source of IPS-causing T cells.

Bronchoalveolar Lavage

Mice were sacrificed on Day 7 post-BMT after an intraperitoneal injection of sodium pentobarbital and the thoracic cavity was partially dissected. The trachea was cannulated with a 19-gauge needle and infused with 1 ml of ice-cold sterile PBS and withdrawn. This was repeated two times and return fluid was combined. The bronchoalveolar lavage fluid (BALF) was immediately centrifuged at 500 × g for 10 min at 4° C to pellet cells. BALF total protein was determined with the bicinchoninic acid (BCA) method, with bovine serum albumin (BSA) used as a standard.

ELISA for SP-A

SP-A concentration (µg/ml or µg/mg total protein) in cell-free BALF was determined by enzyme linked immunosorbent assay (ELISA) using polyclonal rabbit anti-sheep SP-A (provided by Dr. Samuel Hawgood, UCSF, San Francisco, California), and polyclonal rabbit anti-human SP-A (provided by Dr. David Phelps, Pennsylvania State University, Hershey, PA). Equal aliquots (0.5 µl) of BALF were serially diluted using 50 mM Na₂CO₃-NaHCO₃ buffer at pH 9.6, coated to ELISA plates, and allowed to bind for at least 18 h at 4° C. Nonspecific binding sites were blocked with 1% BSA for 1 h at room temperature. The wells were then incubated with the primary antibody (1:10,000 dilution) at 37° C for 1 h. Unbound antibody was removed by a series of washes with PBS-Tween 20 buffer. Horseradish peroxidase-conjugated goat anti-rabbit IgG (1:2,500 dilution; Sigma Co., St. Louis, MO) was added as the secondary antibody. After serial washes color was developed by adding o-phenylenediaminedihydrochloride (OPD; Sigma Co.) and hydrogen peroxide to each well, and absorbance was read at 490 nm. Purified SP-A (0.125-3 ng), isolated from patients with alveolar proteinosis as previously described (16), was used as standard. Concentration of SP-A was calculated from each sample's slope (absorbance/µl BALF) and slope of standard curve (absorbance/ng of SP-A). The lowest concentration of SP-A detectable by this method was approximately 0.2 ng/µg of total protein. Because human SP-A was used as standard, results are considered relative and not absolute concentrations.

Western Blots for SP-A

Specificity of the antibodies to human SP-A and mouse SP-A was assessed by Western blotting using the anti-sheep and anti-human SP-A antibodies. Equal volumes (20 µl) of BALF were solubilized in 0.1 M Tris buffer containing 50 µM dithiothreitol, 0.01% bromophenol blue, 1% sodium dodecyl sulfate (SDS), and 10% glycerol, and boiled for 5 min. The proteins were resolved by 12% SDS polyacrylamide gels, transferred to nitrocellulose paper, and probed with the SP-A antibodies (1:10,000 dilution) followed by alkaline phosphatase-conjugated goat anti-rabbit IgG (1:7,500 dilution) as the secondary antibody. Bound antibody was detected with nitro blue tetrazolium and 5-bromo-4-chloro-3indolyl-1-phosphate kit (Sigma Co.). Additional experiments also showed that the polyclonal rabbit anti-sheep and anti-human SP-A antibodies we used for SP-A quantification had similar affinities for both normal and nitrated/oxidized SP-A (data not shown).

Northern Blots

In some animals, Day 7 post-BMT lungs were extracted without lavage and immediately frozen in liquid nitrogen. Total RNA was extracted using the guanidium thiocyanate-phenol-chloroform method (Tri-Reagent; Sigma). RNA samples (10 µg) were electrophoresed on a formaldehyde denaturing agarose gel, transferred to nitrocellulose filter, and cross-linked by exposure to ultraviolet radiation (120,000 µJ). The membrane was hybridized with [³²P]dATP-labeled murine SP-A cDNA (provided by Dr. Jefferey Whitestt, University of Cincinnati, OH) overnight at 65° C. The hybridized filter was washed and autoradiography performed by exposure to Kodak X-OMAT film at 70° C. mRNA levels were quantified by scanning densitometry, using 28 S ribosomal RNA as an internal loading control.

Semiquantitative RT-PCR

Reverse transcriptase (RT) was performed using a cDNA synthesis kit (First-Strand cDNA Synthesis Kit; Amersham Pharmacia Biotech, Uppsala, Sweden). The RT mixture consisted of 2 µg RNA; 5 µl Moloney murine leukemia virus reverse transcriptase containing RNase/ DNase-free BSA, dATP, dCTP, dGTP, and dTTP; 1 µl random hexadeoxynucleotides primer (pd[N]₆; 0.2 µg/µl); and 1 µl dithiothreitol (200 mM). The RT mixture was incubated for 1 h at 37° C. The products were further amplified by polymerase chain reaction (PCR) with Klen Taq DNA polymerase (Clontech, Palo Alto, CA). The oligonucleotide upstream and downstream primer sequences, annealing temperatures, and cycle numbers were as follows: mouse -actin, 60° C/25 cycles, 5'-AAGTGTGACGTTGACATCCGT-3' and 5'-CTCATCGTACTCCTGCTTGC-3'; mouse KGF, 60° C/25 cycles, 5'-CTGCTCTACAGATCATGC-3' and 5'-GCTGTGTGTCCATTTAGC-3'. The PCR products were electrophoresed through 1% agarose gel and amplified cDNA bands were visualized by ethidium bromide staining. To ensure experiments were in the linear range of amplification cycle experiments were performed on a representative sample from each group of PCR products measured. Aliquots were removed during RT-PCR amplification after each few cycles starting with cycle 18 and ending at cycle 30. Densitometry was used in relative semiquantitative assessment of RT-PCR product.

Detection of Nitrotyrosine in Immunoprecipitated SP-A

Day 7 post-BMT BALF from TBI mice given T cell-depleted bone marrow cells (BM) and Cy/TBI mice injected with donor spleen T cells at time of BMT (BMS + Cy) was precleared by incubation with 25 µl/ml of protein G agarose and spun at 6,000 × g for 1 min. The supernatant was incubated with rabbit polyclonal antibody against sheep SP-A (30 µl/ml) at 4° C overnight. Immune complexes were precipitated with 50 µl of Gamma Bind Plus Sepharose at 4° C for 90 min and solubilized in 50 µl of SDS-PAGE loading buffer, heated at 95° C for 5 min, and fractioned by SDS-PAGE in 12% gel, electrophoretically transferred onto nitrocellulose membranes, and probed with the polyclonal rabbit antibody against nitrotyrosine (NTAb; Upstate Biotechnology, Lake Placid, NY), or the anti-sheep SP-A antibody. Negative control included samples treated with dithionite at pH 9.5 for 30 min to reduce nitrotyrosine to aminotyrosine (not recognized by NTAb) prior to Western blotting.

One-way Mixed Lymphocyte Reaction (MLR)

To simulate the in vivo alloreactivity model, H-2^b-anti-H-2^k MLR were set up by incubating spleen T cells and monocytes from C57BL/6 mice with an equal number (5 × 10⁶ cells/ml) of irradiated (1000 cGy via a ¹³⁷Ces source) spleen cells from B10.BR mice. After 5 d in culture in MLR media consisting of Dulbecco's minimal essential medium (Bio-Whittaker, Walkersville, MD), 10% fetal calf serum (FCS), 2-mercaptoethanol (5 × 10⁵ M; Sigma), 10 mM HEPES, 1 mM sodium pyruvate (Gibco BRL, Grand Island, NY), and amino acid supplements (1.5 mM L-glutamine, L-arginine; Sigma) at 37° C and 5% CO₂, the MLR reaction was terminated. Unbound cells (activated T cells) were added to the bottom of 96-well plates (2 × 10⁶ cells/well suspended in 250 µl medium) for 24 h at 37° C in the presence of human SP-A (0, 5, 10, or 50 µg/ml), or KGF (100 ng/ml). Interleukin-2 (IL-2) levels in the cell-free supernatant of control and experimental wells were measured by ELISA (Pharmingen, San Diego, CA).

Statistical Analysis

Results are expressed as means ± SEM. Data were analyzed by ANOVA. Statistical differences among group means were determined by Tukey's Studentized test. Probability (p) values less or equal to 0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pre-BMT Systemic KGF Increases SP-A Level in BALF of Mice Injected with T Cells, But Not T Cells and Cy

The percentage recoveries of BALF volume collected on Day 7 after transplantation were similar in all groups (> 90%). Day 7 post-BMT was chosen because our previous data indicate that lung inflammation in irradiated mice given allogeneic T cells at time of BMT ± Cy starts after Day 3 and peaks on Day 7 post-BMT (5). Furthermore, we have reported that systemic KGF treatment suppresses allogeneic T cell-dependent production of inflammatory mediators measured on Day 7 post-BMT (10).

Using polyclonal rabbit anti-sheep SP-A antibody, SP-A content in the BALF of control and experimental mice was quantified by ELISA (Figure 1A). SP-A concentration (µg/ml BALF) was increased in the BALF of KGF-treated bone marrow plus spleen T cell-recipient mice (BMS). KGF alone without injection of allogeneic T cells, or T cells in the absence of KGF did not significantly increase SP-A level on Day 7 post-BMT. More importantly, KGF failed to enhance SP-A level in BMT mice given Cy conditioning in addition to T cells (BMS + Cy). The difference between SP-A content of BMS and BMS + Cy KGF-treated mice was even more pronounced when values were expressed as µg SP-A/mg of BALF total protein (Figure 1B). Values of BALF total protein are shown in Table 1. Consistent with our published data (10), KGF treatment suppressed permeability edema in mice given allogeneic T cells, but not allogeneic T cells and Cy.

View larger version (42K): [in this window] [in a new window]   Figure 1.   (A) ELISA for SP-A (µg/ml) using equal volume (0.5 µl) of bronchoalveolar lavage fluid (BALF) obtained on Day 7 post-BMT from control (not transplanted, not irradiated), irradiated BMT mice (BM), BMT mice given allogeneic donor spleen T cell (BMS), and BMT mice given T cells and Cy (BMS + Cy). KGF (5 mg/kg subcutaneously) was given on Days 6, 5, and 4 pre-BMT. (B) Results of ELISA expressed as µg SP-A/mg of BALF total protein. The primary antibody used for ELISA was the polyclonal rabbit anti-sheep SP-A antibody followed by horseradish peroxidase-conjugated goat anti-rabbit IgG as the secondary antibody. Return BALF volume was the same in all groups. Values are mean ± SEM obtained from at least 10 mice in each group. *p < 0.05 compared with controls. +p < 0.05 comparing the effect of KGF within the same group of mice.

Specificity of the anti-sheep SP-A antibody was confirmed by Western blotting of equal BALF volume (Figure 2). The SP-A dimer at 66 kD characteristic of SP-A extracted from patients with alveolar proteinosis was absent in murine SP-A. Similar results were obtained using the rabbit anti-human SP-A antibody, although the anti-sheep SP-A antibody exhibited higher affinity to mouse SP-A (data not shown).

View larger version (46K): [in this window] [in a new window]   Figure 2.   Western blot of equal BALF volume (20 µl) obtained on Day 7 post-BMT from the indicated BMT mice in the presence (+) or absence () of KGF given subcutaneously on Days 6, 5, and 4 pre-BMT. The amount of lavage fluid instilled and returned from the lungs was similar in all groups of mice. SP-A was detected with a rabbit anti-sheep SP-A antibody that cross-reacts with mouse SP-A. Lane 1 represents purified human SP-A (1 µg). Shown is a representative figure that was repeated two times with similar results.

Pre-BMT KGF Increases SP-A mRNA in BMS, But Not BMS + Cy Mice

Northern blots of total RNA extracted from lung tissue on Day 7 post-BMT were probed with mouse SP-A cDNA. KGF treatment up-regulated SP-A mRNA expression in allogeneic T cell-recipient mice (Figure 3). In contrast, KGF failed to enhance SP-A mRNA of mice receiving both T cells and Cy (BMS + Cy). KGF alone or Cy alone did not modify lung tissue SP-A mRNA assessed on Day 7 post-BMT (data not shown).

View larger version (46K): [in this window] [in a new window]   Figure 3.   Northern blot analysis of SP-A mRNA expression in mouse lungs on Day 7 after allogeneic BMT. Total RNA was extracted and Northern blot analysis was performed using mouse SP-A cDNA. Each lane on the gel was run with 10 µg of RNA (A). Quantitative assessment of SP-A mRNA using densitometry. Results, expressed relative to 28 S rRNA band intensity, are means ± SEM (B). Shown is a representative blot, which was repeated two times with similar results. *p < 0.05 compared with control (BM).

Up-regulation of Endogenous KGF in BMS, But Not BMS + Cy Mice

Because SP-A was increased only in KGF-treated BMS mice, we hypothesized the presence of an additive/synergistic effect between allogeneic T cells and exogenous KGF leading to increased SP-A expression. Therefore, we assessed whether the injection of allogeneic T cells directly or indirectly induces endogenous KGF mRNA in lungs of BMS and BMS + Cy mice, as assessed by semiquantitative RT-PCR. Reactions were terminated during linear range below saturation of PCR products. Results indicate that lungs of T cell-injected mice, but not mice given T cells and Cy, contained a 3-fold higher level of endogenous KGF mRNA (Figure 4). The increased endogenous KGF mRNA in mice given allogeneic T cells was not modified by exogenous KGF treatment (data not shown).

View larger version (46K): [in this window] [in a new window]   Figure 4.   Expression of mRNA for KGF and -actin in lung tissues of irradiated BMT mice that did not receive exogenous KGF using RT-PCR. Linear range for PCR products was determined by preparing a PCR master mix from RNA obtained from irradiated BMT mice given allogeneic T cells. Aliquots were removed during RT-PCR amplification after indicated cycles starting with cycle 18 (A). Amplified KGF cDNA (60° C/ 25 cycles) indicated on the figure were run on agarose gel and stained with ethidium bromide (B). Semiquantitative assessment of amplified cDNA using densitometry (C ). Results, expressed relative to -actin band intensity, are means of two samples/group obtained from two independent experiments. *p < 0.05 compared with control (BM).

Potential Structural Injury, But Not Nitrotyrosine, in Immunoprecipitated SP-A from the BALF of BMS + Cy Mice

SP-A was precipitated using anti-sheep SP-A antibody and nitrated residues visualized on Western blots probed with the antinitrotyrosine antibody (NTAb). Increased immune reactivity of NTAb with SP-A samples from BMS + Cy compared with BM mice was observed (Figure 5A). However, this increased binding was nonspecific as treatment of samples with dithionite at pH 9.5 for 30 min to reduce nitrotyrosine to aminotyrosine (not recognized by NTAb) failed to eliminate the binding of NTAb with SP-A samples (Figure 5B). Of interest is that the SP-A antibody recognized a band of similar molecular weight, suggesting that the band recognized by the NTAb represents a fragment of SP-A (Figure 5C). These data are consistent with degradation of SP-A from mice given allogeneic T cells plus Cy (BMS + Cy).

View larger version (107K): [in this window] [in a new window]   Figure 5.   SP-A contained in Day 7 post-BMT BALF from BM and BMS + Cy mice was immunoprecipitated using polyclonal anti-sheep SP-A antibody. Immune complexes were precipitated with 50 µl of Gamma Bind Plus Sepharose and solubilized in 50 µl of SDS-PAGE loading buffer, heated at 95° C for 5 min, and fractioned by SDS-PAGE in 12% gel, electrophoretically transferred onto nitrocellulose membranes, and probed with the polyclonal nitrotyrosine antibody (NTAb) (A), and SP-A antibody (C ). Negative control included samples treated with dithionite at pH 9.5 for 30 min prior to Western blotting (B). SP-A fragmentation consistent with structural injury to SP-A was observed in samples obtained from BMS + Cy mice (arrow). Shown is a representative Western blot which was repeated two times with similar results.

SP-A, But Not KGF, Suppresses IL-2 Production by Alloactivated T Cells

Human SP-A, incubated with MLR-alloactivated T cells for 24 h, suppressed IL-2 production in a dose-dependent manner (Figure 6). In contrast, KGF (100 ng/ml) failed to modify IL-2 production by activated T cells, suggesting the production of a mediator was required for the observed KGF-mediated down-regulation of inflammatory cell activation in vivo.

View larger version (51K): [in this window] [in a new window]   Figure 6.   IL-2 production by activated T cells (2 × 106) cultured for 24 hr in the presence of SP-A (0, 5, 10, or 50 µg/ml), or human recombinant KGF (100 ng/ml). Spleen cells form C57BL/6 mice were activated in mixed lymphocyte reaction by incubation with irradiated BR.B10 spleen cells for 5 d at 37° C. SP-A was isolated from the BALF of patients with alveolar proteinosis. IL-2 was measured in cell-free medium by ELISA. Values are means of at least four samples in each condition. *p < 0.05 compared with activated T cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The major findings of this study are that (1) the pre-BMT systemic KGF treatment up-regulates SP-A protein and mRNA during peak T cell-mediated inflammation measured on Day 7 post-BMT and (2) the ability of KGF to increase SP-A production is impaired in mice injected with T cells and Cy. The high SP-A level correlates with our previously published data showing that systemic KGF suppresses the production of proinflammatory mediators in irradiated mice receiving allogeneic T cells (10). Furthermore, failure of KGF treatment to up-regulate SP-A protein and message in mice given T cells and Cy is associated with persistence of inflammation implicated in IPS pathogenesis.

Prior studies observed alveolar type II cell hyperplasia and the induction of SP-A 1-3 d after the last dose of intratracheal KGF in normal and injured lungs with return toward baseline by Day 7 (12, 17). In our BMT model, we recently reported the induction of alveolar type II cell hyperplasia in all mice treated with KGF (18). Alveolar type II hyperplasia was assessed on Day 3 post-BMT, prior to the activation of allogeneic T cells and generation of ONOO and other strong oxidants. In the present study, we measured SP-A concentration 10 d after the last systemic KGF dose (Day 7 post-BMT) at time of in vivo allogeneic T cell activation and found consistently high SP-A levels only in the BALF of KGF-treated mice given T cells. KGF injections were insufficient to enhance SP-A protein in irradiated mice that did not receive supplemental allogeneic T cells. Although we may have missed an earlier time point during which a transient increase in SP-A occurred in all KGF-treated mice, the data suggest additive/synergistic effects between exogenous KGF and allogeneic T cells in SP-A protein regulation in the lungs at Day 7 post-BMT. Indeed, injection of allogeneic T cells at the time of BMT up-regulated Day 7 lung mRNA for endogenous KGF (Figure 4). The most likely mechanism by which allogeneic T cells may increase endogenous KGF mRNA is by induction of cytokines known to up-regulate KGF expression (19, 20). Despite the abundance of endogenous KGF mRNA, SP-A mRNA and protein were not increased in BMS recipients, most likely because of the inhibitory effects of proinflammatory cytokines on SP-A expression (21, 22). However, in the presence of additional KGF in KGF-treated BMS recipient mice, up-regulation of SP-A production and down-regulation of inflammation are favored.

Why Did the Same Regimen of KGF Given to Irradiated Mice Injected with T Cells and Cy Fail to Enhance SP-A Expression?

The first mechanism by which Cy may impair the ability of KGF to enhance SP-A expression is by prevention of KGF- induced proliferation of alveolar type II pneumocytes, the main cell type responsible for SP-A secretion. However, this is less likely to be the main mechanism for impaired KGF responses because, as discussed above, we have reported Day 3 post-BMT alveolar type II cell hyperplasia in both BMS and BMS + Cy mice after treatment with KGF (18). In addition, Ulich and coworkers showed that the ability of KGF to induce the proliferation of bladder epithelial cells is preserved in rats given toxic doses of Cy (23). The second potential mechanism for impaired ability of KGF to up-regulate Day 7 post-BMT SP-A expression in BMS + Cy is by Cy-induced depletion of antioxidants (24), leading to enhanced generation of oxidants. We have observed the simultaneous production of ·NO and O₂·, and therefore ONOO, after Day 3 post-BMT mainly in the lungs of mice given allogeneic T cells and Cy, but not T cells alone, or Cy alone (5). ONOO is a potent oxidant and nitrating agent known to inhibit surfactant production by alveolar type II cells, and to damage the structure and function of SP-A (25, 26). We observed that the ability of KGF to up-regulate SP-A mRNA was impaired in mice given donor T cells and Cy (BMS + Cy). Although we noted protein degradation, we were unable to detect specific nitration of immunoprecipitated SP-A obtained from BALF of BMS + Cy mice (Figure 5). Taken together, our data favor the hypothesis that the generation of reactive oxygen/nitrogen species, and not the antiproliferative effects of Cy, is responsible for the delayed ability of KGF to up-regulate Day 7 post-BMT SP-A expression in BMS + Cy mice mainly by suppressing the KGF-induced up-regulation of SP-A production. Despite the high levels of immunoreactive nitrotyrosine observed in the lung of BMS + Cy mice on Western blotting (5) and immunohistochemistry (10), the reason for the absence of specific nitration of SP-A remains unclear.

Is SP-A the Mediator Responsible for the Antiinflammatory Effects of KGF Post-BMT?

Our previous data suggest that KGF up-regulated the production of an epithelial-derived, oxidant-sensitive mediator with antiinflammatory properties (10). We observed a strong correlation between BALF SP-A level and the ability of KGF to ameliorate the inflammatory response on Day 7 post-BMT. SP-A is a known immunoregulator molecule (27). Consistent with our in vitro data, Borron and coworkers (13) and Wang and coworkers (28) showed that SP-A is a potent suppressor of T cells proliferation and activation in a dose-dependent manner. Inhibition of macrophage activation has also been reported (29, 30), however, this effect of SP-A remains controversial (31). The method of SP-A isolation and contamination with lipopolysaccharide (LPS) alters the response of alveolar macrophages to SP-A (32). Emerging evidence utilizing SP-A gene targeted knockout mice (SP-A^/) supports a role for SP-A in attenuating inflammation induced by the tracheal instillation of a replication-deficient recombinant adenovirus (33). In an in vivo model of LPS-induced lung inflammation, Borron and colleagues reported increased levels of ·NO and TNF- in SP-A^/ mice compared to SP-A^+/+ mice (34). Treatment with exogenous human SP-A inhibited LPS-induced production of inflammatory mediators in SP-A^/ mice.

Our data support, but do not establish, a causal relationship between the KGF-mediated up-regulation of SP-A production and the antiinflammatory effects of KGF after allogeneic transplantation. KGF may prevent IPS by multiple mechanisms including restoration of the integrity of epithelial cells (35), enhancing DNA repair (36), and up-regulation of antioxidant enzymes (37). BMT experiments using SP-A gene targeted knockout mice treated with KGF are necessary to establish whether SP-A plays a role in mediating at least some of KGF responses. Similarly, the biological effects of KGF may be mediated by several pathways, and the generation of strong oxidants and nitrating agents may impair some but not all of the KGF-mediated protective pathways (18).

Significance

KGF is currently in clinical trials for the prevention of mucositis post-BMT (38). Because BMT is always scheduled, our data predict that pre-BMT systemic administration of KGF may also prevent or reduce lung dysfunction in some patients after allogeneic BMT. Low levels of SP-A have been found in patients at risk for developing acute lung injury (39), and our approach to increase the concentrations of SP-A by KGF treatment may prove useful in reducing lung inflammation while preserving graft-versus-leukemia effects of allogeneic T cells, as was recently shown in an animal model of GVHD (40).

In summary, we find high levels of SP-A mRNA and protein in KGF-treated BMT mice during the time when KGF is known to prevent allogeneic T cell-mediated lung injury and inflammation. The same KGF treatment in BMT mice given T cells and Cy during the generation of strong oxidants and nitrating species failed to up-regulate SP-A expression. Further studies will explore the relative significance of SP-A and oxidant-modified SP-A in mediating the antiinflammatory effects of KGF in vivo.