INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IDIOPATHIC PNEUMONIA SYNDROME (IPS) is a significant cause of non-graft-versus-host disease (non-GVHD) deaths after bone marrow (BM) transplant (BMT) and accounts for the majority of complications involving the lung in the early post-BMT period (9). Intense conditioning regimens lead to a higher incidence of pulmonary toxicity in BMT recipients but are beneficial in preventing relapse and promoting BM engraftment (11, 20, 26, 43, 47, 55). Greater severity of GVHD post-BMT also increases the risk of developing IPS. In a mouse model study evaluating the contributions of preconditioning with total body irradiation (TBI) with and without cyclophosphamide (Cy) and allogeneic T cells to the generation of IPS, we reported that the severity of early post-BMT IPS injury was dependent on allogeneic T cells and potentiated by Cy (42). IPS injury was associated with the recruitment of host monocytes and donor T cells into the lung in response to tissue injury (42). The association of IPS with host monocyte infiltration and the dependence on allogeneic T cells has since been confirmed independently by two other laboratories (see Refs. 10, 13). The manifestations of lung injury we previously described included epithelial cell injury, increased wet and dry lung weights, and decreased specific lung compliance and lung capacity. Because almost all of the macrophages in the lung are host derived at this early time point post-BMT (42), measures to hinder the activation of immune effectors of the recipient may present alternative strategies for the prevention of IPS.

In response to injury in the lung, type II alveolar epithelial (ATII) cells proliferate and differentiate to replace dying type I epithelial pneumocytes (27a). The degree of lung injury ultimately manifested is to a large extent dependent on the ability of type II cells to effectively carry out this process to reepithelialize the alveolar membrane. Acute injury also releases intracellular contents that trigger acute-phase reactants, cytokines, and chemokines to initiate the inflammatory response and tissue repair process. This cascade can involve the induction of selectins and adhesion molecules leading to transmigration of inflammatory cells into the alveolar space. Acute lung inflammation is correlated with increased levels of proinflammatory cytokine mRNA in the lung (31, 44, 46, 53, 54, 58). Studies of bronchoalveolar lavage (BAL) fluid in mice following BMT across minor histocompatible differences demonstrated that increased levels of tumor necrosis factor (TNF)- and endotoxin [lipopolysaccharide (LPS)] at 6 wk post-BMT were associated with lung damage and IPS generation (12). In addition, we also described that the frequency of cells expressing T-cell costimulatory B7 molecules increased in the lung, as did the frequency of cells expressing mRNA for transforming growth factor (TGF)- (a monocyte chemoattractant), IL-1, and TNF- (42). This sets the stage for an immune-mediated attack on lung tissue by allospecific donor T cells, since activation of monocytes and their subsequent increased expression of B7 regulates the generation of cytotoxic T lymphocytes (CTL) that express the cytolysins granzymes A and B and, therefore, may contribute to the amplification of tissue injury.

KGF is a mediator of epithelial cell proliferation (24) and mesenchymal-epithelial interactions (23) as well as a growth factor for type II pneumocytes (39, 52). KGF is protective against chemotherapy- and radiation-induced injury in various rodent models (22, 51). We recently reported that in vivo administration of exogenous KGF, completed before conditioning, ameliorated GVHD-induced weight loss and mortality following allogeneic BMT in mice (41). In addition, KGF diminished GVHD-induced lesions in the target organs, especially the skin and lung, of long-term allogeneic BMT survivors. The protective effect of KGF on GVHD also has been confirmed recently in another murine model (34). In investigations targeting pulmonary injury, KGF was protective in lethal models of radiation-, hyperoxia-, acid-, and bleomycin-induced lung injury in rats (28, 60, 62), possibly by facilitating repair of DNA damage in alveolar epithelial cells (50, 57). Furthermore, KGF induces increased lung surfactant levels (49), potentiates alveolar fluid clearance by increasing Na⁺-K⁺-ATPase activity (6), decreases hyperoxia-induced apoptosis of (ATII) cells, and may detoxify reactive oxygen species generated by injured cells (reviewed in Ref. 56). Here, we report the effects of KGF on the IPS-inducing inflammatory events in the lung. The goal of this study is to investigate some potential mechanisms of KGF-mediated amelioration of lung injury in the peri-BMT period. Our data indicate that KGF diminishes IPS injury, at least in part, by dampening the immune system response to chemoradiotherapy-induced lung damage and by accelerating repair of the damaged tissue.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Mice. B10.BR (H2^k) and C57BL/6 (H2^b) mice were purchased from Jackson Laboratories (Bar Harbor, ME). Mice were housed in microisolator cages in the 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.

KGF production. Recombinant human KGF produced in Escherichia coli was prepared as previously described (52) at Amgen (Thousand Oaks, CA).

Pre-BMT conditioning. B10.BR mice received PBS or KGF (5 mg · kg¹ · day¹ sc) on days 6, 5, and 4 pre-BMT. Mice were then segregated into those receiving either PBS or Cy (Cytoxan; Bristol Myers Squibb, Seattle, WA), 120 mg · kg¹ · day¹, as a conditioning regimen pre-BMT on days 3 and 2. All mice were lethally irradiated on the day before BMT (7.5-Gy TBI) by X-ray at a dose rate of 0.41 Gy/min as described (3).

BMT. Our BMT protocol has been described previously (5). Briefly, donor C57BL/6 BM was T cell depleted (TCD) with anti-Thy 1.2 monoclonal antibody (MAb) (clone 30-H-12, rat IgG2_b, kindly provided by Dr. David Sachs, Charlestown, MA) plus complement (Nieffenegger, Woodland, CA). Recipient mice were transplanted via caudal vein with 20 × 10⁶ TCD C57BL/6 (H-2^b) marrow with or without 15 × 10⁶ NK cell-depleted (PK136, anti-NK1.1 + complement) spleen cells (BMS) as a source of IPS-causing T cells.

Electron microscopy. This was performed as previously described (42). After inflation to ~20 cmH₂O by hand, 2- to 3-mm³ pieces of lung tissue were fixed in 2% glutaraldehyde for 1-2 days at 4°C followed by postfixation in 1% osmium tetroxide (EM Sciences, Fort Washington, PA) in 0.1 M sodium cacodylate buffer for 1 h, dehydrated in graded ethanol and propylene oxide, and embedded in Epon 812 (EM Sciences). Sections were cut at a thickness of 600 nm, stained with uranyl acetate-lead citrate (EM Sciences), and examined with a Philips 301 electron microscope. A minimum of 17 prints (maximum 39) taken at a magnification of ×4,200 from multiple sections from 2 representative mice of each group at day 0 and day 3 post-BMT were examined by three observers in coded fashion. The number of type II cells present (based on morphological appearance and presence of lamellar bodies) was expressed as a percent of total nucleated cells.

Frozen tissue preparation. After death of the mouse at either day 0, 3, or 7, a mixture of 0.5 ml of optimal cutting temperature compound (Miles, Elkhart, IN) and PBS (3:1) was infused via the trachea into the lungs. Lungs were snap-frozen in liquid nitrogen and stored at 80°C.

Immunohistochemistry. After fixation in acetone, cryosections (4 µm) were immunoperoxidase stained using biotinylated MAbs with avidin-biotin blocking reagents, ABC-peroxidase conjugate, and diaminobenzidine chromogenic substrate purchased from Vector Laboratories (Burlingame, CA) essentially as described (4). The biotinylated MAbs used were as follows: anti-CD4 (clone GK1.5), anti-CD8 (clone 2.43), anti-Mac-1 (clone M1/70), anti-Gr-1 (clone RB6-8C5), anti-I-A^k (clone 11-5.2), anti-B7.1 (clone 1G10), and anti-B7.2 (clone GL1), all purchased from PharMingen (San Diego, CA). Representative sections from each tissue block were stained with hematoxylin and eosin for histopathological assessment. The number of positive cells in the lung was quantitated as the percent of nucleated cells under ×200 magnification (×20 objective lens). Four fields per lung were evaluated.

In situ hybridization. This procedure has been described in detail elsewhere (40). Cryosections (4 µm) were hybridized with digoxigenin-labeled antisense RNA probes. The corresponding ribonucleotide sequences used were 80-910 bp for granzyme A and 239-775 bp for granzyme B. Immunological detection of digoxigenin-labeled RNA duplexes was accomplished with anti-digoxigenin antibody (alkaline phosphatase conjugated; Boehringer Mannheim). After color development, sections were mounted in Crystalmount (Biomeda, Foster City, CA). Positive cells were quantitated as described above.

Serum cytokine level determination. At the time of death of the mouse, blood was collected by cardiac puncture and placed immediately at 4°C; the serum was separated at 4°C and stored at 80°C. Serum levels of IL-1, IL-4, IL-6, IL-10, IL-13, interferon (IFN)-, and TNF- were determined by ELISA using commercial kits (R&D Systems, Minneapolis, MN).

Statistical analysis. Data were analyzed by ANOVA (Dunnett's test) or Student's t-test. Probability (P) values less than or equal to 0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

KGF preconditioning preserves ATII cells in the lung after allogeneic BMT. To determine whether KGF administration (subcutaneous on days 6, 5, and 4 pre-BMT) would affect early lung injury parameters documented previously by this laboratory (42), experiments were set up to compare the effect of KGF in the context of TBI or Cy/TBI conditioning as well as with (BMS) or without (BM) the addition of a 100% lethal dose of allogeneic spleen cells (see Fig. 1). B10.BR recipients were infused with C57BL/6 cells and killed on either day 0 (pre-BMT) or day 3 post-BMT, and lungs were examined by electron microscopy. We chose days 0 and 3 for two reasons: 1) because the allogeneic cells are infused on day 0, they would be exposed to early manifestations of conditioning-induced injury and any potential tissue-altering effects of KGF at this time; and 2) endothelial and epithelial cell injury are evident at least as early as day 3 by electron microscopy and represent the end of the first wave of host monocyte infiltration (42).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Keratinocyte growth factor (KGF) treatment and conditioning schedule for bone marrow (BM) transplant (BMT) experiments. KGF (5 mg · kg1 · day1) or PBS was administered subcutaneously to B10.BR recipients on days 6, 5, and 4. Mice were then further segregated into those receiving cyclophosphamide (Cy; 120 mg · kg1 · day1) or PBS intraperitoneally on days 3 and 2. All recipients were irradiated (7.5 Gy) on the day before transplant (day 1). Groups were segregated into those receiving C57BL/6 BM alone or BM with allogeneic spleen cells (BMS) intravenously. Lungs were harvested for analysis on days 0 (before BMT), 3, and 7 post-BMT. EM, electron microscopy; IHC, immunohistochemistry; ISH, in situ hybridization; TBI, total body inrradiation.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   KGF pretreatment does not affect the number of alveolar type II (ATII) cells by the day of BMT (day 0) but does preserve ATII cell integrity. B10.BR recipient mice were preconditioned with TBI (day 1) with and without Cy (days 3 and 2). Groups received KGF (5 mg · kg1 · day1) on days 6, 5, and 4 pre-BMT. Lung tissues were harvested on day 0 before BMT. Data are expressed as percent of nucleated cells displaying ATII characteristics (rounded nucleus, abundant cytoplasm, located at alveolar septal vertices) as determined by counting hematoxylin and eosin-stained lung cryosections under light microscope using ×40 objective lens with ×10 ocular. Mean values ± SD are for 3 mice per group from 2 representative experiments. The proportion of noninjured vs. injured ATII cells was determined by counting from electron micrographs such as those shown in Fig. 4 that were generated from 600-nm sections photographed under ×4,200 magnification.

View larger version (108K): [in this window] [in a new window]   Fig. 3.   KGF-pretreated mice exhibit small focal clusters and diffuse cuboidal proliferation of ATII cells on day 0. B10.BR recipient mice were preconditioned with TBI (day 1) with and without Cy (days 3 and 2). Groups received KGF (5 mg · kg1 · day1) on days 6, 5, and 4 pre-BMT. Lung tissues were harvested on day 0 before BMT. Lung cryosections (4 µm) were stained by hematoxylin and eosin, and black arrows point to cells displaying ATII characteristics (rounded nucleus, abundant cytoplasm, located at alveolar septal vertices). Blue arrows point to small focal ATII cell clusters; green arrows point to diffuse cuboidal proliferation. Resolution power is of ×20 objective lens for wide-view image under light microscope.

View larger version (100K): [in this window] [in a new window]   Fig. 4.   Representative electron micrographs of lungs from day 0 before BMT of TBI or Cy/TBI conditioned mice with (B and D) or without (A and C) KGF pretreatment. KGF-pretreated mice exhibited fewer injured type II cells compared with control vehicle recipients. Salient features and treatment groups are indicated. a, Alveolar space; 1, type I epithelial cell; 2, type II epithelial cell; m, macrophage (magnification ×4,200; bar = 5 µm). E and F are examples of injured and noninjured ATII cells, respectively (magnification ×22,000; bar = 1 µm).

View larger version (79K): [in this window] [in a new window]   Fig. 5.   KGF-pretreated mice exhibit hyperplasia of ATII cells on day 3 post-BMT. B10.BR recipient mice were preconditioned with TBI (day 1) with and without Cy (days 3 and 2) and given C57BL/6 BM with spleen cells (BMS) on the day of BMT (day 0). Groups received KGF (5 mg · kg1 · day1) on days 6, 5, and 4 pre-BMT. Lung tissues were harvested on day 3 post-BMT. Lung cryosections (4 µm) were stained by hematoxylin and eosin, and black arrows point to cells displaying ATII characteristics (rounded nucleus, abundant cytoplasm, located at alveolar septal vertices). Resolution power is of ×20 objective lens for wide-view image under light microscope.

View larger version (67K): [in this window] [in a new window]   Fig. 6.   Representative electron micrographs of lungs from day 3 post-BMT of Cy/TBI conditioned BMS recipient mice with and without KGF pretreatment. Type II cell hyperplasia was apparent in all mice pretreated with KGF (Cy/TBI BMS shown). Salient features and treatment groups are indicated. a, Alveolar space; 1, type I epithelial cell; 2, type II epithelial cell; m, macrophage. Magnification ×4,200, bar = 5 µm.

View larger version (33K): [in this window] [in a new window]   Fig. 7.   KGF-pretreated mice have increased numbers of ATII cells on day 3 post-BMT. B10.BR recipient mice were preconditioned with TBI (day 1) with and without Cy (days 3 and 2) and given C57BL/6 BM alone or with spleen cells (BMS) on the day of BMT (day 0). Groups received KGF (5 mg · kg1 · day1) on days 6, 5, and 4 pre-BMT. Lung tissues were harvested on day 3 post-BMT. Data are expressed as percent of nucleated cells displaying ATII characteristics (rounded nucleus, abundant cytoplasm, located at alveolar septal vertices) as determined by counting hematoxylin and eosin-stained lung cryosections under light microscope using ×40 objective lens with ×10 ocular. Mean values ± SD are for 3 mice per group from 2 representative experiments. * P < 0.05 vs. non-KGF counterpart.

KGF preconditioning attenuates B7 costimulatory molecule expression and the induction of the cytolysin granzyme B in the lung on day 7 post-BMT. In the initial description of our murine IPS model, we showed that the pulmonary cellular infiltrate was composed of CD4⁺ and CD8⁺ donor T cells, neutrophils, and host monocytes/macrophages. Therefore, we sought to determine the effects of KGF on the cellular inflammatory response to Cy-, TBI-, and T cell-mediated lung injury on day 7 post-BMT. Day 7 was chosen because we previously demonstrated that lung dysfunction and inflammatory infiltrates were evident at this time (42). Although conditioning caused a transient increase in host monocytes, the prolonged second wave of host monocyte infiltration and induction of B7 expression was T cell dependent. Figure 8 shows that KGF pretreatment does not significantly affect the composition of the cellular infiltrate in the lung at day 7 post-BMT as assessed by immunohistochemical staining of cryosections. Regardless of conditioning regimen, KGF attenuated the expression of B7.1 (CD80) and B7.2 (CD86), molecules, which are costimulatory for T cells, while not affecting the frequency of cells expressing host major histocompatibility complex (MHC) class II (I-A^k) as shown in Fig. 9 (from the same pool of experiments). KGF pretreatment did not affect the lack of donor MHC class II (I-A^b) cells, characteristic of our IPS model at this day 7 time point (42) (data not shown). Representative photomicrographs of B7.1 staining are shown in Fig. 10.

View larger version (26K): [in this window] [in a new window]   Fig. 8.   KGF does not affect the cellular composition of lung infiltrates post-BMT. Expression of Mac-1, CD4, CD8, and Gr-1 was determined by immunoperoxidase staining with biotinylated monoclonal antibodies. B10.BR recipient mice were preconditioned with TBI (day 1) with and without Cy (days 3 and 2) and given C57BL/6 BM with spleen cells (BMS) on the day of BMT (day 0). Groups received KGF (5 mg · kg1 · day1) on days 6, 5, and 4 pre-BMT. Lung tissues were harvested on day 7 post-BMT. Data are expressed as percent of nucleated cells expressing the surface marker in the lung as determined by counting 4 fields per lung section under light microscope. Mean values ± SE are for 3-5 mice per group from 3 representative experiments.

View larger version (24K): [in this window] [in a new window]   Fig. 9.   KGF attenuates expression of B7.1 and B7.2, while not affecting host major histocompatibility complex (MHC) class II (I-Ak), as determined by immunoperoxidase staining with biotinylated monoclonal antibodies. B10.BR recipient mice were preconditioned with TBI (day 1) with and without Cy (120 mg · kg1 · day1, days 3 and 2) and given C57BL/6 BM with 15 × 106 spleen cells (BMS) on the day of BMT (day 0). Groups received KGF (5 mg · kg1 · day1) on days 6, 5, and 4 pre-BMT. Lung tissues were harvested on day 7 post-BMT. Data are expressed as percent of nucleated cells expressing the indicated surface marker in the lung as determined by counting 4 fields per lung section under light microscope. Mean values ± SE are for 4-6 mice per group from 3 representative experiments. Staining for donor MHC class II at these time points in the lung was negative for all groups. *P < 0.05 vs. non-KGF counterpart. ^P = 0.07 vs. non-KGF counterpart.

View larger version (83K): [in this window] [in a new window]   Fig. 10.   Immunoperoxidase staining for B7.1 expression of lung tissue taken on day 7 post-BMT of lethally irradiated B10.BR recipients pretreated with PBS (A) or KGF (days 6, 5, and 4; B) and transplanted with C57BL/6 BMS cells. Frozen sections were incubated with biotinylated anti-B7.1 monoclonal antibody (clone 1G10) and developed with peroxidase-conjugated avidin-biotin complex and diaminobenzidine chromogen (methyl green counterstain). Positive cells are indicated by solid arrows. Magnification ×100; resolution power is equivalent to ×40 objective lens.

View larger version (29K): [in this window] [in a new window]   Fig. 11.   KGF decreases granzyme B expression as assessed by in situ hybridization using antisense digoxigenin-labeled riboprobes for granzymes A and B mRNA on lung cryosections taken day 7 post-allogeneic BMT. B10.BR recipient mice were preconditioned with TBI (day 1) with and without Cy (120 mg · kg1 · day1, days 3 and 2) and given C57BL/6 BM with 15 × 106 spleen cells (BMS) on the day of BMT (day 0). Groups received KGF (5 mg · kg1 · day1) on days 6, 5, and 4 pre-BMT. Data are expressed as percent of nucleated cells expressing the indicated mRNA in the lung as determined by counting 4 fields per lung section under light microscope. Mean values ± SE are for 2 or 3 mice per group from 2 representative experiments. *P < 0.05 vs. non-KGF counterpart. ^P = 0.07 vs. non-KGF counterpart.

View larger version (77K): [in this window] [in a new window]   Fig. 12.   In situ hybridization using anti-sense digoxigenin-labeled riboprobes for granzyme B mRNA on day 7 after allogeneic BMT of TBI BMS recipients pretreated with PBS (A) or KGF (B). Positive cells are indicated by solid arrows and were detected with alkaline phosphatase-conjugated antidigoxigenin antibody and 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium chromogen (no counterstain). Magnification ×100; resolution power is equivalent to ×40 objective lens.

KGF increases the expression of IL-4 and IL-13, suppressors of monocyte differentiation. Reduction of GVHD-induced lethality can be associated with the induction of counterregulatory cells (Th2) producing anti-inflammatory cytokines (such as IL-4 and IL-10) that skew the T-cell response away from inflammatory Th1 cytokine production (such as IL-2 and IFN-) (25). We wished to determine whether the beneficial effects of KGF [i.e., amelioration of GVHD and lung pathology (41), lowered B7 and granzyme B expression in the lung] were related to effects on circulating cytokine levels at the time of BMT when allogeneic donor T cells are infused (i.e., day 0). If so, one may expect either reduced levels of proinflammatory cytokines (IFN-, TNF-, and IL-1) or increased levels of anti-inflammatory cytokines (such as IL-4, IL-10, and IL-13). Because the BM inoculum is administered intravenously on day 0, the cells would be exposed immediately to the cytokine milieu at this time point, which then may translate into a subsequent effect on the cells infiltrating the lung. Table 1 (pooled data from 4 experiments) shows the circulating levels of IL-4, IL-13, and IL-6 on day 0 immediately before BMT. One day postirradiation, TBI conditioning did not significantly increase cytokine levels. Interestingly, KGF pretreatment of TBI-conditioned mice did not affect IFN-, TNF-, IL-1, or IL-10 on day 0 (data not shown) but did significantly increase levels of several Th2 (anti-inflammatory) cytokines including IL-4, IL-6, and IL-13. The addition of Cy to the TBI conditioning increased the level of IL-6 on day 0 and decreased the level of IL-13 compared with control mice. KGF significantly increased the level of IL-13 in Cy/TBI recipients back to a level comparable to normal control mice. The increases in IL-4 and IL-13 were due solely to KGF and not an interaction with irradiation, since KGF-treated controls that were not irradiated had equivalent, elevated levels of these cytokines. There was, however, a multicomponent interaction to the increase in IL-6 seen in the TBI/KGF group, since the increased IL-6 level was significantly higher than the non-BMT KGF control group. These data indicate that at this day 0 time point, KGF pretreatment induced the production of Th2-like (or anti-inflammatory) cytokines, especially IL-13. In addition to Th2 cells, the other major producer of IL-13 is the alveolar macrophage (30); therefore, it is possible that these cells may have been the source of this serum IL-13. However, analysis of BAL fluids harvested on day 0 or day 3 post-BMT showed negligible levels of IL-13 (data not shown). Therefore, it appears that KGF is inducing Th2 cells to produce IL-13 (and perhaps IL-4), which can serve to downregulate macrophage differentiation and function.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

We examined the use of KGF as a pretreatment therapy for IPS. IPS results from a combination of tissue damage and allogeneic T cell-dependent inflammatory events that occur in the early peri-BMT period in our established murine model. The effects of KGF pretreatment on lung repair entailed preservation of noninjured ATII cells on day 0 and apparent type II cell hyperplasia on day 3 post-BMT. More strikingly, the effects of KGF pretreatment on the immune response to lung injury were manifest as the attenuation of the expression of the T-cell costimulatory molecules B7.1 (CD80) and B7.2 (CD86) and a decrease in the frequency of cells transcribing mRNA for the inducible cytolytic molecule granzyme B. These decreases were not due to reduced cellular infiltration, since KGF did not change the composition of the lung-infiltrating cells on day 7 post-BMT and, furthermore, did not affect the frequency of cells transcribing the constitutive granzyme A mRNA.

KGF-treated mice exhibited lower frequencies of injured ATII cells on day 0, consistent with the cytoprotective effects of KGF on these cells. Because this is the time of BMT, the infused donor cells would be exposed to fewer injured cells, perhaps blunting their ensuing activation that normally occurs in IPS. Because the ability to repair the injured alveolar epithelium is dependent on the ability of type II cells to proliferate, differentiate, and replace type I cells, an epithelial cytoprotective agent such as KGF could potentially enhance this process. The apparent ATII cell hyperplasia seen in the KGF-pretreated mice at day 3 post-BMT is consistent with this hypothesis and with the lack of evident pathology in the lungs of long-term survivors of Cy/TBI-conditioned recipients of BMS (usually exhibiting the most severe injury and mortality) that were pretreated with KGF. We did not observe overt multifocal knobby proliferation of alveolar epithelial cells as shown by others (52) by light microscopy of hematoxylin and eosin-stained lung cryosections, but we did see occasional small clusters and diffuse cuboidal growth of ATII cells. These features have been demonstrated by Ulich et al. (52) and have been described as the later stages of KGF-mediated ATII cell proliferation by KGF given intratracheally. The route of KGF injection is most likely the reason for the lack of this overt type II cell sequela of KGF because we administered it subcutaneously as opposed to the intratracheal injection of KGF, which causes the formation of knobby cuboidal cell growth along the alveolar septa composed of proliferating type II cells. Consistent with our hypothesis, a recent study by Guo et al. (28) has shown that intravenous administration of KGF, while not causing overt type II cell proliferation, is still effective at preventing bleomycin- and hyperoxia-induced lung injury but not as potently as the intratracheal route. There is also recent evidence by Borok et al. (7) that KGF may cause a reversion of type I cells into cells with type II pneumocyte characteristics, making them more resistant to injury, a process termed reversible transdifferentiation. In vivo, this would translate into an increased number of type II cells.

The relevance of the decreased B7 expression in KGF-pretreated BMS recipient mice may relate to the decrease in cytolytic T-cell granzyme B expression. In the absence of costimulatory signals such as those mediated by B7 (on antigen-presenting cells) on binding to the CD28 counterreceptor (on T cells), antigen-triggered T cells become nonresponsive or anergic (8, 45). It is well established that B7 expression augments CTL generation (35). In our KGF-pretreated IPS model, the colocalization of donor T cells with monocytes expressing low levels of B7 may lead to a failure of conducive costimulation for an allo-MHC response resulting in the observed lower frequency of cytolytic cells in situ. This is consistent with our recent observations that preclusion of the CD28-B7 interaction by anti-B7 MAb reduced the generation of granzyme B-positive cells and the in vitro cytolytic activity of T cells obtained from MHC-disparate irradiated recipients of allogeneic T cells (2). Because of the apparent preservation of the type II cells by KGF, the lower level of B7 and granzyme B expression may be sequelae of the reduced damage resulting in less immune activation.

Sera from KGF-treated mice exhibited increased levels of the Th2 cytokines IL-4, IL-6, and especially IL-13 4 days after cessation of KGF administration (day 0) even in the absence of irradiation or BMT. This argues that at least some of the effects of KGF may be independent of the sequelae of reduced lung injury. Because day 0 is the time at which the BM inoculum is administered, these cells would be immediately exposed to this Th2 type of cytokine environment in the circulation into which the cells are infused. It has recently been demonstrated that the functional response of bone marrow-derived macrophages is determined by the first exposure to cytokine, thereby rendering them unresponsive to subsequent exposure to alternate cytokines (21). Several studies have demonstrated that IL-13 can suppress many monocyte activities, similar to IL-4 (1, 14, 16-18, 36, 37, 48, 59). The downregulation of CD14 (LPS receptor) expression by IL-13 (14) may, in part, explain the survival of KGF-pretreated BMT recipients in the face of GVHD-induced colitis (41). Damage to other GVHD target organs, particularly the gastrointestinal tract, induces the systemic release of endotoxin (LPS) that primes monocytes to release proinflammatory cytokines that exacerbate GVHD and IPS (12). Because our data suggest that KGF pretreatment leads to decreased monocyte activation, KGF also may protect the BMT recipient against adverse consequences of monocyte activation that contribute to lung injury.

IL-13 does not affect the chemotactic properties of monocytes (61), and we found no difference in the infiltration of monocytes into the lung on day 7 post-BMT with KGF treatment. We do not yet know whether the decreased expression of B7 molecules by these monocytes is due to direct KGF effects or indirect effects via the inhibitory effects of IL-13; the effect of IL-13 on monocyte-B7 expression has not been addressed in the literature. Decreased monocyte function would explain the decreased inflammatory (IFN-, data not shown) and cytotoxic (granzyme B) T-cell mediators we found in the lungs of KGF-pretreated BMS recipients at day 7 post-BMT. These recipients also had significantly lower levels of serum TNF- than non-KGF-treated counterparts on day 7 post-BMT (data not shown) consistent with the recent observations of Krijanovski et al. (34) using another murine GVHD model. The association of lower TNF- and LPS levels with less severe manifestations of GVHD and lung injury post-BMT suggests that TNF- and LPS contribute to IPS injury. The source of the Th2-type cytokines found in the serum of KGF-treated mice on day 0 just before BMT is most likely Th2 cells for the following reasons: 1) aside from alveolar macrophages, Th2 cells are the only known major producers of IL-13, and BAL fluid IL-13 levels did not parallel serum levels; 2) IL-4 is produced primarily by activated Th2 cells; and 3) IL-1 and TNF-, monocyte products, were not increased at this time point. Although IL-6 also was elevated in response to KGF, it can be produced by numerous cells including keratinocytes, which likely were activated in response to the subcutaneous injection of KGF. In addition, T cells costimulated by activated keratinocytes preferentially produce IL-4 (27), and IL-4 was elevated in response to KGF pretreatment in our study. Of note, IL-13 can also have proinflammatory properties depending on the target cell (19, 29) and the timing of the exposure of the monocyte to IL-13 in relation to the stimulus (i.e., antigen) (15). In fact, high levels of monocyte-derived IL-13 are found in the BAL fluid of asthmatics (32) and patients with pulmonary fibrosis (30). Perhaps the latter are examples of failed attempts at repair. Numerous reports have shown that type 2 alloreactive T cells, generally considered to be anti-inflammatory, have a reduced capacity to induce GVHD (25, 33). However, this may not be a universal finding in GVHD (38). We do not know whether the amelioration of IPS by KGF is Th2 dependent.

Taken together, these data suggest that KGF can hinder IPS by two modes: 1) enhancement of alveolar epithelialization and 2) attenuation of B7 and cytolytic molecule (granzyme B) expression in situ, possibly via induction of Th2 cytokines, in particular IL-13, a potent macrophage downregulator, thus enabling lung repair.