Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The lung is a major target for oxidant injury, and, among the structures of the respiratory system, the surface of the alveolus is directly exposed to the various forms of reactive oxygen (O₂) species that are generated within the lung. Several studies have described the deleterious effects of oxidants on the two cell types that compose the alveolar epithelium, namely type 1 and type 2 cells. It is now well established that type 1 cells are highly sensitive to oxidative stress and that once damaged, they cannot be replaced by division of adjacent noninjured sister cells. Consequently, repair of the damaged epithelium is dependent on the ability of type 2 cells to initiate proliferation, providing additional cells that have the potential to undergo transition into type 1 cells, thereby replacing injured alveolar epithelial cells (1). Any impairment in type 2-cell replication may lead to an altered repair of the respiratory epithelium (2). Therefore, determining factors capable of preventing epithelial cell damage and/or stimulating an optimal process of re-epithelialization appears essential for the protection of the pulmonary alveolar structure against various forms of oxidant injury. This is especially important in diseases such as respiratory distress syndromes that require treatment with elevated O₂ and that involve epithelial damage.

Retinoids, including retinol (vitamin A) and retinoic acid (RA), are known to regulate morphogenesis, cell proliferation, and differentiation in a variety of organs. In the respiratory system, there is now evidence that retinoids can affect a number of processes associated with lung development and maturation, as well as with lung repair after injury to maintain lung integrity (3, 4). Importantly, Massaro and Massaro (5) reported that treatment with RA was able to inhibit elastase-induced emphysema in adult rats. These authors also provided data indicating that RA could antagonize the possible deleterious effect of glucocorticoids on lung growth by preventing the lowered number of alveoli caused by dexamethasone treatment in rats (6).

The repair potentials of vitamin A and carotenoids led us to wonder whether RA could interfere with the proliferative response of alveolar epithelial cells exposed to oxidants. In the present study we focused on the effects of RA on cell-cycle progression of type 2 cells exposed to hyperoxia. Experiments were performed with a rat type 2 cell line that has been shown to regulate some aspects of proliferation in a fashion similar to that in primary type 2 epithelial cells (7). We observed that pretreatment by RA of oxidant-exposed type 2 cells was able to prevent the growth-arrest and the cell loss induced when cells were placed under hyperoxia. To gain some insights into the mechanisms involved, we studied the effects of RA on the protein-regulators of the cell cycle, the cyclin-dependent kinase (CDK) system. Having previously shown that oxidants block the entry into S phase of alveolar epithelial cells by acting on a subset of late G₁ events, in the present work we focused on cyclin E, CDK2, and p21^CIP1 expression as well as on activity on cyclin E-CDK2 complexes (8). Our results reveal a significant decrease in cyclin E-CDK2 complex activity in oxidant-exposed cells, which was no longer observed upon RA treatment. We show that RA mediated its effect on the proliferative response of oxidants-exposed cells through an inhibition of p21^CIP1 expression, which resulted in a reduced interaction of p21^CIP1 with cyclin E-CDK2 complexes. We also provide data suggesting that RA action may involve interference with the transforming growth factor (TGF)- signaling pathway.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and Cell Culture Conditions

The type 2 cell line used in this study was derived from rat primary neonatal type 2 cells and has been extensively characterized (7). Cells were grown in Eagle's minimum essential medium with Earle salt (GIBCO BRL, Grand Island, NY) supplemented with 4 mM glutamine, 50 U penicillin/ml, 50 µg streptomycin/ml, and 10% fetal bovine serum in a 5% CO₂/95% air atmosphere at 37°C. For hyperoxic conditions, cells were cultured in 5% CO₂/ 95% O₂ atmosphere at 37°C.

For the study of RA effects, exponentially growing cells were cultured for 24 h in medium with or without all-trans RA (Sigma Chemical Co., St. Louis, MO). Cells were then exposed to hyperoxia in the presence or absence of RA for the indicated durations with medium changed every 24 h. Stock solutions of RA were prepared at a concentration of 10 mM in 100% ethanol and stored at 80°C in the dark. Cells cultured without RA in hyperoxic conditions were cultured in medium containing the same amount of ethanol.

Proliferation studies were performed by determination of cell number using a hemocytometer, and by DNA synthesis assay using autoradiography of ³H-thymidine-labeled nuclei (9).

For each protocol, three or four independent experiments were performed.

RNA Isolation and Northern Blot Analysis

Total cellular RNA was isolated using the guanidium isothiocyanate procedure as described previously (10). The quantity of 20 µg of RNA was fractionated by electrophoresis through 1% agarose-2.2 M formaldehyde gels and blotted onto nylon membranes (Stratagene, Inc., La Jolla, CA). The integrity of RNA was assessed by visual inspection of the ethidium bromide-stained 28S and 18S ribosomal RNA (rRNA) bands. The blots were prehybridized and hybridized to ³²P-labeled probes, washed, and exposed to film as previously described (10). The relative intensity of bands was quantified by scanning densitometry using comparison with 18S rRNA band intensity.

The probes were generated by labeling plasmid inserts with -³²P-deoxycytidine triphosphate using random oligonucleotide priming (Amersham, Buckinghamshire, U.K.). Plasmids containing inserts for rat p21^CIP1 and TGF-1 were obtained as previously described (8, 10). Plasmids containing the rat type I and the type II TGF- receptor complementary DNAs were kindly provided by Dr. J. S. Brody (Boston University, Boston, MA).

Protein Studies

Cell extract preparations. Total cell extracts were prepared as previously described (8) and stored at 80°C until analysis by immunoprecipitation and immunoblotting assays.

For the preparation of nuclear extracts (11), cells were harvested with trypsin-ethylenediaminetetraacetic acid (EDTA) (0.5 g/liter trypsin and 0.5 mM EDTA; GIBCO BRL) after three washings in cold phosphate-buffered saline (PBS) (20 mM Tris-HCl [pH 7.6] and 137 mM NaCl). Cells were then counted using a hemocytometer, and pelleted by centrifugation at 2,000 × g for 5 min. The cell pellet was resuspended in a cell number-adjusted volume of buffer A (10 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid [Hepes], 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM dithiothreitol [DTT], and 1% Nonidet P-40 [NP-40]) with freshly added protease inhibitors for 5 min at 4°C with agitation. Membrane lysis was accomplished by the presence in buffer A of 1% NP-40. The nuclei were then collected by centrifugation at 2,000 × g for 5 min, as well as the supernatant containing the cytosolic extracts. The nuclear pellet was resuspended in a cell number- adjusted volume of buffer B (20 mM Hepes, 1.5 mM MgCl₂, 300 mM KCl, 0.5 mM DTT, 0.2 mM EDTA, and 25% glycerol) with freshly added protease inhibitors as described earlier and shaken vigorously at 4°C for 30 min. Nuclear debris were collected by centrifugation at 15,000 × g for 30 min, and the supernatant (nuclear extract) was stored at 80°C until analysis by electrophoretic mobility shift assays (EMSAs). Protein concentration was determined using the Bio-Rad Protein Assay Kit (Bio-Rad, Richmond, CA).

Protein electrophoresis and immunoblotting. Equal volumes of samples corresponding to equal protein concentrations were loaded for each experimental condition, and proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (11% acrylamide). Western blots were prepared by transferring the proteins onto 0.45 µm nitrocellulose membranes (Bio-Rad), and immunoblotting was performed as previously described (8). The antisera used were the following: rabbit anti-cyclin E, anti-p21^CIP1, and mouse anti-CDK2 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:1,000 in 5% milk-PBS. Horseradish peroxidase-conjugated goat antirabbit immunoglobulin (Ig)G or antimouse IgG (Amersham) diluted 1:6,000 in 5% milk-PBS was used for revelation.

Immunoprecipitation. Cellular extracts (3 × 10⁶ cells) were incubated at 4°C overnight with either anti-cyclin E antibody or anti-CDK2 antibody. Cyclin-CDK complexes were isolated by incubation at 4°C for 1 h with 50 µl protein A-sepharose beads 6MB (Pharmacia, Piscataway, NJ). The beads were then washed and resuspended in 40 µl reaction buffer (50 mM Tris-HCl [pH 7.4], 10 mM MgCl₂, and 1 mM DTT) and 40 µl 2× SDS sample buffer (62.5 mM Tris-HCl [pH 6.8], 2% SDS, 10% glycerol, 0.025% bromophenol blue, and 5% -mercaptoethanol). The samples were then boiled for 5 min, separated by SDS-PAGE (11% acrylamide) and analyzed by immunoblotting.

Kinase assays. Cellular extracts (8 × 10⁵ cells) were incubated at 4°C overnight with either anti-cyclin E antibody or anti-CDK2 antibody and kinase assays were studied as previously described (8). The substrate used was 5 µg histone H1 (Boehringer Mannheim, Mannheim, Germany). The samples were then boiled for 5 min and analyzed by 10% SDS-PAGE. ³²P-labeled proteins were detected by autoradiography.

EMSAs. EMSAs were performed using Smad probe corresponding to consensus binding site for Smad3/4 transcription factor complex. Smad oligonucleotides were obtained commercially (Oligo-Express, Paris, France) and used in the EMSAs. The oligonucleotides included three-CAGA binding motif, the minimal recognition site for Smad3/4 transcription factor complex; 5'-TCG AGA GCC AGA CAA AAA GCC AGA CAT TTA GCC AGA CAC-3'. EMSAs were also performed using nuclear factor (NF)-1 probe (5'-AAG AAA AAC CGT TCC TAC CAT ACT AAA-3') (a gift from Dr. Y. de Keyzer, Hôpital Cochin, Paris, France). The double-stranded probes were obtained from annealing of two synthesized oligonucleotides and end-labeled using T4 polynucleotide kinase in the presence of [-³²P]adenosine triphosphate. Unincorporated nucleotides were removed by filtration through a Bio-spin column (Bio-Rad). Binding reaction was carried out in 20-µl binding mixture (20 mM Hepes, 0.1 mM EDTA, 50 mM NaCl, 50 mM KCl, 5 mM MgCl₂, 4 mM spermidine, 2 mM DTT, 100 µg/ml albumin, and 17.5% glycerol) containing 1.5 µg of the competitor poly(dI-dC) and 4 µg of nuclear proteins (11). After addition of nuclear extracts, reaction was allowed to run for 15 min at 4°C. Thereafter, 20,000 counts per min (cpm) of the probes were added and the incubation was pursued for 15 min at 4°C. The competition experiments were performed using increasing amounts of unlabeled specific Smad probe (1-, 10-, 50-, 100-, and 200-fold molar excess) and nonspecific poly(dI-dC), which were mixed with extracts before the addition of labeled DNA probe. Samples were fractionated by electrophoresis on 5% nondenaturing polyacrylamide gel in 1× Tris Borate EDTA (10 mM Tris, 9 mM boric acid, and 0.1 mM EDTA). Gels were run at 120 V for 2.5 h. After electrophoresis, gels were dried under vacuum at 80°C and exposed to X-ray Hyperfilm (Amersham) at room temperature.

Statistical Analysis

Results are reported as means ± standard error of the mean (SEM). Data were analyzed using analysis of variance, followed, when applicable, by Mann-Whitney U test for multiple comparisons against control conditions (12). Significance was assigned for P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of RA on the Proliferative Response of Alveolar Type 2 Cells Exposed to O₂

In previous studies we have reported that alveolar type 2 cells were reversibly growth-arrested by oxidant exposure (95% O₂) (9), and that RA treatment of serum-deprived type 2 cells led to a stimulation of cell proliferation with an increase in cell number in a dose-dependent manner (13). These results prompted us to ask whether RA treatment could prevent the growth inhibition of type 2 cells exposed to O₂. Exponentially proliferating cells were pretreated for 24 h in medium containing either RA (at a concentration of 1 µM) or 0.001% ethanol (solvent of RA) and then exposed to 95% O₂ for 24 or 48 h in the same media containing RA or not. When hyperoxic cells were treated with RA, no decrease in cell number was observed throughout the studied durations of O₂ exposure (Figure 1A). This was associated with increased percentages of thymidine-labeled nuclei (data not shown). To document the dose effect of RA on type 2 cell proliferation, exponentially growing cells were pretreated for 24 h with various RA concentrations (1 nM to 10 µM) and exposed to O₂ for 48 h. As shown in Figure 1B, a stimulating effect of RA on proliferation of type 2 cells exposed to hyperoxia, evidenced by an increase in cell number, was observed for concentrations as low as 10 nM (P < 0.001). A maximal proliferative response was obtained when cells were treated with 1 µM RA.

View larger version (18K): [in this window] [in a new window]   Figure 1.   Effects of RA on type 2 cell proliferation. (A) Proliferative cells were pretreated or not with 1 µM RA for 24 h and then exposed to hyperoxia (O2) for the indicated durations in the absence of RA (O2) or in the presence of 1 µM RA (O2 + RA). At the end of culture, cells were harvested and cell number was determined using a hemocytometer. *P < 0.005 versus 0 h; # or § P < 0.01, ##P < 0.001 versus O2 corresponding condition. (B) RA dose- response: cells were pretreated or not with various concentrations of RA and exposed to hyperoxia in the absence of RA (O2) or in the presence of RA at the same concentrations. Then, cell number was determined. The means ± SEM for three experiments, each performed in triplicate, are shown.

Effects of RA on Cyclin E-CDK2 Complex Activity in Alveolar Type 2 Cells Exposed to O₂

In a previous study focusing on the effects of hyperoxia exposure on the various G₁ cyclin-CDK complexes, we provided data indicating that oxidants block the entry into S phase by specifically impairing the activation of cyclin E- CDK2 complexes (8). To determine whether RA prevention of growth-arrest of cells exposed to O₂ was associated with changes in the activities of cyclin E-CDK2 complexes, we performed in vitro kinase assays using histone H1 as a substrate after immunoprecipitation of CDK2 or cyclin E. As shown in Figure 2, in the protein extracts from cells exposed to O₂ in the presence of RA for 24 to 48 h, the levels of kinase activities remained high throughout O₂ exposure.

View larger version (20K): [in this window] [in a new window]   Figure 2.   Effects of RA on cyclin E- CDK2 complex activities. Cell lysates were prepared from exponentially proliferating cells cultured under air (control) or from cells pretreated or not with 1 µM RA during 24 h and exposed to hyperoxia for 24 or 48 h in the absence of RA (O2) or the presence of 1 µM RA (O2 + RA). Cyclin E-associated kinase and CDK2 complex activities were evaluated using histone H1 as a substrate. Reaction products were separated on denaturing gels, and phosphorylated proteins were detected by autoradiography as described in MATERIALS AND METHODS. Autoradiograms from a representative assay of CDK2 complex and cyclin E-associated kinase activities are shown. Histograms show quantitative representation of complex activities obtained from laser densitometric analysis of six independent experiments. Densitometry results are expressed as percentage of control. *P < 0.02 versus control; #P < 0.05 versus O2 condition.

Effects of RA on Cyclin E and CDK2 Expression and on Cyclin E-CDK2 Complex Formation in Alveolar Type 2 Cells Exposed to O₂

The observations that the activities of the complexes formed with cyclin E and CDK2 were enhanced upon RA treatment in cells exposed to O₂ raised the possibility that the expression of cyclin E and/or CDK2 was changed. To address this question, Western blotting analysis was performed. Results are shown in Figure 3A. The anti-CDK2 antibody detected a 33-kD band. As described previously (14), cyclin E migrated as a series of bands consistent with its normal processing. No significant changes in the levels of CDK2 and cyclin E were observed in the extracts of cells exposed to oxidant for 24 to 48 h either in the absence or presence of RA.

View larger version (34K): [in this window] [in a new window]   Figure 3.   Effects of RA on CDK2 and cyclin E protein levels and on cyclin-CDK complex formation. Proteins were extracted from exponentially proliferating cells cultured under air (control) or from cells pretreated or not with 1 µM RA for 24 h and exposed to hyperoxia for 24 or 48 h in the absence (O2) or the presence (O2 + RA) of 1 µM RA. (A) Immunoblotting analysis. Proteins were fractionated on SDS-PAGE, transferred to nitrocellulose, and probed with the corresponding antibody, as described in MATERIALS AND METHODS. Autoradiograms from a representative experiment of CDK2 and cyclin E protein expression are shown. Histograms show quantitative representation of CDK2 and cyclin E protein levels obtained from laser densitometric analysis of three independent experiments. Densitometry results are expressed as percentage of control. (B) Proteins were immunoprecipitated (IP) by anti-CDK2 antibody, then immunoprecipitates were analyzed by immunoblotting through fractionation on SDS-PAGE, transfer to nitrocellulose, and probing with antibody specific to cyclin E as described in MATERIALS AND METHODS. Immunoblot from a representative experiment is shown. Histogram shows a quantitative representation of cyclin E-CDK2 complex amounts obtained from laser densitometric analysis of three independent experiments. Densitometry results are expressed as percentage of control. *P < 0.04 versus control; #P < 0.05 versus O2 condition.

The observation that the levels of cyclin E and CDK2 proteins were similar in the various conditions of O₂ exposure led us to question whether RA treatment modified the formation of cyclin E-CDK2 complexes. To answer this question, we performed immunoprecipitations with anti-CDK2 antibody followed by immunoblotting with anti- cyclin E antibody. Surprisingly, the results shown in Figure 3B evidenced a striking increase in cyclin E-CDK2 complex formation in cells exposed to O₂ that reached three-fold in control conditions at 48 h (P < 0.05). By contrast, this effect was no longer observed when oxidant-exposed cells were treated with RA.

Effects of RA on p21^CIP1 Expression in Alveolar Type 2 Cells Exposed to O₂

We have previously reported an increased expression of p21^CIP1 in growth-arrested cells exposed to O₂ (8). The stimulatory effect of RA on the activity of cyclin E-CDK2 complexes prompted us to determine whether changes in the expression of p21^CIP1 occurred in O₂-exposed cells. We first examined the consequences of RA treatment on p21^CIP1 expression at the level of messenger RNA (mRNA) through Northern blotting, depending on the presence or absence of RA. As shown in Figure 4A, the level of p21^CIP1 mRNA was reduced when oxidant-exposed cells were cultured in the presence of RA.

View larger version (26K): [in this window] [in a new window]   Figure 4.   Effects of RA on p21CIP1 expression. Total RNA and cellular proteins were extracted from exponentially proliferating cells cultured under air (Air) or from cells pretreated or not with 1 µM RA for 24 h and exposed to O2 for 24 or 48 h in the absence (O2) or the presence (O2 + RA) of 1 µM RA. (A) Expression of p21CIP1 mRNA was analyzed by Northern blotting with specific rat probe as indicated in MATERIALS AND METHODS. Autoradiogram of hybridization signals from a representative experiment is shown. Ethidium bromide staining of the gel is shown. Histogram shows a quantitative representation of p21CIP1 mRNA levels obtained from laser densitometric analysis of three independent experiments. Results are expressed in arbitrary densitometric units after normalization for RNA loading on the basis of hybridization with 18 S rRNA probe, and are reported as percentage of control. (B) p21CIP1 protein levels were analyzed by immunoblotting through fractionation on SDS-PAGE, transfer to nitrocellulose, and probing with corresponding antibody, as described in MATERIALS AND METHODS. Immunoblot of signal for p21CIP1 protein from a representative experiment is shown. Histogram shows a quantitative representation of p21CIP1 protein levels obtained from laser densitometric analysis of three independent experiments. Densitometry results are expressed as percentage of control. *P < 0.04 versus control; #P < 0.05 versus O2 condition.

To determine whether the decreased amount of p21^CIP1 mRNA in RA-treated cells was associated with similar changes in the protein itself, we determined the relative amount of p21^CIP1 in cell extracts by Western blotting. As shown in Figure 4B, the level of p21^CIP1 protein was markedly reduced when oxidant-exposed cells were treated with RA (P < 0.05 versus O₂ condition).

Effects of RA on the Association of p21^CIP1 with Cyclin E-CDK2 Complexes in Alveolar Type 2 Cells Exposed to O₂

To determine whether the decrease in cyclin E-CDK2 complex activities in growth-arrested oxidant-exposed cells despite increased complex formation could be explained, at least in part, by the induction of the CDK inhibitor (CKI) p21^CIP1, we studied the association of p21^CIP1 with cyclin E-CDK2 complexes. For these experiments, proteins in cell extracts were first immunoprecipitated with either anti-CDK2 antibody or anti-cyclin E antibody, and the immunoprecipitates were then analyzed by Western immunoblotting with anti-p21^CIP1 antibody. As shown in Figure 5, a dramatic increase in the amount of p21^CIP1 associated with either cyclin E or CDK2 was found in cells exposed to O₂ for 24 or 48 h (P < 0.05 versus control conditions). As expected from results shown in Figure 4, when oxidant-exposed cells were treated with RA the presence of p21^CIP1 in cyclin E and CDK2 complexes was barely detectable.

View larger version (20K): [in this window] [in a new window]   Figure 5.   Effects of RA on the association of p21CIP1 with cyclin E-CDK2 complexes. Proteins were extracted from exponentially proliferating cells cultured under air (Air) or from cells pretreated or not with 1 µM RA for 24 h and exposed to O2 for 24 or 48 h in the absence (O2) or the presence (O2 + RA) of 1 µM RA. Proteins were immunoprecipitated (IP) by anti-CDK2 antibody or anti-cyclin E antibody, and immunoprecipitates were then analyzed by immunoblotting through fractionation on SDS-PAGE, transfer to nitrocellulose, and probing with antibody specific to p21CIP1 as described in MATERIALS AND METHODS. Immunoblots from a representative experiment are shown. Histograms show a quantitative representation of CDK2-p21CIP1 and cyclin E-p21CIP1 complex amounts obtained from laser densitometric analysis of three independent experiments. Densitometry results are expressed as percentage of control. *P < 0.04 versus control; #P < 0.05 versus O2 condition.

Effects of RA on TGF- Pathway in Alveolar Type 2 Cells Exposed to O₂

On the basis of the observations by several authors that TGF- is a potent stimulator of p21^CIP1 expression (15), we wondered whether the TGF- pathway could be involved in the changes in p21^CIP1 expression induced by RA treatment. We first investigated the expression of TGF-1 at the mRNA level. For these experiments, RNA from oxidant-exposed cells incubated in the absence or presence of 1 µM RA was extracted and studied by Northern blotting. As shown in Figure 6, and consistent with a previous report (10), a strong induction of TGF-1 expression in cells growth-arrested by O₂ exposure was observed. When O₂-treated cells were cultured in the presence of RA, similar induction in TGF-1 mRNA expression was observed.

View larger version (65K): [in this window] [in a new window]   Figure 6.   Effects of RA on the TGF- pathway. Total RNA was extracted from exponentially proliferating cells cultured under air (Air) or from cells pretreated or not with 1 µM RA for 24 h and exposed to O2 for 24 or 48 h in the absence (O2) or the presence (O2 + RA) of 1 µM RA. Expression level of type I and type II TGF- receptor mRNAs was analyzed by Northern blotting using specific rat probes as indicated in MATERIALS AND METHODS. Autoradiograms of the hybridization signals from a representative experiment are shown. Ethidium bromide staining of the gel is shown indicating minimal variation in gel loading.

It is now well documented that TGF-1 interacts with a set of cell-surface receptors, including the type I and type II receptors, that are essential for signal transduction. In a previous study (8), we documented an induction of type I and type II receptors expression in cells exposed to O₂. To determine whether RA treatment was associated with changes in the expression of these receptors, RNA from oxidant-exposed cells incubated in the absence or presence of RA was extracted and studied by Northern blotting. As shown in Figure 6, the induction of these receptors was reduced when oxidant-exposed cells were treated with RA.

Effects of RA on Smad Activity in Alveolar Type 2 Cells Exposed to O₂

Smad proteins have been identified as mediators of intracellular signal transduction by members of the TGF- superfamily (16). In TGF- signal transduction, Smad2 and Smad3 are phosphorylated and form complexes with the common mediator Smad4, and then complexes translocate to the nucleus to regulate gene transcription. Evidence for a role of Smads as transcription factors has been presented in several studies with the identification of Smad binding elements (SBEs), composed of a Smad box, the sequence CAGA (17). To gain more insight into the mechanisms involved in the changes of the proliferative response of type 2 cells exposed to O₂ upon RA treatment, we questioned whether RA could antagonize TGF- action by interfering with Smad activity. To answer this question, gel mobility shift assays were performed with a Smad consensus double-stranded oligonucleotide containing three Smad boxes. Band-shift analysis revealed that this oligonucleotide was able to bind type 2 cell nuclear factors and that binding increased in O₂-treated cells. Interestingly, the formation of the DNA-protein complexes was reduced when oxidant-exposed cells were cultured in the presence of RA (Figure 7A). To confirm activation of Smad proteins in oxidant-exposed cells, the specificity of gel-shift complexes was assessed using competition experiments (Figure 7B). When nuclear proteins were incubated in the presence of a 1- to 200-fold molar excess of unlabeled oligonucleotides, the bands were decreased. This suggests that there are several complexes that can be formed with Smad3/Smad4, as reported in the literature (18, 19). To further confirm the specificity of Smad activation in O₂-treated cells, gel shift was also performed using an oligonucleotide to which binding is invariant. For this experiment, we used the ubiquitous NF-1 oligonucleotides, as previously reported (11). Gel retardation analysis revealed that the binding was similar in the nuclear extracts obtained from cells cultured under normoxia or under hyperoxia either without or with RA (data not shown).

View larger version (33K): [in this window] [in a new window]   Figure 7.   Effects of RA on Smad activity. Nuclear proteins were extracted from exponentially proliferating cells cultured under air (air) or from cells pretreated or not with 1 µM RA for 24 h and exposed to O2 for 48 h in the absence (O2) or the presence (O2 + RA) of 1 µM RA. (A) A total of 20,000 cpm of the -32P end-labeled Smad probe were incubated either without protein or with 4 µg of nuclear-extract proteins, and subjected to gel retardation assays as described in MATERIALS AND METHODS. The uncomplexed DNA probe is shown at the bottom. (B) A total of 20,000 cpm of the -32P end-labeled Smad probe was incubated without protein or with 4 µg of nuclear-extract proteins from cells exposed to O2 for 48 h with increasing amounts of unlabeled Smad probe (1-, 10-, 50-, 100-, and 200-fold molar excess) used as competitor, and subjected to gel retardation assays as described in MATERIALS AND METHODS. The uncomplexed DNA probe is shown at the bottom.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Oxidative stress represents a major cause of lung injury, and characterization of the key factors involved in damage induced by oxidants is crucial for the determination of molecules capable of favoring a proper remodeling of the respiratory structures. In the present work, we focused on the cell surface of the alveolar structure of the lung, which is a major target of oxidant injury. We showed that RA was capable of modulating the inhibitory effect of O₂ exposure on the proliferative response of alveolar epithelial cells. Exploration of the mechanisms involved revealed that molecular targets of RA action were cyclin E-CDK2 complexes and p21^CIP1. RA-induced downregulation of p21^CIP1 was associated with a reduced interaction of this CKI with cyclin E-CDK2 complexes, resulting in increased complex activities. In addition, analysis of Smad activity strongly suggests that the mechanisms by which RA prevents O₂-induced growth inhibition may interfere with the TGF- signaling pathway.

In a previous study, we showed that RA is mitogenic for lung alveolar epithelial cells cultured in serum-free medium (13). Data reported herein indicate that, in addition to being a mitogenic factor for lung alveolar epithelial cells in situations of growth-factor deprivation, RA could also protect these cells from the antiproliferative action of O₂ exposure. Our present findings provide new insights into the repair potentials of retinoids reported in previous studies. Cohen-Addad and colleagues (20), studying the effects of vitamin A-deficient diets in mice exposed to hyperoxia, showed evidence of increased toxic effects of O₂, including a thickening of the alveolar barrier. These data support the assumption of a role of vitamin A deficiency in the development of chronic pulmonary disease after respiratory distress syndrome. In addition, Jackson and associates demonstrated that RA was able to enhance the survival of PC12 neuron-like cells after oxidative injury generated by H₂O₂ treatment (21). The mechanisms by which RA may favor repair processes in response to oxidative stress are likely to be numerous. They certainly include increased oxidant defenses. Thus, Manzano and coworkers recently showed that RA could prevent H₂O₂ cytotoxicity in human renal mesangial cells by increasing both catalase activity and glutathion content (22).

As documented by a number of studies in various cell systems, RA appears to affect cell-cycle progression by acting mainly on regulatory proteins involved in the G₁ phase and in the G₁/S-phase transition (23, 24). Interestingly, in the present work we showed that the proliferative response of O₂-exposed cells induced by RA treatment did not involve significant changes in G₁ cyclin or CDK levels, but was associated with a reactivation of the complexes formed with cyclin E and CDK2. These results are consistent with findings indicating effects of RA on the cyclin-CDK system. As an example, in human lung-cancer cells, Weber and colleagues observed a G₁ phase arrest after RA treatment associated with a selective decrease of CDK2 complex activity without changes in cyclin E and CDK2 protein levels (25). Reciprocally, Crowe and Shuler showed in human squamous carcinoma cell lines an RA-induced increase in CDK2 complex activity (26).

Our investigation of the mechanisms involved in the activation of cyclin E and CDK2 complexes by RA led us to point out the key role of p21^CIP1. We showed a dramatic decrease in p21^CIP1 when O₂-exposed cells were treated with RA, and a subsequent reduction of p21^CIP1 interaction with cyclin E-CDK2 complexes. These data are consistent with previous studies indicating that p21^CIP1 could mediate the effect of RA on the proliferative response of various cells (27).

From the current understanding of the mechanisms that participate in the regulation of p21^CIP1 expression, the downregulation of this CKI by RA can be discussed in several ways. It is well established that RA effects are mediated through nuclear RA receptors (RARs) and retinoid X receptors (RXRs) that are members of the steroid/thyroid/retinoid hormone receptor family. These receptors form either heterodimers (RAR-RXR) or homodimers (RXR) and act as DNA-binding proteins, which can activate or repress transcription of a multitude of target genes. It has been shown that the gene coding for p21^CIP1 possesses an RAR/RXR-responsive element in its promoter (30). Recently, much attention has been focused on accessory transcription factors that interact with RARs and RXRs to modulate RA-responsive gene response (31). The repression of p21^CIP1 expression by RA reported herein could be explained by the presence of factor(s) bound to RARs and/or RXRs, which may affect their ability to heterodimerize and/or to interact with their cognate promoter element.

In addition to a possible direct action exerted through RARs, interference between TGF- and RA pathways in the regulation of p21^CIP1 should be considered. As described in a number of reports, including those of our previous studies, oxidative stress is associated with the induction and activation of the TGF- pathway (8, 10). Interestingly RA has been reported to downregulate TGF- receptors in preosteoblastic RCT-1 cells as well as in liver epithelial cells (32, 33). These data share similarities with our findings indicating a reduced TGF--receptor expression in O₂-exposed cells after RA treatment. Recently, much interest has been focused on the Smad proteins, which have been identified as mediators of intracellular signal transduction for members of the TGF- superfamily (16). The mechanisms for Smad-mediated gene activation include direct interaction with SBE, as well as cooperation of Smads with other transcription factors. Studies led to documenting the role of Smad4 in the transcriptional activation of p21^CIP1 through direct binding to SBE sequences present in the regulatory region of the gene, and/or through interaction with the transcription factor Sp-1 (34, 35). Consistent with the decreased levels of TGF- receptors was our observation of a reduced nuclear protein binding to the Smad box in extracts from RA-treated cells. Taken together, our data strongly suggest interactions between the TGF- and RA pathways in the regulation of p21^CIP1 likely to involve Smads. A demonstration of a link between TGF- pathway and a pathway involving another member of the nuclear receptor superfamily, the vitamin D receptor (VDR), has similarly been reported recently by Yanagisawa and associates through ligand-dependent interaction of Smad3 with RXR/VDR heterocomplex (36). According to their conclusions, this cooperative action may be either synergistic or antagonistic, depending on the tissue.

In conclusion, our results provide evidence of a protective effect of RA against oxidant-induced growth-arrest of lung alveolar epithelial cells. This effect appears to be linked to downregulation of p21^CIP1. Studies are in progress to further characterize the signaling pathways and factors involved in the transcriptional regulation of p21^CIP1, as well as the mechanisms through which RA mediates its proliferative response in alveolar epithelial cells.