INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: congenital diaphragmatic hernia; lung hypoplasia; lung development; retinoic acid; surfactant

Congenital diaphragmatic hernia (CDH) remains the most frustrating cause of respiratory failure and persistent pulmonary hypertension of the newborn, affecting about 1/2,000 live births. Despite advances in perinatal care, the outcome of CDH remains unsatisfactory (1). This anomaly, which was first thought to be merely a hole in the diaphragm potentially curable by surgical closure of the defect after birth, appears today as a disease with a complex pathophysiology, including lung hypoplasia (2) with structural (3) and functional (4) anomalies of the pulmonary vascular bed. There is an increasing body of clinical and experimental evidence suggesting that the lungs in CDH are not only small but also immature (5). The hypoplastic lung in CDH appears to be delayed in its progress through the normal developmental stages. An apparent consequence of this developmental delay is that in comparison with age-matched normal lungs, CDH lungs have both fewer bronchial branches and a delay in the differentiation of epithelial cells resulting in primary or secondary (because of decreased surfactant pool size) surfactant deficiency (6). In a previous work, we reported that vitamin A treatment decreased the incidence of CDH and reduced the degree of fetal lung hypoplasia in a rat model of CDH induced by exposure to the herbicide nitrofen (2,4-dichloro-phenyl-p-nitrophenyl ether) (7). In the present work, we explored whether this treatment affected surfactant storage.

Surfactant is a phospholipid-protein complex produced by alveolar type II cells that lines alveolar walls and prevents their collapse (8). Its characteristic components responsible for surfactant's unique properties are disaturated phosphatidylcholine (DSPC) and surfactant proteins (SP), -A, -B, -C, and -D. SP-A and SP-D play a major role in host defense (9). SP-B is a critical component essential for the formation of tubular myelin, and (together with SP-C) for achieving the rapid spreading of surfactant phospholipids and alveolar stability (10). Respiratory function at birth depends to a large part on a sufficient amount of lung surfactant.

Histologic, morphologic, and biochemical similarities have been established between the fetus/newborn with CDH and the surfactant deficient premature newborn with respiratory distress syndrome. Lungs from human newborns with CDH have reduced phosphatidylcholine content and a smaller number of lamellar granules in type II pneumocytes (11). Moreover, lecithin/sphingomyelin ratio (12) and concentrations of SP-A and saturated phosphatidylcholine (13) are decreased in the amniotic fluid of full-term babies with CDH. Lung immaturity is also a characteristic feature in experimental animal models of CDH. The surgical CDH fetal lamb model displays reduced lecithin/sphingomyelin ratio and reduced surface tension properties in bronchoalveolar lavage fluid (14). Structural developmental delay (15) and decreased DSPC (16) and SP expression (17) are also experienced by fetal mice and rats in another well-documented experimental model of CDH induced by nitrofen.

Retinoic acid, a biologically active derivative of vitamin A, plays an important role in cell proliferation, differentiation, and lung organogenesis (18). Retinoic acid binds to specific retinoic acid receptors (RAR and RXR) that are members of the steroid-thyroid-retinoid receptor superfamily. Retinoids influence the biochemical maturation of the developing lung: vitamin A and retinoic acid enhance the synthesis of phospholipid surfactant components (19) and, at least in some conditions, the expression of SP in vitro (20), whereas vitamin A- deficient rats display decreased surfactant-phospholipid and SP expression (21). Furthermore, interesting relations between retinoids and CDH have been established. Wilson and colleagues (22) found a high incidence of CDH in rat-pups from vitamin A-deficient rats. Major and coworkers (23) showed that markers of vitamin A status (i.e., blood levels of retinol and retinol-binding protein) were decreased about 50% in human babies with CDH, as compared with healthy newborns. Mendelsohn and colleagues (24) reported agenesis of the left lung and hypoplasia of the right lung in transgenic mice bearing double deletions of RAR genes. We have shown that antenatal vitamin A administration stimulates lung growth in the nitrofen model by promoting distal lung development, as measured by the enhanced radial saccular count (7). This is consistent with an earlier report by Massaro and Massaro (25) showing that intraperitoneal administration of retinoic acid to 3-d-old rats increased the number of alveoli. Finally, more recently, Greer and colleagues (26) showed in genetically engineered mice that have the lacZ reporter gene linked to the retinoic acid response element (RARE), that nitrofen decreases -galactosidase labeling, suggesting that nitrofen decreased activation of the RARE.

In the light of these findings and based on our previous observations (7), we hypothesized that antenatal administration of vitamin A in the nitrofen-induced model of CDH may stimulate surfactant synthesis. Because respiratory function at birth depends to a large part on a sufficient amount of lung surfactant, there is considerable interest in identifying regulatory substances capable of enhancing epithelial maturation in CDH.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal Model

Wistar rats were mated overnight in the laboratory. Observation of positive vaginal smears was considered as Day 0 of pregnancy (term: 22 d). The latter was confirmed by palpation on Day 11. The herbicide 2,4-dichlorophenyl-p-nitrophenyl ether (Nitrofen, 100 mg; Rohm Haas Company, Philadelphia, PA) was suspended in olive oil (1 ml) and administered to pregnant rats with the aid of an oral-gastric tube at Day 12 of gestation. Preliminary studies showed that 100 mg nitrofen given to the mother at gestational Day 12 resulted in lung hypoplasia associated with a 60 to 70% incidence of right-sided CDH. This model reliably mimics the pulmonary abnormalities observed in human babies with CDH (27, 28).

Vitamin A Administration

Vitamin A (Avibon; Rhône-Poulenc, Paris, France), 15,000 IU, was similarly given intragastrically on Day 14 to pregnant rats, either control rats or those previously given nitrofen. The dose of 15,000 IU was chosen according to Wilson and colleagues (22). Vitamin A treatment was performed on Day 14 corresponding to the stage when its effect on lung growth was the most efficient as previously shown (7).

Experimental Design

Pregnant animals were randomly assigned to four groups: (1) control rats receiving the vehicle (olive oil); (2) control rats receiving oral vitamin A on Day 14; (3) rats given nitrofen on Day 12; and (4) rats given nitrofen on Day 12 and vitamin A on Day 14. Cesarean section was performed on Day 21. Six experimental groups of fetuses were studied: (1) controls; (2) vitamin A-exposed fetuses; (3) nitrofen (N)- exposed fetuses without diaphragmatic hernia (N/no DH); (4) N/no DH fetuses from rats given vit A on Day 14; (5) nitrofen exposed fetuses with DH (N/+DH); (6) N/+DH fetuses from rats given vit A on Day 14. Fetuses were collected under maternal pentobarbital anesthesia. Each fetus was weighed and exsanguinated, and the presence of CDH was assessed via inspection through a median sternotomy using a dissecting microscope. The lungs were excised, cleared of surrounding tissues, weighed, frozen in liquid nitrogen, and stored at 80° C until processing.

Isolation of Surfactant Material and Disaturated Phosphatidylcholine (DSPC) Determination

In order to specifically analyze DSPC of the surfactant compartment, surfactant material was fractionated from lung tissue through an ultracentrifugation method previously shown to be quantitative (29). After addition of a trace amount of labeled dipalmitoylphosphatidylcholine (Amersham Pharmacia Biotech, Les Ulis, France) for recovery determination, DSPC was separated by osmium tetroxide treatment and thin-layer chromatography according to Patterson and colleagues (30). DSPC was eluted from the gel, and aliquot fractions served for scintillation counting and for phosphate determination after sample mineralization (29).

Lung DNA Content

DNA was determined in the pellets of surfactant fractionation (that contain cell nuclei) by the colorimetric diphenylamine method (31).

Isolation of RNA and Northern Blot Analysis

Total RNA was isolated from lung tissue with aid of the Trizol reagent (Gibco BRL, Grand Island, NY). Northern blot analysis of SP transcripts was performed by successive blotting with the corresponding cDNA probes as described previously (21). Quantitative analysis of signal intensity was performed by densitometric analysis of autoradiograms (NIH image software), using hybridization with an 18S rRNA probe to allow correction for variations in RNA loading.

Immunofluorescence Study

To determine the percentage of SP-B immunoreactive cells in fetal lungs, a double labeling for SP-B and for cell nuclei was performed. Lungs were fixed for cryosections in freshly prepared 4% paraformaldehyde in phosphate buffered saline (PBS) at pH 7.4. These were fixed overnight, cryoprotected with 15%, then 30% sucrose in PBS for several hours, and surrounded in O.C.T. Compound from Tissue-Tek (Sakura Finetek Europe). Frozen sections, cut on a Leica cryostat at 22° C, were rehydrated in PBS, blocked with 1% BSA in PBS for 30 min at room temperature. Labeling of alveolar type II cells was carried out with anti-SP-B antibody (1:1,000) (kindly provided by Dr. J.A. Whitsett, Cincinnati, OH) for 1h. After rinsing with PBS, the sections were stained for 45 min with a Texas red-conjugated antirabbit IgG antibody from Amersham Pharmacia Biotech. The sections were then rinsed and treated for nuclear labeling with the bisbenzimide dye (Hoechst 33258) from Sigma at 0.25 µg/ml for 30 min, rinsed twice, and coverslipped. Slides were photographed with a Nikon epifluorescent microscope through appropriate filters. For each sample, counts were made on photographs of different fields of lung parenchyma, excluding the large bronchial areas. As many as 4,000 cells were counted for each fetal lung.

Statistical Analysis

Data are expressed as mean ± SEM. Multiple comparison between experimental groups was done by ANOVA (Fisher PLSD) using p = 0.05 as the limit of statistical significance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Number of Rats

A total of 40 pregnant rats was used, consisting of seven controls, nine vitamin A-treated, 11 nitrofen-treated, and 13 nitrofen + vitamin A-treated animals. Lung samples were taken at random among fetuses of the six experimental groups for the various determinations.

Lung Growth

Fetal lung weight was not significantly different between the control and the control + vitamin A groups, whereas lung weight to body weight (LW/BW) ratio was slightly increased in the latter group (Figure 1). Both parameters were markedly decreased in the nitrofen-exposed fetuses and, consistent with previous observation (7), they were enhanced by subsequent vitamin A treatment, although normal levels were not recovered (Figure 1).

View larger version (34K): [in this window] [in a new window]   Figure 1.   Lung wet weight and lung to body weight (LW/BW) ratio in the various experimental groups. (1) control fetuses; (2) fetuses from rats given vitamin A on gestational Day 14; (3) fetuses from rats given nitrofen on Day 12, no diaphragmatic hernia (N/no DH); (4) fetuses from rats given nitrofen on Day 12 and vitamin A on Day 14, no DH; (5) fetuses from rats given nitrofen on Day 12, presence of DH (N/ +DH); (6 ) fetuses from rats given nitrofen on Day 12 and vitamin A on Day 14, presence of DH. The sole administration of vitamin A to normal rats had no incidence on lung weight. Lung weight and LW/ BW ratio were markedly decreased in nitrofen-exposed fetuses. Vitamin A increased both parameters, although normal levels were not recovered. Mean ± SE on 15 to 30 animals per group. Significant difference with control fetuses for: *p < 0.05; **p < 0.01; ***p < 0.001. Significant difference with the preceding group (no vitamin A treatment) for: +p < 0.05; +++p < .001.

DNA Content

Lung DNA content was unchanged in vitamin A-treated control fetuses (Figure 2). In nitrofen-exposed fetuses, total lung DNA was about 25% lower than in control animals and brought up to a level not significantly different from that in controls by vitamin A treatment (Figure 2). This appeared to result from correction of lung hypoplasia in N/no DH since DNA concentration did not change but was accompanied by increased DNA concentration in N/+DH, implying some reduction of mean cell size in the latter (Figure 2).

View larger version (34K): [in this window] [in a new window]   Figure 2.   Lung DNA concentration (µg/mg ww) and DNA content (µg/ lung) in the various experimental groups. (Same groups as in Figure 1.) Lung DNA concentration was unchanged by treatments, except for fetuses exposed to both nitrofen and vitamin A in which it was increased. By contrast, whole lung DNA content, that is representative of lung cell number, was markedly decreased by nitrofen either in the absence or presence of DH, and retrieved levels no longer different from that of controls following vitamin A treatment. Mean ± SE on six animals per group. Significant difference with control fetuses for: *p < 0.05; **p < 0.01; ***p < 0.001. Significant difference with the preceding group (no vitamin A treatment) for: +p < 0.05.

Surfactant DSPC

DSPC was unaffected by vitamin A treatment of control fetuses (Figure 3). DSPC concentration was slightly increased in N/no DH (Figure 3A) and DSPC pool size was increased by vitamin A treatment in this group (Figure 3C). In the N/+DH group, although DSPC concentration was unchanged (Figure 3A), the surfactant pool of DSPC per lung, was reduced by 34% as compared with that in control animals (Figure 3C). Vitamin A treatment compensated this deficit (+85% versus N/+DH, no significant difference from controls; Figure 3C), as a consequence of increased DSPC concentration (Figure 3A). DSPC was proportionally more increased than DNA since the DSPC/DNA ratio was increased 38% over that in control animals (Figure 3B).

View larger version (36K): [in this window] [in a new window]   Figure 3.   Disaturated phosphatidylcholine (DSPC) in surfactant material isolated from fetal lung in the various experimental groups. Results are expressed as surfactant DSPC concentration (nmol/mg ww and nmol/mg DNA), and as surfactant DSPC pool size in the whole lung (nmol/lung). (Same groups as in Figure 1.) The sole administration of vitamin A to normal rats had no incidence on DSPC concentration or pool size. DSPC concentration was slightly increased in nitrofen-exposed fetuses without DH and unchanged in nitrofen-exposed fetuses with DH. Because lung hypoplasia, surfactant DSPC pool size was decreased in the latter group, however, and more than restored by vitamin A treatment. DSPC pool size was also increased by vitamin A treatment in nitrofen-exposed fetuses without DH. DSPC was proportionally more increased than DNA by vitamin A treatment (increased DSPC/DNA ratio) in the presence of DH, suggesting specific stimulation of DSPC synthesis. Mean ± SE on 6 animals per group. Significant difference with control fetuses for: *p < 0.05; **p < 0.01; ***p < 0.001. Significant difference with the preceding group (no vitamin A treatment) for: +p < 0.05; ++p < 0.01; +++p < 0.001.

Surfactant Protein Expression Level (Figure 4)

SP-A. In control animals, Vitamin A decreased SP-A expression about 60%. SP-A mRNA relative amount was decreased to about the same extent in nitrofen-exposed fetuses, either with or without DH, compared with the control group. Vitamin A had no effect in N/no DH fetuses, whereas it significantly increased SP-A expression in N/+DH fetuses to about 90% of control level.

SP-B. In control fetuses, vitamin A did not change SP-B expression significantly. SP-B expression was significantly decreased by nitrofen only in N/no DH, and antenatal vitamin A significantly increased SP-B expression in this group. In the N/+DH fetuses treated or not by vitamin A, SP-B expression was not significantly different from that in control fetuses.

SP-C. In control animals, vitamin A markedly decreased SP-C expression. SP-C mRNA expression was decreased by 56% in N/no DH and 37% in N/+DH compared with the control group. Similarly to its effects on SP-A expression, vitamin A had no effect on SP-C expression in N/no DH fetuses, but increased it in N/+DH fetuses up to normal levels.

SP-B-positive Cell Count

SP-B labeling was used for characterizing SP-B producing cells, i.e., type II cells. The number of SP-B immunoreactive cells per microscopic field considered to be representative of each corresponding fetal lung was referred to the total number of parenchymal cells as determined through nuclear labeling (Figure 5). Two fetuses in each group were studied. A total of 3,000 to 4,000 cells were counted per lung. The proportion of SP-B-positive cells was found to be decreased in N/+DH (10.8%, Figure 5E) as compared with control fetuses (17.2%, Figure 5A). Vitamin A increased the percentage of SP-B-positive cells in fetuses with DH (15.5%, Figure 5G), whereas it decreased it in control fetuses from rats given vitamin A (10.5%, Figure 5C).

View larger version (147K): [in this window] [in a new window]   Figure 5.   Immunofluorescent detection of SP-B as a marker of type II cells in fetal rat lung on gestational Day 21 (A, C, E, G) and corresponding nuclear staining with Hoechst dye (B, D, F, H ) (original magnification: ×200). (A, B) Control lung showing alveoli lined with interspersed immunoreactive type II cells. (C, D) Control+vitamin A lung showing a decrease of immunoreactive cell number. (E, F ) Lung in nitrofen-induced DH (N/+DH fetuses) showing a marked decrease of immunoreactive cell number; the lung has a compact aspect. (G, H ) Lung in N/+DH fetuses after vitamin A treatment showing an increased number of immunoreactive cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CDH results in profound lung developmental disturbances with marked hypoplasia. The hypoplastic lung appears to be delayed both in its process of normal outgrowing and parenchymal differentiation. As a consequence, fewer bronchial branches and delayed differentiation of epithelial cells are documented in the literature (6). Surfactant deficiency has been evidenced in the surgical model of CDH (32), although this is disputed in human subjects with CDH (33, 34). Results reported herein show that lung immaturity is present also in the experimental nitrofen-induced model of CDH that is therefore suitable to study this particular aspect of lung development. The present study was undertaken to determine whether vitamin A treatment might oppose the delaying effect of nitrofen upon the storage of the various surfactant components. Our findings indicate that a beneficial effect was obtained for the major surfactant phospholipid DSPC, as well as for the level of expression of SP-A and SP-C.

The rationale for using vitamin A is based on previous observations from our group showing that vitamin A enhanced DSPC synthesis in vitro (35) and in vivo (19), whereas mild vitamin A deficiency in the pregnant rat led to decreased lung DSPC content and relative level of SP transcripts in the offspring (21). Moreover, type II cells isolated from the lungs of vitamin A-deprived adult rats also display a reduced rate of DSPC synthesis (36). In the present study, surfactant DSPC was found to be decreased in nitrofen-exposed fetuses with CDH. Consistent with previous observations, vitamin A treatment increased DSPC concentration and pool in CDH fetuses over the control levels. Although enhanced growth of the organ may account to some extent for the increase in surfactant pool size, the increase of DSPC concentration and DSPC/ DNA ratio argue for a specific stimulation of DSPC synthesis (Figure 3).

Consistent with previous observations, vitamin A decreased SP-A and SP-C mRNA levels when administered to normal fetuses. Decreasing effects of vitamin A or of its metabolite, retinoic acid, on SP-A or SP-C expression have thus been reported in vivo (19) and in vitro (37, 38), although in vitro stimulating effects have also been reported for low concentrations (20). Although SP-A and SP-C expression was decreased in nitrofen-exposed fetuses both with and without CDH, vitamin A restored normal levels in fetuses with CDH only, suggesting that mechanisms leading to lowered SP expression could be different in these two experimental groups. In contrast with SP-A and SP-C, nitrofen-exposure significantly affected SP-B expression only in fetuses without CDH, although a trend towards a decrease was observed in fetuses with CDH. Vitamin A increased SP-B in fetuses without CDH and tended to increase it in fetuses with CDH, consistent with the enhancing effects previously reported for retinoic acid in lung explants (37). The presence of decreased SP-B expression only in fetuses without CDH reinforces the assumption that alterations in SP expression proceed through different mechanisms in nitrofen-exposed fetuses with and without CDH.

As evidenced by SP-B immunolabeling, the nitrofen-induced decrease of DSPC and SP expression in fetuses with CDH, and the vitamin A-induced restoring effects correlate with decreased and increased proportion of alveolar type II cells in these experimental groups, respectively. It therefore appears that decreased surfactant synthesis in this CDH model results from an altered process of alveolar type II cell differentiation. Because vitamin A increased the proportion of type II cells and the global expression of their markers to similar extents, vitamin A appears to have favored differentiation of a larger number of epithelial cells rather than having enhanced surfactant synthesis on a per cell basis.

With regard to the mechanism of action of vitamin A, a first assumption is a possible competition with nitrofen for various receptors, including RAR or RXR. Nitrofen has indeed been reported to bind to thyroid hormone receptors (39) that belong to the same nuclear receptor family. More recently, demonstration that nitrofen interferes with RARE activation has been provided (26). Although it may account for a part of vitamin A effects, such a mechanism, however, would not explain why nitrofen-exposed fetuses responded differently to vitamin A according to the presence or absence of CDH. Alternatively, the apparent discrepancy between vitamin A effects on intact, N/no DH, and N/+DH fetuses could be explained by possible vitamin A deficiency in fetuses with CDH. Vitamin A deficiency in CDH has indeed been suggested by several investigators (22, 26). Because vitamin A deficiency is known to decrease SP expression (21), the low level of SP-A and SP-C expression in fetuses with CDH may result from insufficient vitamin A supply, and vitamin A administration may have compensated a deficit at either level in the retinoic acid-signaling pathway. The fact that vitamin A failed to restore SP-A and SP-C expression in N/no DH fetuses suggests that these would have, by contrast, unaltered vitamin A status. Because vitamin A is stored as retinol esters in mesenchymal cells of the fetal rat lung (40), a possible effect of lung compression in the presence of CDH might be to limit the amount of retinol ester storing tissue. This would in turn reduce retinol availability thus accounting for responsiveness to administered vitamin A that may have restored the retinoid status in fetuses with CDH.

Previous attempts to promote surfactant production in CDH had been undertaken by the antenatal administration of glucocorticoids, a well-established therapy to accelerate lung maturation (41). This treatment effectively increased surfactant phospholipid and SP synthesis (42) and improved lung compliance and oxygenation (43) in experimental models of CDH. To date, only glucocorticoids have shown consistent clinical benefits in accelerating functional maturation of the fetal lung (41). Because glucocorticoids may, however, have adverse effects on lung growth and alveogenesis (44, 45) and on general development in the long term (46, 47), alternative or synergistic treatments are therefore desirable. An interesting aspect of the use of retinoids in this respect is that retinoic acid prevents the inhibition of alveolar septation caused by glucocorticoid exposure (25).

We conclude that antenatal treatment with a single dose of vitamin A on Day 14 restores lung maturation in the nitrofen-induced rat model of CDH in terms of surfactant DSPC storage and of SP expression. Although the molecular mechanism underlying the vitamin A-induced beneficial effects calls for further investigation, these findings point out vitamin A as a potential useful agent for correcting lung immaturity in CDH.