Main

The possible role of vitamin A in lung development is recently being explored extensively in lung explant and cell culture models(1–4). Most of these studies use surfactant phospholipids or SP (SP-A, SP-B, SP-C) expression as markers of the response to RA(1–4). It is known that RA and its metabolite, 9-cis-RA, bind to nuclear RAR and retinoid X receptor to affect gene regulation(5–7). These ligand activated DNA-binding proteins belong to the steroid receptor superfamily. Monomer, dimer, and heterodimer forms of these nuclear receptor proteins interact with response elements to affect target genes(5–7). The RARs (α, β, and γ) show a variable expression in developing lung(1, 8), and there appears to be redundancy of function, because only double RAR mutants result in abnormal lung growth and early respiratory death(9–11). In the in vitro studies thus far, exogenously added RA and 9-cis-RA generally decrease SP-A expression, increase SP-B expression, and have variable effects on SP-C mRNA(2–4, 12).

Vitamin A stores (predominately as retinyl esters), are high in fetal rat lung from gestational d 16 to 19, after which a marked decline occurs at 19-21 d(13–16). This decrease can be produced by cesarean birth(14) and programmed a day earlier by antenatal maternal DEX(15). Because the rapid depletion in lung retinyl esters is remarkably different from the accumulation of liver retinyl ester stores occurring in the same time period, it suggests that the depletion of fetal lung retinyl ester stores has an important metabolic function in lung development. Measurements of RA, a short-lived metabolite in tissue, have not been reported in fetal lung.

Vitamin A deficiency in adult animals and human adults and children is associated with characteristic lung pathologic lesions, including a replacement of the mucus-secreting epithelium by stratified squamous keratinizing epithelium in the trachea and bronchi(13, 14). It has been suggested that vitamin A deficiency is part of the pathogenesis of neonatal bronchopulmonary dysplasia, a frequent problem of human neonates who survive prematurity with respiratory distress due to pulmonary surfactant deficiency(13, 14, 16). Various types of pulmonary morbidity and pulmonary complications also occur in vitamin A deficiency during late childhood illnesses(17).

DEX is used with increasing frequency in perinatal medicine and has profound effects on vitamin A(12, 14, 18). Clinical observations years ago suggested that adrenocortical hormones could accelerate retinol mobilization from the liver, and DEX stimulates the release of plasma retinol-binding protein from cultured rat liver cells(18). Antenatal steroids, used to promote fetal lung maturation, also affect vitamin A. DEX given to pregnant rhesus monkeys 33 d before term increases both maternal and fetal serum concentrations of retinol-binding protein at delivery 3 d later in a dose-dependent response(19). Similar observations were made in premature human infants whose mothers were treated with prenatal steroids(20). The prenatal steroids elevated neonatal serum retinol and retinol-binding protein at birth compared with a control group of premature infants. At this time, it is not clear whether these effects of antenatal and postnatal steroids are useful or detrimental to the human newborn premature infant(14–16, 18–23).

These observations reviewed above suggest that vitamin A could have an important role in lung development, but further clarification is necessary. The purpose of the work presented here was to use surfactant protein expression as a marker to determine whether maternal/fetal vitamin A deficiency in vivo has an effect on fetal lung surfactant protein development. The data show that vitamin A deficiency did not affect fetal lung SP-C or SP-A expression. Further, vitamin A deficiency did not inhibit the stimulation of fetal lung SP-C expression that occurs with the administration of antenatal DEX.

METHODS

Vitamin A-deficient pregnancies. Weanling 30-35-g female Sprague-Dawley (Harlan Industries, Indianapolis, IN) rats were obtained on d 21 of life, and randomly assigned to the control (C) group, which was fed diet containing 4 mg of vitamin A/kg diet (Harlan Teklad, Madison, WI) or a vitamin A-deficient (D) group, which was fed the same diet containing 0.06 mg vitamin A/kg of diet. This approach is similar to that used recently by others(24). These weanling rats were weighed individually every 7-14 d. At about 200 g and 75-90 d of life, the females of the D group were usually 5-20 g lighter than those in the C group. At this time, the females on both diets were mated with a male rat fed regular rat chow. The food supplied to the male and dam during mating was either C or D, being the same as the dam's previous diet. Mating was confirmed by observing a vaginal plug, and that morning was classified as d 0 of the pregnancy. At 18 or 20 d, pregnant mothers were anesthesized with intraperitoneal pentobarbital, and fetal and maternal lung and liver samples were obtained and frozen in liquid nitrogen for analyses.

In a second experiment, weanling rats in the C or D groups were grown and impregnated by the methods described above, and then subjected to the antenatal DEX protocol previously reported by others(25). At 15 d, some mothers from the C and D groups were intraperitoneally injected with 1 mg/kg DEX for 3 d, whereas the others were injected in the same manner with an equal volume (0.1 mL) of normal saline. All rats were killed on d 18, and tissue was obtained and immediately frozen in liquid nitrogen as noted above.

RP analysis. Maternal and fetal liver and lung tissue from some of each C and D pregnancies were used for RP analysis. These samples were homogenized and extracted using an established chloroform/methanol procedure(26). The extracts were assayed by HPLC(22, 27–29) (Pecosil C18, 5 μm, 15 cm) at a wavelength of 325 nm with a mobile phase for RP of methanol:chloroform:water (80:18:2). Using these solvents and flow rate of 1.5 mL/min, the peak retention time was 5.7-6.0 min for RP. The total amount of RP was quantitated from HPLC peak areas determined from standard curves that were run with each series of analyses.

Isolation of RNA and Northern analysis. Total RNA was extracted from approximately 100-200 mg of fetal lung tissue using TRIzol reagent (GIBCO BRL, Grand Island, NY) as specified by the manufacturer. Samples from each pregnancy were used. If enough tissue remained, additional analyses on that pregnancy were done. After fractionation of 25 μg of RNA on a 1% agarose gel containing 0.41 M formaldehyde and ethidium bromide, the RNA was transferred to nylon membranes (MagnaGraph; Micron Separation, Inc., Westborough, MA) by downward alkaline capillary transfer(30). The integrity of the ribosomal RNA was assessed by visualization of ethidium bromide stains of the 28 and 18 S bands. The cDNA probes for RAR-β and SP-A and SP-C was the generous gifts of P. Chambon, Strasbourg, France and J. A. Whitsett, Cincinnati, OH, respectively. To correct for RNA levels, filters were also probed with 28 S mRNA (Clontech, Palo Alto, CA). Labeling of probes and hybridization procedures have been previously described(8, 12, 31). Filters were quantitated using a PhosphoImager (Bio-Rad, Hercules, CA). The 28 S normalized mRNA data for each treatment were expressed relative to mean normalized control values. Examples of the species size of these are indicated in Figure 1.

Figure 1
figure 1

Expression of RAR-β, SP-A, and SP-C mRNAs. Extraction, fractionation on gel, transfer to membrane, and hybridization to cDNA probes are described in text. Sizes of the mRNA species are indicated (in kb), as well as the position of the 28 and 18 S mRNA bands.

Full size image

Statistical analyses were made by comparing the means of multiple analyses for each treatment with the mean control values and tested with nonpaired t test. A p value of <0.05 was considered significant.

RESULTS

Dietary vitamin A restriction effect on maternal/fetal vitamin A(RP). The maternal liver RP concentration in control animals at d 20 was 246 ± 32 nmol/g of wet weight compared with those in the D group that had only 6.2 ± 2.9 nmol/g of wet weight (Table 1). The same magnitude of differences was noted at d 18 (data not shown). Control fetal liver RP was 12-fold higher than in D fetal livers, and control fetal lung RP was about 3-fold higher than those fetal lungs from the D group(Table 1). Fetal weight at d 20 was not affected by the dietary vitamin A level (C = 3.5 ± 0.05 g; D = 3.6 ± 0.05 g).

Table 1 Concentration of RP in maternal and fetal tissue on control (C), and deficient (D) vitamin A diets at gestation d 20
Full size table

Effect of vitamin A deficiency on fetal lung SP-C and SP-A mRNA. A sample autoradiograph (Fig. 2) shows that SP-C mRNA was increased at d 20 over d 18 in both C and D fetal lungs. At d 20 gestation, the level of SP-C mRNA was the same in fetal lungs in either diet groups (C = 1.00 ± 0.09; D = 1.03 ± 0.09) (Fig. 3). Also, there was no difference in SP-A mRNA at d 20, regardless of vitamin A deficiency (C = 1.00 ± 0.07; D = 1.09 ± 0.34)(Fig. 3).

Figure 2
figure 2

Representative gels of expression of fetal lung SP-C mRNA. Methods are described in the text. Lane 1, d 18, C group; lane 2, d 20, C group; lane 3, d 18, D group; lane 4, d 20, D group.

Full size image
Figure 3
figure 3

Effect of dietary vitamin A level on d 20 fetal lung SP-C and SP-A mRNA. The number of litters and total gels analyzed for each category were as follows. For SP-C mRNA: C (control) (n = 4, 16 gels); D (deficient) (n = 3, 8 gels); for SP-A mRNA: C (n= 3, 6 gels); D (n = 3, 3 gels) There were no statistical differences between diets.

Full size image

Effect of antenatal DEX on fetal lung SP-C mRNA and SP-A mRNA. The RP levels in control maternal livers were not affected by DEX but were expectedly decreased by the D diet (Table 2). In control fetal liver, RP increased due to DEX injection. This was not found in the D group, which had greatly reduced RP levels. The higher fetal weight of the D/saline group most likely indicates that the fetuses were closer to 18.5-d gestation (Table 2).

Table 2 Fetal weight and maternal and fetal liver RP concentration of control (C) and deficient (D) groups receiving antenatal saline or DEX injections
Full size table

Antenatal DEX increased fetal lung SP-C mRNA 2-fold both in the control and D groups (Figs. 4 and 5). Fetal lung SP-A mRNA also appeared to increase in response to DEX, but statistical significance could not be established due to the smaller available sample size (data not shown).

Figure 4
figure 4

Representative gel of expression of d 18 fetal lung SP-C mRNA after antenatal maternal saline or DEX treatment from gestation d 15-17. Lane 1, C/saline; lane 2, C/DEX; lane 3, D/saline; lane 4, D/DEX.

Full size image
Figure 5
figure 5

Effect of dietary vitamin A levels and DEX on SP-C mRNA. Analyses of mRNA and probing were performed as described in“Methods.” Fetal lung tissues from four separate litters in each treatment were analyzed. Significant differences: between C/saline vs C/DEX (*p < 0.02); C/saline vs D/DEX(*p < 0.01); C/DEX vs D/DEX (x = p< 0.01) and D/saline vs D/DEX (a = p < 0.05).

Full size image

Effect of DEX on fetal lung RAR-β mRNA. The level of mRNA for RAR-β, a transcription factor generally responsive to vitamin A, was not statistically affected by the D diet in the saline injected pregnancies (Figs. 6 and 7). Antenatal DEX decreased fetal lung RAR-β mRNA in the D diet group. No effect of DEX on C fetal lung RAR-β mRNA could be statistically demonstrated in this series of experiments.

Figure 6
figure 6

Representative gel of expression of d 18 fetal lung RAR-β mRNA after antenatal maternal saline or DEX treatment from gestation d 15-17. Lane 1, C/saline; lane 2, C/DEX; lane 3, D/saline; lane 4, D/DEX.

Full size image
Figure 7
figure 7

Effect of dietary vitamin A level and DEX on RAR-β mRNA. Analyses of mRNA and probing were as described in “Methods.” Fetal lung tissues from four separate litters in each treatment were analyzed. Significant differences were noted between: C/saline vs D/DEX;*p < 0.01; C/DEX vs D/DEX, x = p< 0.02; D/saline vs D/DEX, a = p < 0.05.

Full size image

DISCUSSION

To study effects of vitamin A deficiency in the fetus, an appropriate model is necessary. RP, the predominate ester of stored retinol(18), was used as the marker for the dietary effect on vitamin A status. Our dietary vitamin A-deficient model using 0.06 mg of vitamin A per kg of diet (D) had an effect on fetal liver vitamin A that was similar to that reported previously(24). We found this lower intake level was required to decrease the fetal lung RP level. When an intermediate level diet (0.18 mg of vitamin A/kg of diet) was used, the fetal liver RP was less than half the level of those on the control diet, but the fetal lung RP was not affected (data not shown). At the D level of vitamin A, in this strain of rats, there was no effect on the number of abortions or the fetal weights at 18 or 20 d of gestation. Decreased fetal weight and pregnancy losses due to more extreme vitamin A deficiency has been reported(32). In the strain of rats used here, pregnancy losses regularly occurred with a diet containing 0.02 mg of vitamin A/kg. The adult dams at the time of pregnancy in the present experiment showed no signs of infection, hemorrhage, or neurologic manifestations seen in adult animals with more extensive and clinically apparent vitamin A deficiency. Therefore, the dietary model used here seems to be reasonable.

Inasmuch as several publications have shown that RA stimulates SP-C expression in lung tissue in vitro(2–4, 11) it was surprising to find that fetal lung SP-C and SP-A mRNA of vitamin A-deficient animals (D) was at similar levels to that in fetal lungs from vitamin A-sufficient animals (C). It appears that vitamin A has a more noticeable role in the regulation of the phospholipid component of surfactant than SP-C mRNA. Maternal administration of RA increases fetal lung surfactant phospholipids and choline incorporation into PC(33). In isolated fetal rat type II cells, which produce surfactant, RA stimulates choline incorporation into PC, despite its ability to inhibit type II cell proliferation(34). Type II cells isolated from vitamin A-deficient adult rats incorporate less choline into PC and disaturated PC compared with controls, whereas adding RA stimulates choline incorporation into both PC and desaturated PC in control and deficient cells(35), similar to the finding in fetal type II cells.

Vitamin A deficiency was also clearly established in the second experiment by the assessment of maternal and fetal liver RP (Table 2). After 3 d of antenatal maternal DEX, SP-C expression was higher in fetal lungs of DEX-treated animals than it was in saline-injected controls (Figs. 4 and 5). DEX stimulated SP-C mRNA in fetal lungs of the D group to the same or possibly greater extent as it did in the C group (Fig. 5). The observation that the D/saline fetal lung SP-C mRNA is greater than the C/saline is likely due to the fact that the fetuses of the D/saline group were apparently 12-18 h older in gestation(based on fetal weights) than the other three groups in this experiment(Table 2). An increase in SP-C mRNA would be expected with increasing fetal age. Again, either vitamin A has a minor role in the fetal rat lung surfactant protein development, as measured by SP-C mRNA, or a still lower amount of the vitamin than present here is required for surfactant protein expression. DEX stimulation of fetal lung SP-C mRNA in both control and D suggests that the stimulation of surfactant protein mRNA by DEX possibly does not involve pathways that require promotion by retinoids. On the other hand, because the D/DEX SP-C mRNA is considerably higher than that of C/DEX, the data also suggest that vitamin A deficiency potentiates an increased response to DEX. This might be an indirect effect, through a transcription factor that is enhanced by RA and which inhibits or stimulates SP-C mRNA stability or synthesis, but none of this is certain at this point.

The RAR-β mRNA was statistically lower in D/DEX versus the C/DEX fetal lung pairs (Figs. 6 and 7). This is consistent with previous observations showing that lung RAR-β mRNA is decreased in vitamin A deficiency and that it responds to exogenously added RA(27, 36). This finding was demonstrated easier under the more extreme deficiency states of those studies than were used here. DEX has specifically decreased the expression of lung RAR-β in all experiments in vivo(31) and in vitro(12) conducted thus far. This phenomenon is now demonstrated here in developing fetal lung. In previous studies, RAR-α, and RAR-γ, and RXR-β mRNA were not affected by DEX. One might speculate that the DEX suppression of RAR-β is somehow related to the effect of DEX on SP-C expression. An interaction of DEX and RA at the nuclear receptor level is possible because both are ligands for the superfamily of nuclear receptors. Additionally, it is possible that vitamin A deficiency might cause an increase in endogenous steroid production, and this may result in the observation of apparently normal levels of SP-C and SP-A mRNA in the D fetal lungs. However, we have no endogenous steroid measurement of the mothers or fetuses. The mechanism of this interaction needs to be further investigated for clarification.

The results obtained in this rat model of vitamin A deficiency cannot be directly applied to human fetal lung development. However, it seems probable that mild vitamin A deficiency in humans would not affect in utero fetal lung surfactant protein maturation. However, infants born prematurely do have stores of vitamin A that are lower than what they would be at term gestation(18), so the relative lack of vitamin A could still be important during postnatal pulmonary growth and adaptation. Although still controversial, there is considerable work published suggesting a role of vitamin A postnatally in bronchopulmonary dysplasia(13, 14, 16) and more recently, in infant respiratory syncytial virus lung infection(37, 38). Perhaps vitamin A is more necessary for postnatal lung integrity than during fetal lung development. Perhaps also, the potential affects of postnatal DEX on vitamin A levels(21–23) and functional elements like RAR-β(12, 29, 31) should be considered and investigated further.