Main

Although increases in glucocorticoids normally accompany parturition(1, 2), there is growing concern that excessive prenatal exposure to corticosteroids may have deleterious effects on development. Increased glucocorticoid levels have been reported in cases of intrauterine growth retardation, thereby offering a hormonal mechanism for the decreased size of the infant(3). Low birth weight has been reported in rat pups prenatally exposed to glucocorticoids, with subsequent elevations of systolic blood pressure in rats as adults(4). Low birth weight may also predict higher blood pressures in humans(5, 6). This line of evidence suggests that prenatal glucocorticoids may produce teratogenic effects linked to low birth weight. The effects of prenatal glucocorticoid exposure on cardiac development have not been fully examined, but there is growing evidence that prenatal glucocorticoid exposure may have profound effects on the cardiovascular system.

Although glucocorticoids are known to waste muscle mass, it has long been held that these corticosteroids spare the heart(7). For example, glucocorticoids increased the mass of embryonic chick heart(8). However, data derived from rodents treated prenatally with glucocorticoids suggest that glucocorticoids can alter development of mammalian hearts by restricting growth. Prenatal exposure to glucocorticoids can decrease DNA content of the heart(9). Neonatal dexamethasone treatment of rat pups produced decreased left ventricular mass at maturity(10). Our laboratory recently described the effects of dexamethasone exposure on embryonic myocardium grafted into the anterior eye chamber and developing without hemodynamic load. Specifically, we found that dexamethasone exposure promoted cardiac growth when sympathetic innervation to the grafted tissue was intact but suppressed growth when sympathetic innervation was prevented(11). In light of the evidence suggesting that prenatal glucocorticoid exposure could compromise myocardial growth, we tested the effects of glucocorticoid exposure on normally developing myocardium.

We investigated whether prenatal glucocorticoid exposure would modulate cardiac development by evaluating factors that reflect the maturity of the tissue. After parturition, there are changes in a variety of parameters associated with cardiac growth and function(12). Heart/body weight ratios show a developmental decrease. There is also a decline in the rate of cell division as the myocytes withdraw from the cell cycle and the heart begins to grow by increasing myocyte volume. This transition to hypertrophic growth is accompanied by an increase in ECM components, such as collagen, providing greater structural support. Biochemical markers can also reflect maturity. In rats there is a shift in the predominant MHC isoenzyme from the β-MHC isoform to α-MHC[e.g. see Lompre et al.(13)]. Evidence supporting a delay in maturation would, therefore, include higher heart/body weight ratios and rates of proliferation coupled with a relative decrease in ECM and α-MHC mRNA expression.

METHODS

Subjects. Sprague-Dawley dams (8-9 wk old) were obtained from our breeding colony derived from stock originally purchased from Taconic Farms. Embryonic d 0 (E-0) was designated on the day a sperm plug was found. Dams were housed 4/cage until E-17. At this point dams were separated, randomly assigned to experimental conditions, and individually housed to await parturition. All rats had free access to rat chow and tap water in a room maintained on a 12-h light:dark cycle. Housing conditions and surgical manipulations conform to standards set by the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and the research protocol was approved by the University of Alabama at Birmingham Animal Care and Use Committee. The animal facility is fully accredited by AAALAC.

Drug administration. At E-17, 20 dams were palpated to establish pregnancy. Half were randomly designated into the dexamethasone treatment group with the remainder designated as unmanipulated controls. Of the 20 dams, four never gave birth. The control dams were left untouched except to be placed in individual cages. A s.c., slow release, dexamethasone pellet (1 mg released over 21 d, approximately 48 μg/d, Innovative Research, Toledo, OH) was inserted at the nape of the neck of lightly etherized dexamethasone-treated dams. This dose provides a concentration of drug that is in the lower range of doses used in animal research that are considered comparable to the doses used to promote lung maturation in humans(14).

Harvesting tissue and heart/body weight ratios. Pups were removed from the dams within 12-24 h after birth, with the criteria of having milk in their stomach. The sampling of male and female pups in each litter was equally distributed, except when there were disproportions in sex. Pups were sexed and weighed to the nearest 0.01 mg. Those pups contributing to histologic studies were injected with the labeling agent. Pups were killed by decapitation. The hearts were excised and placed in ice cold Tris-Tyrode's solution to trim excess tissue. Hearts processed for histologic studies (one male and female pup from each litter) were placed in Carnoy's solution for 24 h before being blotted and weighed. These hearts provided our measures of heart weight and heart/body weight ratios. The ventricles that were designated for MHC expression (one male and female from each litter) were frozen in liquid nitrogen and stored at -80 °C until assayed.

Histologic preparations and stains for nuclei and ECM. After pups designated for histologic analysis were weighed, BrdU (100 mg/kg, i.p.; Sigma Chemical Co., St. Louis, MO), a thymidine analog, was administered. Pups were then marked and replaced with their dam. After 2 h, the pups were killed, and the hearts were prepared as described above. After fixation in Carnoy's solution, the hearts were embedded in paraffin, sectioned at 2 μm, and randomly selected sections throughout the heart were mounted on glass slides. Each heart had sections that were stained with hematoxylin/eosin for obtaining nuclear counts, stained with picrosirius red for ECM analysis, and left unstained for BrdU incorporation.

BrdU labeling of DNA-synthesizing cells. BrdU incorporation was evaluated in slides that were deparaffinized, rehydrated, blocked for endogenous peroxidase activity with 3% H2O2/methanol, and incubated with 3.5 N HCl to denature DNA. Sections were then incubated with a monoclonal mouse anti-BrdU (1:100; monoclonal mouse anti-BrdU M744, IgG1, DAKO) for 1 h at 37 °C. Linking antibodies were then used from a streptavidin-biotinylated peroxidase kit (K681, DAKO) with 30-min incubations at 37 °C. BrdU labeling was visualized using 0.05% 3,3′-diaminobenzidine tetrahydrochloride (Sigma Chemical Co.) as the chromagen with nickel-cobalt enhancement. This produced a gray residue on labeled cells.

Image analysis to determine proliferation index. A computerized image analysis system (Image 1; Universal Imaging, Philadelphia, PA) was used to determine total nuclear density from hematoxylin/eosin slides and BrdU nuclear density from BrdU-labeled slides. These numbers provided estimates of DNA synthesis that should correspond with cell proliferation. To determine the classifiers for the computer protocol, a set of pilot slides were evaluated by comparing the computer counts with manual counts. Classifiers were adjusted until there was maximal agreement between the two techniques. The standard area designating a nucleus was 28 μm2. The program rejected all particles under 5 μm2 and over 500 μm2.

Total nuclear density was estimated by counting hematoxylin-stained nuclei present in nonoverlapping fields randomly sampled throughout the ventricles at 200×, which yielded approximately 40 fields per slide. Adjacent slides were used to provide estimates of DNA synthesis as the hematoxylin interfered with the detection of BrdU-labeled and vice versa. BrdU-labeled nuclei were counted in adjacent, nonoverlapping fields for the entire ventricle at 50×, yielding approximately 30 fields per slide. The fragile neonatal atria did not provide adequate samples for evaluation. The percentage of cells synthesizing DNA was calculated by dividing the BrdU nuclear density by the total nuclear density per square mm of tissue.

Image analysis to determine ECM. Slides were stained with picrosirius red to define areas of ECM in the ventricles, of which the major component is collagen. The areas in the left ventricle, right ventricle, and septum that were reflecting red pixels were determined from pixel counts of the image at 50× by a color camera used in the Image I analysis system. The percent of of the ventricular region occupied by ECM was calculated by the area of red pixels divided by the total area of the ventricular region.

MHC isoform expression by S1 nuclease protection assay. RNA was extracted from tissue using guanidinium isothiocynate-CsCl(15). The amount of RNA was measured by OD with values of 1.5-2.0 deemed acceptable. Sufficient probe (cDNA clone Rat5 labeled at Hin fI site with [32P]ATP) to provide 5 × 104 dpm was hybridized to approximately 12 μg of RNA (one ventricle) in sealed capillary tubes for 16 h. The probe included 65 bp of 3′-coding and all of the 3″-untranslated sequences of the β-MHC mRNA. Because the coding sequences of the α- and β-MHC mRNA are nearly identical(>95%), the mRNA for both isoforms anneal to the probe, but in fragments that differ in size (the β-MHC mRNA fragment is larger than theα-MHC mRNA fragment). Aliquots of the hybridization mixture were incubated with increasing concentrations of S1 nuclease. The triplicates were incubated for 4 h at 37 °C, and the labeled probe was precipitated with ethanol and fractionated on a 5% polyacrylamide urea sequencing gel. The protected fragments of the labeled probe were detected by autoradiography. The proportion of MHC mRNA of the α and β isoforms was determined by analyzing the two resulting bands through laser densitometry (Gelscan XL, version 2.0 software). Values from three concentrations of S1 nuclease were averaged to yield the percentage of α-MHC mRNA present in the sample.

Statistical analysis. Data were analyzed using analysis of variance from the Statistical Analysis System (SAS, Cary, NC). Heart weights, body weights, and ratios and the percent α-MHC mRNA were initially analyzed by comparing the effects of treatment and sex. If no sex differences were obtained, the data were collapsed for analysis of treatment effects. Estimates of cell number were initially evaluated in terms of treatment, sex, and region of the ventricle. When no differences were found in regions, the data were collapsed to produce one estimate per slide to compare treatment and sex. Measures for ECM required analysis for treatment, sex, and repeated measures on region. Differences were considered significant at p< 0.05.

RESULTS

Effects on growth. There were 5 deaths of 81 births in 8 litters in the control group compared with 6 deaths of 82 births in 8 litters with dexamethasone treatment; thus, the average litter size and mortality was not different between the groups. Prenatal dexamethasone treatment decreased body weight by approximately 22% [F(1, 29) = 73.79, p < 0.0001 (Table 1)]. Heart weight was also decreased in dexamethasone-treated pups by 14% [F(1, 29) = 9.75, p < 0.004 (Table 1)]. However, dexamethasone increased the heart/body weight ratio by 14% [F(1, 29) = 8.77, p < 0.006 (Table 1)]. There were no sex differences in body weight or heart/body weight ratios.

Table 1 Effect on heart growth and MHC mRNA expression

Estimates of cell number and DNA synthesis. Analysis of different regions of the ventricles failed to reveal differences across regions; therefore, the data were collapsed to produce a single estimate per ventricle. The samples reflect only those slides that contributed to both total nuclear counts and BrdU labeling. Estimates of proliferative index indicate that DNA synthesis was affected by both treatment and sex[F(3, 21) = 14.51, p = 0.0001 (Fig. 1A)]. Dexamethasone treatment produced a nearly 40% greater BrdU-labeling index in dexamethsone treatment compared with controls [F(1, 24) = 5.67, p = 0.027 (Fig. 1A)]. Additionally, a higher proliferative index was observed for females [F(1, 24) = 39.12, p = 0.0001 (Fig. 1A)], but there was no interaction between sex and treatment. The number of BrdU-labeled nuclei was also greater in dexamethasone-treated pups [F(1, 24) = 4.59,p = 0.0441 (Fig. 1B)] despite no differences in the number of cells present in the ventricles of dexamethasone-treated pups compared with controls [F(1, 24) = 1.72, p > 0.20(Fig. 1C)]. Again, ventricles from female pups had a greater number of BrdU-labeled nuclei [F(1, 24) = 49.53, p= 0.0001 (Fig. 1B)], but there was no interaction between these factors.

Figure 1
figure 1

Proliferative index and cell number. Estimates of DNA synthesis indicate dexamethasone exposure and female gender are independently associated with significant increases in cell proliferation (panel A, p < 0.05). These differences are derived from the same effects evident in BrdU-labeled nuclei (panel B) where one again sees significant effects for dexamethasone exposure and female gender, without interaction. No differences in the number of total nuclei existed (panel C). The number of observations per sample is indicated in parentheses. Dex, dexamethasone.

ECM. Repeated measures analysis of variance for region of ventricle was performed on the picrosirius red-stained slides. Dexamethasone decreased the proportion of the ventricle occupied by ECM by approximately 25%[F(2, 20) = 15.26, p = 0.0001 (Fig. 2)]. Significant differences were found between the regions of ventricle, with lowest ECM content in the septum of both groups [F(1, 21) = 22.15, p = 0.0001 (Fig. 2)].

Figure 2
figure 2

ECM content. Estimates of ECM, whose main constituent is collagen, in three ventricular regions reflect decreased content when pups were exposed prenatally to dexamethasone (Dex) (p < 0.05). Lowest values were found in the septum.

MHC mRNA expression. Analysis of the percent of α-MHC mRNA represented in the ventricle samples revealed a sex by treatment interaction [F(1, 13) = 5.88, p < 0.031(Table 1)]. Specifically, prenatal dexamethasone treatment decreased the percent of α-MHC mRNA in male pups by 20%, but did not affect female pups.

DISCUSSION

By measuring a variety of indicators of maturation, our data strongly suggest that prenatal glucocorticoid exposure may delay cardiac maturation. For each measure obtained, the data are consistent with the interpretation that cardiac development, including biochemical differentiation, was delayed by prenatal glucocorticoid exposure. Prenatal dexamethasone treatment increased heart/body weight ratios, indicating a cardiac sparing effect despite somatic growth retardation. Developmentally, this ratio declines with age. In contrast to untreated neonatal pups, dexamethasone-treated pups, had a higher proliferative index and a lower ECM content. These measures could reflect cells that have not reached the transition from hyperplastic to hypertrophic growth. In normal development decreases in proliferative rate are accompanied by increases in ECM. Finally, the ventricles of dexamethasone-treated male pups had less α-MHC mRNA, again suggesting that the expected developmental transitions in biochemical markers were delayed. We were unable to follow individual pups from these litters into adulthood where we could assess whether the heart recovered from the early effects. However, in light of the research associating intrauterine growth retardation with elevated blood pressures in the adult(5), one can speculate that some of the effects may leave persistent changes in the cardiovascular system.

Our gross measures of cardiac growth indicated a cardiac sparing effect, an effect which is common when adult mammals are treated with dexamethasone(7). This sparing refers to the depletion of other muscle tissue whereas the cardiac myocytes remain undisturbed. Fluid retention is an unlikely alternative explanation as dexamethasone is typically described as having no to little potency in sodium retention(16). Promoting lung maturation is the goal of prenatal glucocorticoid therapy in women in danger of premature delivery(17). The dose used in this study is comparable to the lower doses used in humans to stimulate the maturation of prenatal lungs(14). Glucocorticoid doses used prenatally in humans have not been observed to produce adverse effects on IQ, impaired growth rate, or increased rates of infection [e.g. see Ward(17)], but specific data on cardiac development is lacking. It may be necessary to examine the effects of corticosteroid therapy on cardiac development in greater depth.

Cell Proliferation. Proliferative index can also indicate maturational processes in the developing heart. As with other organs, the myocardium in the fetus grows initially by cell division, then changes to growth through increases in cell volume(12). A previous report(9) indicated that prenatal dexamethasone exposure reduced cardiac DNA content, a phenomenon that could be attributed to actively decreasing proliferation in the heart or delaying hyperplastic growth. In the present study, we provide evidence that prenatal dexamethasone exposure increased proliferation in the developing heart by tracing cells synthesizing DNA with BrdU. Thus, the data can be interpreted as consistent with delayed maturation of the heart. It is possible that dexamethasone initially suppressed cell division such that the higher proliferative index found for the present study indicates a recovery process from an earlier suspension in hyperplastic growth. Alternatively, dexamethasone may have been actively stimulating cell division in the neonatal heart. Studies that examine dexamethasone treatment at several developmental stages may be able to distinguish among these alternative explanations. The significant effect of sex on cell proliferation remains puzzling. We found that females maintained a higher level of DNA synthesis than males. There are no previous reports of this to our knowledge; however, the sex of neonatal pups has not always been designated in research reports. It is possible that sex steroids can modulate cardiac development and that these would be a factor in the sex effects described in this study. Although hearts from both sexes were collected at equivalent times after birth, it is possible that differences in the interval and size of the first feeding could contribute to changes in myocardial tissue. This documentation of sex differences in proliferation rate of the heart merits further investigation.

Glucocorticoids appear to have the capacity to modulate cardiac development during phases of hyperplastic and hypertrophic growth. Cytosolic glucocorticoid receptors can be found in embryonic chick hearts(8). The glucocorticoid type II receptor has been identified in the rat heart during the perinatal period(18). These data suggest that mechanisms are in place whereby corticosteroids could modulate ongoing cell division or be involved in withdrawal from the cell cycle. A rise in circulating glucocorticoid levels has been documented in the third postnatal week(19), a period when myocytes grow primarily by increasing volume. Thus, glucocorticoids could influence cardiac development through a variety of mechanisms.

Glucocorticoids and growth factors. Although generally seen as mitotic inhibitors, glucocorticoids have the capacity to modulate several growth factors that either promote or suppress cell proliferation. Glucocorticoids can have anabolic effects on fetal mouse hearts in organ culture if insulin is present, but the same steroids produce catabolic effects in insulin-free media(20). Glucocorticoids decrease the expression of insulin-like growth factors which are putative growth factors for the heart(21, 22). Future studies that trace the relationship of insulin-like growth factors to the maturational markers used in this study may define the mechanism for the glucocorticoid effects on cardiac development.

ECM. If the cells in an organ are proliferating, the supporting ECM would be expected to be less pronounced. Again, in keeping with identifying delays in maturation, the areas of the ventricle exposed to prenatal dexamethasone had approximately 25% less area of ECM than the control group. A major constituent of cardiac ECM is collagen. Corticosteroids can induce dermal atrophy by interfering with mechanisms that produce collagen(23). Glucocorticoids can reduce collagen synthesis by decreasing production of collagen I, III, and IV mRNA [e.g. see Oikarinen et al.(23, 24)]. This glucocorticoid-induced reduction in collagen has also been demonstrated in embryonic chick hearts(25). Although our data support the larger body of work describing the pathogenic effects of glucocorticoid exposure on ECM components, the reduced area of ECM is also consistent with a delay in maturation.

MHC. The data we collected on MHC expression were surprising in light of research on glucocorticoid modulation of MHC in adults. If the same processes were operating in neonatal pups as in adults, one would have predicted an increase in α-MHC with glucocorticoid exposure because adrenalectomy in adult rats decreases the ratio of this isoform(26). The two isoforms, α-MHC and β-MHC, are developmentally regulated such that the β-MHC dominates in the fetal rat, whereas the α-MHC dominates in the neonatal and young adult rat[e.g. see Zak(27)]. However, pathologic conditions that increase the demand on the heart can cause a shift to theβ-MHC(28). Thus, the lower expression ofα-MHC we observed in dexamethasone-treated pups may reflect a pathologic process, but it again supports the interpretation of a delay in maturation.

SUMMARY

Prenatal dexamethasone treatment produced a higher heart/body weight ratio and proliferative index in association with a relative decrease in ECM content and α-MHC mRNA. These data indicate a heart that is immature compared with those not exposed to prenatal dexamethasone. Although at present a“delay in maturation” does not provide an explanation of the mechanisms through which the glucocorticoids exert their effects on the heart, the evidence is sufficient to warrant further investigations. The data on sex differences in the proliferative index should be pursued. Furthermore, the data justify experiments that vary doses of glucocorticoids from those lower than used in this study to the maximal doses used to promote lung maturation. Is maturation merely delayed such that after a period of time the heart will exhibit normal parameters? Are there persistent deficits that underlie the pathology for cardiovascular problems that can be found in the adult? By what mechanisms (e.g. modulation of growth factors) is maturation delayed? The research that answers these questions may shed light on the relationship of corticosteroids to low birth weight and hypertension in the adult.