Maternal alcohol consumption during pregnancy can affect fetal development, but little is known about the effects on the developing kidney. Our objectives were to determine the effects of repeated ethanol exposure during the latter half of gestation on glomerular (nephron) number and expression of key genes involved in renal development or function in the ovine fetal kidney. Pregnant ewes received daily intravenous infusion of ethanol (0.75 g/kg, n = 5) or saline (control, n = 5) over 1 h from 95 to 133 days of gestational age (DGA; term is ∼147 DGA). Maternal and fetal arterial blood samples were taken before and after the start of the daily ethanol infusions for determination of blood ethanol concentration (BEC). Necropsy was performed at 134 DGA, and fetal kidneys were collected for determination of total glomerular number using the physical disector/fractionator technique; at this gestational age nephrogenesis is completed in sheep. Maximal maternal and fetal BECs of 0.12 ± 0.01 g/dl (mean ± SE) and 0.11 ± 0.01 g/dl, respectively, were reached 1 h after starting maternal ethanol infusions. Ethanol exposure had no effect on fetal body weight, kidney weight, or the gene expression of members of the renin-angiotensin system, insulin-like growth factors, and sodium channels. However, fetal glomerular number was lower after ethanol exposure (377,585 ± 8,325) than in controls (423,177 ± 17,178, P < 0.001). The data demonstrate that our regimen of fetal ethanol exposure during the latter half of gestation results in an 11% reduction in nephron endowment without affecting the overall growth of the kidney or fetus or the expression of key genes involved in renal development or function. A reduced nephron endowment of this magnitude could have important implications for the cardiovascular health of offspring during postnatal life.
- nephron number
- ethanol exposure
alcohol (ethanol) consumption during pregnancy remains common in most developed countries, but recommended guidelines on safe levels of alcohol consumption vary significantly (42). Excessive alcohol consumption during pregnancy is known to result in the fetal alcohol syndrome with characteristic craniofacial dysmorphology, growth restriction, and intellectual dysfunction as the principal diagnostic features manifesting after birth (13, 30, 34, 48). Even moderate levels of alcohol exposure are now known to adversely affect learning and behavior in infants and children. While there is ample evidence that prenatal ethanol exposure can affect central nervous system development, few studies have investigated the effects on other critical organs, such as the kidney. Studies of children diagnosed with FAS demonstrate that some have renal malformations (32, 37); their kidneys are often small, misshapen, and frequently fail to rotate correctly during development (22, 37). However, the effects of less severe prenatal exposures to ethanol are relatively unexplored.
Animal studies suggest that prenatal ethanol exposure can affect the developing kidney. For example, rats that were prenatally exposed to ethanol had decreased renal function compared with controls at 90 days of postnatal age (2). Although these findings suggest tubular dysfunction, there were no overt histological differences between the kidneys of the ethanol-exposed offspring and the kidneys of the controls (2). Another study reported decreased renal protein and DNA content in 9-day-old rats prenatally exposed to ethanol, but these effects were not evident in adulthood (18). In instrumented near-term fetal sheep, maternal ethanol administration (1 g/kg maternal body wt) during gestation led to a transient decrease in urinary output (10). However, none of these previous studies have examined the effect of prenatal ethanol exposure on the number of glomeruli in the kidneys. This is important as nephrogenesis in the human is complete before birth and because a reduced nephron endowment is permanent and has been linked to the development of adult-onset diseases, such as hypertension (7, 23). Therefore, we have used an animal model, the sheep, in which kidney development is similar to that of humans to determine whether prenatal ethanol exposure has the capacity to decrease nephron endowment. As in the human, nephrons are formed only during fetal life in sheep and no new nephrons develop after birth (48).
In the present study, the effects of daily maternal ethanol administration during the period of peak nephrogenesis on glomerular number were examined in the near-term ovine fetus. Our primary objective was to test the hypothesis that repeated prenatal ethanol exposure, via maternal ethanol administration, results in smaller fetal kidneys with decreased nephron endowment. The kidneys were studied at 134 days of gestational age (DGA), as nephrogenesis is complete at 130 DGA (term is about 147 DGA) in the sheep fetus. Signaling-molecule members of the renin-angiotensin system (RAS) are involved in influencing the development and/or function of the fetal kidney and have been shown to have altered expression in the sheep (29) and rodents following developmental perturbations (35, 45). The renal RAS is important for normal kidney development as well as the regulation of arterial blood pressure in the adult. Expression of other genes important for renal function, including sodium channels have also been shown to be altered following a prenatal perturbation (5, 43). Thus, a second objective was to determine the effects of prenatal ethanol exposure on the expression of genes involved in renal development and/or function.
MATERIALS AND METHODS
All experiments were approved by the Animal Ethics Committee of Monash University and were conducted according to the National Health and Medical Research Council of Australia guidelines. At 90 DGA, 10 Merino Border-Leicester ewes carrying single fetuses were anesthetized with halothane (2%) and underwent surgery for implantation of arterial and venous catheters (11). After surgery, ewes had free access to food and water. As the alcohol infused to ewes provides a source of calories, ∼120 g (∼250 calories) of additional food were provided to ewes in the saline cohort to compensate for calories derived from ethanol.
From 95 to 124 DGA, five randomly chosen ewes received a daily 1-h intravenous infusion of 0.75 g/kg of ethanol diluted in saline starting at 9:00 AM. The five control ewes received a 1-h intravenous infusion of an equivalent volume of saline. At 126 DGA, the animals were again anesthetized and underwent fetal surgery (11). Briefly, general anesthesia of the ewe and fetus was induced by 1 g sodium thiopentone iv and was maintained by inhalation of 1–2% isofluorane in oxygen. The ewe's abdomen and uterus were incised to expose the fetus, avoiding loss of amniotic fluid. A catheter was implanted into a fetal brachial artery for the collection of fetal blood, and two catheters were sutured to the fetal skin to allow collection of amniotic fluid. After closure of all incisions, catheters were exteriorized through an incision on the flank of the ewe. After recovery of the ewe, the daily ethanol infusions were resumed on 127 DGA and were continued until necropsy at 134 DGA. From 131 to 133 DGA, fetal and maternal arterial blood samples (3 ml) were obtained immediately before starting each daily ethanol or saline infusion, and then at +1, 2, 4, 8, 12, and 24 h (i.e., immediately before the next daily dose). Blood ethanol concentration (BEC) was measured in these blood samples using an established method (6).
Amniotic fluid and plasma sample analysis.
Amniotic fluid (1 ml) was collected from one of the catheters in the amniotic fluid compartment 1 h before the start of the infusions (−1 h) and 1 h after the infusions (2 h) on 133 DGA. Maternal and fetal blood (2 ml) samples were collected from the vascular catheters and centrifuged to obtain plasma. All samples were analyzed for Na+, K+, Cl−, urea, creatinine, and total protein concentrations using a Synchron CX-5 Clinical System (Beckman). Osmolality was measured using freezing point depression (Advanced Osmometers).
Postmortem examination and fetal kidney collection.
At 134 DGA, ewes and fetuses were euthanized with a maternal overdose of pentobarbital sodium (100 mg/kg maternal body wt iv). The fetus and its kidneys were removed and weighed. The right kidney was cut into 1-mm-thick slices containing samples of cortex and medulla, which were then snap frozen in liquid nitrogen. The left kidney was cut into approximately equal halves before immersion fixation in 300 ml of 10% formalin for 24 h. After fixation, the kidney halves were washed in 70% ethanol in preparation for further sampling.
Fetal kidney sampling.
Each half of the left fetal kidney was cut into quarters, and each quarter was cut into 1.5-mm slices. Every fifth slice was collected, with the first slice chosen at random; each collected slice was cut into blocks of tissue of approximately equal size. Blocks were arranged from smallest to largest, and every sixth block was sampled, with the first being chosen at random. Sampled blocks (10–12 per animal) were embedded in glycolmethacrylate (Technovit 7100; Heraeus Kulzer, Germany) and sectioned at 20 μm thickness with a Leica DM2165 Supercut rotary microtome. The 10th and 11th sections were collected for further analysis, with the first section chosen at random, and were stained with periodic acid-Schiff (PAS) reagent.
Total kidney volume.
Total kidney volume was estimated using the Cavalieri principle. In brief, every 10th section was viewed on a Fuji Minicopy reader with a superimposed orthogonal grid (3 × 3 cm), and points that hit kidney tissue were counted. We used the formula Vkid = 5 × 6 × 10 × Ps × a(p) × t, where 5 is the inverse of the first sampling fraction, 6 is the inverse of the second sampling fraction, 10 accounts for the fact that every 10th pair of sections was analyzed, t is average section thickness, a(p) is the area associated with each grid point, and Ps is the total number of points hitting kidney tissue.
Total glomerular number.
Total glomerular number (Nglom,kid) was estimated using the physical disector-fractionator method (4). This method is considered the gold standard methodology for determination of nephron endowment (3, 31). Each kidney was randomly analyzed “blind” by the same person. Slides with complete kidney sections (and their corresponding pair) were projected at a magnification of ×298 with Olympus BH-2 light microscopes modified for projection. The fields were projected onto an orthogonal grid (6 × 6 cm). We used the formula Nglom,kid = 5 × 6 × 10 × Ps/2Pf × Q−, where 5 is the inverse of the first sampling fraction, 6 is the inverse of the second sampling fraction, 10 is the inverse of the third sampling fraction, Ps is the total area of kidney sections, Pf is the area of sections used for counting glomeruli (2 refers to the fact that counting was performed in both directions to double counting efficiency), and Q− is the actual number of glomeruli counted. Glomeruli were only counted if they were sampled within an unbiased counting frame and were not present in the adjacent projected section.
Glomerular tuft and renal corpuscle volumes.
Grid points overlying glomerular tufts (Pglom) and renal corpuscles (Pcorp) were counted to estimate mean and total glomerular and corpuscle volumes. The following formulas were used Vglom = Vv(glom,kid)/Nv(glom,kid); Vglom.tot = Vglom × Nglom,kid; Vcorp = Vv(corp,kid)/Nv(corp,kid); and Vcorp.tot = Vcorp × Nglom,kid, where V is volume and kid is kidney.
Fetal kidney morphology.
Samples of the left kidney not taken for glomerular counting were embedded in paraffin and sectioned at 5 μm thickness. Sections were stained with hematoxylin and eosin, periodic acid-Schiff, or Masson's trichrome, and were then analyzed by light microscopy.
Real-time PCR quantification.
Total RNA was extracted from samples containing both renal cortex and medulla using RNeasy extraction kits (Qiagen). A 1-μg amount of RNA was reverse transcribed into cDNA (16). Gene expression analysis was performed for angiotensin I and II, angiotensinogen, and renin, and members of the insulin-like growth factor family and their receptors (IGF1, IGF1R, IGF2, IGF2R). Six sodium ion channels (ENaCα, ENaCβ, ENaCγ, NaKα, NaKβ, and NaKγ) were also analyzed (Table 1). A comparative cycle of threshold fluorescence (CT) method was used, with 18S as an internal control (housekeeping gene). For each individual sample, the CT value for 18S was subtracted from the CT value for the gene of interest to give a ΔCT for each sample. The ΔCT of the calibrator (in this case, the mean ΔCT of the saline group) was subtracted from each sample to give a ΔΔCT value. This was inserted into the equation to give a final expression relative to the calibrator.
Data are expressed as means ± SE, except where otherwise indicated. Fetal plasma and amniotic fluid data were analyzed using repeated-measures two-way ANOVA with treatment and time being factors. Comparison of data between the ethanol and saline treatment groups was performed using two-tailed unpaired t-tests. Statistical significance was accepted at P < 0.05.
Maximal maternal and fetal BECs, between 131 and 134 DGA, were 0.12 ± 0.01 g/dl and 0.11 ± 0.01 g/dl, respectively, occurring at the end of the 1-h maternal ethanol infusion. Eight hours after starting the ethanol infusion, fetal and maternal BECs were nondetectable (Fig. 1).
Amniotic fluid composition.
There were no differences between the ethanol-exposed and control groups in amniotic fluid concentrations of Na+, K+, Cl−, urea, creatinine, and total protein at 133 DGA (Table 2). In addition, there was no difference between groups in the amniotic fluid concentrations of these substances after the 1-h maternal infusion compared with levels before the infusion.
Maternal and fetal plasma analysis.
Data on maternal and fetal plasma composition at 133 DGA are given in Table 3. Before the start of the infusions, maternal plasma osmolality was similar in the two groups; it increased in both groups after the infusions, with the elevation tending to be greater in the ethanol-infused ewes (Ptreatment = 0.005, Ptime = 0.001, Ptreatment*time = 0.164). Maternal plasma concentrations of Na+, K+, urea, creatinine, and total protein were similar in the saline and ethanol treatment groups, both before starting and after completing the 1-h maternal infusion.
Fetal plasma osmolality was not different between the two treatment groups before the start of the infusions. In both groups, fetal plasma osmolality increased after the infusions; the increase was greater in the ethanol fetuses compared with controls (Ptreatment = 0.011, Ptime < 0.001, Ptreatment*time = 0.044). Fetal plasma concentrations of Na+, K+, urea, creatinine, and total protein were similar in the saline and ethanol treatment groups.
Maternal and fetal body weights, and fetal kidney weight.
At necropsy on 134 DGA, there was no difference in maternal body weight between the ethanol and control groups (57 ± 3 kg and 59 ± 4 kg, respectively). Similarly, there were no differences in fetal body weight or fetal kidney weight (Table 4).
Fetal total nephron number.
Data for fetal kidney volume, total nephron number, and mean and total glomerular volumes at 134 DGA are shown in Fig. 2. Fetal mean and total renal corpuscle volumes are shown in Table 4. Fetal total nephron number was 11% lower in the ethanol treatment group (377,585 ± 8,325) than the saline group (423,177 ± 17,178) (P < 0.0001). Values for all other parameters were similar in the two treatment groups.
Real-time PCR analysis.
The relative mRNA levels of all genes examined were similar in the fetal kidneys of the saline and ethanol treatment groups at 134 DGA (Table 5).
Fetal kidney morphology.
There was no evidence of fetal kidney pathology or mal-development at 134 DGA in either of the treatment groups. Glomeruli showed no signs of hypercellularity, sclerosis, or enlargement. There was no evidence of renal interstitial fibrosis or cellular infiltration, and renal blood vessels appeared to be normal. In addition, there was no evidence of a nephrogenic zone in either group, indicating that the process of nephrogenesis was complete in all animals (Fig. 3).
This study in sheep has shown that repeated fetal ethanol exposure during late gestation, at a time when nephrogenesis is occurring, results in an 11% lower nephron endowment in the fetus. However, our hypothesis that the ethanol exposure would result in decreased fetal body and kidney weights was not supported, nor did we find changes in gene expression for the renal RAS, IGFs, or sodium channels in ethanol exposed fetuses. As nephrogenesis was complete at the time of nephron number analysis, the low nephron number identified in the ethanol-exposed fetuses is permanent and would persist into postnatal life. At present it is unclear whether an 11% reduction of nephron number would affect renal or cardiovascular function in postnatal life. However, it is known that a decrease in nephron number of 25–30% in the sheep results in hypertensive offspring (44). Most studies in rodents that have observed compromised nephrogenesis following development in a suboptimal fetal environment also report nephron deficits of 25–30%, and some of these studies demonstrate increased blood pressure in adult life (25, 35, 45). Of particular relevance to the present study, Langley-Evans et al. (25) reported a deficit in nephron number of 13% in rats exposed to a low-protein diet in utero. These rats subsequently developed elevated mean arterial blood pressure. Future studies to assess renal function and blood pressure in adult sheep repeatedly exposed to ethanol during fetal life, especially during the critical period of nephrogenesis, are required to determine whether the observed reduction in nephron endowment is of long-term physiological importance.
Epidemiological studies in humans affected by FAS show that fetal ethanol exposure over prolonged periods can lead to fetal growth restriction and low birth weight (1). Low birth weight is often associated with reduced kidney weight and a deficit in nephron number. In such studies it is often assumed that the lower nephron number is a result of reduced kidney growth. For example, a maternal low-protein diet throughout pregnancy in rats results in lower offspring body and kidney weights, together with a nephron deficit of ∼30% (25, 47). However, as in the present study, decreases in nephron endowment can occur in the absence of changes in fetal body weight or kidney weight (14, 35, 44). While the mechanisms causing the reduction in nephron number following ethanol exposure observed in the present study need further investigation, we have preliminary data using rat metanephric organ culture showing ethanol exposure reduces ureteric branching morphogenesis (20). Other models of reduced nephron number have shown changes in expression levels of genes regulating branching morphogenesis (14, 35), suggesting this may be a key mechanism underlying a nephron deficit.
A recent study of ethanol exposure in pregnant sheep, involving administration of the same daily maternal ethanol dosing regimen as in the present study from 116 to 118 DGA, found a 19% decrease in fetal body weight associated with transient suppression of maternal plasma IGF1 concentration (35%) and fetal plasma IGF2 concentration (28%) (19). The authors attributed the decreased fetal body weight to the suppression of maternal and fetal IGF concentrations. In the present study, although the period of alcohol exposure was much longer, we did not find any reduction in fetal body weight, fetal kidney weight, or fetal renal expression of IGF1 or IGF2. While we have not measured maternal and fetal plasma IGFs, a comparison of the present study with that of Gatford et al. (19) suggests that fetal exposure to ethanol during the third-trimester equivalent over a longer gestational period may have different effects on the IGF system than does short-term exposure. The effects on fetal growth are also likely to be dependent on the fetal BEC and the stage of gestation during which exposure occurs (1).
Several studies have suggested that exposure to a compromised intrauterine environment results in alterations in the expression of certain genes, including those involved in kidney development and/or fluid and electrolyte balance. Such alterations in gene expression may contribute to the nephron deficit or may alter renal function and thus predispose the offspring to the development of hypertension in postnatal life. The best studied system is the RAS. In rats exposed to a low protein diet throughout pregnancy, the RAS is suppressed in newborn offspring (46, 39), a time at which nephrogenesis is continuing in this species. Suppression of the RAS may directly contribute to the nephron deficit as inhibition of the RAS with angiotensin converting enzyme inhibitors (ACE-I) can impair renal development (17). However, by 4 wk after birth, when nephrogenesis is complete, most studies in rats have found either no change or an upregulation of the RAS in offspring exposed to low protein in utero (28, 33), suggesting there may be a compensatory increase in the RAS following completion of nephron formation. A similar finding has been made in fetal sheep, in which an increase in ANG I receptor expression was found during late gestation or the early neonatal period following either glucocorticoid exposure (29) or maternal undernutrition (45). In both of these sheep studies, gene expression was analyzed, after the completion of nephrogenesis. In another recent study we have shown no change in mRNA expression for components of the RAS in the fetal kidney at 130 days following growth restriction due to placental embolization even though nephron endowment was reduced (50). Another study has shown changes in the expression of fetal kidney sodium channels following maternal undernutrition (43). However, in the present study there was no altered gene expression in the ovine fetal kidney after repeated ethanol exposure at 134 days of gestation. However, it is important to recognize that these findings do not exclude the possibility that changes in the protein levels did not occur. Further studies examining gene and protein expression during the period of nephrogenesis, as well as in the adult kidney, would be of interest to determine whether there is altered expression produced by repeated fetal ethanol exposure during development or in mature offspring.
Analysis of amniotic fluid composition can provide an indication of renal function in the fetus. The amniotic fluid composition in the present study suggests that fetal renal function was unlikely to be compromised by ethanol exposure. In a previous study in catheterized fetal sheep, acute ethanol exposure led to a transient decrease in fetal urine production (10). In the rat, prenatal exposure to ethanol has been shown to produce postnatal defects in urine concentration, renal sodium conservation, and potassium handling (2). These findings raise the possibility that the ovine offspring in the present study, if studied as adults, may have compromised renal function, especially if challenged with a high sodium diet or dehydration.
Analysis of plasma samples from the ewe and fetus demonstrated that maternal and fetal osmolalities were higher in both groups after the 1-h treatment; however, the increase in osmolality was greater in the ethanol-infused animals. Ethanol has a low molecular weight and would be expected to have a marked osmotic effect in blood. Changes in maternal and fetal osmolality occurred without any changes in the concentration of sodium (or other electrolytes), indicating that the osmolality changes observed may have resulted directly from the osmotic contribution of ethanol itself. However, the effects on the saline-treated ewes were surprising. It is possible that ewes were eating dry feed during the infusion, causing water to move from blood into the rumen to aid in digestion (15, 40). Therefore, the increased plasma osmolality in the saline-infused ewes may have been due to loss of water from blood; hence the greater increase in plasma osmolality in the ethanol-infused ewes may be due to this effect in combination with the osmotic contribution of ethanol itself. However, the effects on plasma osmolality were not persistent, as both the ewe and the fetus had apparently normal values before the ethanol or saline infusion on 133 DGA after more than 30 days of maternal treatment.
Our findings on the fetal renal effects of ethanol exposure during late gestation have relevance to the human in terms of timing of fetal ethanol exposure and maternal ethanol intake. Recent data suggest that women who consume alcohol during pregnancy are more likely to drink during the third trimester than throughout gestation (12). In another study, 13.5% of pregnant women reported that they consumed alcohol regularly during pregnancy, with 6.8% being more likely to consume alcohol in the last trimester (27). In light of these data and the public misperception that alcohol consumption is less harmful to the fetus in the third trimester than in the first trimester (9, 26), knowledge of the effects of late gestational ethanol exposure on organ development is important. Our study clearly shows that alcohol exposure during nephrogenesis, which occurs largely during the third trimester in humans, can result in a reduced nephron endowment.
In relation to the maternal ethanol dose used in the present study, a mean maximal maternal BEC of 0.12 g/dl was achieved, which would require a woman of average weight to consume about three standard drinks in 1 h (8). Although this may seem to be a substantial amount of alcohol to consume, it has been shown that ∼15% of pregnant women in the USA consume more than three standard drinks in one sitting more than three times per week (8). In addition, 28.5% of pregnant women in the USA report engaging in a binge-drinking session at some point during their pregnancy, defined as consuming a minimum of four standard drinks within 3 h (38). This may be an underestimate (41) as many do not recall accurately how much alcohol they consumed (24, 36).
Perspectives and Significance
Repeated fetal ethanol exposure during the period of nephrogenesis results in an 11% reduction in nephron endowment, without causing any overt growth restriction of the fetus or its kidney. This nephron deficit occurred in the absence of altered fetal renal gene expression of components of the RAS, IGFs, or sodium channels. Further studies will be required to examine postnatal renal and cardiovascular outcomes following prenatal ethanol exposure via maternal ethanol administration. Our findings have important implications for pregnant women who drink alcohol during the latter half of pregnancy as they indicate that the developing kidney is susceptible to ethanol teratogenicity, which may result in persistent dysmorphology.
This work was supported by a Canadian Institutes of Health Research–New Emerging Team Grant NET-54014.
The authors acknowledge the expert technical assistance of Natasha Blasch, Alex Satrango, Rebecca Douglas-Denton, Debbie Arena, Andrew Jefferies, and Sue Connell.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society