The exposure of C57BL/6 females to a high-fat diet containing 60% calories from fat induced maternal obesity, and fetal resorption and SGA status in fetuses at gestational day 17 (E17). Although litter size was identical in both feeding groups, more than 10% of fetuses were resorbed in the HF group at the time of sacrifice (day E17, ). When only viable fetuses were accounted, maternal exposure to a high-fat diet also induced intra-uterine growth restriction, as determined by fetal body weight (). According to published data, the most plausible mechanism responsible for fetal resorption is related to the induction of a pro-inflammatory environment by maternal diabetes, during embryo implantation (Jawerbaum and Gonzalez, 2006
). Another aspect to be considered is the potential alteration in energy expenditure, related to gestational outcomes. In C57BL/6J mice a higher basal metabolic rate (BMR) during gestation was associated with gestational weight loss indicative of fetal resorption (Johnston et al., 2007
). The increased BMR in mice receiving high-fat diets is probably due to induced uncoupling protein-1-mediated thermogenesis (Kus et al., 2008
In the present study, hippocampal development was assessed only in fetal males. Although no data are available to suggest that a maternal high-fat diet alters fetal brain development in a sex-specific manner, studies involving intrauterine exposure to maternal hyperglycemia (expected in our study) indicated sex-dependent deficits in learning and memory, suggesting that the neurodevelopment in male and female progeny may be differently affected (Kinney et al., 2003
The assessment of hippocampal development using markers of cell proliferation (phosphorylated histone H3, pH3), apoptosis (activated Caspase-3), and early neuronal differentiation (Calretinin) revealed marked differences in hippocampal formation between male fetuses from HF and CT mothers. These outcomes followed a total exposure to the HF diet of approximately 12.5 weeks (10 weeks prior to pregnancy and until embryonic day 17). The assessment of cell proliferation revealed interesting, divergent outcomes in the VZ/SVZ (increased neurogenesis) versus DG area (decreased proliferation) in the HF group.
Several studies using rodent models have addressed the relationship between high-fat dietary intakes or obesity, and hippocampal development. However, most of them did not refer to prenatal, but rather to postnatal development, indicating that high-fat diets negatively alter the neurogenesis within the dentate gyrus of adult rodents, and these changes are persistent (Lindqvist et al., 2006
; Yu et al., 2009
; Lee et al., 2000
Only few studies have addressed the hypothesis that high-fat maternal intakes, prior and during gestation, may alter fetal brain development. The short exposure of pregnant Sprague Dawley rats to a high-fat diet (50% fat with mainly lard, beginning embryonic day E6) increased the proliferation of cells within the hypothalamic areas of E11 to E15 rat fetal brains, which is the period of peak cell birth in rat hypothalamus (Chang et al., 2008
). In a second study, which is the closest to our model, Tozuka et al reported that, when 4-week old C57BL/6 mice were fed with a similar high-fat diet (57.5% calories from fat) for 6 weeks prior, and during gestation, the neurogenesis in the DG was decreased in the offspring, beginning with postnatal day P21 and until P70, as determined by BrdU incorporation (Tozuka et al., 2009
). However, in opposition to our findings, Tozuka et al reported no differences in DG neurogenesis at embryonic day E18 (one day later than our time point at E17), and no changes in fetal body weight at E18 (Tozuka et al., 2009
). Several factors could account for differences in our reported outcomes. The most significant difference is the duration of dietary exposure prior to gestation (4 weeks longer exposure in our study, resulting in a bigger difference in maternal body weight between the HF and CT groups). A second factor could be the age difference when the exposure to the high-fat diet started (a 4-week difference).
Here we report that a maternal high-fat diet alters hippocampal and cortical neurogenesis, and neuronal differentiation in E17 mouse fetal brains. Interestingly, we report opposite changes in the hippocampal and cortical VZ/SVZ areas (increased proliferation of neural progenitors) versus decreased cell proliferation within the DG area. These changes associated with decreased apoptosis within selected hippocampal areas (Ammon’s horn and fimbria), and decreased calretinin levels within DG, the later indicating decreased early neuronal differentiation.
Since the peak of hippocampal and cortical cell proliferation is between embryonic days E13 and E16 (Rodier, 1977
; Rodier, 1980
), followed by a decrease in cell proliferation, one could speculate that the increase in cell proliferation within the VZ/SVZ areas of HF fetal brains represents developmental delay (earlier stage of development with increased proliferation). Because the neural progenitors in DG migrate from the hippocampal neuroepithelium, specifically from the dentate notch (Altman and Bayer, 1990b
; Altman and Bayer, 1990a
; Pleasure et al., 2000
), the decreased cell proliferation within the DG in HF brains could also be interpreted as a second mark of hippocampal developmental delay. However, in the absence of a time-course examination of neurogenesis, the hypothesis of developmental delay should be taken into consideration cautiously.
In this study, apoptosis in HF fetal brains was decreased in specific hippocampal areas (Ammons’ horn, CA and Fimbria, Fi, ). During fetal brain development, apoptosis has been documented to be a contributing factor enabling the selection of appropriate cells before they complete their differentiation in postnatal life (Blaschke et al., 1996
). In this context, decreased apoptosis reported here in the HF male fetal hippocampi could be interpreted as constitutive to developmental delay. However, other factors like psychological and pharmacological stressors have been implicated in decreased apoptosis within the CA hippocampal region (reviewed in (Lucassen et al., 2006
Decreased calretinin protein levels within the DG of HF fetal brains could be related to reduced early neuronal maturation. In both fetal and adult brains, calretinin is a transient marker for newly generated neurons and, thus, is considered a marker for early postmitotic neuronal differentiation (Soriano et al., 1994
; Brandt et al., 2003
This study has important limitations. The highly unbalanced diet containing 60% calories from saturated fat mainly is less representative for the obesogenic diets in humans (35% energy from fat (Miller et al., 1990
)), and further animal studies should validate whether the alterations reported here are still present using diets with a more moderate fat content. Secondly, the experimental design did not allow differentiating between the potential roles that either maternal obesity, or the diet per se
, had in inducing the described outcomes.
The relevance of rodent studies to the human epidemiological data on the relationship between maternal obesity, or high-fat diets intakes, and fetal outcomes is yet unclear. A recent meta-analysis indicated that, overall, obese mothers are at increased risk of having babies with various birth defects, including spina bifida, cardiovascular anomalies, cleft lip and palate, anorectal atresia, hydrocephaly, and limb reduction anomalies (Stothard et al., 2009
). Two epidemiological studies indicated that obese mothers are also at increased risk of having babies that are either large (LGA) or small for gestational age (SGA). In a prospective study, nulliparous women having singleton babies were at increased risk of having SGA or LGA babies (Rajasingam et al., 2009
). Although the results of this referenced study were in contrast with a previous study from the Swedish Birth Registry, which reported that obesity in nulliparous women protected against fetal growth restriction (Cnattingius et al., 1998
), a second study in Chinese overweight women indicated that women who had BMI > 23 were at increased risk of having an SGA baby (Leung et al., 2008
). Maternal overweight and obesity increased both the risk for fetal growth restriction and sudden intrauterine unexplained death (Froen et al., 2004
In mouse models, different research groups reported discordant results on the relationship between maternal obesity and fetal growth, even when using the same mouse strain (C57Bl/6). When 4-wk old females were fed a high-fat diet (57% calories from fat) for 6 weeks until mating and during pregnancy, no differences in body weight were reported at gestational day 18 and at birth, between the male pups from overweight mouse females and controls (Tozuka et al., 2009
). Meanwhile, an independent study using 8-wk old females exposed to a high-fat diet (32% energy from fat), reported 43% higher fetal weights at E18.5 in the high fat-fed group than in the control group, although there were no statistically significant differences in maternal weight at the time of mating (Jones et al., 2009
). In an earlier study, Samuelsson et al reported no changes in weight at birth, when 3-wk old females were exposed to a high-fat diet 6 weeks prior, and during gestation (Samuelsson et al., 2008
). Interestingly, the first two studies mentioned above (Jones et al., 2009
; Tozuka et al., 2009
) reported opposite outcomes regarding the relationship between maternal insulin resistance and fetal growth, and in opposition to the classical paradigm (i.e. maternal diabetes associates with increased birth weight). While Tozuka et al reported increased insulin resistance associated with no changes in birth weight of pups ((Tozuka et al., 2009
), Jones et al reported no changes in maternal insulin sensitivity, yet bigger pups at birth (Jones et al., 2009
). Potential causes for such different outcomes may include the age of pregnant mice and the duration of dietary treatment, the difference in fat composition and amount, suggesting that different fetal outcomes may be induced by different windows of opportunity for maternal exposures.
Whether the described alterations in hippocampal neurogenesis are transient or not is subject of further research. A previous study indicated that such changes may be long-lasting until post-natal day 70 (Tozuka et al., 2009
Several mechanisms could account for the described outcomes. Alterations in hippocampal neurogenesis were related to peroxidized lipid accumulation in the dentate gyrus (Tozuka et al., 2009
). Other studies indicated that maternal supplementation with n-6 PUFA led to significantly greater protein kinase C (PKC) activity in the hypothalamus and moderately less PKC activity in the whole brain of mouse offspring (Raygada et al., 1998
). When on a high-fat, dextrose rich diet, adult male rats and mice have decreased brain-derived neurotrophic factor (Bdnf) levels in the cortex and hippocampus (Yu et al., 2009
), associated with alterations in discrimination reversal (Kanoski et al., 2007
) and with reduced hippocampal spatial learning performance (Molteni et al., 2002
). Another possible mechanism involved in the dietary regulation of hippocampal development is related to the leptin receptor, which is expressed in hippocampus (Louis and Myers, 2007
), and may have an important role in facilitating memory and learning (Oomura et al., 2006
). Obesity reduces the expression of leptin receptor in liver and hypothalamus (Liu et al., 2007
), while lean mice over-expressed it in the dentate gyrus of the hippocampus (Lin and Huang, 1997
In conclusion, we report, for the first time to our knowledge, that prenatal and gestational exposure to a high-fat diet induces SGA status which associates with fetal hippocampal developmental alterations at embryonic day 17, as determined by markers for cell proliferation, apoptosis, and early neuronal differentiation. The alterations in cell proliferation within the two main neuronal proliferation areas (VZ/SVZ and DG) are opposite and follow the pattern of hippocampal development, while apoptotic changes are confined to the Ammon’s horn and Fimbria. These outcomes associate with decreased early neuronal differentiation within the fetal dentate gyrus. Further studies are warranted to determine the mechanisms responsible for these outcomes.