Introduction

Throughout fetal development, the main nutritional source for the fetus comes from the mother. As a result, the quality of the maternal diet affects growth and development of the fetus during this period. Barker et al. [1, 2] have shown that in humans, maternal under nutrition leads to increased risk of cardiovascular disease in offspring. The mechanisms involved however are poorly understood. Nevertheless, experimental animal models whereby the maternal diet is manipulated have provided insight into the links between in utero environments to susceptibility to disease in later life. Rodent offspring from dams that were fed a protein-restricted diet during pregnancy later developed hypertension in adulthood [3, 4]. However, developmental programming of cardiovascular disease in adulthood is not limited to maternal dietary restriction alone. More recently, we and others have shown in rodents that long-term feeding dams on high fat (HF) diet can also lead to the offspring developing vascular dysfunction and hypertension [5, 6]. And also that decreasing the cholesterol levels in these HF diet consuming hypercholesterolemic pregnant dams with drugs called statins reduces cardiovascular risk factors in the offspring [7]. Nonetheless, the results of our study [7] on statins treatment during late pregnancy should be interpreted with caution because statins are still contraindicated in pregnancy due to their interruption in total cholesterol synthesis which is potentially harmful to the pre-developed fetus. Yet pravastatin, the statin used in our study [7], is hydrophilic and has previously been used during this period as it is least likely to cross the placenta into the fetal bloodstream [8].

Previous studies have also shown that changes in the maternal diet during pregnancy are linked with an increase in fetal left ventricular mass [9]. The cardiomyocytes go through terminal differentiation at or after birth and thus characterised by the transition from hyperplastic growth to hypertrophic growth [10, 11]. Many genes involved in cell cycle progression and associated with cardiac hypertrophy are reported [13].

Quiescent cells are found in G0 phase of the cell cycle and remain in a state where mRNA and protein syntheses are minimal and can re-enter the cycle in the G1 phase when it is stimulated, for example, by a growth factor binding to its extracellular receptor [14]. However, external factors do influence the cell cycle and not all activity occurs in the latent phase. Throughout G1 the cell synthesises a sequence of mRNAs and proteins that are needed for DNA synthesis in S phase following which the cell enters a second gap phase, G2 phase. In the G2 stage of the cell cycle, the cell makes extra mRNA and proteins in preparation for mitosis. This is where the cell divides into two daughter cells [14]. There are many checkpoints in the cell cycle that ensure cell division continues as usual. As soon as the cell has passed this point, a further round of DNA replication and cell division is completed. This is except in fully differentiated cells as in adult cardiac myocytes where binucleation and polyploidy cause S phase to occur in the absence of cell division [14].

Particular molecules involved in various stages of the cell cycle determine the progression of the normal cell cycle to assist with cell replication. Cyclins and cyclin dependent kinases (CDK) are a group of catalytic kinases which play a critical role in formation, activation and inactivation of the progression of cell from a quiescent state, G1 phase into the cycle [14]. Cyclins are a group of proteins that are synthesised and destroyed during each cell cycle, as their name implies. Cyclin G1, one of the G1 phase cyclins, is up-regulated in nutrient restricted fetal heart and has been found to be involved with cardiac hypertrophy in adults. This gene is shown to encourage entry into the cell cycle therefore increasing cell numbers. Protein synthesis is initiated by cyclin G1 to cause cardiac hypertrophy, rather than DNA synthesis through entry into the cell cycle in an adult heart [14]. It has been reported that cyclin G1 is expressed in the only phase of the cell cycle more prone to be influenced by external stimuli; the G1 phase [15]. It seems that changes in this gene expression may be either a cardio-protective response due to a limited nutrient supply, a response to myocyte stretch from increased systemic vascular resistance or possibly a response to a changed endocrine situation.

Amongst the molecules which are considered vital in ventricular dysfunction, brain natriuretic peptide (BNP) is an important marker of cardiac function. It is actually an indicator of increased intra-cardiac pressure not taking into account whether the raised pressure is caused by left ventricular (LV) systolic dysfunction, LV hypertrophy or valve disease [16]. However, BNP is secreted to a lesser degree in fetal life compared to adult life [17]. BNP is also thought to be a marker along with a modulator of cardiac hypertrophy [17, 18] with a significantly increased expression of BNP in cardiac hypertrophy and cardiac stress.

SPRR 1A, a small proline-rich repeat protein is known to be a stress inducible cardio-protective protein in cardiomyocytes responding to either ischemic or biochemical stress [19]. Histological results discovered that SPRR1A induction after mechanical stress from pressure overload was restricted to myocytes that surrounded necrotic lesions. In post-infarcted rat hearts, a similar expression pattern was found. In vitro and in vivo over-expression of SPRR1A protected cardiomyocytes from ischemic injury [20]. Therefore SPRR1A has cardio-protective effect against ischemic stress. A different role for SPRR1A may be its association with actin cytoskeleton. It has been reported that ectopic expression of SPRR1A may prevent disruption of myofibrils and accumulation of nuclear actin after ischemic stress [21]. It further suggested that the cytoprotective action of SPRR1A may be due to mechanical stabilization of myofilament/Z disc structure in cardiomyocytes, therefore preventing cells from permanent damage.

It remains, however, to be determined whether prenatal exposure to maternal HF diet results in changes in the expression of genes involved in cell cycle progression in the developing heart resulting in cardiac remodelling. We hypothesize that hypercholesterolemic condition during pregnancy leads to long term changes in these genes in the adult offspring heart; and that the cholesterol-lowering effects of pravastatin given to pregnant dams consuming a HF diet influence the expression of these genes in the remodelling that may contribute to cardiac stress and increased blood pressure and cardiomegaly of the offspring heart.

We therefore investigated the consequences of lowering maternal cholesterol with pravastatin on these high fat-fed dams in late gestation and look at the expression level of genes involved in cell cycle progression (cyclin G1), ventricular remodelling brain natriuretic peptide (BNP), and environmental stress response small proline-rich protein 1A (SPRR 1A) in the offspring LV.

Methods

Animal and experimental protocol

The blood pressure data, heart weight data and LV tissues in the offspring used in this study came from a previous study [7]. In that study, female C57BL/6 mice (Charles River Laboratories, UK) were maintained under a 12 h light/dark cycle at a constant temperature (22 ± 2 °C) with water and food available ad libitum. All animal procedures in this study were in accordance with the United Kingdom Animals (Scientific Procedures) Act 1986. The female mice were allocated randomly to be fed either a high fat diet (HF 45 % kcal fat Special Diet Services UK) or standard laboratory chow (C 21 % kcal fat) from 4 weeks of age. This HF diet was complemented with added animal lard, vitamins and minerals, choline and protein and this diet has been used in previous studies to induce hypercholesterolemic conditions [7]. At 10 weeks these females were time mated and housed individually after confirmation of successful mating, i.e., the presence of vaginal plug.

Half of the pregnant mice on a HF diet were given a water soluble HMG-CoA reductase inhibitor pravastatin in their drinking water (5 mg/kg body weight/day) from the second half of pregnancy and during lactation (HF − S) to lower their cholesterol levels. The dose of pravastatin was standardised on the basis of our previously published work [7]. Weaned offspring (3 weeks post-partum) were then fed a HF diet until adulthood. This created dam/offspring dietary groups HF/HF and HF − S/HF. These groups were then compared with offspring on standard chow (C) post weaning from mothers that were also on a chow diet (C/C). The blood pressure was measured by tail cuff plethysmography.

The offspring were killed by CO2 inhalation and cervical dislocation at 30 weeks of age. Left ventricular heart tissue samples were then dissected from the male and female offspring weighed and stored at −80 °C until analysed later for changes in the mRNA levels of cyclin G1, BNP and SPRR 1A by RT-PCR.

RNA extraction

Total RNA was extracted from the LV heart tissue using Tri-reagent. Chloroform was then added to separate the mixture into 3 phases containing RNA, DNA and organic material. The aqueous phase was transferred to a fresh tube and RNA was precipitated using isopropyl alcohol. The RNA precipitate was washed with 75 % alcohol and air dried. The RNA precipitate was then dissolved in RNase-free water. A spectrophotometer was then used to determine the A260/280 ratio for every sample and the total RNA concentration for individual samples was calculated. All samples were then diluted to 0.5 μg/μl RNase-free water and agarose gel electrophoresis was used to examine the integrity of RNA. Genes of interest were then analysed by the synthesis of DNA from these RNA samples.

cDNA synthesis

The synthesis of cDNA was carried out for the standard curve, positive and negative controls and the RNA samples to be analyzed. An RNA sample (with the highest quality A 260/280 reading) was used as make up the standards. RNA sample diluted to 0.5 μg/μl was allocated as Standard 7(S7). A series of dilutions were then carried out, double diluting down to produce S6, S5 and so on down to S1. The amount of RNA used for the standard curve varied from 1 μg (S7) to 0.015625 μg (S1). Six positive controls and two negative controls were used. The cDNA for negative controls were made by 1 μg of RNA without the MMLV reverse transcriptase. We diluted the samples to be analyzed from 0.5 μg/μl to 0.05 μg/μl so that they fell approximately within the middle of the standard curve. For all of the samples, cDNA was synthesized using 2 μl of the RNA in a volume of 25 μl containing 0.4 μg random oligo primers, RT-buffer, 12.5 Nm PCR mix (dNTPs), 25 units RNAsin, 200 unit MMLV reverse transcriptase (Promega Southampton UK) and an appropriate amount of ultra-pure water. The RNA was mixed with the random oligo primers and made up to 15 μl with ultra-pure water; then it was then heated at 70 °C for 5 min. The mixture was then rapidly cooled on ice. Following this the remaining reagents were added and made up to 25 μl with ultra-pure water. We then incubated the samples at 37 °C for 1 h, 42 °C for 10 min then an enzyme activation step at 75 °C for 10 min.

Real-time PCR for cyclin G1, BNP and SPRR1A

Specific probes and primers for cyclin G1, BNP and SPRR1A as well as the reference housekeeping gene β-actin were designed based on their published sequences using the Primer Express™ (V1.0) software (See Table for sequences of the probes, forward and reverse primers). For each sample, 3 μl cDNA was mixed with 33 μl PCR master mix (dNTPs, Hot Goldstar DNA polymerase, MgCl2, Uracil-N-glycosidase, stabilisers and passive reference), 2 μl forward and reverse primers and 2 μl probe at the optimum concentration. This was made up to a final volume of 66 μl with 24 μl of ultra pure water. The no-template control (NTC) sample had the cDNA omitted and replaced with 3 μl ultra pure water. The final mixture was then aliquoted in duplicates into a 96 well PCR plate (30 μl per well). PCR amplification was performed for 50 cycles. PCR condition consisted of 2 min at 50 °C and 10 min at 95 °C. Following this was 50 cycles of denaturation at 95 °C for 15 s and then annealing at 60 °C for 1 min (Table 1).

Table 1 Sequences of probes and primers for β-actin, cyclin G1, BNP and SPRR1A for Taqman RT-PCR
Full size table

An increasing fluorescence signal generated an amplification plot for each sample and it was calculated using the ABI PRISM 7700 Sequence Detection System (SDS) software (v1.9). Data were collected at a threshold point in which every sample was in the exponential phase of amplification. A standard curve was created and mRNA values for the samples of interest were then deduced from this, using this equation: mRNA value =10^((Ct − c)/− m). Where c and m come from y = mx + c equation from standard curve. Ct is the threshold cycle used to incur data whereby every sample is in the exponential phase of amplification.

Statistical analysis

All data (n = 6–9 per group) were analysed using Analysis of Variance (ANOVA). The data are presented as Mean ± SEM. p < 0.05 was regarded as statistically significant. Regression analysis was used to ascertain the causal effect of the relationship between gene expression and blood pressure or heart weights.

Results

Blood pressure in male and female offspring

In adult female and male offspring (Fig. 1), the systolic blood pressure values in the HF/HF group were higher compared with C/C offspring (p < 0.001 and p < 0.05, respectively). Maternal Pravastatin treatment during pregnancy and lactation reduced blood pressure in HF + S/HF female offspring versus HF/HF group (p < 0.01) but this was still significantly higher versus the C/C females (p < 0.001). There was no effect of maternal statin treatment on blood pressure in HF + S/HF male offspring and these remained significantly higher versus C/C males (p < 0.05) (Fig. 1).

Fig. 1
figure 1

Blood pressure (mmHg) in female and male adult offspring from HF-fed dams with or without pravastatin treatment during pregnancy and lactation (HF + S and HF, respectively) and fed post weaning a HF diet to adulthood (producing dam/offspring dietary groups HF + S/HF and HF/HF). The groups were compared to offspring from dams fed standard chow and fed post weaning the same chow diet (C/C). Values are expressed as Mean ± SEM.

Full size image

Heart weight in adult male and female offspring

The heart weights, expressed as % body weight (Fig. 2), in the HF/HF females were heavier compared with C/C group (p < 0.05). Maternal Pravastatin treatment during pregnancy and lactation resulted in reduction in heart weights in HF + S/HF female offspring (p < 0.001 respectively vs HF/HF) to levels similar to the C/C group. There was no effect of prenatal and post weaning exposure to the HF diet and of maternal statin treatment on heart weight male offspring (Fig. 2).

Fig. 2
figure 2

Heart weight, expressed as percent of body weight, in female and male adult offspring from HF-fed dams with or without pravastatin treatment during pregnancy and lactation (HF + S and HF, respectively) and fed post weaning a HF diet to adulthood (producing dam/offspring dietary groups HF + S/HF and HF/HF). The groups were compared to offspring from dams fed standard chow and fed post weaning the same chow diet (C/C). Values are expressed as Mean ± SEM.

Full size image

Levels of mRNA expression for cyclin G1 in the adult offspring left ventricle

Left ventricular (LV) mRNA levels for cyclin G1 (Fig. 3) were lower in HF/HF offspring versus C/C groups (p < 0.05 for both sexes). Maternal statin treatment during pregnancy and lactation prevented the reduction in cyclin G1 mRNA levels in HF + S/HF offspring compared to those found in the HF/HF group (Fig. 3).

Fig. 3
figure 3

Levels of mRNA expression for cyclin G1 in the left ventricles of female and male adult offspring from HF-fed dams with or without pravastatin treatment during pregnancy and lactation (HF + S and HF, respectively) and fed post weaning a HF diet to adulthood (producing dam/offspring dietary groups HF + S/HF and HF/HF). The groups were compared to offspring from dams fed standard chow and fed post weaning the same chow diet (C/C). Values are expressed as Mean ± SEM.

Full size image

Levels of mRNA expression for BNP in the adult offspring left ventricle

BNP mRNA levels in the LV of female offspring remained unchanged following maternal and post weaning exposure to the HF diet, and following maternal statin treatment. In male offspring however, LV BNP levels were lower (p < 0.05) in the HF/HF group versus C/C. Maternal statin treatment did not significantly change BNP gene expression in the HF + S/HF offspring compared with HF/HF group (Fig. 4).

Fig. 4
figure 4

Levels of expression for BNP in the left ventricles of female and male adult offspring from HF-fed dams with or without pravastatin treatment during pregnancy and lactation (HF + S and HF, respectively) and fed post weaning a HF diet to adulthood (producing dam/offspring dietary groups HF + S/HF and HF/HF). The groups were compared to offspring from dams fed standard chow and fed post weaning the same chow diet (C/C). Values are expressed as mean ± SEM.

Full size image

Levels of mRNA expression for SPRR 1A in the adult female offspring left ventricle

SPRR 1A mRNA levels in the LV were found to be similar in all groups (Fig. 5).

Fig. 5
figure 5

Levels of expression for SPRR 1A in the left ventricles of female adult offspring from HF-fed dams with or without pravastatin treatment during pregnancy and lactation (HF + S and HF, respectively) and fed post weaning a HF diet to adulthood (producing dam/offspring dietary groups HF + S/HF and HF/HF). The groups were compared to offspring from dams fed standard chow and fed post weaning the same chow diet (C/C). Values are expressed as Mean ± SEM.

Full size image

Regression analysis

A linear regression analysis was performed to determine the relationship between cyclin G1 mRNA levels and blood pressure or heart weight. The cyclin G1 mRNA levels in the LV of female offspring were inversely correlated with blood pressure (p < 0.05). In male offspring, there is also an inverse trend between cyclin G1 mRNA levels and blood pressure but was not correlated (p = 0.06).

The cyclin G1 mRNA levels in the LV of female offspring were also inversely correlated with heart weight (p < 0.05). In male offspring however, there was a positive correlation between cyclin G1 mRNA levels and heart weights (p < 0.001).

Discussion

This study once again confirms our previous published findings that maternal diet high in fat leads to increased risk of cardiovascular disease in its adult offspring. Here, further to that, we have demonstrated that this phenotype is also accompanied by changes in the expression of genes involved in adaptive remodeling in the left ventricles of the offspring’s heart. The results demonstrated that HF/HF male and female offspring had higher blood pressure compared to C/C control offspring. In females, their hearts were also heavier versus C/C. When pravastatin was given to HF-fed dams during the later part of pregnancy and throughout the lactation period, it prevented this increase in blood pressure but only in the HF/HF female offspring. Likewise, statin treatment to the pregnant dams prevented the increased heart weights in these females associated with the maternal and post weaning HF feeding. This suggests that one of the pleiotropic effects of maternal statin treatment is the prevention of cardiac hypertrophy. This is in agreement with a previous study which suggests that one of the cholesterol-independent effects of study is to prevent cardiac hypertrophy [22]. On the other hand, we did not find any effect of maternal statin treatment on blood pressure and heart weights in the HF/HF male offspring. This indicates sex differences in the response of the developing fetus to both maternal statin treatment and HF feeding probably due to differences in sex hormone milieu between females and males [23].

Of particular note is the observation that males offsprings exposed to high-fat diet had relatively smaller hearts than their female counterparts with or without statin therapy. A possible explanation for this observation in this study is problematic since the study was not designed to address this issue, but one could hazard a guess that this may be related to the factors that induce greater pathological remodelling and hypertrophy in the female offsprings as compared with the male offsprings. Indeed, a recent study by Fernandez-Twinn et al. [24] that observed that maternal diet-induced obesity leads to offspring cardiac hypertrophy, which is independent of offspring obesity but is associated with hyperinsulinemia-induced activation of AKT, mammalian target of rapamycin, ERK, and oxidative stress. However, in this earlier study there was no differential sex differences in offprings heart weights reported, an issue that requires further studies.

Another explanation may relate to a recent study by El Akoum et al. [25] that observed that high-fat fed male mice had a significantly reduced circulating adiponectin levels as well as adipose tissue mRNA expression levels as compared with female mice which may lead to reduced cumulative adipose tissue around the heart i.e. absolute and relative fat mass. This observation may in part explain the differences in heart weights observed between male and female offsprings on HF-diet with or without statin treatment.

One of the novel findings of this study was that mRNA expression levels for cyclin G1 in the LV was lower in HF/HF male and female offspring compared with the C/C groups. This is also associated with higher blood pressure in these offspring and heavier hearts in the females. A previous study showed that low cyclin G1 levels were associated with smaller fetal heart volume by mid to late gestation, and that cyclin G1 initiate protein synthesis to cause hypertrophy rather than DNA synthesis through entry into the cell cycle [26]. It is not clear how cyclins are involved in hypertrophic growth in LV during fetal development under HF nutritional feeding. Nevertheless, it is likely that maternal HF diet contributes to suppressing the levels of cyclin G1 as maternal statin treatment reverses this effect.

The prevention in reduction of cyclin G1 mRNA levels in both sexes of HF/HF offspring by treating the HF-fed dams with statin during pregnancy may indicate that cyclin G1 is involved in cell proliferation in the cell cycle much earlier in life. The process of cardiac remodelling may already be happening much earlier than at 30 weeks of age we measured this gene, and that the cyclin G1 levels we are seeing at this time could be a residual effect of that process. That is probably the reason why maternal pravastatin treatment is able to reverse the mRNA levels of cyclin G1 [27]. It would therefore be interesting to measure these genes in the LV of the offspring at an early time point such as during fetal development or before the offspring enters puberty.

We did not find any effects of maternal and post weaning HF-feeding, and maternal statin treatment on LV BNP mRNA levels in female offspring. However, BNP levels in the LV were lower in HF/HF male offspring versus the C/C groups, suggesting reduced cardiac stress associated with no change in heart weight. Furthermore, maternal statin treatment did not significantly change BNP gene expression in the HF + S/HF offspring compared with HF/HF group. It remains to be determined whether changes take place much earlier on in life or even during fetal development.

The SPRR 1A gene is another marker of cardiac dysfunction, and is known to be a stress inducible cardio-protective protein responding to biochemical stress [28]. However, we did not find this to change in the female LV as a result of exposure to maternal and post weaning HF-feeding, and maternal statin treatment during pregnancy and lactation. However we cannot discount the possibility that changes in gene expression may have taken place much earlier on in life or even during fetal development.

In conclusion, the process of cardiac remodeling and cardiac hypertrophy or gender-specific cardiac dysfunction may have origins from pre and post-weaning exposure to a high fat nutritional environment. The data produced from this study provides an insight into possible mechanisms underlying the developmental origins of adult disease. Therapeutic intervention during pregnancy using statins is beneficial to hypercholesterolemic pregnant dams, and a similar benefit may accrue in their offspring possibly by preventing the damaging effects of prenatal HF dietary exposure on cardiac remodeling and development of hypertension in adulthood.