The foetus, though protected by the placental barrier, is highly susceptible to changes in both maternal diet and the hormonal milieu. In particular, a chronically poor maternal diet resulting from either protein/calorie restriction or excess maternal nutrients, as well as elevated hormones, such as insulin and glucocorticoids, have major consequences for foetal development and metabolic disease in adult life. Maternal under-nutrition, for example, can lead to intrauterine growth retardation, which is linked to an increased risk of diabetes, hypertension and cardiovascular disease in the adult offspring [1–3]. Emerging evidence also suggests that maternal over-nutrition may have similar long-term metabolic consequences in the offspring [4–6], but the mechanisms underlying the foetal origins of these diseases have only begun to be investigated.

The concept of foetal ‘programming’ describes a change in gene expression due to an environmental exposure in utero, resulting in a persistent altered metabolic phenotype in the adult offspring [7]. During development, glucocorticoids are essential for organ maturation, particularly the foetal lung. Administration of synthetic glucocorticoids is currently recommended for mothers at risk of preterm and delivery between 24–36 weeks, in order to promote proper foetal lung maturation, and is successful in reducing neonatal mortality and chronic lung disease [8]. Numerous laboratories have, however, provided compelling evidence that foetal exposure to inappropriate amounts of glucocorticoids has profound effects on foetal growth, placental function, and foetal and post-natal brain development, and can result in persistent hyperglycaemia throughout life [9–11]. Suboptimal maternal nutrition [12] and maternal stress [13, 14] are also thought to expose the foetus to excess glucocorticoids. Regardless of the source, elevated glucocorticoid levels during pregnancy predispose to in utero growth retardation and low birthweight [15, 16].

In this issue of Diabetologia, Nyirenda et al. [17] report an exciting new potential mechanism for glucocorticoid-induced foetal programming of hyperglycaemia. The authors’ previous work demonstrated that exposure to prenatal dexamethasone resulted in persistent upregulation of the mRNA and activity of the cytosolic form of the gluconeogenic enzyme phosphoenolpyruvate carboxykinase (PCK) in adult offspring [18]. Furthermore, these changes in PCK expression in adult offspring could not be attributed to altered postnatal maternal behaviour, suggesting that excess foetal glucocorticoid exposure has a permanent effect, programming increased gluconeogenesis [19]. Nyirenda et al. [17] have now followed up this work by investigating the effects of supraphysiological levels of glucocorticoids, administered to female rats during the last week of pregnancy, on key hepatic transcription factors known to regulate PCK expression in the rat foetus and adult offspring. Prenatal dexamethasone resulted in an early increase in the transcription of the gene encoding foetal hepatic nuclear factor 4 (Hnf4a), which remained elevated into adulthood, and paralleled the rise in Pck1 expression and hyperglycaemia. Interestingly, dexamethasone treatment not only increased Hnf4a mRNA expression, but altered the expression of the of Hnf4a isoforms, such that the adult isoforms (Hnfa1/2) were favoured over the foetal isoforms (Hnf4a7/8). Expression of the different isoforms results from alternative promoter usage and differential splicing. An early promoter switch to produce the adult HNF4A isoform thus represents a novel mechanism by which early exposure to glucocorticoids may potentially induce Pck1 mRNA and result in a premature increase in gluconeogenesis.

The gluconeogenic genes encoding PCK and glucose-6-phosphatase, which, respectively, catalyse the first and final steps in hepatic gluconeogenesis, together with the transcription factors that regulate their expression, are likely targets for foetal programming of hyperglycaemia. These genes are usually not expressed before birth, since the maternal glucose supply is sufficient to meet the needs of the foetus during development. In the immediate postnatal period, however, neonatal glucose falls and there is a rapid increase in stress hormones (cAMP and glucocorticoids), triggering increased transcription of the Pck1 gene, which codes for the first enzyme in the gluconeogenic pathway [20]. Glucocorticoids stimulate hepatic Pck1 gene expression by altering the protein levels of specific transcription factors and co-activators and promoting a complex series of interactions between these factors and the extended glucocorticoid regulatory unit (GRU) in the Pck1 promoter [21, 22] (Fig. 1). Glucocorticoids have been shown to enhance the binding of specific transcription factors, including FOXO1, peroxisome proliferative activated receptor alpha (PPARA), and HNF4A to the extended glucocorticoid regulatory unit of the Pck1 promoter [22]. Additionally, both Pck1 and the gene for glucose-6-phosphatase contain glucocorticoid receptor binding elements, allowing glucocorticoids to enhance gene expression [23, 24]. The co-activator peroxisome proliferative activated receptor, gamma, coactivator 1 alpha (PPARGC1A) also activates Pck1 expression through specific interaction with HNF4A, without binding directly to the Pck1 promoter [25]. HNF4A has been associated with recruiting coactivator protein (CREB binding protein/p300) and the ability to mediate chromatin remodelling [26–28]. Whether these other regulators are also involved in stimulating PCK gene transcription directly in response to early glucocorticoid exposure was not explored in the report by Nyirenda et al. [17]. A more thorough and direct analysis of the DNA binding factors at the protein level could be explored using foetal rat liver nuclear extracts and DNA binding assays. In addition, research into the potential mediators that regulate alternative splicing may also further our understanding of the mechanisms whereby glucocorticoids alter HNF4A and its transcriptional activation pattern.

Fig. 1
figure 1

Extended model for the effect of excess glucocorticoid exposure on PCK gene transcription and persistent hyperglycaemia in adult offspring. Prenatal exposure to excess glucocorticoids leads to a specific increase in the expression of Hnf4a but not the gene for glucocorticoid receptor (Nr3c1), Hnf1a or Ppargcla in foetal rat liver. Isoform analysis of Hnf4a mRNA revealed that the glucocorticoid exposure induces promoter switching that prematurely upregulates the adult isoforms (HNF4A1/2) and downregulates the foetal isoforms (HNF4A7/8). These changes in Hnf4a are persistent and parallel the increases in Pck1 expression and hyperglycaemia in the exposed adult offspring. These new findings suggest that changes in Hnf4a expression in utero may be a driving force for foetal programming of glucocorticoid-induced insulin resistance, possibly through premature assembly of transcription factors on the Pck1 promoter. AF Accessory factor, GRE glucocorticoid regulatory element

Full size image

Animal models of excess glucocorticoid during late pregnancy are often used as a means of understanding how maternal stress can induce changes in the foetus. The foetus is normally protected from excess maternal glucocorticoid, as 11-beta-hydroxysteroid dehydrogenase type 2 (11β-HSD2) acts as a foeto-placental barrier, inactivating circulating maternal cortisol to inert cortisone, and thereby preventing foetal exposure to elevated glucocorticoid. In rats, inhibition of 11β-HSD2 causes postnatal development of hyperglycaemia, increased blood pressure, and increased hypothalamic–pituitary–adrenal (HPA) axis activity in adult offspring [29–31]. Additionally, prenatal betamethasone exposure resulted in a persistent increase in glucose-6-phosphatase activity in sheep adult offspring [32] and PCK in rat adult offspring [18], supporting the concept that changes in gluconeogenic enzyme expression may lead to the chronic hyperglycaemia in adult offspring associated with prenatal glucocorticoid exposure.

Although the paper from Nyirenda et al. does not provide direct evidence that the change in Hnf4a expression is the primary effect of dexamethasone, it does provide insight into what may be a common underlying mechanism for foetal programming. Different foetal insults (e.g. over- and under-maternal nutrition, placental insufficiency, maternal stress, and excess glucocorticoid exposure) can all result in a similar outcome: offspring predisposed to adult hyperglycaemia and insulin resistance. In the liver, HNF4A is thought to regulate the transcription of over half of all hepatic genes [33]. How HNF4A is targeted to genes is under active investigation, and is probably related to its interactions with other regulatory transcriptional complexes. However, it is important to note that its activation and expression are altered in response to both dietary and hormonal factors, thus making HNF4A an important intersection for nutritional and/or hormonal regulation of gene expression and cell function [34–36]. We have recently found, in a non-human primate model, that a high-fat, high-calorie maternal diet also results in a premature upregulation of gluconeogenic gene expression and increased HNF4A protein expression in the foetal liver (Unpublished results). A more focused study may elucidate whether HNF4A acts as the primary mediator or a corollary of foetal programming.

It is also important to note that prenatal glucocorticoid exposure in rats has intergenerational effects, resulting in adult offspring with increased PCK and hyperglycaemia in the subsequent F2 generation, despite normal pregnancies [37]. Transcriptional control of gene expression depends on DNA accessibility, which is epigenetically regulated by histone modification, DNA methylation and chromatin remodelling. The notion that maternal nutrition and glucocorticoids may lead to epigenetic changes that underlie the foetal programming of adult disease, is increasingly gaining acceptance [38, 39]. The mechanisms involved in generating such responses are not well characterised, however, opening up new study areas that promise new insights into foetal programming of adult disease.

Diabetes is a polygenic disorder, and its onset is influenced as much by our inherited genes as by our exposure to environmental factors, starting at our earliest developmental stages in utero. While studies of families with disease clusters may be informative for identifying highly penetrant gene variants, other approaches are needed to identify factors that determine environmental susceptibility to diabetes, including prenatal and postnatal nutrition, stress, and physical activity. The study by Nyirenda et al. [17] highlights a new window of opportunity for investigation of the origins and mechanisms of adult hyperglycaemia.