Reversing Fetal Undernutrition by Kick-Starting Early Growth

The relationship between poor maternal nutritional status and enhanced chronic disease risk in the offspring is now abundantly established in both experimental models and in epidemiological data from human populations worldwide. The classic studies of Barker et al (1), demonstrating a relationship between low birth weight (LBW) and increased risk of cardiovascular disease, hypertension, and dyslipidemia, paved the way for extensive research into the influence of the intrauterine environment on development of chronic diseases. In these original studies, men in their sixties who were smallest at birth had a 3-fold higher death rate from ischemic heart failure compared with those with high birth weights (1). Likewise, incidence of type 2 diabetes and glucose intolerance was also significantly higher in men with LBWs (2). To add to an expanding list of ailments, it is now well accepted that the propensity to develop obesity, type 2 diabetes, immunologic dysfunction, insulin resistance and related metabolic diseases is, in part, influenced by the maternal in utero environment (3). The focus of many investigations on LBW and conditions that lead to small babies arises from the fact that environmental factors and nutrient supply play an important role in determining birth weight and alterations in birth weight represent a gross deviation in the fetal experience. Nonetheless, changes in birth weight should not be confused to be a sine quo non in developmental programming, because it is quite evident that long-term alterations in risk occur well below the threshold of alterations in body weight at birth.

Over the last decade and half, an explosion of studies using a variety of approaches has shown that developmental programming occurs in the contexts of both maternal undernutrition and overnutrition (3). Although one's genetic makeup, postnatal diet, and lifestyle factors such as physical activity indisputably are important proximate causes of obesity, the influence of gestational experiences is evident in the U- or J-shaped relationship between birth weight and later obesity risk, suggesting that both extremes of nutritional environments lead to detrimental changes in the offspring. Certainly, this is relevant to the current global epidemic of obesity and ubiquitous consumption of calorically dense diets, high in fat and simple sugars both before and during pregnancy (4, 5).

The conceptual framework provided by the developmental origins of health and disease hypothesis posits that unfavorable prenatal and postnatal environments, specifically during sensitive periods in development, lead to permanent alterations in organ structure and function that may confer increased risk towards a variety of chronic diseases (6). As elaborated in the predictive adaptive response hypothesis (7, 8), these adaptive changes, facilitated by the process of developmental plasticity are designed to be beneficial in ensuring short-term survival and preparing for a “predicted” postnatal environment. However, when presented with a highly discrepant postnatal environment, mismatch ensues and these changes may prove disadvantageous or maladaptive, thereby increasing predisposition to disease. Evidence from follow up of individuals affected during the Dutch Hunger Winter famine during the latter part of the Second World War and the Chinese famine (1959–1961) lend strong support to this hypothesis (9, 10). Furthermore, high rates of obesity and diabetes in populations where poor maternal nutrition is historically prevalent in combination with current trends of urbanization, as evidenced in many developing nations, also presents an ongoing natural experiment. Despite these trends and studies of populations, it is difficult to assess fetal nutritional status directly in human studies. Hence, much of our mechanistic understanding related to maternal nutrition and developmental programming comes from studies in experimental animal models (11).

Studies in altricial species such as rats and mice have been the mainstay models to examine various aspects of developmental programming in many contexts. Studies in sheep have largely contributed to the understanding of maternal changes on placental function and fetal development, and much less on long-term persistent effects, presumably due to greater logistical and financial costs involved. Likewise, studies in nonhuman primates have been pivotal in elucidating both the immediate and long-term effects of maternal diet on metabolic and behavioral aspects of offspring development. In general, these models can be broadly classified as those examining the consequences of maternal undernutrition or overnutrition. In the former category, models of placental insufficiency, induced via uterine or umbilical artery ligation, produce clinically relevant phenotypes of intrauterine growth restriction (IUGR). However, by far fetal undernutrition is most frequently achieved by either limiting maternal protein or overall caloric intake. In models of maternal low-protein diet, pregnant animals are fed isocaloric diets with either low (5%–8% wt/wt) or sufficient protein (18%–20% wt/wt) during gestation alone or through gestation and lactation. Offspring of animals from protein restricted dams are lighter, and display catch-up growth when protein restriction is limited to gestation alone. However, if protein restriction is continued into lactation the growth deficits are permanent. Numerous studies have documented a host of phenotypic alterations including increased pathogenesis of diabetes, insulin resistance, endocrine changes, and associated structural and functional changes in multitude of tissues (11–13). Along the same lines, global undernutrition employed at varying levels, from moderate to severe (20%–70% of ad libitum intake), also recapitulates growth restriction in utero followed by catch-up growth, which is generally detrimental to the offspring (14, 15). These findings collectively show that maternal undernutrition induces long-term detrimental changes in the offspring ranging from hyperinsulinemia, hypertension, obesity, disruption of glucose homeostasis, β-cell secretory dysfunction, adipose expansion, mitochondrial dysfunction, hepatic steatosis, and hyperphagia (16, 17). The mechanisms relating to programming of these diverse end points have been the subject of continued work from numerous groups. Alterations in set points of metabolic and endocrine pathways, reprogramming of stem cell pools and differentiation potential, structural changes induced by reduced nutrient transport and epigenetic alterations of transcriptional pathways have been explored to varying degrees. Strikingly, recent reports have also demonstrated that at least some of these phenotypes are transmissible to the F2 generation via the male offspring, associated with alterations in the sperm methylome (18, 19). Although continued research attempts to elucidate the mechanistic underpinning of developmental programming, an equally pivotal but less appreciated focus has been on interventions to reverse or mitigate the effects of developmental programming. This laudable goal is the focus of work by Li et al, appearing in the current issue of Endocrinology (30).

The list of approaches directly aimed at mitigating the effects associated with fetal undernutrition is fairly short. Previous studies have heavily focused on the ability of neonatal leptin treatment in modifying gestational influences (20). Much of this well-deserved excitement stems from the fact that perturbations in leptin levels during neonatal life is associated with altered susceptibility to obesity and metabolic dysfunction later in life. It is worth noting that leptin levels are highly dynamic during the early postnatal period, as part of the “leptin surge,” and constitutes a critical window of normal neural development. Extensive studies both in the context of maternal nutritional sufficiency and undernutrition have revealed that neonatal leptin treatment in the context of maternal undernutrition does reduce the rate of neonatal growth and ameliorates programmed metabolic dysfunction (21). Nonetheless, these effects of leptin were found to be sexually dimorphic and dependent on postnatal nutritional status. More importantly, treatment with leptin in control mice has conflicting effects, with some showing modestly increased risk of obesity and glucose intolerance, while others showing little or no effect. Other researchers have focused on maintaining insulin secretion and β-cell function, with the aim to mitigate glucose intolerance and type 2 diabetes risk programmed by gestational deprivation. Work by Simmons and coworkers (22, 23) showed that neonatal treatment with glucagon-like peptide 1 analog exendin-4 reversed the effects of IUGR and prevented the development of type 2 diabetes in IUGR offspring. This was associated with maintenance of β-cell mass and patent expression of the key pancreatic transcription factor, pancreatic and duodenal homeobox 1 (PDX-1), previously shown to be a target of developmental programming (24). Other approaches have employed interventions during pregnancy, using nutritional approaches ranging from supplementation of amino acids (taurine, glycine) to other dietary components (folic acid, choline). An important take-home message from these studies appears to be the notion that, at least in rodent models, the critical window of developmental plasticity extends well into the early postnatal period, allowing for interventions to be effective in resetting developmental trajectories, set astray by fetal undernutrition. An important unanswered question is the identification of similar critical windows in human infants where meaningful interventions might be deployed to reverse gestational growth or nutrient restriction.

In this context, the present work along with recently published related evidence (25–27) from the same group highlights a fascinating role of the somatotropic axis in neonatal life in mediating metabolic aspects of developmental programming. In this report, Li et al use a well-characterized and frequently employed model of global maternal undernutrition, which reproducibility has been shown to induce obesity and metabolic dysfunction in adult offspring (30). Offspring of rat dams either fed ad libitum or restricted calories (50% of controls) during pregnancy alone, were treated with either saline vehicle or GH from postnatal day 3 until weaning. Offspring were then maintained on chow diets provided ad libitum until age 5 months. As anticipated, fetal undernutrition reduced birth weights and display catch-up growth compared with control offspring by the time of weaning (d 22). These offspring also show greater body weights and adiposity later in life. Although early postnatal GH treatment led to increased body weight compared with vehicle treated counterparts in the preweaning period, it also normalized postnatal growth trajectories, development of excessive adiposity and improved insulin sensitivity in females. This report has several strengths. First, the investigators performed detailed analysis of growth, body composition, endocrine, and metabolic assessments in both male and female offspring. Second, these studies provide novel insights into the modulation of hepatic and systemic GH/IGF-1 axis after preweaning GH treatments. Interestingly, although body composition related effects of maternal undernutrition were reversed in both males and females, alterations in markers of glucose metabolism improved in a sex-specific manner. Finally, in the broader context of developmental programming, the authors have recently demonstrated that other detrimental phenotypes associated with maternal undernutrition, including ventricular hypertrophy (26), bone marrow macrophage inflammation (25), and adipose tissue inflammation and insulin resistance (27), were also ameliorated after intervention of preweaning GH. These findings taken together present a robust body of seminal evidence supporting the efficacy of preweaning GH in reversing developmental programming.

Not unlike other potentially far-reaching research, the observations described by Li et al also raise a number of interesting questions regarding the mechanisms underlying developmental programming in the context of maternal undernutrition, the role of catch-up growth and the specific roles of GH and IGF-1 in the critical preweaning window in mediating long-term metabolic homeostasis. Many of these questions will require additional experimentation and have the potential to refine our understanding of fundamental development. An important observation from these studies is perhaps the distinction that not all growth after fetal undernutrition and LBW may be bad. Although certainly considerable documented evidence exists linking rapid catch-up growth to worse outcomes in animal models and human studies (28), an appropriate paradigm to balance the beneficial aspects of postnatal growth and prevention of excessive rapid growth is lacking. While the consequences of rapid growth have been well studied, the underlying basis is likely to be complex and much less understood. By directly modulating one specific aspect of the growth axis, the present studies may have provided an avenue to balance these contrasting objectives. It is also worth noting that unlike GH supplementation in adults, preweaning GH did not lead to alterations in insulin sensitivity in adulthood. It is also quite intriguing how the pleiotropic effects of GH in remediating developmental programming on such diverse tissues are mediated. Because the GH/IGF-1 axis accounts for much of postnatal growth, this is an obvious target. Studies of growth hormone receptor-null and IGF-1-null mice suggest that approximately 83% of postnatal growth is attributable to the GH/IGF-1 axis (29). Although the current studies point to modulation of IGF-1 bioavailability as a potential mechanism, further studies to probe the specific role of IGF-1 would be illuminating. In addition, both extrahepatic actions of GH and interactions with other systemic endocrine axes are also worthy of further evaluation. The findings also raise interesting and novel questions about the fundamental roles of insulin-like growth factor binding proteins in developmental programming. Although insulin-like growth factor binding proteins control free IGF-1 levels and serve as a transport mechanism, they also have independent signaling effects. Whether the observed effects of GH supplementation are mediated via modulation of signaling pathways independent of IGF-1 will be an important question for future research to address. Overall, the findings by Li et al further validate the preweaning period as a malleable window during which the consequences of gestational undernutrition can be partly undone. Treatment of offspring with GH appears to be a new tool in the as yet limited arsenal but may present a broad strategy in preventing detrimental consequences. Although the findings are imminently applicable to humans, there is still ways before their true translational value can be realized. For that to happen we need a more careful understanding of the developmental trajectories of GH and IGF-1 in infants and children, and a better appreciation of the developmental window of opportunity in humans during which interventions might be permissible.