Perinatal diet influences health and survival in a mouse model of leukemia

Abstract

Introduction

Diet-induced obesity is one of the most notable public health concerns that leads to increased risk of type 2 diabetes and cardiovascular disease (Must et al. 1999). Poor quality diet is a prominent but modifiable risk factor for obesity and many chronic illnesses (Hall and Guo 2017). Prolonged and repeated consumption of high amounts of refined sugars, salt, and fats often leads to adverse health outcomes (Myles 2014). Specifically, consumption of high levels of carbohydrates, in the form of sucrose or fructose, has detrimental health effects (Kroemer et al. 2018; McDonald 1995). Consumption of a Western diet is often associated with low physical activity (Bortolin et al. 2018), suggesting that the quality and food sources of nutrients may influence a variety of physiological parameters and behaviors. Conversely, caloric restriction (CR) without malnutrition has been shown to improve cardiovascular parameters and body composition in humans (Stekovic et al. 2019), indicating that dietary composition as well as diet quantity plays an important role in outcomes related to human health.

Nutrition at various life stages can have lasting effects on health (Tarry-Adkins and Ozanne 2014). In humans, the nutritional status of mothers significantly contributes to birthweight and is associated with the offspring’s risk for future health complications (Danielewicz et al. 2017). Thus, maternal nutrition has been proposed as an opportunity to prevent type 2 diabetes (Silva-Zolezzi et al. 2017). Conversely, controlling maternal caloric intake may prevent gestational diabetes and excessive weight gain during pregnancy but strict guidelines need to be followed in order to avoid malnutrition, which continues to be a significant global public health issue (Phelps and Metzger 1992). Children born in low-income countries are more likely to not reach their full developmental potential due to maternal and early life nutrient deficiencies resulting from malnutrition (Symington et al. 2018).

Animal models provide valuable tools to study maternal exposures, particularly since studies involving pregnant women pose specific challenges (Dominguez et al. 2012). A 10-week maternal consumption of a Western diet (40% kcal from fat and 43% from carbohydrates, mostly sucrose) during pregnancy increased mortality rate in neonatal mice (Perez and DiPatrizio 2018). Interestingly, exposure to high-fat feeding before and not during pregnancy, significantly reduced fetal and neonatal blood glucose concentrations by enhancing fetal pancreatic β cell development (Qiao et al. 2019). Female offspring of mothers that were fed a high-sucrose diet showed insulin resistance, obesity, and hypertension (Samuelsson et al. 2013). Offspring behavior can also be programmed by maternal high-fat diet feeding during the perinatal period (Sullivan et al. 2014). On the other hand, implementation of early life CR-like strategies including litter size manipulation, crowded litters, and lower maternal milk production has been shown to enhance pup lifespan (Davis et al. 2016). Whether maternal exposure to various diet types during fetal development can reprogram health leading to modulations in survival of the offspring later in life is currently an area that has not been explored in depth.

The AKR/J mouse strain has a relatively short lifespan of about 326 days for males and 276 days for females, and is susceptible to developing leukemia around 6 months of age (Kaptzan et al. 2004). Previous studies have shown that AKR mice are responsive to CR and have lower incidence of leukemia leading to better survival compared to ad-libitum-fed controls (Shields et al. 1991). Although AKR/J mice develop obesity in response to high-fat diet (West et al. 1994), it is unclear whether the manipulation of maternal diets can influence the health and lifespan of their offspring. Mouse maternal diets affect postnatal growth and expression of proteins involved in lifespan regulation (Chen et al. 2009), but reports examining the effect on progeny lifespan itself are limited. Some studies suggest that response to CR is not universal in all mouse strains and the beneficial outcomes are dependent on interactions between degree of CR, diet composition, sex and genetic background, as well as age of onset (Liao et al. 2010; Ingram and de Cabo 2017; Mitchell et al. 2016, 2019). Furthermore, age of onset of CR plays an important role in health and lifespan benefits (Vaughan et al. 2017). Thus, the exact effects of maternal diets, including CR, on the healthspan and lifespan of offspring remain largely unknown. The goal of the present study was to perform a side-by-side comparison of the effect of various types of diet given to female AKR mice to address this gap in knowledge.

Materials and methods

Mating and pregnancy

Based on concerns from the Interventions Testing Program (Nadon et al. 2008), pups from the second pregnancy were used for the experiment to avoid external factors, such as poor maternal care of the first litter. For the initiation of the study and generation of first litter, female mice were housed two per cage and were started on the experimental diets for 1 week before introducing one male mouse per cage for mating. Dams ate the respective diets followed by a period of normal chow during gestation and nursing of first litter. After removal of the first litter, the females were placed on their respective diets for 7–10 days prior to returning to the male (Fig. 1a). After obvious signs of pregnancy, males and females were separated into individual cages. Females remained on the experimental diet throughout the breeding, gestation, and nursing period, until the pups reached 2 weeks of age, at which point the maternal diet was switched to standard mouse chow diet (2018 Teklad Global 18% Protein Rodent diet) and fed ad libitum. After an additional 1-week nursing period, the offspring were weaned into new cages and were fed standard chow for the rest of their lives (Fig. 1a). Food consumption and bodyweight of the pups were measured every week.

Open in a separate window

Fig. 1

Effects of maternal dietary composition on litter size, early pup survival and bodyweight trajectories. a Schematic of study design. Females were started on the diets 7–10 days prior to mating and were maintained on the diets through mating, gestation, and up until 2 weeks after giving birth. Mothers were switched to a normal chow at this point. After nursing for 3 weeks, pups were weaned to a normal chow for the reminder of their lives. Scale bar, 1 week (b) Maternal body weights at baseline. n = 25–28 mice per dietary group. c Percentage of females that became pregnant. Stacked bars represent the fraction of litters that survived (filled bars) at day 1 and those that did not (hatched bars). Statistical analysis was conducted by Fisher’s exact test. *p < 0.05. d Number of pups per dam at day 1. **, p < 0.01 versus SD group. e Percentage of pups that survived up to 21 days after birth. Number of pups whose dams were on either SD (n = 100), high sucrose diet (83), Western diet (74), or CR (123). **, ***, ****p < 0.01, 0.001, 0.0001 versus pup survival at day 1. f Bodyweights of pups at day 14 post-birth.****p < 0.0001 versus SD. g Representative images of 14-day-old pups. Offspring of dams fed sucrose diet had a body size that was approximately between SD and CR (not shown)

Results and discussion

Maternal body weight measured immediately after giving birth was not significantly impacted by the diet type or dietary regimen (Fig. 1b). Approximately, 40–55% of females became pregnant where sucrose-fed females showed a significantly different number of litters that survived until day 1 compared to SD (Fig. 1c). Given that the viability of litters is an important determinant of early survivability, the litter size per dam in the four maternal diet groups was quantified. On day 1 of giving birth, the number of pups per litter was significantly lower when dams were fed a high sucrose diet and also trended lower in the Western diet and 15% CR groups compared to the SD-fed dams (Fig. 1d). A progressive increase in pup mortality was observed postnatally within the first 21 days when nursed by dams that were fed Western diet > CR > high sucrose > SD. At 21 days, survival was ~ 65% for the Western group, ~ 75% for the CR group, and ~ 80% for the sucrose group, compared to 95% survival for the SD group (Fig. 1e). At 14 days from birth, pups showed significantly different body weights compared to SD where pups of sucrose-fed and CR dams showed lower body weights while pups of Western-fed dams showed higher body weights (Fig. 1f). Figure ​Figure1g1g depicts representative images of 14-day-old pups nursed by dams fed SD, Western diet, or CR. Offspring of dams fed sucrose diet generally had a body size that was between SD and CR (image not shown). Pups of dams fed a Western diet were larger in body size, had a scruffier coat and weighed more than SD controls, similar to observations made in previous reports (Du et al. 2012) whereas pups from the other groups were more or less comparable to SD.

Given that maternal diets can have lasting effects on lifespan and health of offspring, we examined the impact of maternal diet composition on several health markers. As anticipated, offspring from dams fed Western diet had a trend to gain more body weight regardless of sex, while those from CR-fed dams showed a trend for least body weight gain during the first 6 months of life (Fig. 2a). However, the body weight differences between groups were not statistically significant. The body weight trajectory of pups of dams maintained on high sucrose trended lower than pups nursed by dams that were maintained on SD (Fig. 2a). Maternal diets had minimal effects on the amount of food consumed by the offspring during the first 4 weeks after weaning (Fig. 2b). Fasting blood glucose (FBG) levels were measured in 15–17-week-old offspring mice and found to be comparable between all four experimental groups (Fig. 2c). Lastly, NMR spectroscopy was employed to measure body composition in 14–16 week-old offspring. Sex-specific differences in percent fat mass and associated lean-to-fat ratio were observed, with male offspring of dams maintained on high sucrose diet or CR being significantly leaner than males of SD-fed dams (Fig. 2d, e). No alterations in these morphological measures were noted in female progeny (Figs. 2d, e). By and large, there was little effect of maternal diet on motor coordination in the progeny regardless of sex on the accelerating rotarod. However, female progeny from Western diet-fed mothers showed a significant decrease in performance, which may be related to the trend toward increased body weight in this group (Fig. 2f).

Open in a separate window

Fig. 2

Implications of maternal diet composition on markers of health and survival of the pups. a Body weight trajectories over 25 weeks. Inset, area under the curve (AUC). b Trajectories of food consumption for both female and male pups over a 4-week period. c Fasting blood glucose (FBG) levels in 15–17-weeks-old female (n = 14–32) and male (n = 11–49) pups. d Percent fat in 14–16-weeks-old female (n = 14–15) and male (n = 11–15) pups as assessed by nuclear magnetic resonance spectroscopy (NMR). e Lean-to-fat ratio. f Latency to fall from the accelerating rotarod. Data are represented as whisker plots (c-f) and analyzed using one-way ANOVA with Dunnett’s post-test. *,**, ***p < 0.05, 0.01, and 0.001 vs. SD. g Kaplan-Meier survival curves for female and male offspring from dams that were maintained on either SD [n = 38(F)-59(M)], high sucrose diet [n = 22(F)-47(M)], Western diet [n = 33(F)-16(M)], or 15% CR [n = 49(F)-47(M)]. h Kaplan-Meier survival curves for female and male offspring on SD (upper panel) or after combining all groups (lower panel). F, female; M, male. See Table ​Table11 for additional information. g, h Orange arrow represents time at weaning (21 days) when mice were enrolled into the survival analysis

Previous studies have shown that almost all ad libitum-fed AKR mice die from luekemia (Shields et al. 1991). When the effect of maternal diet was investigated, we found notable differences in survival among groups of AKR progeny mice (Fig. 2g). Median survival shown in Table ​Table11 revealed that all 3 maternal diet groups resulted in higher median lifespan in female progeny (284.5, 293, 271 days for sucrose, Western and CR, respectively) compared to SD controls (257.5 days), while in males the sucrose group showed the highest median survival (343 days) followed by SD (317 days), CR (293 days), and Western (268.5 days) groups. Differences were observed in maximum lifespan as well, which is defined as the remaining 10% survival. In females, the sucrose (485 days) and CR (467 days) groups exhibited an extended maximum lifespan compared to SD (326 days) or Western diet (390 days) (Fig. 2g). Similarly, differences in maximum lifespan were observed in male progeny: sucrose (518 days), CR (472 days), SD (383 days), and Western diet (358 days) (Fig. 2g). Out of these, sucrose females and males showed a significantly enhanced lifespans compared to SD (p = 0.0429 and p = 0.0110, respectively). Significant differences were observed in median lifespan between female and male progeny from SD-fed dams (Fig. 2h, upper panel, and Table ​Table1)1) and from dams maintained on all four dietary regimens (Fig. 2h, lower panel, and Table ​Table1).1). It is currently unknown if maternal diet elicits directly or indirectly an alteration in the viral transcriptomic machinery that drives lifespan in this model. Interestingly, compared to observations from Sheilds and colleagues where AKR mice showed a maximum lifespan of ~ 52 weeks (Shields et al. 1991) and other reports on AKR lifespan (~ 44 weeks) (Festing and Blackmore 1971), our AKR mice survived much longer. Differences were also noted where male offspring lived longer than females regardless of the maternal diet. Mechanisms underlying these sex differences are poorly understood and warrant further investigation. Collectively, in both female and male offspring, maternal sucrose-rich diet and CR resulted in median and maximum lifespan extension. Altogether, it would appear that the maternal diet, consumed during the pre and perinatal period has a significant effect on metabolic health measures and lifespan of offspring.

An unexpected finding of the study is that female and male progeny of both sucrose-fed dams and dams on CR showed enhanced lifespan (both median and maximum lifespan, only former statistically significant). Although the effects of CR on lifespan and health biomarkers have been previously reported by us and others (Mitchell et al. 2018), it is unclear how a maternal diet that is high in sucrose can positively influence health-related outcomes and extend longevity. One possibility is that the consistency, taste, and smell of the sucrose diet is unpalatable for mice, causing them to undergo a CR-like effect (de Cabo lab unpublished observation). Previous reports have indicated that fructose is the main driver of the spike in metabolic disorders in humans (Bray et al. 2004). Complementary to these findings, adult female and male mice fed a diet high in fructose/glucose experienced a decline in survival that was more dramatic than mice fed a diet high in sucrose (Ruff et al. 2015), a phenotype likely attributable to differences in glucose uptake/absorption. Diets low in protein and high in carbohydrates lead to longer lifespans in adult ad libitum-fed mice (Le Couteur et al. 2016). Indeed the method of sucrose delivery has been shown to be an important determinant of health outcomes where liquid sucrose rather than solid sucrose resulted in body weight gain in C57BL/6 mice (Togo et al. 2019). Whether a different sugar type, such as fructose, or a different delivery method would result in alternative lifespan and health outcomes is currently not known and must be investigated in future studies.

Fetal malnutrition, resulting from undernourished mothers, has detrimental effects on the offspring and can derail the trajectory of an otherwise potentially healthy fetus. Undernourished mothers expose the fetus to a lack of nutrients, resulting in low birth weight and increased susceptibility to diseases. Moreover, maternal overnutrition exposes offspring to an excess of nutrients, resulting in high birth weight, susceptibility to diseases, and adverse metabolic phenotypes in the offspring (Sun et al. 2012; Martin-Gronert and Ozanne 2006). In the present study, we attempted to represent conditions across this feeding spectrum. The degree and timing of maternal CR plays an important role in health outcomes in pups. Previous studies have shown that maternal 30% CR leads to fetal hyperphagia, hypertension, and hyperinsulinemia (Vickers et al. 2000) whereas maternal 20% CR causes no differences in litter size or body weight in the offspring (Stone et al. 2019). CR has been shown to elicit age-dependent effects on health (Chen et al. 2019) and therefore, the age at which maternal CR was conducted may directly influence pup health as well. However, the effect of various degrees of dietary exposures including CR, on pups from younger or older mothers was not examined in the present study. The exact mechanisms by which maternal nutrition during early pregnancy adversely affects offspring health and development remain poorly understood; however, epigenetic and developmental reprogramming appears to play a crucial role (Dearden and Ozanne 2015) as evidenced by the fact that undernourished children have a distinct methylation signature (Uchiyama et al. 2018). Various maternal factors such as diet, environmental exposure, and stress are likely to impact offspring health (Preston et al. 2018) through a variety of different pathways including stem cell alteration, gut microbiota, and immunomodulation (Friedman 2018). Prior studies have also shown that CR increases the number of primodial follicles and enhances ovarian reserve (Garcia et al. 2019) which can significantly contribute to early life health. Data from our study showed that maternal diet during the gestational and nursing periods had the most profound long-term implications for offspring lifespan, likely through alteration in molecular landscapes related to survival.

There are some limitations in the present study. First, the AKR mouse model used here has the endogenous AKV murine leukemia retrovirus in all tissues that renders them predisposed to lymphatic leukemia at early ages, and it is unknown how this factors into health and survival matrices. While leukemia incidence was not an endpoint that was quantified per se in the present study, previous reports have shown that 60–70% of mice of this strain develop lymphomas (Chrisp et al. 1996). Whether these findings can be extrapolated into other strains of mice that are more representative of mouse models used in biomedical research studies should be a focus of future research. Second, it is unclear how differences in early pup survival could contribute to the overall survival of each experimental group. Because different numbers of mice were dying post-natally prior to day 21 for each group, a survival bias may have been introduced when monitoring the remaining mice, thus affecting lifespan outcomes. Pup survival can dictate litter size and lead to crowded litters and food restriction (Sadagurski et al. 2015). Third, we did not investigate possible diet-mediated alterations in nutrient uptake or absorption, which could have a profound influence on the ability of macronutrients to enter systemic circulation in the dams and impact whole-body metabolic and physiological responses.

Acknowledgements

Authors’ contributions

All experiments were designed by K.J.P. and R.d.C. Lifespan and whole-body experiments were carried out by A.M.S., K.N.L, C.Y., and K.J.P. Data analysis and interpretation was conducted by D.L.P, A.R., M.B., and R.d.C. The manuscript was written by D.L.P., A.R., M.B., and R.d.C.

Funding information

This work was supported by funding from the Intramural Research Program of the National Institute on Aging/NIH.

Data availability

Not applicable.

Code availability

Not applicable.

Compliance with ethical standards

Footnotes

References