Maternal Obesity-Induced Endoplasmic Reticulum Stress Causes Metabolic Alterations and Abnormal Hypothalamic Development in the Offspring

The steady increase in the prevalence obesity and associated type II diabetes is a major health concern, particularly among children. Maternal obesity represents a risk factor that contributes to metabolic perturbations in the offspring. Endoplasmic reticulum (ER) stress has emerged as a critical mechanism involved in leptin resistance and type 2 diabetes in adult individuals. Here, we used a mouse model of maternal obesity to investigate the importance of early life ER stress in the nutritional programming of metabolic disease. Offspring of obese dams displayed increased body weight, adiposity, food intake and developed glucose intolerance. Moreover, maternal obesity disrupted the development of melanocortin circuits associated with neonatal hyperleptinemia and leptin resistance. ER stress-related genes were upregulated in the hypothalamus of neonates born to obese mothers and neonatal treatment with the ER stress-relieving drug tauroursodeoxycholic acid improved metabolic and neurodevelopmental deficits and reverses leptin resistance in neonates born to obese dams.


44
A major shift in our nutritional environment has greatly contributed to the recent obesity 45 epidemic. There is growing evidence that adverse fetal and early postnatal environments

55
The adipocyte-derived hormone leptin directly targets these neuronal populations to cause

79
In the present study, we investigated whether maternal diet-induced obesity induces ER stress

90
Maternal obesity causes metabolic disturbances in the offspring 91 A mouse model of maternal obesity induced by high-fat high-sucrose (HFHS) feeding during 92 pregnancy and lactation was used to study the effects of maternal obesity on the offspring's 93 metabolism and development. Adult female mice were either fed a HFHS (58% kcal fat w/ 94 sucrose) or a control diet (6% calories from fat) six weeks before breeding. Dams were kept on 95 their respective diet throughout pregnancy and lactation. A significant increase in dams' weight 96 gain was observed as early as 4 weeks after HFHS diet began and persisted throughout the 97 HFHS exposure (Fig 1A). This elevated body weight was associated with increased fat mass 98 ( Fig 1B). Moreover, dams fed a HFHS diet displayed altered glucose tolerance during gestation 99 (Fig 1C).

100
The offspring of HFHS-fed dams had heavier body weights at weaning and this elevated 101 body weight persisted into adulthood (Fig 1D). We also evaluated body composition and found 102 that adult animals born to obese dams displayed elevated fat and lean mass compared to 103 control mice (Fig 1E). Moreover, neonatal exposure to HFHS caused adipocyte hypertrophy as 104 revealed by a 1.5-fold increase in adipocyte size in epididymal white adipose tissue (Fig 1F).

105
There was also an increase in food intake, and decreases in oxygen consumption (VO 2 ) and 106 energy expenditure in adult animals born to obese dams (Fig 1G-I). Respiratory exchange ratio 107 and locomotor activity were not significantly different compared to controls (Fig 1J and K).

108
However, adult mice born to obese dams displayed impaired glucose and insulin tolerances 109 compared to mice born to lean dams (Fig 1L and M). To examine if maternal obesity was associated with activation of ER stress response in the 114 offspring, we measured the expression levels of the following ER stress markers in 6 metabolically-relevant tissues: activating transcription factor 4 (Atf4), 6 (Atf6), X-box binding 116 protein (Xbp1), glucose regulated protein GRP78 (referred to as Bip), and CCAAT-enhancer-117 binding protein homologous protein (Chop), in P10 and adult animals born to chow-or HFHS-118 fed dams. The mRNA levels of Atf4, Atf6, Xbp1, Bip, and Chop were significantly elevated in the 119 ARH of P10 mice born to obese dams (Fig 2A). Moreover, expression of Atf4, Atf6, Xbp1, and 120 Chop mRNAs were significantly higher in the ARH of adult mice born to HFHS-fed dams (Fig   121   2B). In contrast, only Atf4 mRNA was significantly increased in the paraventricular nucleus 122 (PVH) of P10 mice (Fig 2C and D). Atf4, Atf6, and Xbp1 mRNAs were elevated in the pancreas 123 of P10 pups of HFHS-fed dams (Fig 2E), but these markers were not significant changed in 124 neonatal liver and fat tissues (Fig 2G and I). In addition, Atf4, Atf6, Xbp1, and Chop mRNA 125 levels were higher in the pancreas of adult mice born to obese dams (Fig 2F), but only Xbp1 126 and Xbp1 as well as Chop mRNAs were significantly elevated in the liver and fat tissues, 127 respectively, of adult mice of HFHS-fed dams (Fig 2 H and J).

128
To investigate the importance of early life ER stress, we treated pups born to HFHS-fed 129 dams with daily peripheral injections of tauroursodeoxycholic acid (TUDCA) from P4 to P16, 130 which represents a critical period for growth and development, including of the hypothalamus

131
[2]. TUDCA is a chemical chaperone of low molecular weight that is well-known to alleviate ER 132 stress [14,16]. Neonatal treatment with TUDCA in animals born to obese dams reversed 133 induction of most ER stress markers in the postnatal and adult ARH, pancreas, liver, and 134 adipose tissue (Fig 2A-J), with the exception of Xbp1 in P10 ARH and pancreas (Fig 2A and E) 135 and in adult pancreas (Fig 2F), and Chop in adult adipose tissue (Fig 2J). Neonatal TUDCA 136 treatment also reduced normal mRNA levels of Xbp1 in P10 PVH and liver (Fig 2C and G), Bip 137 in P10 liver (Fig 2G), and Atf6 in adult liver (Fig 2H). Physiologically, neonatal TUDCA 138 treatment in animals born to obese dams reversed alterations in body weights, body 139 composition, adipocytes, food intake, energy expenditure, and glucose and insulin tolerances 140 (Fig 1D-G, I, and L-M), with only VO 2 not being improved (Fig 1H).  [8,17,18,19,20]. We therefore measured circulating 150 leptin levels in animals exposed to maternal obesity. Maternal HFHS feeding was associated 151 with a marked increase in serum leptin levels in dams at gestational day 16 and in E16.5 152 embryos (Fig 3A). Serum leptin levels were also elevated in P10 pups born to obese dams, 153 which were normalized upon neonatal TUDCA treatment (Fig 3A). However, serum leptin levels 154 were unchanged in adult mice born to HFHS-fed mothers (Fig 3A).

175
of P14 mice born to obese dams was 2-fold lower than that observed in control mice (Fig 4A).

176
In contrast, the density of AgRP-labeled projections innervating the PVH appeared normal in 177 P14 pups born to HFHS-fed dams (Fig 4A). Also, the number POMC and NPY positive cells in 178 the ARH of offspring of obese mice was comparable to that of control mice (S1  Our results show that overconsumption of a western-style diet rich in fatty acids during 191 pregnancy and lactation is associated with abnormal hypothalamic development. We also 192 measured circulating fatty acid concentration during pregnancy and found that dams fed a 9 HFHS diet have a 4-fold increase in serum fatty acid levels compared to control dams (Fig 5A).

194
Offspring born to obese dams also displayed higher levels of circulating fatty acids at P10 that compared to vehicle-treated cells (Fig 5B). In contrast, expression of ER stress markers was 205 not affected when cells were treated with the monousaturated fat oleic acid (Fig 5B).

206
We next assessed fatty acids intracellular transport in hypothalamic cells using BODIPY, 207 a fluorescent long-chain fatty acids analog. Exposure of hypothalamic N43/5 cells to a 208 combination of palmitic, lauric, and myristic acids resulted in greater BODIPY labeling in N43/5 209 cell bodies compared to vehicle-treated cells (Fig 5C). In order to determine if these saturated 210 fatty acids also impacted ARH axon growth and whether it involves ER stress, we also 211 performed a series of in vitro experiments in which ARH explants were microdissected, placed 212 in a collagen matrix, and then exposed to combination of saturated fatty acids (i.e., palmitic, 213 lauric, and myristic acids), or saturated fatty acids with TUDCA, or vehicle alone. After 48 hours, 214 the density of TUJ1-labeled neurite, neuron-specific class III beta-tubulin, from ARH explants 215 treated with saturated fatty acids was approximately 10-fold lower than that of vehicle-treated 216 explants (Fig 5D). Moreover, pre-incubation of ARH explants with TUDCA improved disrupted 217 axon outgrowth after saturated fatty acids treatment (Fig 5D).

10
Together, these data indicate that maternal obesity caused elevated circulating fatty acid 219 levels in the dams and offspring and that direct exposure to saturated fatty acids induced ER 220 stress gene expression in hypothalamic cells. They also show that saturated fatty acids can be 221 transported in hypothalamic cells blunting axon growth, and that this phenomenon appears to 222 involve ER stress pathways.

224
Although the link between perinatal overnutrition and lifelong metabolic regulation has been 225 clearly shown, little is known about the mechanisms underlying this programming effect. In this 226 study, we show that maternal obesity causes lifelong metabolic alterations associated with 227 abnormal development of hypothalamic feeding circuits in the offspring. We also report that

398
Slides were counterstained using bis-benzamide (Invitrogen; 1:10,000) to visualize cell nuclei.  stained with TUJ1 (b III tubulin) (rabbit, 1:5,000, Covance) as described previously [37].     For the quantitative analysis of fiber density (for POMC, AgRP, and TUJ1 fibers) and 429 BODIPY fluorescence, each image plane was binarized to isolate labeled materials from the 430 background and to compensate for differences in fluorescence intensity. The integrated 431 intensity, which reflects the total number of pixels in the binarized image, was then calculated 432 for each image as previously described [8,37]. This procedure was conducted for each image

448
We thank Brad Wanken and the CHLA Rodent Metabolic Core for metabolic studies. We also 449 thank the CHLA Cellular Imaging Core for confocal imaging studies. We are also grateful to 450 Gricelda Vasquez for her assistance with animal husbandry. This study was supported by the 451 National Institutes of Health (Grants DK84142, DK102780, and DK118401 to SGB).