The authors have declared that no competing interests exist.
Current address: Henry M. Jackson Foundation for the Advancement of Military Medicine, Expeditionary and Trauma Medicine, Combat Casualty Care, Naval Medical Research Unit-San Antonio (NAMRU-SA), 3650 Chambers Pass, JBSA-Ft. Sam Houston, San Antonio, TX, United States of America
Current address: Department of Pediatrics, Child Neurology/Neurodevelopmental Disabilities Residency Program, Baylor College of Medicine, Houston, TX, United States of America
Offspring of murine dams chronically fed a protein-restricted diet have an increased risk for metabolic and neurobehavioral disorders. Previously we showed that adult offspring, developmentally exposed to a chronic maternal low-protein (MLP) diet, had lower body and hind-leg muscle weights and decreased liver enzyme serum levels. We conducted energy expenditure, neurobehavioral and circadian rhythm assays in male offspring to examine mechanisms for the body-weight phenotype and assess neurodevelopmental implications of MLP exposure. C57BL/6J dams were fed a protein restricted (8%protein, MLP) or a control protein (20% protein, C) diet from four weeks before mating until weaning of offspring. Male offspring were weaned to standard rodent diet (20% protein) and single-housed until 8–12 weeks of age. We examined body composition, food intake, energy expenditure, spontaneous rearing activity and sleep patterns and performed behavioral assays for anxiety (open field activity, elevated plus maze [EPM], light/dark exploration), depression (tail suspension and forced swim test), sociability (three-chamber), repetitive (marble burying), learning and memory (fear conditioning), and circadian behavior (wheel-running activity during light-dark and constant dark cycles). We also measured circadian gene expression in hypothalamus and liver at different Zeitgeber times (ZT). Male offspring from separate MLP exposed dams had significantly greater body fat (P = 0.03), less energy expenditure (P = 0.004), less rearing activity (P = 0.04) and a greater number of night-time rest/sleep bouts (P = 0.03) compared to control. MLP offspring displayed greater anxiety-like behavior in the EPM (P<0.01) but had no learning and memory deficit in fear-conditioning assay (P = 0.02). There was an effect of time on
Maternal diet has a significant impact on fetal growth, with maternal malnutrition being a major cause of intrauterine growth restriction (IUGR) in developing countries [
Neural development begins in the early stages of gestation, a critical developmental period demonstrated to be sensitive to environmental, physiological and nutritional modifications [
Furthermore, the hypothalamus plays a central role in the establishment and maintenance of circadian rhythms [
We previously showed that male offspring from dams chronically fed MLP from 4 weeks prior to pregnancy onwards display reduced body weight, reduced size of specific hind limb muscles, lower serum levels of liver enzymes from weaning up to one year of age and altered expression of cohesin-mediator complex genes which may play a role in epigenetic regulation [
In the current study, we assessed whole body composition and energy expenditure, together with an extensive neurobehavioral examination of male offspring born to dams chronically fed a low protein diet. The goal was to investigate how chronic low protein diet in dams influences metabolism, circadian rhythm and neurobehavior in male mice offspring. We found that offspring of dams on low protein diet were smaller, had higher body fat content, showed nocturnal hypoactivity and lower levels of energy expenditure (EE). These offspring had no circadian rhythm alterations but exhibited mild anxiety related behavioral differences.
C57BL6/J mice were obtained from Center of Comparative Medicine at Baylor College of Medicine (BCM) and colonies were maintained at BCM. Dams were fed a protein-restricted diet before and during gestation, and throughout lactation as previously described [
Upon detecting a copulatory plug, pregnant females were housed individually. Litters were culled to a maximum of 6 male pups/dam on postnatal day (PND) 3. On PND21, male pups were weaned to a laboratory non-purified diet (20% protein; PicoLab Rodent Diet 20–5053, Lab diet) and housed individually. These singly housed pups were considered separate units for purpose of data analysis. Offspring from 2–5 litters were used in all the tests. Nestlets were provided to all offspring for environmental enrichment purposes. The study protocol was approved by the Institutional Animal Care and Use Committee (IACUC) at BCM. All experiments were conducted according to institutional and governmental regulations concerning the ethical use of animals in research. All animal facilities are approved by the Association for Assessment and Accreditation for Laboratory Animal Care International (AAALAC).
Measurements of body weight composition, energy balance, food intake and activity were performed in the Mouse Metabolic Research Unit at the USDA/ARS Children’s Nutrition Research Center BCM at 12 weeks of age on 7 male offspring of each group.
The Comprehensive Laboratory Animal Monitoring System (CLAMS) (Columbus Instruments) was used to monitor food intake, energy expenditure (EE) and activity in MLP and control mice (n = 7 each) as described previously [
The fat and lean content of the mice were measured with a Quantitative Magnetic Resonance (QMR) analyzer (EchoMedical, Houston Texas) at the completion of the CLAMS measurements.
All neurobehavioral assays were carried out in the Neurobehavioral core facility of the BCM Intellectual and Developmental Disabilities Research Center (IDDRC). The mice were acclimated to the procedure room 30 minutes before commencing the behavioral test. Unless otherwise noted, the lighting in the procedural room was 800–900 lux intensity, and the background noise maintained at 60±5 dB with the use of a white-noise generator. All the assays were conducted at approximately the same time of the day in batches, but for each batch, approximately equal numbers of controls and MLP-exposed offspring were simultaneously analyzed. Behavioral assays were simultaneously conducted on groups of 20 male MLP-exposed offspring and 20 control offspring at 8–12 weeks of age, except where other numbers are specified.
Mice were evaluated for anxiety-like behavior on a plus-shaped platform elevated 50 cm above the floor as previously described [
A total of 31 MLP and 13 control mice were evaluated by tail suspension test [
Mice were evaluated for depression-like behavior in the forced swim test [
Social behavior was evaluated using the three-chamber test [
Repetitive behavior was assayed using the marble burying test [
Fear-associated learning and memory was evaluated with a standard contextual and cued fear-conditioning assay as previously described [
Mice were assessed for exploratory and anxiety-like behavior using in open-field assay according to a standard protocol [
The light/dark exploratory assay was conducted per standard protocol [
The wheel-running assay was used to evaluate circadian behavior [
At 18–20 weeks of age, offspring housed under a L:D cycle for more than two weeks were randomly assigned to one of four tissue collection Zeitgeber time (ZT) points: ZT0, ZT6, ZT12, or ZT18. Animals were deeply anesthetized with isoflurane and rapidly decapitated. Hypothalamus was harvested and flash frozen in liquid N2, and stored at -80°C until needed. Tissues were homogenized with a pestle (P7339-901; Argos Technologies). Total RNA was isolated using miRNeasy Mini Kits (217004; Qiagen) following manufacturer’s instruction. An on-column DNase treatment was performed using RNase-Free DNase Set (79254; Qiagen). cDNA was synthesized using qScript cDNA Supermix (95048; Quanta Biosciences). qRT-PCR was performed using PerfeCTa® SYBR® Green FastMix (95072; Quanta Biosciences) on the Bio-Rad CFX Connect Real-Time instrument. The ΔΔ
Statistical analysis was performed considering each offspring as a single unit, as they were housed as one mouse/cage. Energy expenditure and food intake data were analyzed by ANCOVA using lean and fat mass or body weight as covariates. Tukey’s test was applied to
Because we have shown previously that adult male MLP offspring had lower overall body weight and decreased weight of selected hind leg muscles [
(A): Body weight; (B) Body fat as percent of total body mass in male offspring (8–12 weeks age) from dams exposed to either control or MLP diet (n = 7 each). Bars are mean ± SEM and P<0.05 statistically significant by student t-test.
We next compared food intake and energy balance. Overall, total daily food intake, adjusted for variations in body weight was similar between control and MLP offspring (P = 0.28), with MLP displaying a lower intake during the dark phase (P<0.1;
Control | MLP | P | ||
---|---|---|---|---|
n | 7 | 7 | ||
Food Intake |
||||
24 h | 4.07 ± 0.22 | 3.71 ± 0.22 | 0.28 | |
12 h light | 1.27 ± 0.08 | 1.33 ± 0.08 | 0.59 | |
12 h dark | 2.81 ± 0.16 | 2.38 ± 0.16 | 0.10 | |
Total Energy expenditure (EE) |
||||
24 h | 9.93 ± 0.18 | 9.05 ± 0.18 | <0.02 | |
12 h light | 4.41 ± 0.10 | 4.10 ± 0.10 | 0.08 | |
12 h dark | 5.53 ± 0.09 | 4.95 ± 0.09 | <0.01 | |
Respiratory Exchange Ratio (RER) | ||||
24 h | 0.868 ± 0.004 | 0.861 ± 0.004 | 0.53 | |
12 h light | 0.819 ± 0.008 | 0.830 ± 0.008 | 0.12 | |
12 h dark | 0.917 ± 0.003 | 0.892± 0.003 | <0.001 |
1Values are least square means, adjusted for body weight, ± SEM
2 Values are least square means, adjusted for lean and fat mass, ± SEMP<0.05 was considered statistically significant by ANCOVA with Tukey’s post hoc test.
In addition to feeding, spontaneous activity and sleeping are two behaviors that are known to differ in frequency between the dark and light phases, and which could have a significant influence on EE. The CLAMS system provides quantitative measures of total activity in the horizontal plane (X total) comprised of ambulatory activity (X ambulatory) and non-ambulatory activity associated with stereotypical behaviors such as grooming and twitching (X fidgeting), and rearing in the vertical plane (Z total). Although all mean values were numerically smaller in MLP mice, only Z total counts were significantly different (P = 0.04,
Control | MLP | P | |
---|---|---|---|
n | 7 | 7 | |
24 hours | |||
X total | 41708 ± 226 | 37801 ± 241 | 0.26 |
X ambulatory | 21097 ± 155 | 18542 ± 166 | 0.28 |
X fidgeting | 20611 ± 821 | 19259 ± 878 | 0.28 |
Z total | 6377 ± 656 | 4270 ± 701 | 0.05 |
Total activity (X total +Z total) | 48085 ± 284 | 42070 ± 303 | 0.17 |
Light/resting phase | |||
X total | 8493 ± 410 | 8053 ± 438 | 0.48 |
X ambulatory | 3163 ± 265 | 3144 ± 265 | 0.96 |
X fidgeting | 5329 ± 194 | 4910 ± 208 | 0.16 |
Z total | 733 ± 147 | 477 ± 157 | 0.26 |
Total activity (X total + Z total) | 9225 ± 563 | 8531 ± 563 | 0.38 |
Dark/Active phase | |||
X total | 33215 ± 206 | 29747 ± 223 | 0.28 |
X ambulatory | 17933 ± 145 | 15398 ± 155 | 0.25 |
X fidgeting | 15282 ± 723 | 14349 ± 772 | 0.39 |
Z total | 5645 ± 547 | 3792 ± 585 | 0.04 |
Total activity (X total + Z total) | 38860 ± 257 | 33539 ± 275 | 0.18 |
Values are means ± SEM. P<0.05 was considered statistically significant by ANCOVA with Tukey’s post hoc test
Given the lower rearing activity during the dark phase, we assessed the amount of time mice spent resting/sleeping. Over 24 hours, MLP offspring showed a non-significant trend for overall more time resting/sleeping (P = 0.07). This trend results from MLP offspring spending significantly more time resting/sleeping during the dark phase (P = 0.033), when mice are normally more active, but there was no difference between the groups in the time spent resting/sleeping during the light phase (
(A): Total resting/sleeping time/day; (B): Sleep bouts/day; (C): Sleep bout duration; (D): Maximum sleep bout duration (n = 7/group, 12 weeks age). Data shown is mean ± SEM with P<0.05 considered statistically significant by student t-test.
Taken together, the data suggest that the altered behavior of MLP offspring during the dark phase described above contributed to their lower overall EE. This could be due to circadian activity abnormality or behavioral changes, including anxiety or depression, which has been reported before [
To examine if circadian rhythm abnormalities are at the origin of the altered rest/sleep pattern described above, we carried out circadian behavior testing. At the end of the two-week light:dark (L:D) acclimation phase, (
(A) Representative double-plotted actogram for control (left panel) and MLP (right panel) offspring. Zeitgeber times (ZT) (0, 6, 12, 18) are represented on top. Black and white blocks below ZT’s indicate dark and light conditions, respectively. Mice were individually housed first in 12 h light:12 h dark for acclimation (L:D) (white background, top half) then in constant darkness (D:D) (gray background, bottom half). Black notches represent activity. The total wheel revolutions during final 10 days of the (B) acclimation (Control, n = 9; MLP, n = 15), (C) constant darkness (Control, n = 8; MLP, n = 15), and (D) Periodicity (Tau) (Control, n = 8; MLP, n = 15) is shown. (E) Total RNA (mRNA) from hypothalamus of MLP and control offspring (n = 3/time point/group) after two weeks housing under L:D cycle were extracted at ZT 0, 6, 12, or 18. Data shown is relative gene expression for indicated circadian genes (numerator) over the housekeeping genes
Multiple studies have linked food intake with circadian transcriptome changes [
Studies have shown that dietary modifications during gestation can influence neurodevelopment in male [
There was no significant difference between MLP and control offspring in any of the tested parameters in the open field assay (
Elevated plus maze test (MLP and Control, n = 9 each); (A) Time spent in open arm, center and closed arm, (B) Latency to enter closed arm. (C) Tail suspension test (MLP, n = 31 and Control, n = 13) and (D) Forced swim test (MLP and Control, n = 20 each). Data is presented as mean ± SEM. P<0.05 was considered statistically significant by student t-test.
To investigate the potential anxiety-like phenotype observed in the EPM more completely, we next performed the light/dark exploration assay, but saw no significant difference between the two groups in latency to enter the dark compartment (P = 0.79) (
Because anxiety-like traits were modest and MLP offspring also displayed increased resting/sleeping during the dark phase (
We next evaluated sociability, repetitive behavior, and associative learning and memory of the MLP and control offspring by the three chamber (3CH), marble burying and fear conditioning test, respectively. As shown in
(A) Three-chamber test (MLP and Control, n = 9 each), (B) Marble burying (MLP and Control, n = 9 each). Data is presented as mean ± SEM. P<0.05 was considered statistically significant by student t-test for (A) and (B).
Fear conditioning test in MLP and Control male offspring mice. (A) Training phase (MLP and Control, n = 20 each), (B) Context phase, (C) pre cue phase, and (D) cue phase. MLP, n = 18; Control, n = 19 for (B-D). Data is presented as mean ± SEM. P<0.05 was considered statistically significant by Mann Whitney test.
Taken together, the above data suggests that exposure to chronic MLP diet may affect offspring behavior, manifested as some increase in anxiety but no deficit in learning and memory. We also did not find objective evidence of depression-like behaviors or deficits in sociability.
The primary objective of this study was to comprehensively characterize the metabolic consequences, neurobehavioral and circadian rhythm effects on male offspring of C57/Bl6J dams chronically exposed to a protein-restricted diet from four weeks before gestation and continued throughout lactation. Previously, we have shown in this model that male MLP offspring have reduced body weight gain compared to controls [
To evaluate the cause of body composition difference, we measured the major components of energy balance. There was no difference in food intake between MLP and control offspring, but there was a decrease in total daily energy expenditure (EE), that occurred during the dark phase (
To understand the differences in EE between groups, we considered the various components of EE. Activity and resting EE make the largest contribution to total EE. We did not measure resting EE in this study and, therefore, cannot be certain whether it may have contributed to the difference in total EE between groups. However, under the experimental conditions of this study, differences in resting EE would exhibit minimal circadian variation and, thus, are unlikely to account for the observed decrease in EE of the MLP group which mostly occurred in the dark phase. The evaluation of activity patterns and sleep patterns, however, were more insightful. MLP offspring displayed more hypoactive behaviors in the CLAMS only during the dark/active phase, manifested as a reduction in rearing, and an increase in time spent inactive, i.e., resting/sleeping, (
Because the largest discrepancy in total EE between groups occurred at a time associated with lower activity, we explored which behavioral abnormalities might underlie these observations. Specifically, we set out to examine if the MLP offspring had a disrupted circadian rhythm or altered neurobehavior, such as a depressive-like state or anxiety. Therefore, we performed comprehensive set of neurobehavioral assays for depression like features, anxiety, sociability, and learning and memory.
We did not find circadian rhythm abnormalities as measured by wheel-running activities in the L:D, D:D test (
We found a modest effect on offspring anxiety (
Although our findings were more subtle, a stronger anxiety and/or depressive phenotype in MLP offspring in the CF1 strain of mice [
Housing conditions can also drastically impact behavior. In this model, offspring from both groups were continually, single-housed beginning at weaning (day 21 of age) to have a more controlled environment for individual food intake standardization, but mice are naturally social animals, and previous studies have demonstrated that chronic exposure to single housing, even with environmental enrichment, can by itself alter behavior [
In this mouse model of exposure to chronic maternal protein malnutrition, we have demonstrated that offspring are smaller, but have higher body fat content. This difference in body composition was in conjunction with lower levels of EE and nocturnal hypoactivity during the normally active time for mice. This difference was not explained by circadian rhythm alterations, but the mice exhibited mild anxiety-related behavioral differences along with nocturnal hypoactivity. Future studies will focus on epigenetic, neuroanatomical and neurophysiological correlates of these effects, which can uncover potential future therapeutic targets that can modify this complex phenotype.
MLP and Control offspring (18–20 weeks age, n = 9 each) examined by open field test. The test examined (A) total distance traveled; (B) center distance traveled; (C) total time moving; (D) speed of mobility; (E) stereotypy; (F) vertical activity; (G) total revolutions. Data is presented as mean ± SEM with P<0.05 considered statistically significant by student t-test.
(DOCX)
MLP and Control offspring (18–20 weeks age, n = 20 each) tested in light/dark exploration assay. Data presented are (A) latency to enter dark chamber, (B) time spent in light chamber, and, (C) total transitions between light and dark chambers. Data shown is mean ± SEM with P<0.05 considered statistically significant by student t-test.
(DOCX)
(DOCX)
The authors acknowledge the expert assistance of Mr. Firoz Vohra at the USDA/ARS Children’s Nutrition Research Center, Baylor College of Medicine for CLAMS analyses, Dr. Corinne Spencer of Baylor Mouse Neurobehavior Core for behavior tests and Dr. Susan Hilsenbeck, Biostatistics and Informatics group, Lester & Sue Smith Breast Center, Baylor College of Medicine, for statistical consultation.