Abstract
Dietary malpractice is a risk factor for obesity. This study tested the hypothesis that consumption of a high-fat diet alters mammary metabolome in pubertal mice. We performed untargeted metabolomic analysis of primary metabolism on mammary glands from pubertal mice fed the AIN93G standard diet or a high-fat diet (HFD) for 3 weeks. We identified 97 metabolites for statistical comparisons. The HFD altered the amino acid metabolism considerably. This included elevated expression of branched-chain amino acids, non-essential amino acids (aspartic acid and glutamic acid), and methionine sulfoxide (oxidized methionine) and an alteration in the aminoacyl-tRNA biosynthesis pathway. Furthermore, elevations of fumaric acid and malic acid (both are citrate cycle intermediates) and glyceric acid (its phosphate derivatives are intermediates of glycolysis) in HFD-fed mice suggest an acceleration of both citrate cycle and glycolysis. Lower expression of glycerol, oleic acid, and palmitoleic acid, as well as decreased mammary expression of genes encoding lipid metabolism (Acaca, Fads1, Fasn, Scd1, and Srebf1) in HFD-fed mice indicate an attenuated lipid metabolism in the presence of adequate dietary fat. In conclusion, consumption of the HFD for 3 weeks alters metabolic profile of pubertal mammary glands. This alteration may affect mammary development and growth in pubertal mice.
Introduction
Childhood obesity has reached epidemic proportions worldwide. 1 It remains one of the most prevalent public health problems. Childhood obesity in girls is associated with an increased risk of breast cancer later in life. Studies have shown that obese girls exhibit an early onset of puberty2,3 and breast development4,5 with elevations of ovarian steroid hormones. Early menarche in girls is associated with an increased risk of adulthood breast cancer.6-8
Mammary gland development mainly occurs during puberty. With the onset of puberty, ductal morphogenesis creates epithelial mammary trees that fill the mammary fat pad. Ovary-derived steroid hormones, in addition to pituitary hormones, play important roles in the growth of pubertal mammary glands. Estrogen acts directly on mammary glands; estrogen receptor 1 regulates ductal morphogenesis9,10 and estrogen receptor 2 facilitates terminal differentiations. 11 Obese girls with early onset of breast development have elevated estrogen in saliva compared to girls with normal body weights. 5 Laboratory studies have shown that higher concentrations of estrogen and its receptors occur in mammary glands from pubertal mice fed an obesogenic diet compared to those fed a standard control diet. 12
Dietary malpractice is a risk factor for obesity. Feeding weanling mice a high-fat diet increases body fat mass,13,14 disrupts the pattern of metabolism (eg, respiratory exchange ratio), 14 and changes plasma concentrations of cytokines that are involved in metabolism (eg, leptin and adiponectin).13,14 Furthermore, the high-fat diet alters metabolic profile of mammary tumors 15 and enhances mammary tumorigenesis 16 in MMTV-PyMT mice. However, there is a lack of knowledge of the impact of dietary fat on pubertal mammary metabolism. We hypothesized that consumption of a high-fat diet alters mammary metabolome in pubertal mice. In this study, we performed an untargeted metabolomic analysis of primary metabolism on mammary glands from pubertal mice fed a high-fat diet in comparison to that from their age-matched counterparts fed the AIN93G standard diet.
Materials and Methods
Animals and diets
Three-week-old female C57BL/6NHsd mice (Envigo, Madison, WI, USA) were group-housed (3 per cage) in a pathogen-free room with a 12:12-hour light/dark cycle and a temperature of 22 ± 1°C. The standard AIN93G diet 17 and a high-fat diet (HFD) containing 16% and 45% of energy from soybean oil, respectively, were used in this study (Table 1). Diets (powder form) were kept at −20°C. Fresh diets were provided to mice every other day.
Diet composition.
Quantified by using an oxygen bomb calorimeter, Model 6200 (Parr Instrument, Moline, IL, USA), n = 3 per diet (means ± SEM).
Experimental design
The study design was reported previously. 12 Briefly, mice were acclimated with the AIN93G diet for 2 days before they were randomly assigned to 2 groups of 70 each and fed the AIN93G diet or HFD and deionized water ad libitum for 3 weeks. Mice were weighed weekly. At the end of study, 5 mice from each group were euthanized (intra-peritoneal injection of ketamine and xylazine) and followed by exsanguination every 4 hours at Zeitgeber time (ZT) 0, 4, 8, 12, 16, 20, and 24 over a period of 48 hours. Mammary glands were collected and stored at −80°C.
Metabolomic analysis
Twelve mammary samples from each group (harvested in the dark phase at ZT 16 and 20) were used for metabolomic analysis. The analysis18,19 was performed at the West Coast Metabolomics Center (University of California-Davis, Davis, CA, USA). Briefly, mammary tissues were extracted by the acetonitrile/isopropanol/H2O (3:3:2) buffer, resuspended in a diluted acetonitrile solution (acetonitrile/H2O, 1:1), and analyzed by gas chromatography time-of-flight mass spectrometry (GC-TOF-MS) for untargeted metabolomics of primary metabolism. Obtained data were processed by using the BinBase database. 20 For identified compounds, quantifier ion peak heights were normalized to the sum intensities of all known compounds. Compounds representing less than 0.02% of total signal intensity were excluded from statistical analysis. Additional compounds were excluded from statistical analysis if they could not be identified as metabolites or metabolic intermediates common to mammalian metabolism based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database or the Human Metabolome database.21-23
RNA isolation and real-time quantitative PCR
Total RNA from mammary glands was isolated by using the QIAzol Lysis reagent following the protocol from RNeasy Lipid Tissue Mini Kit (Qiagen, Germantown, MD, USA). The quality and quantity of the RNA were analyzed by using the NanoDrop 8000 spectrophotometer (Thermo Scientific, Wilmington, DE, USA). cDNA was synthesized by using the high-capacity cDNA reverse transcription kit (Applied Biosystems, Waltham, MA, USA). Real-time qPCR of acetyl-CoA carboxylase (Acaca) (Mm01304257_m1), fatty acid desaturases (Fads) 1 (Mm00507605_m1) and 2 (Mm00517221_m1), fatty acid synthase (Fasn) (Mm00662319_m1), stearoyl-CoA desaturase 1 (Scd1) (Mm00772290_m1), and sterol regulatory element-binding protein 1 (Srebf1) (Mm00550338_m1) was analyzed and normalized to the 18 S rRNA (Mm03928990_g1) by using the TaqMan Assay of Demand primers on the ABI QuantStudio 12K-Flex Real-time PCR system (Applied Biosystems). The 2−ΔΔCT method was used to calculate the relative changes in gene expression. 24
Statistical analyses
The GLM procedure was used to compare differences in metabolites between the AIN93G and HFD groups with the false discovery rate (FDR)-corrected p values (obtained by PROC MULTTEST) reported (SAS 9.4, SAS Institute, Cary, NC, USA). Student t-test was performed to compare differences in genes encoding lipid metabolism between the 2 groups (SAS 9.4). For metabolomic analysis, data were normalized by the Pareto scaling method and analyzed by sparse partial least square-discriminant analysis (sPLS-DA) (MetaboAnalyst 5.0, McGill University, Sainte Anne de Bellevue, Quebec, Canada).25,26 Obtained results from the HFD group are reported as fold changes to that from the AIN93G group. Results are reported as means ± standard error of the mean (SEM); differences with a P ⩽ .05 are considered significant.
Pathway and network analyses were conducted by using MetaboAnalyst 5.0 (McGill University). 27 Pathway analysis of alterations in metabolic pathways by the HFD was performed by using the pathway library for Mus musculus according to the KEGG database. 28 Pathway enrichment analysis coupled with pathway topology analysis was conducted to identify the altered metabolic pathways with the Holm-adjusted P values reported. 29 Network analysis was performed to map the functional relationships of identified metabolites by using the KEGG global metabolic network and the metabolite-metabolite interaction network analyses. Differences with a P ⩽ .05 are considered significant.
Results
Metabolomic analysis
Feeding pubertal mice the HFD for 3 weeks resulted in a slight, but significant increase in body weight (19.0 ± 0.2 and 18.2 ± 0.2 g; P < .05, n = 70 per group). 12 We identified a total of 149 compounds from 361 discrete signals detected by GC-TOF-MS (Supplemental Table 1). Ninety-seven of the 149 compounds met the criteria for statistical analysis. They were grouped into 4 categories related to amino acid, lipid, energy, and nucleotide and vitamin metabolism (Tables 2–5). Twenty-two of the identified metabolites differed significantly between the AIN93G and HFD groups (Tables 2–5).
Identified metabolites related to amino acid metabolism in mammary glands from pubertal mice fed the AIN93G or high-fat diet.
Values (means ± SEM) of the high-fat diet group are standardized to that of the AIN93G group (n = 12 per group). Differences at P ⩽ .05 are considered significant (FDR-adjusted P values).
Identified metabolites related to lipid metabolism in mammary glands from pubertal mice fed the AIN93G or high-fat diet.
Values (means ± SEM) of the high-fat diet group are standardized to that of the AIN93G group (n = 12 per group). Differences at P ⩽ .05 are considered significant (FDR-adjusted P values).
Identified metabolites related to energy metabolism in mammary glands from pubertal mice fed the AIN93G or high-fat diet.
Values (means ± SEM) of the high-fat diet group are standardized to that of the AIN93G group (n = 12 per group). Differences at P ⩽ .05 are considered significant (FDR-adjusted P values).
Identified metabolites related to nucleotide and vitamin metabolism in mammary glands from pubertal mice fed the AIN93G or high-fat diet.
Abbreviation: N-methyl-UMP, N-methyl-uridine-5-monophosphate.
Values (means ± SEM) of the high-fat diet group are standardized to that of the AIN93G group (n = 12 per group). Differences at P ⩽ .05 are considered significant (FDR-adjusted P values).
The hierarchical clustering heatmap analysis of the identified metabolites in mammary glands produced 2 responsive clusters of the 20 metabolites that differed significantly between the 2 groups (Figure 1). The expression of 7 metabolites was lower in HFD-fed mice than in AIN93G-fed mice. These included oleic acid (18:1), palmitoleic acid (16:1), hypotaurine, methionine, lactose, glycerol, and α-tocopherol. The expression of 13 metabolites was higher in the HFD-fed mice. These included fumaric acid, uracil, ornithine, asparagine, methionine sulfoxide, malic acid, valine, xanthine, glyceric acid, tyrosine, glutamic acid, leucine, and isoleucine.
Hierarchical clustering heatmap of the 20 metabolites in mammary glands that differed significantly between the AIN93G-fed and high-fat diet (HFD)-fed pubertal mice (P < .05 with applied FDR). Heatmap squares represent the mean of 12 samples. The red color indicates greater signal strength and blue color indicates weaker signal strength by GC-TOF-MS.
The sPLS-DA scores plot showed a clear separation between the AIN93G and HFD groups (Figure 2A). The loadings plot for component 1 showed that amino acids and amino acid metabolites (leucine, methionine sulfoxide, isoleucine, glutamic acid, hypotaurine, and tyrosine), lipid metabolites palmitoleic acid (16:1) and glycerol, the citrate cycle intermediate malic acid, and vitamin α-tocopherol are major determinants of separation (Figure 2B).
The scores plot by the sparse partial least square discriminant analysis (sPLS-DA) of mammary metabolites (A) and loadings plots of the 10 metabolites that are most influential in treatment separation between the AIN93G and high-fat diet groups for components 1 (B), 2 (C), and 3 (D) (n = 12 per group). The x-axis shows that variables are ranked by the absolute values of their loadings. The red color indicates greater signal strength and blue color indicates weaker signal strength by GC-TOF-MS. Adnosine-5-MP, adenosine-5-monophosphate.
The loadings plot for component 2 demonstrated that metabolites related to energy metabolism (maltotriose, 2-hydtoxyglutaric acid, 1,5-anhydroglucitol, succinic acid, and maltose), amino acids (threonine, asparagine, and lysine), nucleotide adenine, and fatty acid linoleic acid (18:2) were major determinants of separation (Figure 2C).
The loadings plot for component 3 showed that lipid metabolites (lauric acid, ethanolamine, and myristic acid), metabolites related to energy metabolism (3-hydroxybutyric acid, N-acetylmannosamine, citric acid, and sorbitol), amino acid metabolites lysine and β-alanine, and nucleotide adenosine-5-monophosphate were major determinants of separation (Figure 2D).
Pathway analysis
The pathway analysis examined alterations in metabolic pathways by the HFD. It mapped the identified metabolites into 57 metabolic pathways (Supplemental Table 2) based on the KEGG database.22,28 Five pathways were altered significantly by the HFD (Figure 3 and Table 6). These pathways included aminoacyl-tRNA biosynthesis, alanine, aspartate, and glutamate metabolism, arginine biosynthesis, pantothenate and CoA biosynthesis, and branched-chain amino acid (BCAA, valine, leucine, and isoleucine) biosynthesis.
Metabolic pathways from identified metabolites in mammary glands from pubertal mice fed the AIN93G or high-fat diet (n = 12 per group). Each node represents a pathway. The nodes with larger sizes and darker color (from white-yellow to red) positioning toward the top right region represent greater pathway impact values and pathway enrichment. Pathways that are altered significantly are presented with their names next to their nodes. All metabolic pathways identified in this study are available in Supplemental Table 2.
Metabolic pathways identified by the pathway analysis and network analysis that are significantly altered by the high-fat diet.
Number of identified metabolites that match to pathway metabolites (calculation based on total number of metabolites identified in this study).
Obtained by the over-representation analysis and adjusted by the Holm method.
Obtained by the pathway topology analysis.
Number of identified metabolites that match to pathway metabolites (calculation based on the number of identified metabolites that differ significantly between the AIN93G and high-fat groups).
False discovery rate-adjusted P values.
Network analyses
The KEGG global metabolic network analysis identified 31 metabolic pathways that were altered by the HFD (Supplemental Table 3). Among them, 6 were altered significantly (Table 6). These included aminoacyl-tRNA biosynthesis, arginine biosynthesis, BCAA biosynthesis, alanine, aspartate, and glutamate metabolism, pantothenate and CoA biosynthesis, and glycerolipid metabolism. The metabolite-metabolite interaction network analysis showed the functional relationship of the identified metabolites (Figure 4).
Metabolic network of identified metabolites in mammary glands from pubertal mice fed the AIN93G or high-fat diet (n = 12 per group). Colors (from white-yellow to red) represent levels of impact the metabolites have to the network in an ascending order (the number of connections a node has to other nodes and the number of shortest paths going through the node).
Expression of genes encoding lipid metabolism
The HFD, compared to the AIN93G diet, decreased mammary expression of genes encoding enzymes that are involved in lipid metabolism. These genes were Acaca, Fads1, Fasn, Scd1, and Srebf1 (Table 7). There was no significant difference in Fads2 between the 2 groups (Table 7).
Expression of genes encoding enzymes involved in lipid metabolism in mammary glands from pubertal mice fed the AIN93G or high-fat diet.
Values are means ± SEM (n = 12 per group).
Discussion
The present study showed that consumption of an HFD for 3 weeks altered metabolomic profile of mammary glands in pubertal mice.
Both pathway and network analyses showed that 4 of the metabolic pathways that were significantly altered by the HFD were related to amino acid metabolism. The aminoacyl-tRNA biosynthesis pathway is important in protein synthesis. 30 It is a group of aminoacyl-tRNA synthetases that catalyze aminoacylations by covalently linking an amino acid to its cognate tRNA in the first step of protein translation. The other 3 pathways are alanine, aspartate, and glutamate metabolism, arginine biosynthesis, and BCAA biosynthesis pathways.
The alteration of the BCAA biosynthesis pathway is supported by elevations of isoleucine, leucine, and valine in mammary glands from mice fed the HFD. BCAAs account for one-third of dietary essential amino acids. 31 They activate mammalian target of rapamycin complex (mTORC1) signaling pathway, which is essential for initiation of protein synthesis.32,33 The BCAAs increase protein synthesis in lactating mammary glands 34 and breast cancer tissue 35 through the activation of mTORC1 signaling pathway. Increases of BCAAs in bloodstream occur in human obese subjects,36,37 indicating abnormalities in BCAA metabolism in obesity. Findings from our study suggest that the HFD may impair BCAA metabolism or utilization in pubertal mammary glands.
The observed elevations of aspartic acid and glutamic acid in mammary glands and the alteration of the alanine, aspartate, and glutamate metabolism pathway by the HFD suggest an increase in biosynthesis of non-essential amino acids in pubertal mammary glands. These changes may be related to the altered BCAA metabolism. BCAAs are an indirect source of nitrogen for non-essential amino acid synthesis in mitochondria through α-ketoglutarate, a citrate cycle intermediate.38,39 BCAAs interact with α-ketoglutarate to form glutamate, which in turn interacts with oxaloacetate to form aspartate and α-ketoglutarate.
Observations of decreased mammary expression of methionine and increased expression of methionine sulfoxide in HFD-fed mice indicate alterations in methionine metabolism. Methionine sulfoxide is the oxidized form of methionine; this oxidation can be reversed by methionine sulfoxide reductases (Msr). 40 Increases in methionine sulfoxide have been associated with breast cancer development in in vitro and animal studies. For example, silencing MsrA in MDA-MB-231 breast cancer cells increases cell proliferation and extracellular matrix degradation and consequently results in more aggressive phenotype in nude mice. 41 High levels of MsrB1 transcript and promoter activity occur in lower metastatic MCP7 breast cancer cells, whereas low levels of both occur in high metastatic MDA-MB-231 cells. 42 Furthermore, the catabolism of methionine is linked to that of BCAAs through branched-chain keto-acid dehydrogenase. 43 Thus, the observed elevation of BCAAs by the HFD may be interrelated with the lower expression of methionine, as methionine-sparing effect of BCAAs has been reported in animal studies. 44
Greater mammary expression of fumaric acid and malic acid occurred in HFD-fed mice than in their AIN93G-fed counterparts. Furthermore, the HFD increased glyceric acid and decreased glycerol in mammary glands. Both fumarate and malate are intermediates of the citrate cycle. Glyceric acid is an oxidative product of glycerol. Several phosphate derivatives of glyceric acid are important intermediates of glycolysis (eg, 1,3-biphosphoglycerate). These findings suggest that the HFD may have facilitated anabolic processes through both citrate cycle and glycolysis to supply ATP to pubertal mammary glands.
The mammary expression of genes encoding fatty acid and triacylglycerol metabolism (Acaca, Fads1, Fasn, Scd1, and Srebf1) was lower in HFD-fed mice than in AIN93G-fed controls. Furthermore, oleic acid, palmitoleic acid, and glycerol were lower in mammary glands from HFD-fed mice. Oleate and palmitoleate are metabolic products of stearoyl-CoA desaturase 1 (encoded by Scd1) from stearoyl-CoA and palmitoyl-CoA, respectively. Glycerol provides the carbon backbone for synthesis of triacylglycerols, which is catalyzed by sterol regulatory element-binding transcription factor 1 (encoded by Srebf1). Our findings suggest that the HFD may have satisfied the anabolic need for lipids by providing adequate fatty acids from the digested triacylglycerols, which attenuates de novo lipid synthesis in pubertal mammary glands.
The finding of lactose in pubertal mammary glands was unexpected. Lactose is the primary carbohydrate in milk. Its synthesis is catalyzed by lactose synthetase in mammary epithelial cells during late pregnancy and lactation. Pubertal mammary development undergoes hormonal regulation. Puberty initiates branching morphogenesis, which requires growth hormone and estrogen to create ductal trees that fill the fat pad. 45 The AIN93G diet is formulated for growth, pregnancy, and lactation for rats and mice. 17 It conforms to or exceeds the nutrient requirements for laboratory rodents set forth by the National Research Council. 17 We quantified estrogen, progesterone, and their receptors in mammary glands from these pubertal mice. 12 Our findings suggest that the AIN93G formulation may play a role in ovary steroid hormone-mediated regulation in pubertal mammary development, which may be responsible, at least partly, for the lactose found in pubertal mammary glands.
Both lactose and glycerol (an alternative source of carbon for lactose synthesis) in mammary glands were lower in HFD-fed mice than in AIN93G-fed controls. This observation suggests that the HFD may have disturbed the mechanism through which the AIN93G diet elevated lactose in pubertal mammary glands. However, higher mammary concentrations of estrogen and estrogen receptor 1 occur in these HFD-fed pubertal mice. 12 Thus, the roles of the HFD-mediated elevation of ovary steroid hormones in pubertal mammary development, particularly that related to lactose production, warrant further investigation.
Mammary expression of α-tocopherol in HFD-fed mice was lower than that in AIN93G-fed mice. α-Tocopherol protects polyunsaturated fatty acids in cell membranes and lipoproteins against oxidation. Intestinal absorption α-tocopherol depends on the quantity of fat in a diet. Tissue storage of α-tocopherol relies on α-tocopherol transfer protein46,47 and tocopherol-associated protein 48 to transfer α-tocopherol between membranes. Catabolism of α-tocopherol is initiated by cytochrome P450 4F2/cytochrome P450 3A4. 49 The lower mammary α-tocopherol in HFD-fed mice does not seem to be due to a decrease in intestinal absorption because the HFD is a lipid-rich diet and soybean oil is a good source of α-tocopherol. 50 Rather, it may be related to a decrease in transport, an increase in catabolism, or both. α-Tocopherol reduces cell proliferation by inactivating protein kinase C in different cell types.51,52 Thus, the impact of the HFD-mediated alteration in α-tocopherol metabolism on pubertal mammary development remains an interest for further investigation.
Laboratory rodents are nocturnal animals. They are active in the dark phase of the day (ZT 12-24). In a recent transcriptome analysis of liver samples from these pubertal mice, we found that the HFD-fed mice, compared to the AIN93G-fed controls, had greater numbers of differentially expressed genes and KEGG metabolic signaling pathways in the dark phase (particularly ZT 16-20) and fewer numbers in the light phase. 53 The present metabolomic analysis was performed by using mammary samples collected during the dark phase (ZT 16-20). Taken together with results from the transcriptome analysis, our findings indicate that the HFD alters metabolism of nocturnal rodents the most in the dark phase.
In conclusion, feeding pubertal mice the HFD for 3 weeks alters metabolic profile of mammary glands, with considerable alteration in amino acid metabolism. This alteration may affect mammary development and growth in pubertal mice. The impact of this metabolomic alteration by the HFD on breast health in adult ages remains an interest for further investigation.
Supplemental Material
sj-xlsx-1-nmi-10.1177_11786388221148858 – Supplemental material for Metabolomic Alterations in Mammary Glands from Pubertal Mice Fed a High-Fat Diet
Supplemental material, sj-xlsx-1-nmi-10.1177_11786388221148858 for Metabolomic Alterations in Mammary Glands from Pubertal Mice Fed a High-Fat Diet by Lin Yan, Bret M Rust, Sneha Sundaram and Michael R Bukowski in Nutrition and Metabolic Insights
Supplemental Material
sj-xlsx-2-nmi-10.1177_11786388221148858 – Supplemental material for Metabolomic Alterations in Mammary Glands from Pubertal Mice Fed a High-Fat Diet
Supplemental material, sj-xlsx-2-nmi-10.1177_11786388221148858 for Metabolomic Alterations in Mammary Glands from Pubertal Mice Fed a High-Fat Diet by Lin Yan, Bret M Rust, Sneha Sundaram and Michael R Bukowski in Nutrition and Metabolic Insights
Supplemental Material
sj-xlsx-3-nmi-10.1177_11786388221148858 – Supplemental material for Metabolomic Alterations in Mammary Glands from Pubertal Mice Fed a High-Fat Diet
Supplemental material, sj-xlsx-3-nmi-10.1177_11786388221148858 for Metabolomic Alterations in Mammary Glands from Pubertal Mice Fed a High-Fat Diet by Lin Yan, Bret M Rust, Sneha Sundaram and Michael R Bukowski in Nutrition and Metabolic Insights
Footnotes
Acknowledgements
The authors gratefully acknowledge Dr. Kathleen Yeater (Plains Area Administrative Office, Agricultural Research Service) and Daniel Palmer for assistance in statistical analyses and the assistance from Lana DeMars and Jack He for technical support and vivarium staff for preparing experimental diets and providing high-quality animal care. USDA is an equal opportunity provider and employer. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the USDA.
Funding:
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by the USDA Agricultural Research Service Projects #3062-51000-050-446 00D and #3062-51000-056-00D.
Declaration of Conflicting Interests:
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Author Contributions
LY and SS designed the study, conducted experiments, collected data, and wrote the manuscript. All authors contributed to result interpretation and manuscript revision.
Ethics Statement
The Institutional Animal Care and Use Committee of the Grand Forks Human Nutrition Research Center approved this study.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
