ABSTRACT: Purpose: Dietary fat-driven hyperphagia increase intestinal growth and permeability and create microbial imbalance in the gut. Over-nutrition coupled with increase in intestinal nutrient absorbance capacity then lead to nutrient overload in the liver, visceral adipose tissue followed by ectopic fat deposition with consequential lipotoxicity and inflammation. We investigated how protein quantity (10-30%) and quality (casein and whey) interact with dietary fat (20-55%) to affect these outcomes in adult mice. High (30%) protein groups showed differential response to high fat diets (40 and 55% fat) and these groups were selected for RNA-seq experiments. RNA-seq was performed on RNA extracted from the jejunum, liver and hypothalami of mice exposed to the different diets. Methods: C57BL/6J male mice used in this work. All mice were fed a 20% kcal casein, 10% kcal fat and 70% kcal carbohydrate diet (Research Diets; USA; D12450Bi), until they were aged 20 weeks. Following 15 weeks of intake of the latter diet (at age 20 weeks), all animals were weight matched and randomly allocated to different groups and switched to experimental diets for 12 weeks. After 12 weeks of dietary interventions all animals were sacrificed and dissected. Methods: The dietary treatments consisted of 3 different levels of dietary fat (20, 40 and 55%) combined with 3 different levels of protein (10, 20 and 30%), each replicated with the protein coming from either casein (CAS) (diet codes: D17052701- D17052709) or whey protein isolate (WPI) (diet codes: D17052710- D17052718). The levels of cellulose and sucrose were fixed at 5% by weight and energy respectively. The carbohydrate (starch) amount was reduced in parallel with increase of dietary protein and fat. All these diets can be ordered direct from research diets (www.researchdiets.com) using the diet codes provided. The 18 different treatments with 12 mice per group, where mice were housed in cages of 3 animals. Methods: The hypothalamus, liver, and jejunum of all animals from 30% protein groups (6 diets [D17052703, D17052706, D17052709, D17052712, D17052715, D17052718], 71 animals) were collected. Total RNA was extracted from hypothalamic blocks using Tri-reagent (Sigma) and from liver and jejunum using the RNeasy Mini Kit (QIAGEN, 74104) according to manufacture instructions. Hypothalamic RNA was paired-end (PE) sequenced by PE100 on BGISEQ-500 platform at Beijing Genomic Institute (PE 2 x 75 bp, 150 bp per fragment, 20M read-pairs per sample). Sequencing of liver and jejunal RNA was performed using the NovaSeq 6000 platform (chemistry V4.0 Illumina) at Macrogen Inc Seoul, South Korea (PE 2 x 75 bp, 150 bp per fragment, 25M read-pairs per sample). Raw sequence reads were obtained in FASTQ format and these sequence reads were quality assessed using the software FastQC (v 0.111.58; http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Sequences from all samples were quality trimmed, and cleaned of adaptor sequences using BBDuk java package to trim Illumina adapter sequences. On average 0.1% of the bases were trimmed per sample. Trimmed reads were aligned to the Mus musculus reference genome assembly GRCm38 (ftp://ftp.ensembl.org/pub/release-96/fasta/mus_musculus/dna/) using STAR RNA-seq aligner v2.5.2, and uniquely mapped read counts per Ensembl annotated gene/ transcript were estimated using the STAR -- quantMode option. Genes with zero read counts across all samples as well as non-protein coding genes were removed prior to subsequent analysis. Differential gene expression analysis and data transformations and visualization were carried out using DeSeq2 v1.18.1 in R 3.5.2. Sample clustering was carried out on variance stabilizing transformed data and visualised using PCA. Differentially expressed genes lists were generated using a negative binomial generalized linear model and pairwise comparisons using each combination of fat and protein content in each tissue. P values were adjusted for multiple comparisons using a Benjamini and Hochberg (B-H) method. DE genes with an adjusted P value < 0.05 were used for further DE gene data exploration and pathway analysis. The fold changes and p values were loaded into the Ingenuity Pathway Analysis (IPA) program (Ingenuity Systems; http://www.ingenuity.com/) for core analysis. IPA Knowledge Database was used to analyse the dietary impacts on canonical pathways and upstream regulators available in IPA. Results: Differential gene expression and IPA results revealed that jejunal immunity of the high (30%) CAS fed groups at 40 and 55% fat was oriented toward pro-inflammatory responses, whereas the corresponding WPI fed groups showed a reduction of inflammation. The liver transcriptome reveals genes involved in FA/lipid uptake and synthesis to be upregulated and genes involved in FA/lipid degradation were downregulated following a proportional increase of dietary fat (20 to 55%) in 30% CAS fed animals and these effects were not seen in WPI fed animals. In the hypothalamus, the intake of 40 and 55% fat significantly increased the number of differentially expressed genes compared to 20% fat. The effect of protein quality on hypothalamic gene expression (CAS vs WPI) was only seen at the highest fat content (55%), where several changes in signalling pathways were seen in WPI relative to CAS. These include a decreased reactive oxygen species production, reduction interleukin and chemokine inflammatory pathways and increased oxidative phosphorylation compared with mice fed CAS and 55% fat. Conclusions: Analysis of jejunal, liver and hypothalamic transcriptomic revealed that the intake of high WPI impeded the negative effects of high fat feeding (increased lipogenesis in the liver, inflammation, activation of disease-related pathways etc.), whereas the intake of high CAS exacerbated these effects. The data highlight the importance of selecting proteins rich in essential amino acids, for improving the metabolic outcomes of high fat diet.
