Page 35 - Read Online

Qu et al. J Transl Genet Genom 2023;7:3-16 https://dx.doi.org/10.20517/jtgg.2022.16 Page 5

NJ). The LFD contained 10.5% kcal/g of fat, 73.1% kcal/g of carbohydrate, and 16.4% kcal/g of protein
(D12328; Research Diets; New Brunswick, NJ). Diets were otherwise matched for all micronutrients. At
30 days of age, male WT andApc Min/+ mice were assigned to either the HFD or LFD, resulting in four groups
according to genotype and diet: WT-LFD, WT HFD,Apc Min/+ -LFD,Apc Min/+ -HFD. Mice were fed experimental
diets ad libitum for three days (30-33 days of age). Biological replicate mice were used for each combination
of genotype and diet type. All animal procedures were evaluated and approved by the Institutional Animal
Care and Use Committee of CWRU School of Medicine, Protocol Number 2020-016.

Intestinal epithelia isolation
After three days on experimental diets, mice were euthanized by cervical dislocation. Small intestines were
immediately removed, flushed with ice-chilled phosphate-buffered saline (PBS), and cut longitudinally.
Mesenteric tissue was removed from the intestines. Each small intestine was then cut transversely into four
equally sized strips. The intestinal strips were gently agitated in EDTA-based Cell Dissociation Buffer
(13151014; Thermo Fisher Scientific; Waltham, MA) at 4 °C for 45 min, rinsed in cold PBS, and then
agitated again in fresh dissociation buffer at 4 °C for another 45 min. Tissue strips were cut into small
pieces, approximately 2 mm × 2 mm. Tissue pieces were rapidly pipetted up and down in cell dissociation
buffer with a 10 mL serological pipet for 5 min or until the solution became cloudy. Intestinal pieces were
removed from the solution using a 150 µm sterile paint straining mesh (2650; Mutual Dropcloth Company;
Monroe, NC). The suspension was centrifuged at 200× g for 5 min, and the supernatant was aspirated,
leaving a pellet of intestinal epithelia. The pellets were then suspended in 10.5 mL of cold PBS in
preparation for ChIP-Seq (10 mL) and RNA-Seq (500 µL). The samples were immediately flash frozen with
liquid nitrogen and stored at -80 °C for later use.

ChIP-Seq
[10]
ChIP-Seq was performed on each 10 mL intestinal epithelia sample as previously described . The Covaris
TM
truChIP Chromatin Shearing Kit (520154; Covaris; Woburn, MA) was used to cross-link the intestinal
epithelia and extract their cell nuclei according to the manufacturer’s protocol. Samples were sheared using
the Covaris model S2 AFA focused ultrasonicator with the following settings: duty cycle-5%, intensity-4,
cycles/burst-200, time-seven minutes). Chromatin immunoprecipitation using 9 µg of rabbit anti-H3K27ac
antibody (ab4729; Abcam; Cambridge, UK) followed by recovery of DNA and then preparation of ChIP-Seq
libraries were performed as previously described . ChIP-Seq libraries were sequenced on an Illumina
[11]
NextSeq 550 system (Illumina; San Diego, CA) at the CWRU Genomics Core.

The FASTX-Toolkit v0.013 was used to remove adapter sequences and trim read ends using a quality score
cutoff of 20 (Available from: http://hannonlab.cshl.edu/fastx_toolkit/). ChIP-seq data were aligned to the
mm9 genome assembly using Bowtie 2 v2.3.4.3, discarding reads with at least one mismatch and reporting
the best alignment if multiple alignments were present . Output SAM files were converted to binary
[12]
[13]
(BAM) format, sorted, indexed, and PCR duplicates were removed using SAMtools v1.10 . Peaks were
detected using MACS2 v2.1.2 with default settings and the broad flag set . Peak lists were filtered to
[14]
remove all peaks overlapping ENCODE blacklisted regions [15,16] . DeepTools v3.2.0 was used to generate
RPGC-normalized bigWig tracks with 50 bp bin sizes of the final sample BAM files . bigWig tracks were
[17]
visualized on the Integrative Genomics Viewer to assess samples for any track irregularities or low signal-to-
noise ratio .
[18]
Identification of variant enhancer loci
H3K27ac peaks called across samples were filtered for significance (Benjamini-Hochberg corrected
P ≤ 0.001). Overlapping peaks were merged, and the read depth for each peak region across samples was
determined using BEDTools v2.17.0, generating a count matrix . DESeq2 v1.34, a tool used to analyze
[19]