Aerobic exercise effects on tumor and skeletal muscle transcriptome in a DMBA-induced carcinogenesis

Here we sought to investigate changes caused by aerobic exercise in the skeletal muscle and tumor during the progression of a DMBA-induced carcinogenesis and cachexia. Also, we investigate the cross-talk between the tissues. We used Sprague-Dawley rats starting at 50 days of age. Cancer was induced using a single dose of DMBA carcinogen (65 mg/kg), and the induction time took 3 months. After induction, we performed aerobic exercise training (treadmill running) 5 times a week for 8 weeks. Tibialis anterior muscle (n=4 per group) and tumor tissue (n=4 per group) were collected. Snap-frozen tibialis anterior and tumor tissue were homogenized in liquid nitrogen via mortar and pestle. RNA was isolated from ∼50 mg of tissue using the miRNeasy Mini Kit (Qiagen, cat: 217004). The concentration and quality of total RNA samples were assessed using an Agilent 2100 Bioanalyzer, and an RNA Integrity Number (RIN) of 8 or higher was required to pass the quality control. Library preparation and sequencing was completed at the Massively Parallel Sequencing Shared Resource at Oregon Health and Science University using the Illumina NovoSeq6000. 150ng of total RNA was used for each sample. cDNA library preparation includes mRNA purification/enrichment, RNA fragmentation, cDNA synthesis, and ligation of index adaptors using the KAPA mRNA Hyperprep Kit prep (cat: KK8581). Each resulting indexed library was quantified, and its quality was assessed by NovaSeq 6000. 5μL of 2nM pooled libraries per lane was denatured, neutralized, and applied to the cBot for flow cell deposition and cluster amplification before loading to 100bp paired-end sequencing (Illumina, Inc.). Approximately 30-50 million reads per library were generated. The processed data were obtained using the software Partek Flow. The quality trimming level (Phred) was 35 (3' end). We aligned the reads to the UCSC Rattus norvegicus (rn7) reference genome using STAR (Spliced Transcripts Alignment to a Reference). To quantify gene expression, we used the annotation model Rn07. The gene counts were filtered using a noise filter reduction filter <=10. The normalization was made using Median ratio (DESeq2) and differential analysis was made using DESeq2.