RNAseq analysis of gene counts and expression levels in diabetic foot ulcers

In this study we examine the temporal changes in the transcriptome of chronic wounds to identify dynamic differences between healing and non-healing wounds. Wound tissue samples were collected over a 12-week period from human subjects with chronic, diabetic foot ulcers. Some healed during the observation period, and some remained unhealed. Bulk RNAseq analysis was performed to find differences in dynamic patterns.

Methods
Seventeen patients, 18+ years from Northern California VA Medical Center (Mather, CA) with a diabetic foot ulcer that had not improved > 50% for 4+ weeks, were enrolled in the study from 2021 to 2022. Wounds were monitored over 12 weeks to classify them as "healers" or "non-healers". Tissue from wound margins was collected during the standard of care and debridement, and preserved in RNALater (Life Tech). Tissue samples were homogenized, total RNA was extracted with the Quick-RNA Mag Bead kit (Zymo Research), and the integrity of the RNA was evaluated using an Agilent Tape Station (Agilent). For sequencing, indexed libraries were constructed using the SMARTer Stranded Total RNA-Seq Kit v3 (Takara Bio). Both the quantity and quality of these libraries were appraised by a Qubit fluorometer and an Agilent 2100 bioanalyzer. Molar concentrations of the libraries were confirmed by qPCR prior to pooling. The sequencing process was conducted on the Illumina NovaSeq 6000 platform, utilizing PE150 chemistry (Illumina).

The Cogent NGS analysis pipeline (CogentAP) from Takara Bio (v2.0) was used to de-multiplex and analyze each fastq file. The analysis process involved the "cogent analyze" wrapper function, which carried out the following tasks: (1) trimmed reads using cutadapt (version 3.2, doi:https://doi.org/10.14806/ej.17.1.200), (2) aligned genomes to the Homo sapiens genome GRCh38 using STAR (version 2.7.2a, doi: 10.1093/bioinformatics/bts635), and (3) performed read counting for exonic, genomic, and mitochondrial regions in Homo sapiens genes from ENSEMBL gene annotation version 103 (https://www.ensembl.org/Homo_sapiens/Info/Index) using featureCounts (version 2.0.1, doi: 10.1093/bioinformatics/btt656).   Normalization was performed to adjust for differences in sequencing depth between samples, and the size factors were estimated by computing the median ratio of counts for each gene across all samples.  The resulting dataset comprises 117 sample points, each representing 58,735 genes.