Profiling of Human milk cell and lipid microRNAs from healthy and during naturally occurring infection breastfeeding mothers

Human milk is highly recommended for infant during the first six month of life by World Healthy Organisation (WHO). Human milk is not only rich in macro-nutritional components, but also rich in cells and molecules. MicroRNAs are small non-coding RNAs, which enriched in human milk. These molecules are vital in enormous biological and cellular functions including immune system and in response to infections. By using deep sequencing method, 770,374,554 raw reads were generated from all samples (n=26). Then, filter analysis was done to remove 81,091,772 (10.5%), and 689,282,782 clean reads (89.5%) were considered as clean reads, which was retained for the subsequent bioinformatics analysis. Annotation and matching reads to miRBase revealed1780 mature known microRNAs identified in human milk cells and lipids derived from healthy, cold and other infection types mothers. In particular, 1680 known microRNAs were determined in infected mothers (n=14), while 1,606 known microRNAs in healthy mothers (n=12). Of these known microRNAs, 453 microRNAs were differentially expressed (p<0.05) between healthy and infected samples. The majority of the highly expressed miRNAs in all samples, in particular top 20 microRNAs, were also differentially expressed between healthy and infections. Further, 592 novel mature microRNA sequences were predicted, with only 65,878 total reads. Amongst the total reads of the novel microRNAs, top 20 novel miRNAs were found to contributed in 73.3% (total reads 48,295) of the total reads (65,878). Overall design: Human milk samples were collected from 11 mothers in healthy (n=6) and in infection (n7) conditions. Human milk samples were fractionated to obtain cell and lipid fractions. Total RNA and miRNA were extracted and quantified from cell and lipid fractions. Next generation sequencing using illumina HiSeq 2000 was used to profile microRNAs in 26 samples, where 13 cells and 13 lipids. Filtered (clean) reads were aligned to the latest version of miRBase 21.0. Also, mirdeep software was used to predict novel miRNAs. DESeq and Linear mixed effect models and ANOVA were used to compare miRNA expression levels between samples and conditions