Comparison of five different cellular fractions to examine the cow mammay gland transcriptome using RNA-sequencing

The objective of this study was to examine the cow mammary gland transcriptome using RNA-Sequencing by comparing five different cellular fractions: mammary gland tissue (MGT), milk somatic cells (SC), laser microdissected mammary epithelial cells (LCMEC), milk fat globules (MFG) and epithelial cells from milk (MEC). These samples were collected from one Normande and two Jersey cows. An average of 32-million short-sequence reads was obtained for each sample, and were mapped to the annotated UMD3.1 bovine genome assembly. In all samples 60-70% of the reads were categorized as mapped reads, with the exception of one MFG and one MEC sample that mapped 44.6% and 32.7% reads, respectively. Gene expression in SC and LCMEC showed the highest correlation with MGT, ranging from 0.99 to 0.92. In general, all the highly expressed genes such as caseins (CSN1S1, CSN1S2, CSN2, CSN3) and lactoglobulins (BLG, LALBA) showed the same expression profile in MGT, SC and LCMEC. However, MFG and MEC showed a different profile, where PLIN2, TBA1B, HMOX1, S100A11 and RL10 genes were highly expressed. These genes represent markers of lipid accumulation and immunity. MFG membrane contains specific components such as phospholipids, proteins and glycophospholipids that are known to have an affinity for the bacterial cell surface. Analyzing all the MFG reads not mapping to the bovine genome (17-million), 500 different kinds of bacteria were identified. Milk SC were the most representative sample of the mammary gland transcriptome and can be used as an effective and easy way to study gene expression in the mammary gland.