The SRC-family kinase LYN acts as a tumour suppressor gene in an in vivo BRCA1 loss-of-function-associated mammary tumourigenesis model

To examine the interaction between LYN kinase and BRCA1 loss-of-function in an in vivo mouse mammary tumour model, we have used conditional knockout Brca1 and Lyn alleles targeted to the mammary epithelium, on a p53 heterozygote background. Comparison of tumour cohorts demonstrated that homozygous conditional Lyn cohorts developed tumours significantly faster than the other genotypes. Transcriptomic analysis to identify and compare tumours with the highest and lowest Lyn levels from across the cohorts showed that 'Lyn high' tumours had a significantly slower doubling time and were enriched in inflammatory pathways and canonical NfkB signalling. 'Lyn low' tumours had a faster doubling time and were enriched in PI3K, WNT and NOTCH pathways. Therefore, these results suggest that, in this model at least, LYN acts as a tumour suppressor in the context of Brca1 loss. Overall design: RNAseq analysis of mammary tumours from mouse cohorts carrying Brca1 conditional (floxed) alleles on a p53 heterozygous background and either wild type, heterozygous or homozygous for the conditional (floxed) Lyn allele. All cohorts also carried the Blg-CRE transgene, expressed in the . Differentially Expressed Genes (DEGs) were determined in three ways. First, by comparison of BlgCre Brca1 floxed p53 heterozygous Lyn wild type tumours to BlgCre Brca1 floxed p53 heterozygous Lyn floxed tumours. Second, by determining the thirteen tumours most strongly expressing Lyn ('Lyn high' tumours) and the thirteen tumours with the weakest Lyn expression ('Lyn low' tumours) according to the normalised RNAseq data and comparing DEGs between them. Third, by carrying out an unbiased principal component analysis on all 39 tumours and allowing them to self-sort into (two) distinct groups (PCA group 1+2 and PCA group 3+4) and comparing DEGs between them.