Characterisation of microRNA expression in post-natal mouse mammary gland development

Abstract

Background: The differential expression pattern of microRNAs (miRNAs) during mammary gland development might provide insights into their role in regulating the homeostasis of the mammary epithelium. Our aim was to analyse these regulatory functions by deriving a comprehensive tissue-specific combined miRNA and mRNA expression profile of post-natal mouse mammary gland development.We measured the expression of 318 individual murine miRNAs by bead-based flow-cytometric profiling of whole mouse mammary glands throughout a 16-point developmental time course, including juvenile, puberty, mature virgin, gestation, lactation, and involution stages. In parallel whole-genome mRNA expression data were obtained.

Results: One third (n = 102) of all murine miRNAs analysed were detected during mammary gland development. MicroRNAs were represented in seven temporally co-expressed clusters, which were enriched for both miRNAs belonging to the same family and breast cancer-associated miRNAs. Global miRNA and mRNA expression was significantly reduced during lactation and the early stages of involution after weaning. For most detected miRNA families we did not observe systematic changes in the expression of predicted targets. For miRNA families whose targets did show changes, we observed inverse patterns of miRNA and target expression. The data sets are made publicly available and the combined expression profiles represent an important community resource for mammary gland biology research.

Conclusion: MicroRNAs were expressed in likely co-regulated clusters during mammary gland development. Breast cancer-associated miRNAs were significantly enriched in these clusters. The mechanism and functional consequences of this miRNA co-regulation provide new avenues for research into mammary gland biology and generate candidates for functional validation.

Figure 1

Normal development of the mouse mammary gland. (a) Schematic of distinct stages of development and characteristic biological processes. Shown are time points used for mRNA expression profiling (time point 12 hours involution was used for miRNA expression profiling only, time points 8 weeks and 10 days lactation were used for mRNA expression profiling only). (b-c) Mean relative log₂ mRNA expression for non-redundant Gene Ontology biological processes (b) and KEGG pathways (c) (shown are gene sets with at least 100 genes represented on the array, numbers of genes are indicated in brackets). Red and blue indicate high and low relative log₂ expression, respectively. (d) Representative H&E stained histological sections of mouse mammary glands obtained at distinct time points during development and used for expression analysis.

Figure 2

Global changes in miRNA and mRNA expression. (a) Hierarchical clustering of time points based on the relative log₂ expression of 102 detected miRNAs. (b) miRNAs and mRNAs exhibit reduced overall expression during lactation and early involution time points. Black dots correspond to independent biological replicates. Median values for each time point are connected by a solid black line. (Top) miRNA expression. Normalised log₂ median fluorescence intensity (MFI) values (see Methods) were averaged for each sample by taking the arithmetic mean. (Middle) mRNA expression. The background corrected Illumina probe level data were transformed by taking logs (base 2) but no between sample normalisation was performed. Non-normalised log₂ intensities were averaged for each sample by taking the arithmetic mean. (Bottom) Mass of total RNA extracted from whole mammary glands, and equally sized partial glands for lactation and involution time points.

Figure 3

miRNA expression during mammopoiesis is highly correlated. Seven clusters obtained by model-based clustering are displayed in separate panels. (a) Grey lines represent the standardized log₂ expression profiles of individual miRNAs. Black lines correspond to mean cluster profiles. (b) Heatmaps of individual clustered miRNAs. Red and blue indicate high and low standardized log₂ expression, respectively. Within each cluster panel, miRNAs are ordered according to miRNA families with identical seed sequence (position 2-8) (indicated by brackets). miRNAs associated with basal or luminal breast cancer are highlighted in red and blue, respectively. Individual panels correspond to the seven clusters in (a).

Figure 4

In-situ hybridisation for miRNAs let-7b (cluster 1), miR-205 (cluster 3), and miR-206 (cluster 6) shows higher expression in the epithelial compared to the stromal cell compartment. let-7b and miR-205 show specificity for the luminal and basal epithelial cell layers, respectively. White scale bars in images for time point 25d indicate 100 microns. Higher magnification images are provided in Additional file 6.

Figure 5

Predicted targets for the miR-29 family show systematic changes in their relative expression. (Top) Relative log₂ expression profiles of individual miR-29 family members are shown in distinct colours. (Bottom) Negative Log₁₀ transformed two-sided P-values are shown for each time point. The positive and negative y-axes indicate increased and reduced mean relative expression levels compared to control genes, respectively. Dashed lines indicate the threshold for Benjamini-Hochberg corrected P-values smaller than 0.05.