Our data show that miRNAs are highly co-regulated during mammary gland development. We found seven temporal clusters with complex expression patterns which did not coincide with single developmental stages. Breast cancer-associated miRNAs were found predominantly in two of these clusters.
MicroRNAs previously identified to be highly expressed in the luminal as compared to the basal molecular subtype of breast cancer [20
] were enriched in cluster 1. These miRNAs showed high levels during puberty and gestation where proliferation and invasion are the predominating biological processes. We hypothesize that these luminal breast cancer miRNAs may be involved in the control of proliferation and invasion during normal development and become deregulated in breast cancer. MicroRNAs associated with the basal breast cancer subtype, including four members of the miR-17~92 family, miR-25 and miR-150, showed a distinctly different pattern, being equally distributed among clusters 1 and 2.
Of particular interest was cluster 4, which in contrast to the global decrease in miRNA expression during lactation and involution showed a specific increase in expression during these stages, paralleled by a decrease in the expression of predicted targets for one of the cluster members, miR-429 (Additional file 7
). The miRNAs included in this cluster have previously been associated with various biological functions and pathways. The cluster includes miR-200 family members miR-200a, miR-141 and miR-429. There is increasing evidence that this miRNA family plays a crucial role in the regulation of epithelial to mesenchymal transition (EMT). All five members of the miR-200 family were markedly down-regulated in cells that had undergone EMT in response to transforming growth factor (TGF)-beta [30
]. Enforced expression of the miR-200 family was sufficient to prevent TGF-beta induced EMT, while inhibition of the miR-200 family was sufficient to induce EMT in cells [30
]. During mammary gland development, a process similar to EMT is known to occur between lactation and involution. The loss of inter-epithelial cell-cell-contacts, including loss of the cell-adhesion molecule E-cadherin, takes place during the early reversible phase of involution [31
]. It is therefore tempting to speculate that the miR-200 family may act to prevent premature loss of E-cadherin and induction of EMT during lactation.
Another miRNA highly expressed during lactation and early involution, miR-146b, has been associated with the regulation of innate immune response and inflammation. miR-146a expression has been reported to be substantially increased in murine T helper (Th) 1 cells compared to Th2 cells and naïve T cells [32
]. The Th1 and Th2 cytokine milieu has recently been implicated in mouse mammary gland development. Mammary epithelial cells were demonstrated to undergo a switch from Th1 to Th2 cytokine production upon the induction of luminal differentiation [33
]. It is an intriguing possibility that miR-146 might be involved in abrogating the Th2 bias in cytokine expression that continues during gestation and lactation, to facilitate a switch back to a Th1 environment upon involution.
Of note, we did not detect obvious changes in the miRNA expression pattern between the stages of lactation and early involution in our data, although there are known morphological and transcriptional changes [34
]. This contrasts with what we observed at the level of mRNA expression, where we could see clear differences between the two stages (Figure ). Subtle differences in miRNA expression between lactation and early involution may have potentially been masked by the global decrease in miRNA expression we observed during these stages.
We identified a significant global down-regulation of miRNAs and mRNAs during the stages of lactation and early involution. The total RNA content per mammary gland was increased during these stages. Since our analysis of miRNA expression was performed in equal amounts of total RNA, this finding suggests that miRNAs and mRNAs represented a smaller fraction of total RNA during lactation and early involution. We hypothesize that the up-regulation of the transcriptional and translational machinery during milk production (with corresponding increases in tRNAs/rRNAs) may lead to a relative rather than absolute depletion of miRNAs during these developmental stages.
During mouse mammary gland development we observed that many let-7
family members showed a peak in expression during puberty, the mature virgin stage, and early gestation, followed by a marked decrease and low levels during lactation and involution. This observation is consistent with recent reports that let-7 expression is depleted in mouse mammary epithelial progenitors [38
] and in breast tumour-initiating cells [39
], and that enforced let-7 expression could inhibit the self-renewal capacities of cells [38
]. The developmental stages of the juvenile gland, puberty, and early gestation, where we detected increasing levels of let-7, are characterised by a marked expansion of the luminal and alveolar cell compartment which presumably contains a high frequency of relatively differentiated cells. The subsequent decrease in let-7 expression during lactation may allow for a relative expansion of the progenitor compartment which is necessary for the reconstitution of the alveolar compartment during the next pregnancy cycle. On the other hand, miR-22 and miR-205, which were reported to be highly expressed in mammary progenitor cells [38
], seemed to be enriched during gestation and again during late involution. These data suggest that miRNAs may contribute to the balance between progenitor cells and their differentiated progeny during mammary gland development.
When interpreting changes in miRNA expression during mammary gland development, changes in the cellular composition of the mammary gland have to be taken into consideration. We observed that the proportion of stromal cells gradually declined until puberty. Epithelial and in particular luminal cells showed a peak at puberty followed by a decrease, which was reflected in gene expression patterns, such as those of luminal and basal cytokeratins. Nevertheless, the variety of biological processes and pathways and their observed temporal expression pattern (Figure ) cannot be solely explained by changes in the cellular composition. Although stromal cells contribute significant numbers of cells to the mammary gland, our mRNA expression data show that the observed differential gene expression mostly reflected mammary epithelium-driven processes, and therefore we infer that changes in miRNA expression are also predominantly epithelial-driven. We were not able to demonstrate by microarray analysis a clear correlation between the expression of a particular miRNA and a predominating cell type. It is an inherent limitation of this study that the impact of cell type heterogeneity on relative expression patterns obtained from whole mammary glands cannot be conclusively assessed. Future analysis of isolated cell populations may help to address this question, and newer technologies for miRNA profiling which have become available in the meantime allow the analysis of smaller numbers of cells. Nevertheless, our analysis of three candidate miRNAs by in-situ hybridisation gives some indication of cell type specific expression in mammary epithelial cells. MicroRNA expression was generally much higher in epithelial cells compared to the stroma, suggesting that the expression patterns found by microarray mostly reflect temporal changes in the epithelial cell compartment. Within the epithelial compartment, cell type-specific expression was demonstrated for let-7 and miR-205, which were predominantly expressed in luminal and basal cells, respectively, confirming previous reports of their expression in human luminal and basal breast epithelium [40
It was shown previously that miRNAs and their targets often exhibit mutually exclusive expression both in the spatial and temporal domain [41
], and that conserved targets and miRNAs are co-expressed but with reduced relative target expression in the presence of the miRNA [7
]. We failed to detect widespread systematic changes in the relative expression of putative targets during mammary gland development. However, we did observe trends for individual miRNA families (Additional file 7
), most notably the miR-29 family, for which miRNAs and targets showed anti-correlated expression (Figure ). These observations are consistent with miRNAs affecting mRNA expression levels [7
]. The down-regulation of miR-29 targets, enriched for focal adhesion genes, would fit with the extensive remodelling and changes during late involution. Future analysis of isolated cell populations will further contribute to investigations of the relationship between miRNAs and their targets.