Background The molecular events underlying mammary development during pregnancy, lactation, and involution are incompletely understood. degradation and cell-environment communication; (6) the involution switch is primarily transcriptionally 56124-62-0 IC50 mediated; and 56124-62-0 IC50 (7) during early involution, the transcriptional state is partially reverted to the pre-lactation state. A new hypothesis for secretory diminution is suggested C milk production gradually declines because the secretory machinery is not transcriptionally replenished. A comprehensive network of protein interactions during lactation is assembled and new regulatory gene targets are identified. Less than one fifth of the transcriptionally regulated nodes in this lactation network have been previously explored in the context of lactation. Implications for future research in mammary and cancer biology are discussed. Background Lactation is one of the most remarkable products of evolution. The signature feature and basis of the competitive emergence 56124-62-0 IC50 of mammals, including humans, is the production of complete early nourishment of neonates by the mother. The processes of lactation include the development of mammary tissue, as well as the synthesis and secretion of milk. At weaning, the mammary gland morphologically returns to a near pre-pregnant state. Thus, in addition to the important nutritional implications, lactation provides a model for basic biological processes such as the proliferation, differentiation, survival and death of cells. Although lactation is believed to be a product of Darwinian selective pressure, little is known of its molecular origins or its regulation. Current knowledge of the molecular regulation of mammary development and lactation has largely been derived from dissection of signaling networks in cell culture systems and phenotypic characterization of genetically altered mice. Some proteins modulated during pregnancy and lactation have been identified and characterized in the context of hormonal and metabolic pathways (reviewed in [1,2]). Beyond these signaling pathways, the regulation of mammary gland development and lactation is incompletely understood. Of particular interest are the major molecular events that govern macroscopic and histological changes in the mammary gland during secretory differentiation, secretory activation (the lactation switch), and the onset of involution (the involution switch). Unbiased genome-wide approaches are likely to identify novel genes and gene products involved in the regulation of lactation, particularly when incorporated into a larger picture of mammary development and function. In this study, bioinformatic techniques are applied to transcriptomic and proteomic data to enhance understanding of how the mammary gland is regulated through pregnancy, hEDTP lactation, and involution. Using non-hypothesis-driven analyses, transcriptional and post-transcriptional trends are described and putative key regulatory targets are identified. Gene products and their interactions unexplored in the current literature are visualized as a network, providing a framework on which to base future research. Such exploratory methods can be applied to other areas of biological inquiry to establish a quantitative representation of current knowledge and to facilitate the generation of new hypotheses. Results Global transcriptional trends during mammary development Using microarray data from the Neville study [3] (see Methods), a statistical analysis of genome-wide transcriptional changes in the mammary gland was applied to identify 4,832 genes differentially expressed (p < 0.001) of the 12,488 measured during a full mouse lactation cycle. To understand the major trends in gene transcription across developmental stages of the mammary gland from initial pregnancy to involution, a principal component analysis with mean centering and scaling was applied to these differentially expressed genes across all ten time points. (For descriptions of these time points, see Materials and methods. ) The top three principal components of the data in the time domain are diagrammed in Figures 1ACC. The first principal component describes 50.0% of the variance in the data. This major trend is a rise in gene expression during late pregnancy that remains high during lactation and falls during involution. A substantial set of genes C 592 C has a standard correlation of 0.90 or better with this first 56124-62-0 IC50 principal 56124-62-0 IC50 component (Additional data file 1). The second and third principal components appear to be minor trends, explaining 13.6 and 11.6% of the variance in the data. In the second principal component (Figure ?(Figure1B),1B), expression is unchanged during pregnancy and lactation, but rises during involution. In the third principal.