To assess AP-2 transcription factor gene expression, HC11 cells were grown under normal conditions (in the presence of EGF) or subjected to differentiation conditions by first culturing to confluency without EGF and then treatment with the lactogenic hormones dexamethasone, insulin and prolactin (DIP) for 72 hours. RNA was isolated from confluent cultures and subjected to RT-PCR analysis. As shown in Figure 1, only cells cultivated with DIP expressed the milk protein genes beta-casein and WAP, indicative of lactogenic differentiation (Figure 1, middle panel). HC11 cells expressed predominantly AP-2alpha, -beta, and -gamma, and at very low levels AP-2epsilon, whereas AP-2delta transcripts were not detectable. By RT-PCR we did not detect changes in expression of these AP-2 isoforms between non-differentiated and differentiated cells (Figure 1, upper panel). Therefore, AP-2 transcript levels do not appear to decrease during lactogenic differentiation of HC11 cells.
To verify this result, we tested the expression levels of AP-2alpha and AP-2gamma in a semi-quantitative manner using Northern blot analysis. We were particularly interested in these two AP-2 isoforms because both are downregulated during lactation in vivo, and overexpression of either isoform in the mammary glands of transgenic mice was interfering with lactation, suggesting a functional role in lactogenic differentiation [5, 6]. As depicted in Figure 2, neither AP-2alpha nor AP-2gamma transcript levels decreased during lactogenic differentiation of HC11 cells. Taking into acount a decrease in GAPDH expression (relative to the amount of RNA shown by the gel photo), rather a slightly increased expression level of both AP-2 transcription factors during differentiation was observed. These results suggest that the lactogenic differentiation of HC11 cells does not require a downregulation of AP-2 transcription factors.
Because these findings were suggesting that differentiated HC11 cells are not representative of mammary epithelium at lactation, we compared milk protein gene expression levels between differentiated HC11 cells and mammary glands taken from mid- or late-pregnant mice or from lactating mice. Using Northern blot analysis, we could readily detect beta-casein transcripts in DIP-treated HC11 cultures. Their amount, however, was much lower than even in mammary glands at day 12.5 of pregnancy (Figure 3, upper panel). Whereas WAP transcripts could be readily amplified by RT-PCR (Figure 1) they were not detectable by Northern analysis in DIP-treated HC11 cells (Figure 3, middle panel). The kinetics of WAP gene expression in vivo is known to be delayed compared to beta-casein expression [10], and accordingly we could detect only low expression levels at day 12.5 of pregnancy. In conjunction with the sustained AP-2 transcript levels, these results clearly demonstrate that lactogenic differentiation of HC11 cells does not fully reflect the in vivo situation. Lactogenic differentiation in vivo can be subdivided into distinct phases [2]: First, secretory differentiation of alveolar epithelial cells is taking place which is characterized by increased biosynthesis of milk proteins and of lipids that are organized in membrane bound milk fat globules residing in the cytoplasm. The transition to actual milk secretion has been termed "secretory activation" and is reflected by a strong induction of genes involved in lipid biosynthesis. Whereas differentiating HC11 cells display milk protein gene induction and also secretory activation [1], several glycoproteins of the milk fat globule membrane are not appropriately expressed [11]. Moreover, expression of beta-Casein and of Xanthine oxidoreductase, a component of the milk fat globule, were shown to be controlled by distinct mechanisms in HC11 cells [12]. It therefore remains possible that the downregulation of AP-2 transcription factor genes plays a role in milk fat globule formation or other aspects of lactogenic differentiation that are not reflected by the HC11 system.