The role of prolactin and growth hormone in the regulation of casein gene expression and mammary cell survival: relationships to milk synthesis and secretion.

We have compared involution of the rat mammary gland, induced by litter removal, where milk accumulation occurs, with involution induced in the presence of the suckling young by combined PRL and GH deficiency. Both treatments induced involutionary processes involving apoptosis, as judged by DNA ladders and resulted in significant decreases in the DNA content of the gland. Surprisingly, the effects of hormone deprivation on protein output in milk were principally explained by the loss of secretory cells, as there were only modest decreases in casein messenger RNA (mRNA) expression and protein synthesis rates per U DNA in vitro. The association of casein mRNA with the polysome fraction was also unaffected by hormone deprivation, whereas involution induced by litter removal resulted in much greater decreases in steady state levels of casein mRNA and an increased association of the mRNAs with the monosome fraction. In PRL- and GH-deficient rats, PRL treatment could prevent all of these effects, GH was partially effective, whereas putative mediators of GH action, insulin-like growth factor I (IGF-I), IGF-II, and IGF-binding protein-3, were ineffective. This lack of effect of IGFs may be due to an inhibitory IGFBP, which we demonstrate to be present in increased amounts in the involuting mammary gland.

Interactions between normal mammary epithelial cells and mammary tumour cells in a model system.

Normal mammary epithelial (NME) cells and MCF-7 cells aggregate and grow as spheroids when cultured on extracellular matrix derived from Engelbreth/ Holmes/Swarth (EHS) tumour. NME cells stop dividing and differentiate but MCF-7 cells continue to proliferate, although growth is counterbalanced by cell death. In mixed cultures of NME cells and MCF-7 cells, the two cell types form mixed aggregates but then segregate to form well separated domains, often joined by only a narrow neck of cells. In these mixed cultures the growth of MCF-7 cells is inhibited by a factor secreted by NME cells into the medium.

Hormonal control of IGF-binding protein (IGFBP)-5 and IGFBP-2 secretion during differentiation of the HC11 mouse mammary epithelial cell line.

The mouse mammary epithelial cell line HC11 upregulates the synthesis of beta-casein (a differentiation marker) following treatment with the lactogenic hormone mix dexamethasone, insulin and prolactin (DIP). We demonstrate that the basal levels of IGF-binding protein (IGFBP)-5 secreted by undifferentiated HC11 cells are upregulated 10-fold during DIP-induced cellular differentiation whereas the level of the other IGFBP species secreted by HC11 cells (IGFBP-2) is downregulated during this process. As previously reported, the combination of all three of these hormones is required for synthesis of the differentiation marker beta-casein, whereas basal IGFBP-5 secretion is evident in the absence of any hormonal treatment and, unlike beta-casein, secretion of this protein can be stimulated by binary combinations of the hormones (although maximal levels of IGFBP-5 are achieved in the presence of all three lactogenic hormones). Additionally, levels of IGFBP-5 can be increased by DIP treatment under conditions (non-competency of HC11 cultures or presence of epidermal growth factor) where DIP treatment does not increase synthesis of beta-casein. For IGFBP-2, dexamethasone is a potent inhibitor of secretion whilst prolactin stimulated the secretion of this binding protein into the medium. For the IGFBP axis in HC11 cells we conclude that, although the levels of IGFBP-5 and -2 are influenced by the state of cellular differentiation, the hormonal regulation of the levels of these IGFBP species can be dissociated from the regulation of beta-casein synthesis. In a further series of experiments we demonstrate that IGF-I is able to replace insulin in the DIP lactogenic hormone mix and by the use of a specific IGF-I receptor blocking antibody indicate that the action of IGF-I is mediated through the cell surface IGF-I receptor and not by cross-reaction of IGF-I ligand at the insulin receptor. We discuss our data in the context of the potential role of the IGF axis in the process of cell differentiation and illustrate the significance of our findings in the context of the physiology and life cycle of the mammary epithelial cell.

Programmed cell death during mammary tissue involution induced by weaning, litter removal, and milk stasis.

Programmed cell death in mammary tissue was studied during natural weaning in lactating mice and after litter removal or milk stasis. All treatments stimulated mammary apoptosis, indicating that this process is an integral part of the tissue's involution after lactation. Induction of apoptosis was slower in natural weaning than after litter removal but occurred earlier when mice were concurrently pregnant during natural weaning. Ipsilateral induction of apoptosis by milk stasis in teat-sealed glands indicates that cell death is under local (i.e., intramammary) as well as endocrine regulation. Apoptosis detected by DNA laddering was associated with changes in expression of p53 and bax, two genes implicated in the regulation of cell death, and was accompanied by structural degeneration characteristic of mammary involution. Reciprocal changes in stromelysin mRNA, and that of its inhibitor TIMP-2, suggested that this structural reorganisation was the result of coordinated changes in gene expression favouring proteolysis of the extracellular matrix.

Apoptosis in lactating and involuting mouse mammary tissue demonstrated by nick-end DNA labelling.

Mammary involution after cessation of milk removal is associated with extensive loss of secretory epithelial cells. Ultrastructural changes and the appearance of oligonucleosomal DNA laddering in ethidium bromide-stained gels indicates that cell loss during involution occurs by apoptosis. In this study, a technique for nick end-labelling of genomic DNA with radiolabelled deoxynucleotide has been used to monitor the induction of programmed cell death in mice after litter removal at peak lactation. This technique proved more sensitive than conventional ethidium bromide staining, and results suggested that apoptosis was induced rapidly by milk stasis, before extensive tissue re-modelling had begun. Oligonucleosomal DNA laddering on agarose gels was detected within 24 h of milk stasis, and increased progressively for at least 4 days. Nick-end labelling also detected laddering before litter removal, suggesting that programmed cell death is a normal feature of the lactating tissue. The DNA end-labelling technique was also adapted for in situ visualisation of apoptotic cells in tissue sections. By this criterion, apoptotic cells were identified in both the secretory epithelium lining the alveoli of the gland and, increasingly with prolonged milk stasis, amongst those sloughed into the alveolar lumen. The results demonstrate the utility of these techniques for study of mammary cell death and suggest that, whilst apoptosis is rapidly induced by milk stasis, it is also a normal physiological event in the lactating mammary gland.

Hypothalamic PGE2 and cAMP production and adrenocortical activation following intraperitoneal endotoxin injection: in vivo microdialysis studies in Lewis and Fischer rats.

Inflammatory disease-susceptible Lewis (LEW) rats exhibit reduced glucocorticoid release in response to inflammatory and neurotransmitter stimuli, compared to histocompatible Fischer (F/344) rats. This compromised hypothalamo-pituitary-adrenal (HPA) axis activity has been ascribed to a primary defect in hypothalamic corticotrophin-releasing factor-41 (CRF) secretion, possibly caused by abnormal signal transduction in the CRF neuron. In the present study, we have used in vivo microdialysis to asses the role of hypothalamic prostaglandin E2 (PGE2) and cyclic adenosine monophosphate (cAMP) in endotoxin-mediated HPA axis activation in adult hyporesponsive LEW and hyperresponsive F/344 rats. Basal plasma corticosterone concentration was significantly higher in F/344 relative to LEW rats; however, the basal levels of PGE2 and cAMP, recovered from microdialysis probes in the anterior hypothalamus, were significantly greater in the LEW rat. Lipopolysaccharide (LPS) (200 micrograms/kg) caused a time-dependent increase in corticosterone secretion, the magnitude of which was markedly greater in the F/344 rat. Both LEW and F/344 rats displayed a similar PGE2 profile in response to LPS, although in absolute terms the response was more pronounced in LEW rats. LPS caused a dose-related increase in cAMP production in the LEW rat and comparison with F/344 animals, following the 200 micrograms/kg dose of LPS, revealed a larger and more prolonged cAMP response in the LEW strain. Simultaneous administration of indomethacin (50 mg/kg) with LPS (200 micrograms/kg) in the LEW rat completely blocked the PGE2 and cAMP responses to the toxin and whilst the corticosterone response to LPS was significantly attenuated at 140 min, no difference was apparent by 240 min. Hence, PGE2 and cAMP participate in the hypothalamic response to endotoxin-mediated adrenocortical activation in both LEW and F/344 adult rats but the steroid and second messenger profiles are strain-specific. The cAMP response to LPS appears to depend on products of arachidonic acid metabolism, such as PGE2, and hence basal and stimulated production of these mediators may be effected by the steroid milieu.

Does an IGF-binding protein (IGFBP) present in involuting rat mammary gland regulate apoptosis?

We examined the effects of GH and prolactin deficiency upon milk production, apoptosis and IGFBP production by the mammary gland. GH deficiency produced a 15% reduction in milk yield, prolactin a 50% reduction and combined prolactin- and GH-deficiency an 85% reduction in milk production. Litter removal led to complete inhibition of milk synthesis within 24 h owing in large part to milk accumulation. GH- and prolactin-deficiency also led to significant loss of mammary cells within 48 h and this was owing at least in part to apoptosis as judged by the appearance of characteristic DNA ladders. Prolactin replacement therapy could prevent all of these changes whilst GH was partially effective. The effects of GH are believed to be mediated via IGF-I, however, we were unable to mimic the effects of GH with IGF-I, IGF-II or a combination of IGF-I, IGF-II and IGFBP-3. We hypothesized that the cell-survival effects of exogenous IGFs might be blocked by an inhibitory IGFBP produced by the gland. Indeed the involuting mammary gland produces large concentrations of an IGFBP which Northern blotting identified as IGFBP-5. There was also a small increase in IGFBP-4 mRNA expression. Both appear to be produced by the secretory epithelial cells as judged by in situ hybridization. Preliminary studies using mouse "mammosphere" cultures suggest that they will be useful for investigating a potential causal relationship between IGFBP-5 synthesis and apoptosis.

Influence of microenvironment on mammary epithelial cell survival in primary culture.

Mammary epithelial cells cultured on Engelbreth-Holm-Swarm (EHS) matrix form multicellular structures termed mammospheres, in which cells and matrix become arranged around a central luminal space. In the presence of lactogenic hormones, cells within mammospheres become polarized, form tight intercellular junctions, and secrete milk proteins vectorially into the luminal space. This study examined the mechanism of lumen formation. Histological examination of developing mammospheres showed that cavitation was associated spatially and temporally with the appearance of fragmented nuclear material in apoptotic bodies, and with the presence of cells positively labeled by terminal deoxynucleotide transferase-mediated deoxyuridine nick end-labeling (TUNEL). Analysis of [(32)P]-deoxynucleotide end-labeled genomic DNA by electrophoresis and autoradiography showed DNA laddering indicative of apoptosis. A transient increase in laddering coincided with both lumen formation and the presence of TUNEL-positive cells. Lumen formation, DNA laddering, and detection of TUNEL-positive cells were all accelerated when matrix composition was altered. They were also impaired coordinately when caspase inhibitor was present during the first two days of culture. Therefore, lumen formation in mammosphere cultures is due to selective apoptosis of centrally located cells. Mammosphere cavitation was accompanied by redistribution of matrix constituents to the mammosphere periphery. Western blotting and Western ligand blotting of culture medium showed that lumen formation was also associated with a transient increase in insulin-like growth factor binding protein-5 (IGFBP5), a factor implicated in mammary apoptosis in vivo. We propose that epithelial cell survival during mammosphere development is induced selectively through stabilization by basement membrane constituents, which may act directly on the epithelial cell or confer protection against autocrine apoptotic factors.