| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology of the Cell I, German Cancer Research Center, D-69120 Heidelberg, Germany
Address all correspondence and requests for reprints to: Professor Dr. Günther Schütz, Molecular Biology of the Cell I, German Cancer Research Center, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. E-mail: g.schuetz{at}dkfz.de.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Glucocorticoids have long been known to act as lactogenic hormones in cell and tissue culture of mammary epithelia. Several studies have demonstrated a synergistic effect of glucocorticoids and prolactin on mammary epithelial cell differentiation and milk protein production (reviewed in Ref.1). From studies done in mammary epithelial cell culture, it has been suggested that this lactogenic activity might be due to a stimulatory effect of the GR on milk protein gene transcription. This action may involve a ligand-dependent interaction of the GR with the prolactin-activated transcription factor STAT5 (signal transducer and activator of transcription 5), leading to enhanced transcription of the ß-casein gene (2). The physiological significance of this cross-talk for mammary epithelial cell function, however, is not known.
Using gene-targeting techniques, researchers have been able to define the role of several signaling molecules in mammary gland development [reviewed by Hennighausen and Robinson (3)]. Mice lacking the GR in all tissues die at birth and, therefore, their use is limited in the investigation of mammary gland function (Refs.4 and 5 and see below). To investigate specific GR functions in mammary gland development, mice harboring a point mutation in the GR gene (GRdim) were analyzed (6). This mutation abolishes DNA-binding-dependent transcriptional regulation by GR but does not affect transcriptional regulation mediated by protein-protein interactions of GR with other transcription factors. These GRdim mice are viable and fertile (7). Mammary gland development in these mice is impaired due to a reduced ductal epithelial cell proliferation in virgin mice. Pregnancy-associated mammary gland development and lactation, however, remain normal in GRdim mice. These results suggested a new function for the GR in mammary gland development, i.e. the control of ductal epithelial cell proliferation in the virgin animal. Therefore, GR was assumed to exert its functions in lobuloalveolar differentiation and milk protein gene expression by GR-STAT5 interaction, which is not impaired in GRdim mice. Using the GRdim mouse model, however, not all functions of the GR molecule in mammary gland development can be dissected.
In a different experimental approach, embryonic mammary gland anlagen of GR-deficient mice (4) was transplanted into the cleared mammary fat pad of syngeneic hosts. Kingsley-Kallesen et al. (8) demonstrated aberrant ductal morphogenesis in transplants of GR-deficient animals, but no defect in cell proliferation, and normal lobuloalveolar development during pregnancy and lactation. Milk protein gene expression was normal in these mice, and the authors suggested that MR could compensate for GR in GR-deficient transplants. They demonstrated that, in explant cultures, the MR ligand fludrocortisone is able to stimulate ß-casein expression. Interpretation of these results may be complicated by two facts: First, although the mammary transplant method is an acknowledged experimental procedure by which to study ductal and lobuloalveolar development, it is not possible to observe a period of lactation and milk production under normal circumstances. Second, it could not be ruled out that the GR-deficient mice used as transplant donors in Ref.8 had residual GR activity, as they had been derived from a mouse line that carries a GR hypomorph allele (4, 9).
To define the biological function of the GR in the adult, intact mammary gland, we generated a mouse model with a mammary-epithelial specific ablation of the GR gene in the adult mouse using the Cre-loxP recombination system. Using a conditional allele for the GR (10) and a transgenic mouse line expressing the Cre recombinase in secretory epithelial cells of the lactating mammary gland (11), we were able to obtain mice in which the GR gene was deleted in the late stages of pregnancy, leading to loss of epithelial GR in mammary glands throughout lactation. In these mice, we analyzed lobuloalveolar morphogenesis, epithelial cell proliferation, and milk protein production.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). We found elevated levels of mammary-specific recombination from d 14.5 of pregnancy (P14.5). Recombination levels reached 60% at L0.5 and remained at 65% during lactation. In line with our previous observations using WAPiCre mice (11), this suggests almost complete recombination of the GR allele in the epithelial compartment, which accounts for about 70% of total cells in the lactating mouse mammary gland. The faint bands seen in the virgin glands correspond to non-tissue-specific background recombination in the WAPiCre mice that we reported earlier (11).
|
, A and B, shows that GR is present in ductal epithelial cells of virgin mammary glands of all genotypes, consistent with the absence of WAPiCre activity at this stage of mammary gland development. At P14.5, GR can be detected in the developing alveolar lobes of control mice (Fig. 2C), but is lacking the majority of epithelial cells in mutant animals (Fig. 2D). These results suggest that the loss of GR takes place after the initiation of lobuloalveolar development. At the onset of lactation, GR is present in all epithelial cell nuclei of control mammary glands (Fig. 2E), but absent from the vast majority of epithelial cells in the mutant glands (Fig. 2F). We could not detect any GR-positive epithelial cell nucleus in mutant glands at L2 and later (Fig. 2, panel H vs. panel G). This suggests that GR inactivation is complete in the epithelial cells of the lactating mammary glands of GRloxP/loxP; WAPiCre mice. In contrast, stromal cells retain GR protein during all stages of mammary gland development in mutant animals (see, for example, Fig. 2H).
|
). Next, we analyzed the histology of the GR mutant mammary glands. In GRloxP/loxP; WAPiCre-mice, secretory alveoli are formed and contain epithelial cells just as in control mice; however, less space is occupied by epithelial structures, and more space is occupied by fat compared with controls (Fig. 4, A–H). This difference is pronounced at the time of pup delivery (Fig. 4, panel B vs. panel A), but persists throughout lactation (d 10: see Fig. 4F and 4H vs. 4E and 4G).
|
|
, A and B). At P16.5, however, there is a clear difference between GR-deficient (Fig. 5D) and control (Fig. 5C) glands in that lobuloalveolar development in GR-deficient glands has not proceeded as far as in the control glands. Analysis of mammary glands from P18.5 (Fig. 5, E and F) and L0.5 (see Fig. 4
, A and B) shows that lobuloalveolar development is not completely blocked, but is significantly retarded in GR mutant mammary glands, enabling some alveolar structures to be observed.
|
). At P14.5, however, less proliferating cells are seen in GR-deficient glands than in control glands (Fig. 6, panels B and D vs. panels A and C). At this stage, we determined the mitotic labeling index (BrdU-positive cell nuclei/ total epithelial nuclei) in wild-type glands to be 8.4%, whereas in the mutants, it was 5.2%, indicating that epithelial cell proliferation is affected by GR loss.
|
, labeling index 33% vs. 6E, labeling index 2.1%). This proliferative burst seems to compensate, in the initial stages of lactation, for the poor alveolar structures in GR mutant glands before parturition. We did not observe, however, increased apoptosis in mutant glands at any time point during pregnancy. Apoptotic cells can be observed only at the time when the epithelial expansion into the fat pad ceases, i.e. at L0.5 in control animals and at L2 in mutant glands (data not shown).
Milk Protein Expression Is Normal in GR Mutant Mammary Glands
To investigate epithelial cell differentiation and function in mammary glands devoid of GR, expression of the two milk protein genes, ß-casein and whey acidic protein (WAP), were analyzed in mutant mammary glands. Northern blot analysis shows that mRNAs for both milk proteins are expressed in GR mutant mammary glands at levels comparable to controls (Fig. 7). We conclude that GR is not essential for the expression of milk protein genes in the lactating mammary epithelium in vivo. An estimation of milk fat percentage, by creamatocrit measurement, and a fatty acid profile did not give any evidence that fat mobilization into milk was affected by the GR loss (data not shown). After weaning, involution proceeds normally, and GRloxP/loxP; WAPiCre-mice were able to go through successive rounds of pregnancy and lactation, albeit with the same alveolar phenotype in later lactational periods as in the first one (data not shown). These results indicate that epithelial GR is not essential for mammary gland function once secretory alveoli have formed.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Most of the alveolar cells lack GR protein during the whole period of observation (i.e. from P14.5 onward); therefore, the GRloxP/loxP; WAPiCre mice can serve as a tool with which to identify GR-dependent functions in the mammary gland after P14.5. Because GRloxP/loxP; WAPiCre mice retain GR in the mammary ductal epithelium of the nonpregnant, nonlactating mouse, possible roles of GR in early mammary (ductal) development cannot be identified with this mouse model. Yet such a role can be inferred from previous work on GRdim mice that are deficient in GR binding to DNA and display altered ductal development (7).
We have shown that, in the absence of GR, mammary alveolar development is retarded. Between P14.5 and P18.5, alveoli fail to develop fully and penetrate the fat pad in GRloxP/loxP; WAPiCre. This phenotype results from impaired cell proliferation but not from increased apoptosis, as no apoptotic epithelial cells can be detected in this phase of mammary development (data not shown). Although these observations concentrate on a different time window in mammary gland development, they are similar to the findings in GRdim mice by Reichardt et al. (6). In this report, a defect in ductal morphogenesis in GRdim mice was observed that was also due to reduced epithelial cell proliferation. However, GRdim can fully support alveolar proliferation, and the structure of the lactating gland is normal in GRdim mice. Therefore, the GR-signaling pathways involved in the control of cell proliferation in the developing ducts seem to be different than those that influence cell proliferation in the alveoli. From the data on GRdim mice and our observations, we conclude that alveolar proliferation in late pregnancy is dependent on GR via a pathway that is independent from dimeric GR-glucocorticoid response element interaction. The molecular mechanism for GR influence on alveolar cell proliferation between P14.5 and P18.5 remains to be elucidated. The serum- and glucocorticoid-induced kinase has been described as a GR target that mediates proliferative and antiapoptotic effects of corticosteroids on cultured mammary epithelial cells (12). Serum- and glucocorticoid-induced kinase was detected at equal levels in mutant and control glands at P14.5 (data not shown).
After parturition, however, a dramatic increase in epithelial cell proliferation takes place in mammary glands from GRloxP/loxP; WAPiCre mice (Fig. 6, panel E vs. panel F). This proliferation is compensating for the poor alveolar structures in mutant glands. It should be noted that this proliferative boost cannot be derived from the residual GR-positive epithelial cells, as the number of BrdU-positive epithelial cells (i.e. cells that have undergone S phase within 2 h before death) at L0.5 vastly exceeds the number of residual GR-positive epithelial cells at L0.5 or P18,5. (compare Fig. 6F with Fig. 2F). Therefore, compensatory mechanisms seem to take place in the absence of GR that restore proliferative capacities upon GR-negative epithelial cells. They do not suffice, however, to restore the penetration of the mammary fat pad by alveolar structures to wild-type levels. Mammary glands from GRloxP/loxP; WAPiCre mice at L10 still can easily be distinguished from wild-type glands by a decrease in alveolar space relative to fat (Fig. 4, panels E and G vs. panels F and H). The dramatic boost of cell proliferation at L0.5 could be linked to the transition of the mammary epithelium from stage I to stage II lactogenesis, triggered by the decrease in circulating progesterone levels after parturition (13). This process has been demonstrated to rely on basal levels of glucocorticoids. As the morphological changes from pregnancy to lactation proceed normally in the absence of epithelial GR, targets for glucocorticoid actions in this process are the mammary stromal cells or endocrine cells outside of the mammary gland.
After parturition, alveolar morphology and milk production are normal in GRloxP/loxP; WAPiCre mice, as judged from histology, from monitoring milk protein gene expression and from the ability of GRloxP/loxP; WAPiCre mothers to nurse their pups. Furthermore, there was no evidence of impaired mobilization of fat into milk in GRloxP/loxP; WAPiCre mice. GR deletion does not impinge on the expression of the milk proteins, ß-casein and WAP, thus indicating that GR is not essential for ß-casein and WAP expression in vivo. The costimulatory action of liganded GR on the prolactin-activated transcription factor STAT5 in the activation of the ß-casein promotor that has been described in several cell culture model systems (2, 14) seems to be dispensable for milk protein synthesis in vivo.
Kingsley-Kallesen et al. (8) have studied epithelial transplants from mice carrying a hypomorphic allele of the GR (4, 5). They observed altered ductal development, but normal lobuloalveolar development and lactation. They suggested that the MR might compensate for the GR in milk protein synthesis. Because we cannot detect MR in mammary glands of GRloxP/loxP; WAPiCre or control mice by immunohistochemistry (data not shown), we have no data supporting a compensation of GR loss by MR in our experimental system. The MR expression observed in transplanted mammary epithelial cells by Kingsley-Kallesen et al. (8) could be an experimental feature (mouse strain, transplantation).
Kingsley-Kallesen et al. (8) showed unchanged lobuloalveolar development in the absence of GR at P14.5 and at L1. Clearly, the conditional mutagenesis model described here is different from transplantation experiments, i.e. in the latter, epithelial cells and their progenitors have been devoid of GR since zygote formation of the donor mice, and compensatory mechanisms with regard to cell proliferation could be established in these cells early on. In the conditional mutagenesis model described here, GR is deleted in mammary epithelial cells only during lobuloalveolar development, and a phenotype is seen in these cells soon after.
In addition, both studies differ in the GR allele used. Kingsley-Kallesen et al. (8) used epithelia from mice carrying a hypomorphic GR allele (4) that retain some activity (9). The conditional GR allele in our study involves a different targeting strategy, which leads to a nonfunctional GR allele after deletion (5, 10).
Apart from these seeming discrepancies, it should be noted that the observations reported here, by and large, corroborate those of Kingsley-Kallesen et al. (8) with respect to lactation, milk protein expression, and involution proceeding normally in the absence of GR. Corticosteroids have been shown to regulate the postlactational involution of the mammary gland by inhibiting the synthesis of matrix metalloproteinases (15). This process probably occurs in cells of the mammary stroma, which remain GR positive in both the cell type-specific deletion model presented here and in the epithelial transplantation model presented by Kingsley-Kallesen et al. (8).
In summary, our observations show that GR regulates mammary epithelial cell proliferation during late lobuloalveolar development, but is dispensable for the differentiation and function of secretory alveolar cells of the lactating mammary gland. As the initiation of lobuloalveolar development takes place in the presence of GR in the model presented here, it may be rewarding, in further studies toward glucocorticoid signaling in the mammary gland, to target GR in ductal epithelial cells or alveolar stem cells. As demonstrated here, loss-of-function experiments can be performed in a highly specific manner in vivo using the Cre-loxP system, once appropriate molecular markers for the respective cell type are identified.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Analysis of Recombination
DNA was prepared from mouse mammary glands by overnight lysis in tail buffer (50 mM Tris HCl, pH 8.0; 100 mM EDTA; 100 mM NaCl; 1% sodium dodecyl sulfate; 0.5 mg/ml proteinase K) and subsequent phenol-chloroform extraction; 10–30 µg of organ DNA were digested with SacI, blotted, and hybridized with a probe to detect the floxed (7 kb) and recombined (4.8 kb) alleles. Signals were quantified using a Phospho Imager (Fuji, Tokyo, Japan). Recombination percentage is calculated as signal intensity of the band representing the recombined allele divided by the sum of the signal intensities of the bands representing recombined and floxed alleles.
Immunohistochemistry and BrdU Labeling
Mice were mated at 8 wk of age and killed on the respective stages of mammary gland development by cervical dislocation. Two hours before death, a solution of BrdU in PBS was injected at 50 mg BrdU/kg body weight. Inguinal mammary glands were removed, fixed overnight in 4% PBS-buffered paraformaldehyde, dehydrated, and embedded in paraffin, and 6-µm sections were prepared. Antigens were detected using specific antibodies diluted in PBS [GR: MC-20, Santa Cruz Biotechnology, Inc. (Santa Cruz, CA); MR: rabbit polyclonal antibody (16); BrdU: DAKO M0744 (DAKO Corp., Carpinteria, CA] and visualized by ABC/peroxidase staining (Vectastain; Vector Laboratories, Inc., Burlingame, CA). To visualize nuclei, sections were stained with hematoxylin after immunodetection. Mitotic labeling index has been determined by counting at least 500 epithelial cell nuclei per animal and determining the number of BrdU-positive nuclei.
RNA Isolation and Analysis
RNA from snap-frozen tissue was isolated using the RNeasy mini kit (QIAGEN, Chatsworth, CA) according to manufacturer’s instructions. Total RNA (15 µg) was run on a denaturing formamide-3[N-morpholino]propanesulfonic acid-1% agarose gel, transferred to a nylon membrane, and hybridized to a probe labeled with 
32P-dCTP by random priming. Probe templates were generated by restriction digestion from cDNA fragments subcloned into plasmids, namely pFLAG-ß-casein (mouse), kindly provided by Dr. N. Hynes, Friedrich-Miescher-Institute, Basel (CH), pEndoB (mouse keratin 18), kindly provided by Dr. W. Franke, German Cancer Research Center, Heidelberg (D), and an expressed sequence tag-cDNA probe for WAP (IMAGE clone 1511486). Signals were detected and quantified using a PhosphoImager.
Milk Fat Analysis
Milk was isolated from mice at L5. Six hours after pup removal, 25 IU oxytocin in PBS was injected ip, and 15 min later, milk was collected by aspiration in a Pasteur pipette. Fat percentage was estimated using the creamatocrit method (17). A fatty acid profile from milk was determined by gas chromatography/mass spectometry (18, 19).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
First Published Online October 7, 2004
Abbreviations: BrdU, Bromodeoxyuridine; GR, glucocorticoid receptor; L3, lactation d 3; MR, mineralocorticoid receptor; P14.5, pregnant d 14.5; STAT5, signal transducer and activator of transcription 5; WAP, whey acidic protein.
Received for publication February 17, 2004. Accepted for publication September 30, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
NURSA Molecule Pages Link:
This article has been cited by other articles:
 |
|
 |
|
  C. Otto, I. Fuchs, G. Kauselmann, H. Kern, B. Zevnik, P. Andreasen, G. Schwarz, H. Altmann, M. Klewer, M. Schoor, et al. GPR30 Does Not Mediate Estrogenic Responses in Reproductive Organs in Mice Biol Reprod, January 1, 2009; 80(1): 34 - 41. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Otto, B. Rohde-Schulz, G. Schwarz, I. Fuchs, M. Klewer, D. Brittain, G. Langer, B. Bader, K. Prelle, R. Nubbemeyer, et al. G Protein-Coupled Receptor 30 Localizes to the Endoplasmic Reticulum and Is Not Activated by Estradiol Endocrinology, October 1, 2008; 149(10): 4846 - 4856. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Otto, I. Fuchs, H. Altmann, M. Klewer, A. Walter, K. Prelle, R. Vonk, and K.-H. Fritzemeier Comparative Analysis of the Uterine and Mammary Gland Effects of Drospirenone and Medroxyprogesterone Acetate Endocrinology, August 1, 2008; 149(8): 3952 - 3959. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hennighausen and G. W. Robinson Interpretation of cytokine signaling through the transcription factors STAT5A and STAT5B Genes & Dev., March 15, 2008; 22(6): 711 - 721. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Genetically Modified Animals in Endocrinology Endocr. Rev., December 1, 2005; 26(7): 985 - 993. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
