| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
1 Institute of Animal Science, ARO, The Volcani Center, Bet-Dagan, Israel; and
2 Georg Speyer Haus, Institute for Biomedical Research, Frankfurt am Main, Germany
Requests for reprints: Itamar Barash, Institute of Animal Science, ARO, The Volcani Center, P. O. Box 6, Bet-Dagan 50250, Israel. Phone: 972-8-9484-418; Fax: 972-8-9475-075. E-mail: barashi{at}agri.huji.ac.il
 |
Abstract |
|---|
 TopNotes Abstract IntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
|
Introduction |
|---|
|
TopNotesAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
Stat5 is one of the seven members of the Stat gene family. It was identified as a prolactin (PRL) activated transcription factor essential for lactogenic hormone response in mammary epithelial cells (1). It shares critical structural and functional features with other members of the Stat gene family (2). In addition to its central function in PRL signal transduction, Stat5 also relays the activity of growth hormone (GH), interleukin-2 (IL-2), IL-7, IL-9, IL-15, thrombopoietin, erythropoietin, epidermal growth factor (EGF), and granulocyte stimulating factor (3–6). Two highly related genes, Stat5a and Stat5b, have been found in the genome of mice (7–9), rats (10), and humans (11, 12). These genes are located on mouse chromosome 11, close to STAT3 (13, 14), and share 96% similarity at the amino acid level (15).
The activation mechanism of Stat5 in mammary epithelial cells has been studied in detail (16). PRL binding causes receptor dimerization and intracellular activation of Jak2 kinase. Tyrosine phosphorylation of the receptor and of Stat5 molecules recruited to the receptor ensues, followed by dimerization through interactions of phosphotyrosines with the SH2 domains and translocation to the nucleus. Transcription is activated through binding to specific DNA elements (5'-TTCNNNGAA-3') located in the promoters of responsive genes (17, 18). At least one Stat5-binding site was found in the promoters of the milk protein genes, ß-casein (19), 
s1-casein (20), whey acidic protein [WAP, (8)], and ß-lactoglobulin [BLG, (21)]. Glucocorticoid receptors can enhance Stat5 transactivation through protein-protein interactions (22, 23). A role for serine/threonine phosphorylation has also been suggested (24, 25). Truncated Stat5 molecules, lacking the carboxyl terminal transactivation domain (TAD), have been shown to exert strong dominant negative effects (26, 27). The natural occurrence of such truncated Stat5 molecules and their ability to recruit transcriptional corepressors defines a role for Stat5 also as a gene repressor (28).
Stat5a expression has been detected in most tissues including the mammary gland in its virgin, pregnant, lactating, and postlactation states (8, 10). Tissues from knockout mice and transplantation of such tissues into wild-type animals have been used to gain additional insights into the role of individual genes in differentiation events (29). These experiments showed that Stat5a is activated by PRL in the epithelium, but GH and EGF activate Stat5a preferentially in the stroma (30). Stat5a-deficient mice are unable to lactate. The gland fails to fully develop and does not undergo functional differentiation during pregnancy (31). A similar phenotype, a little less severe, was observed in Stat5b-/- mice (32, 33).
Inactivation of the Stat5 genes in embryonic stem cells and derivation of knockout mice lines has provided important information. However, mammary gland differentiation is arrested at an early stage in these animals, and inactivation of Stat5a and Stat5b results in infertility (32). To gain additional information on Stat5 gene function at later stages of mammary development, we derived transgenic mice expressing the wild-type Stat5 (STAT5) or a constitutively active Stat5 (STAT5ca) variants (34) under the control of regulatory sequences from the BLG milk protein gene. The expression of both transgenes affects mammary morphology and causes increased ß-casein synthesis and secretion. STAT5ca exhibited stronger effects, enhanced cellular proliferation, and predisposed the mammary cell to transformation.
|
Results |
|---|
|
TopNotesAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
750, not shown). This short form of Stat5 lacks its TAD and acts in a potent dominant negative fashion (27). The coding regions of the three genes were inserted into a BLG gene construct from which sequences between the second and the sixth exons had been deleted. The insertion of Stat5 variants between the second and the sixth BLG exons yields a construct, which subjects Stat5 expression to the hormonal and tissue-specific regulation provided by the BLG regulatory sequences (35).
|
750, and the BLG/STAT5ca constructs in cultured cells. For this purpose, we used HC11 mammary epithelial cells stably transfected with the BLG/STAT5 variants and the ß-casein/luciferase reporter gene. Cells transfected with the ß-casein promoter/luciferase reporter gene showed an 8-fold induction of luciferase activity upon hormonal stimulation (Fig. 2A, lane 1
). This induction was mediated by Stat5, endogenously expressed in HC11 cells. Introduction of BLG/STAT5 resulted in an enhanced induction of luciferase activity (lane 2). The expression of the dominant negative variant STAT5750 resulted in the reduction of luciferase expression upon hormonal stimulation (lane 3). The constitutively active variant of Stat5 caused enhanced expression of luciferase in the uninduced state, which was further increased by hormonal activation of the BLG promoter (lane 4). These results indicate that the recombinant genes are, as expected, properly regulated by lactogenic hormones and are functionally active.
|
-casein antibodies (Fig. 2B). Higher levels of ß-casein were detected in induced cells transfected with BLG/STAT5 and BLG/STAT5ca when compared to nontransfected cells. Expression of the BLG/STAT5750 gene caused a reduction in the inducibility. Expression of BLG/STAT5 was also observed in CID-9 mammary epithelial cells transfected by the biolistic procedure (Fig. 2C). The BLG/STAT5 construct was delivered into the cells using a high-pressure gene gun. CID-9 cells attached to Matrigel form ductal and alveolar structures 2 days after plating (36). When analyzed 2 days after bombardment with the BLG/STAT5 coated particles, expression of the tagged protein could be detected in these structures by immunofluorescence with an antibody to the HA epitope.
Expression of BLG/STAT5 and BLG/STAT5ca in Transgenic Mice
Six lines of transgenic mice carrying the BLG/STAT5 and 13 lines of mice carrying the BLG-STAT5ca gene constructs were generated by microinjection of DNA into the pronuclei of fertilized eggs. The copy numbers and relative expression levels in the individual lines expressing the transgenes are shown in Table 1. Up to 10 copies of the transgene were integrated into the host genome. Two of the six lines carrying the BLG/STAT5 construct expressed the protein in mammary gland cells (Fig. 3A
). Transgene expression was exclusively detected in mammary gland cells and not in salivary gland, kidney, heart, skeletal muscle, and liver (Fig. 3B).
|
|
Five mouse lines expressing the BLG/STAT5ca transgene were derived. The fusion protein with an expected molecular mass of 140 kDa was detected by immunoprecipitation with an antibody specific for the Stat6-derived TAD and immunoblotting with anti-Stat5a monoclonal antibody (Fig. 4A ). The fusion protein was detected in the transgenic lines BLG/STAT5ca-7, 8, 10, 15 (and line 19, not shown), but not in nontransgenic control mice. The migration of the fusion protein was comparable to that from HC11 cells transfected with BLG/STAT5ca. Stat5ca was not present in nontransfected cells.
|
BLG/STAT5 and BLG/STAT5ca Transgene Expression Alters the Morphology of the Mammary Gland at Different Stages of Development
It has been previously shown that Stat5 plays an essential role in the development of mammary epithelial cells and the regulation of milk protein synthesis in the mouse (31, 32). Histological sections from mammary biopsies from wild-type and transgenic mice of lines BLG/STAT5-6 and BLG/STAT5ca-15 are shown in Fig. 5 along with whole mounts from virgins of the same strains. No differences in mammary morphology were observed in the ductal tree of virgin and early pregnant mice (Fig. 5, a–c). By 14 days of pregnancy, significant differences began to appear. The lobuloalveolar structures in line BLG/STAT5-6 (Fig. 5e) were expanded occupying 48 ± 3% of the fat pad compared to 33 ± 7% in the mammary gland of the wild-type mice (Fig. 5d). In these transgenic mice, the alveoli were also more distended and appeared to be more developed compared to the wild-type controls. The alveolar structures in the transgenic lines BLG/STAT5ca-15 (Fig. 5f) and BLG/STAT5ca-19 (not shown) occupied more of the fat pad (44 ± 6%). However, they appeared denser and lumens were smaller. Later in pregnancy (day 18), the differences in size of single alveolar structures were maintained.
|
The alveolar size in wild-type and BLG/STAT5-6 mice increased further at day 6 of lactation (Fig. 5, j and k); alveoli of BLG/STAT5-6 mice exceeded the size of those observed in the wild-type mice. At this stage, the epithelial cells in BLG/STAT5-6 (and BLG/STAT5-8, not shown) appeared even more flattened. The alveolar structures detected in lines BLG/STAT5ca-15 (Fig. 5l) and BLG/STAT5-19 (not shown) again appeared much more condensed. This was not as pronounced when glands of BLG/STAT5ca-15 mice, suckled for 10 days, were analyzed (not shown). These observations suggest that the maturation of fully functional secretory alveoli in BLG/STAT5-6 mice occurs earlier than in wild-type mice.
Differences between the appearance of the mammary gland tissue in wild-type and transgenic mice were also observed during involution. This process was initiated by the removal of the pups from their mothers. Three days after removal, substantial regression of the alveoli was detected in the wild-type mice (Fig. 5m). In contrast, the compaction of the alveoli and the loss of epithelial cells were not found in tissues of BLG/STAT5-6 and BLG/STAT5ca-15 mice (Fig. 5, n and o) in which the involution process was markedly delayed. Five days after pup removal, alveoli from BLG/STAT5-6 and BLG/STAT5ca-15 mice were still filled with secretory proteins and fat droplets and occupied a substantial fraction of the fat pad (Fig. 5, q and r). These protective effects against the progress of involution did not persist. At day 10, after pup removal, the morphologies of all three genotypes appeared similar again (not shown). The morphological consequences of transgene expression were consistently observed in all animals tested. Only minor variations among animals or regions within individual glands were found.
Deregulation of Mammary Epithelial Cell Specific Gene Expression Caused by Transgenic Stat5 Variants
The enhancement of Stat5 expression in BLG/STAT5-6 mice or its forced activation in BLG/STAT5ca-15 mice was accompanied by changes in the timing of developmental events and in the histological appearance of the mammary epithelium. These phenotypes most likely reflect changes in gene expression patterns. We have analyzed a number of genes indicative for individual developmental stages in the mammary glands of wild-type and transgenic mice.
Proliferating cells can be visualized by PCNA staining in the nucleus. The presence of this protein indicates that the cells are in the late S phase of the cell cycle (37). Protein extracts were prepared from mammary tissue of mice from all three genotypes at day 14 of pregnancy. Higher levels of PCNA were observed in extracts obtained from BLG/STAT5ca mice when compared to those of BLG/STAT or nontransgenic mice (Fig. 6A ). Similar observations were also made in mice of day 18 of pregnancy (not shown). This result was corroborated by immunohistochemical analysis of mammary gland sections from these mice (Fig. 6, B and C). Here, PCNA-expressing cells were observed in less than 20% of the mammary epithelial cell population of wild-type and BLG/STAT5 mice. Significantly (P < 0.05) higher levels of stained nuclei (29% in line BLG/STAT5ca-15 and 35% in line BLG/STAT5ca-19, not shown) were observed in mammary sections of transgenic mice expressing the constitutively active variant of Stat5.
|
Stat3 is strongly activated at the time of mammary involution (29). A lower level of Stat3 activity compared to wild-type mice was found in mammary glands of the transgenic lines STAT5-6 and STAT5ca-15 at this stage of development (Fig. 6, E and F).
The milk protein genes are target genes of Stat5. We measured the relative expression levels of the ß-casein gene in the three genotypes. ß-Casein was detected by immunoblot analysis of protein extracts from mammary glands of virgin, pregnant, lactating, and involuting mice (Fig. 7 ). Separate comparisons were made for each stage and the Western blot signals were quantitated by densitometry.
|
Transgenic mice expressing the BLG/STAT5ca construct exhibited even higher levels of ß-casein when compared to BLG/STAT5 or wild-type mice. This was found during most stages of development, differentiation, and mammary regression. On day 1 postpartum, the ß-casein levels reached four times that observed in wild-type animals and twice that found in BLG/STAT5 mice. Considerably higher levels were maintained throughout the time points of lactation and involution examined.
The higher levels of ß-casein protein expression in the mammary gland of transgenic mice were reflected on the mRNA level (38). Transgenic mice of lines BLG/STAT5-6 and BLG/STAT5ca-15 expressed 2- or 2.4-fold higher levels of ß-casein mRNA compared to control mice at day 6 of lactation (not shown).
The transactivating effects of BLG/STAT5 and BLG/STAT5ca on the regulatory sequences of another milk protein gene, BLG, were also investigated. Transgenic mice of line #709, expressing the luciferase gene in their mammary glands under the control of the BLG promoter (39, 40), were mated with transgenic mice of lines BLG/STAT5-6 and BLG/STAT5ca-15. Luciferase activity was measured in extracts from glands obtained 3 days after pup removal. Fig. 8 shows a 25- to 40-fold higher luciferase activity in the extracts obtained from double transgenic mice carrying the BLG/luciferase and BLG/STAT5 or the BLG/luciferase and BLG/STAT5ca genes compared to those carrying only the BLG/luciferase gene.
|
The Effects of the Stat5 Variants on Milk Composition
Despite the altered morphology observed in BLG/STAT5 and BLG/STAT5ca mice, these mice are lactation competent and are able to raise their pups. Since milk protein gene expression was affected by the transgenes, we investigated the abundance of individual proteins in the milk. Milk was collected from wild-type and transgenic animals at day 6 of lactation. Equal volumes of milk were defatted and the milk proteins were analyzed by gel electrophoresis and Western blotting (Fig. 9
). Staining with a ß-casein specific antiserum revealed that ß-casein expression was increased in the milk derived from the transgenic animals (Fig. 9A). -Lactalbumin was slightly affected. BLG/STAT5ca-11 served as a line for comparison. This line has the BLG/STAT5ca gene incorporated in its genome, but does not exhibit STAT5ca expression.
|
-casein, ß-casein, and -casein were detected. -Lactalbumin and WAP were less affected. Serum albumin is not expressed in mammary epithelial cells, but is derived from serum present in the mammary tissue (42). It serves as an internal standard for the quantitation of the transgene effects on the secreted levels of ß-casein and -lactalbumin (Fig. 9C). The higher levels of milk proteins expressed and present in the milk of the transgenic mouse strains affected the growth rate of pups during the early nursing period (Fig. 10 ). The weight of pups from transgenic female of line STAT5-6 increased more rapidly than that of the wild type, from day 3 postpartum. A faster growth rate was maintained during subsequent days of lactation also in pups of lines BLG/STAT5ca-15 and BLG/STAT5ca-19 (not shown).
|
|
|
Discussion |
|---|
|
TopNotesAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
Here, we gained insights into Stat5 functions during later developmental stages through the targeted overexpression of Stat5 and its constitutively active variant to the mammary epithelium of transgenic mice. The targeted expression was achieved by provision with regulatory regions of the BLG gene. The BLG gene-derived expression cassette confers tissue-specific and hormone-dependent expression of recombined sequences in the mammary glands of transgenic mammals. It had been used for the expression of human serum albumin and luciferase genes in mammary epithelial cells (35, 39, 40).
The Stat5 sequences integrated into the BLG cassette have been previously characterized. Ovine Stat5, initially named mammary gland factor (MGF), has a high sequence homology with mouse Stat5a. No differences in the functional properties of these orthologues have been found in experiments in which cells of human, mouse, and rat origin have been employed. The constitutively active variant of Stat5 carries, in addition to the self-activating kinase domain derived from Jak2, a replacement of sequences in the TAD. When this fusion protein was initially derived, it was shown that Stat5's own TAD functions rather weakly in this context. For this reason, it was replaced by the TAD of Stat6, a molecule also capable of inducing the ß-casein gene promoter (26, 34). Our studies have shown that STAT5ca is a potent tool to study the effects of Stat5, independent from cytokine receptor activation (34). Although we initially postulated that the STAT5ca molecule primarily affects promoters with Stat5-binding sites, autonomous functions of the carboxyl terminal kinase domain or recruitment of axillary proteins to the Stat6 TAD cannot be excluded.
The integrity and functionality of the BLG/STAT5 and BLG/STAT5ca constructs were validated by transfer and expression studies in cultured HC-11 cells. These cells are lactogenic hormone inducible and increased Stat5-dependent luciferase activity or ß-casein were detected upon stable transfection with the native or constitutively active Stat5. This induction could be suppressed by the introduction of a Stat5 variant in which the carboxyl terminal TAD domain had been deleted. This variant assumes very strong dominant negative properties (27).
Nontransfected HC11 cells are responsive to lactogenic hormone induction, i.e., they express and activate Stat5 upon PRL treatment, resulting in the induction of the ß-casein gene promoter. The higher levels of ß-casein expression in the transfected cells indicated that the cellular concentration of the endogenously expressed Stat5 is a limiting component in the induction process. In addition to the cellular concentration, the activation state of Stat5 is also limiting the extent of target gene expression. Thus, even higher levels of luciferase activity and ß-casein were detected in HC11 cells transfected with BLG/STAT5ca when compared to those transfected with BLG/STAT5. This is most likely due to effects resulting from lack of receptor-mediated desensitization. Whereas the BLG/STAT5 encoded molecule is still subjected to the mechanisms of induction and down-regulation, the constitutively active variant is independent from such influences and therefore probably even more potent.
A very low level of spontaneous activation of the endogenous Stat5 molecule maintained in HC11 cells even in the absence of cytokine induction could cause residual transcription from the BLG promoter governing STAT5ca expression. The resulting STAT5ca is self-activating and could enhance the transcription from its own promoter, leading in turn to the higher expression levels of the reporter genes observed in the absence of PRL. BLG/STAT5 transfected cells would not exhibit such a phenotype, because Stat5 activation through a cytokine receptor is required in these cells.
The activity of the ß-casein-luciferase reporter gene was more sensitive to transgene effects compared to the level of the native ß-casein. Stronger responses to suckling stimuli have previously been shown for the hybrid gene when BLG-luciferase activity and the level of the native ß-casein were measured (39). This probably results from the shorter half-life of the luciferase protein. However, it is also possible that the luciferase reporter gene does not comprise all the regulatory elements of the endogenous gene promoter.
Transient transfection of BLG/EMGF-STAT5 into hormonally treated CID-9 cells by particle bombardment also resulted in transgene expression. The clonal cell lines CID-9 (45) and HC11 (46) are both derived from the same ancestral COMMA-D cell line, originally isolated from the mammary glands of mice in mid-pregnancy (47). They differ in their dependency on extracellular matrix (ECM) for differentiation (48). Particle bombardment is suitable for delivery of the DNA into these cells which were hormonally primed and grown on laminin-rich ECM.
About 30% of the transgenic lines carrying BLG/STAT5 and 25% of the lines carrying the BLG/STAT5ca genes expressed the encoded Stat5 molecules in a copy number-independent, but tissue-specific and lactation-dependent manner. This is in accordance with our previous observations (40). Our transgenic lines exhibit normal ductal outgrowth during puberty. Phenotypic effects were not observed during early developmental stages when wild-type and transgenic mice were compared. Deviations in the morphology of the mammary gland between wild-type and transgenic lines first became apparent at mid-pregnancy. The extended lobuloalveolar structures, which distinguished the mammary gland of BLG/STAT5 transgenic mice, resemble a more developed stage of the gland, than that found at the same time point in wild-type mice. Condensed alveoli almost devoid of lumenal volume were detected in BLG/STAT5ca mammary sections. The appearance is reminiscent of impaired mammary gland development found in mice derived with dominant negative variants of the erb2 gene (49, 50) or to a lesser extent in Stat5a knockout mice (31). In contrast to these mouse strains, however, the gland in BLG/STAT5ca mice is fully functional. The increased proliferation of mammary epithelial cells in BLG/STAT5ca mice during pregnancy is accompanied by an altered morphology, which affects functional differentiation, enhancement of milk protein gene expression, and secretion of milk proteins into the milk. The enhanced proliferation of the epithelial cells in the STAT5ca mice during pregnancy is probably caused by an earlier and stronger activation of Stat5, independent from lactogenic hormone action, when compared to wild-type mice.
Levels of Stat5 expression and state of activation also affect the extent of milk protein expression and secretion. Higher levels of Stat5 as well as its constitutive activation have consequences for the expression of these genes. The regulation through Stat5 cellular levels might be correlated with an increased sensitivity toward activation by cytokine receptors or with the previously described relief of transcriptional repression in epithelial cells by Stat5 (19). The high inductivity of BLG/luciferase activity during the state of involution, when endogenous Stat5 levels are low, supports the latter assumption. We observed a significant increase in the levels of the ß-casein protein in the milk of lactating females expressing the Stat5 transgene. Such an increase was also observed after GH-induced Stat5 activation in cows (51). The difference between Stat5 effect on the induction of ß-casein and -lactalbumin suggests that not all milk protein promoters are equally regulated. Casein promoters are more sensitive to Stat5 action than other milk proteins. For the -lactalbumin promoter, a more significant role for the MAF transcription factor has been demonstrated (52). The higher levels of the caseins in the milk of the transgenic animals correlated with their pattern of expression in the tissue and enhanced the growth of the litters. Experiments in cultured cells have established the chimeric STAT5-JAK2 molecule as a potent tool to study the effect of Stat5 independent of cytokine receptor activation (34). These experiments showed that the action of the chimeric STAT5-JAK2 gene is mainly restricted to the regulation of genes with Stat5-binding elements, and that Stat3 regulation, e.g., is not affected. Formally, however, we cannot rule out that the kinase domain of the STAT5ca protein might have additional effects.
Stat3 activity is low during lactation and increases sharply during the initiation of the involution phase (41, 53). Conditional knockout of the Stat3 at early stages of lactation led to the observation that Stat3 is required for the onset of involution (41) and might serve as a signal for apoptosis induction in the mammary cells.
We observed a significant delay in the onset of involution in BLG/STAT5 and BLG/STAT5ca mice, associated with lower levels of activated Stat3 activity during this period. It is possible that activated Stat5 serves as a survival factor for mammary epithelial cells at the end of lactation and counteracts the function of Stat3. Stat5 has been recognized as a survival factor in hematopoietic cells (54) where its activation was correlated with Bcl-x induction and the prevention of apoptosis. Constitutively activated Stat5 has been expressed in cytokine-dependent hematopoietic cells and resulted in factor independence (55). A role for Stat5 activation in the process of malignant transformation of these cells has been suggested (56). Overexpression of the transforming growth factor in mammary epithelium resulted in sustained Stat5 activation and showed an increase in cell survival during the involution period (57). Our results are consistent with these observations, and show that Stat5 is an important contributor to the regulation of apoptosis when induced at the end of the lactation phase. Activation of Stat5 by itself, however, is not sufficient to permanently protect the cells, but causes a delay in the onset of apoptosis.
It is also interesting to note that the mammary cells acquire Stat5 dependence in the process of developmental progression. Early stages of mammary epithelial cells (in virgins or at early pregnancy) and cultured mammary epithelial cells (e.g., HC11) do not exhibit survival dependence upon Stat5. The high levels of Stat5 activation during lactation and its precipitous drop with the onset of involution have made it likely that Stat5 is involved in the regulation of this process.
Constitutive activation of Stats has been detected in primary tumors and transformed cell lines. Activation of Stat1 and Stat3 was frequently found in breast cancer (58, 59). Activated Stat5 was mainly observed in myeloid cells, T cells, and lymphocytes, resulting in leukemias and lymphomas (60). Here we showed that forced activation of Stat5 contributes to mammary tumorigenesis.
Our study shows that Stat5 levels and activation states influence at least three seemingly independent functions in the mammary gland. These are the extent of proliferation during pregnancy, the amount of milk protein synthesis and secretion during the lactation phase, and the survival of mammary epithelial cells at the transition from the lactation to the involution stage. This extends our appreciation of the versatile functions of Stat5 gained previously in tissue culture experiments and knockout mouse models.
|
Materials and Methods |
|---|
|
TopNotesAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
750 was prepared by introducing the stop codon at the respective site of the cDNA sequences (27). The construct STAT5ca is comprised of sequences from three genes: amino acids 1–750 of the sheep Stat5, 677–847 of the human Stat6, and 757–1129 of the mouse Jak2 (34).
Constructs STAT5 and STAT5750 were introduced into the EcoRI site in the multiple cloning site of the BLG-based expression cassette (35, 40) and correct orientation was confirmed by restriction analysis. Construct STAT5ca was blunted at its NotI site, cut with SalI, and subcloned into the SpeI-blunted, XhoI-cut (compatible to SalI), BLG multiple cloning site.
The BLG-based expression cassette has been previously described in detail (35, 40). To avoid biphasic expression, the natural BLG ATG translation-initiation codon, as well as a second potential initiation codon in BLG exon 1, was converted into noninitiating ATT and ATC sequences, respectively. BLG sequences between the first part of BLG exon 2 and the middle of BLG exon 6 (including the TAG termination codon) were deleted and replaced by a multiple cloning site.
Cell Culture and Transfections
(a) HC11 mammary cells were grown in RPMI 1640 (Gibco BRL/Life Technologies, Inc., Gaithersburg, MD) containing 10% (v/v) FCS, 5 µg/ml insulin, 10 ng/ml EGF (Sigma Chemical Co., St. Louis, MO), and combined antibiotics (Biolabs, Jerusalem, Israel) diluted 1:1000. Cells, stably transfected with the ß-casein/luciferase reporter gene (61), were cotransfected with additional constructs by the lipofectamine method, using the plasmid pBabe Puro (kindly provided by Prof. Moshe Oren, Weizmann Institute of Science, Rehovot) for puromycin resistance. Selection was performed with 1 µg/ml puromycin. Hormonal induction was performed for 3 days with 1 µg/ml hydrocortisone and 3 µg/ml PRL (NIH PRL 17, kindly provided by the hormone distribution program NIDDK) in confluent cultures grown without EGF. (b) CID-9 mammary cells were grown in DMEM:F12 (1:1, Sigma) containing 5% FCS, 5 µg/ml bovine insulin, 50 µg/ml gentamicin, and combined antibiotics diluted 1:1000. For hormonal and ECM-dependent activation, cells were plated on Matrigel (Collaborative Bio-Medical Products, Bedford, MA) in a six-well plate (0.3 x 106 cells/well) as described (36, 62). Twenty-four hours after plating, the medium was switched to one without FCS. Hormonal induction was performed for 3 days before transfection by the addition of 1 µg/ml hydrocortisone and 3 µg/ml PRL. The BLG/STAT5 construct was transfected into the attached CID-9 cells by the previously described biolistic method (63), except that the bombardment of the coated gold particles into the cells was performed at the reduced pressure of 100 p.s.i. to avoid detachment of the cells from the matrix. Cells were maintained for an additional 48 h in the presence of insulin, hydrocortisone, and PRL at the above-mentioned concentrations.
Transgenic Mouse Lines
Generation of Transgenic Mice. Mice used in this study were of the FBV/N strain. Plasmids containing sequences of BLG/STAT5 and BLG/STAT5ca were restricted with SalI and the appropriate fragments were prepared and microinjected as described (64). Transgenic animals were identified by Southern blot analysis of genomic DNA prepared from tail biopsies using a probe encompassing 3 kb of the BLG promoter. Double transgenic mice were generated by mating mice carrying each of these constructs with transgenic mice of line #709 carrying a BLG/luciferase transgene (40). The double transgenic mice were identified by Southern analysis using a probe encompassing 800 bp of the luciferase cDNA, and PCR analysis with specific primers for BLG/STAT5 located at the 3' end of the influenza virus DNA and the 5' end of the Stat5 sequence, or specific primers for the STAT5ca gene located at the 3' end of the Stat5 and the 5' end of the Jak2 sequences. The two sets of primers were:
Determination of Transgenic Copy Number. Genomic DNA (10 µg) extracted from tail biopsies of transgenic mice was digested with BamHI, subjected to electrophoresis on a 0.8% agarose gel, hybridized first with the above-described BLG, and then reprobed with ribosomal S16 cDNA. The ratio between the densitometric analyses of the signals was related to that of the single-copy transgene (36). Further analyses were performed with females in their first cycle of pregnancy, lactation, or involution.
Analysis of Luciferase Activity in Cellular Protein Extracts
Luciferase activities in stably transfected HC11 cells were determined with the luciferase assay system kit (Promega, Madison, WI) according to the manufacturer's protocol. Briefly, cells were washed twice with cold PBS buffer and extracted with 100 µl of extraction buffer. Extracts were centrifuged at 14,000 x g for 30 s, and supernatants were collected. Luciferase activity was measured using a TD-20e Luminometer (Turner Design Inc., Mt. View, CA) in 20-µl aliquots, supplemented with 100 µl luciferin. The protein concentration of the lysates was determined by the Bradford assay. Luciferase activities were calculated as relative light units per 100 µg of cellular protein.
For the determination of luciferase activity in tissues, mice were killed by cervical dislocation and the six biopsies from each mice, representing mammary glands #1 + #2, #3, #4 + #5, #6 + #7, #8, and #9 + #10 were excised. They were immediately frozen in liquid nitrogen and stored at -70°C. Samples of about 50 mg were extracted with 100 µl of extraction buffer and processed as already described. We have previously shown by real-time imaging (39) that luciferase activity may be altered among glands of an individual mouse, probably as a result of uneven suckling.
Detection of Transgene Products, Milk Proteins, and Activated Stat5 by Immunoblotting
Proteins from mammary glands or cells were extracted with extraction buffer composed of 10 mM Tris-HCl, pH 7.4, 50 mM NaCl, 5 mM EDTA, 0.1% (w/v) BSA, 1% Triton, 50 mM NaF, 1 mM Na3VO4, 1 mM EGTA, 1 mM PMSF, 1 mM aprotinin, and subjected to SDS-PAGE analysis. Fractionated proteins were transferred onto nitrocellulose filters and equal amounts of protein content/sample were confirmed by staining the blot with Ponceau S (Sigma). For casein detection, filters were blocked and reacted with rabbit anti-mouse ß/-casein antibodies (65) diluted 1:250. Goat anti-rabbit IgG complexed with horseradish peroxidase (HRP; Zymed Laboratories, San Francisco, CA) served as the second antibody. For the detection of caseins in milk, samples from females, 6 days after parturition, were diluted 1:5 and aliquots of defatted milk were processed as described in Shani et al. (64). For the detection of epitope derived from HA, filters were reacted with rat monoclonal anti-HA antibodies (Boehringer Mannheim, Mannheim, Germany, or Covance, Richmond, CA) diluted 1:1000. Rabbit anti-rat IgG complexed with HRP, or goat anti-mouse IgG complexed with HRP served as second antibody, respectively. For the detection of PCNA, filters were reacted with anti-PCNA antibodies (Zymed Laboratories) diluted 1:500 and goat anti-mouse IgG complexed with HRP served as second antibody. For the detection of phosphorylated Stat5 and Stat3, filters were reacted with rabbit anti-mouse antibodies produced in rabbits against a peptide containing Tyr-694 or Tyr-750, respectively (Cell Signaling Technology, Beverly, MA) and diluted 1:1000. Goat anti-rabbit IgG complexed with HRP (Zymed Laboratories) served as the second antibody. Signals were generated with an ECL kit (Amersham, Buckinghamshire, United Kingdom) according to the manufacturer's protocol.
For the detection of STAT5ca, proteins were immunoprecipitated with anti-Stat6 polyclonal antibodies (R&D, Minneapolis, MN), analyzed on SDS-PAGE, blotted onto filters, and reacted with anti-Stat5a monoclonal antibodies (Transduction Laboratories, Lexington, KY). Signals were generated with the BLAST blotting amplification system (NEN, Boston, MA) according to the manufacturer's protocol.
Messenger RNA Measurements by RT-PCR
Total RNA was isolated from mammary glands of lactating mice by the LiCl/urea procedure (66). pA+ RNA was separated on oligo(dT) columns with an mRNA isolation kit (Boehringer Mannheim) and 300-ng aliquots were reverse transcribed at 37°C for 50 min in 25 µl of 50 mM Tris-HCl buffer, pH 8.3, containing 75 mM KCl, 10 mM DTT, 3 mM MgCl2, dATP, dTTP, dCTP, dGTP (Stratagene, La Jolla, CA, 10 mM each), 0.5 µg of oligo(dT) primer (Promega), and 200 units of SUPERSCRIPT II (Life Technologies, Inc.). The PCR for STAT5ca was performed in 50 µl of 750 mM Tris-HCl buffer, pH 9.0, containing 1.5 µl of the reverse-transcribed reaction, 200 mM (NH4), SO4, 0.1% (w/v) Tween 20, 1.25 mM MgCl2, 1 unit of Taq polymerase (Life Technologies, Inc.), and 1.5 µl of primers, 10 mM each.
Two sets of primers were used: STAT5ca 1 (5'-CTCCCACTATGGGCAATCTG-3' and 5'-ACATCTCCACACTCCCGAAG-3') and STAT5ca 2 (5'-ACCCACCACACTCTCACTCC-3' and 5'-GAAAGCAGGCCTGAAATCTG-3'), both originating from the 3' sequences of the Stat5 and the 5' sequences of Jak2 and amplifying 398- and 596-bp fragments, respectively, mainly of the Stat6 sequences. The DNA was blotted onto a membrane and hybridized with labeled DNA probe encompassing respective fragments of the STAT5ca gene construct. A sample, which was not reverse transcribed, served as a control for transgenic DNA contamination.
Histological Analysis, Immunohistochemistry, and Immunofluorescence
For whole-mount examination, mammary biopsies from virgin mice were fixed on glass slides with 25% acetic acid/75% ethanol for 1 h and stained with carmine alum overnight as described previously (67).
For section analysis, mammary biopsies from virgin, pregnant, lactating, and involuting glands were fixed in 4% paraformaldehyde, dehydrated in series of 50–100% ethanol, cleared in xylene, and embedded into paraffin. Sections of 5 µm were processed for H&E staining. The area of the lobuloalveolar structures at day 14 of pregnancy was measured by the Image-Pro software (Media Cybernetics, Silver Spring, MD) in five microscopic fields. Immunohistochemical analysis for PCNA was performed with a PCNA staining kit (Zymed Laboratories). Briefly, paraformaldehyde-fixed, paraffin-embedded sections were dewaxed, dehydrated through graded ethanols, treated with 0.3% hydrogen peroxidase for 30 min, boiled in 0.1 M sodium citrate buffer for 10 min, blocked, and reacted with biotinylated PCNA antibody for 90 min. Sections were then washed in PBS, incubated for 10 min with streptavidinated HRP, and reacted with the diaminobenzidine substrate for the generation of signals.
For the detection of STAT5 following biolistic transfections, CID-9 cells were fixed on slides and stained with fluorescein-conjugated anti-HA mouse monoclonal antibody (ICN, Costa Mesa, CA).
|
Acknowledgements |
|---|
|
TopNotesAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
|
Notes |
|---|
|
TopNotesAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
Received February 19, 2002; revised August 2, 2002; accepted August 12, 2002.
|
References |
|---|
|
TopNotesAbstractIntroductionResultsDiscussionMaterials and MethodsAcknowledgementsReferences |
|---|
s1-casein chloramphenicol acetyltransferase constructs. Mol. Endocrinol., 8: 1720–1730, 1994.[Abstract]
activation site-related element. Mol. Endocrinol., 8: 1528–1536, 1994.[Abstract]
