| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Physiology and Pharmacology, and Institute for Molecular Bioscience (S.P.T., P.L., M.J.W.), University of Queensland, St. Lucia, Brisbane 4072, Australia; Unite d’Endocrinologie Moleculaire (J.D.), Institute National de la Recherche Agronomique, 78352 Jouy-en-Josas Cedex, France; and The Walter and Eliza Hall Institute of Medical Research and the Cooperative Research Centre for Cellular Growth Factors (D.J.H.), Parkville 3052, Australia
Address all correspondence and requests for reprints to: M. J. Waters, Ph.D., Department of Physiology and Pharmacology, University of Queensland, St. Lucia, Brisbane 4072, Australia. E-mail: m.waters{at}mailbox.uq.edu.au
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(1, 2, 4, 5). Termination of the tyrosine phosphorylation
signal involves phosphatase action at the level of JAK 2 and the STATs
(6), although the phosphatase(s) are not well defined.
Regulation of target tissue sensitivity to PRL has long been thought to
be controlled through receptor expression, for example, through
up-regulation of mammary gland receptor expression at parturition in
the rat (7), with a corresponding decrease in receptor
expression in the corpus luteum, which is thought to decrease
progesterone secretion (8). Recently, PRL and PRL receptor
knockout mice have been created, and these have emphasized the
important role of PRL in mammary alveolar epithelium development,
ovarian function, and embryo implantation (9, 10). The
importance of the STAT 5s, particularly STAT 5a, in mammary alveolar
and ductal development, was also recently demonstrated by specific STAT
5a and 5b knockout (11, 12), which confirmed the view from
in vitro studies that PRL activation of STAT 5a is a central
feature of its role in mammary alveolar epithelial development and
lactation (1, 4, 13). Tyrosine phosphorylation of the PRL
receptor, demonstrable both in vitro and in vivo
(14), promotes STAT 5 activation by providing SH2 docking
sites, particularly Y580 of the full length receptor
(15).
Our earlier study (14) indicated that the rabbit mammary
gland is refractory to PRL-induced tyrosine phosphorylation of both the
receptor and JAK 2 unless the rabbit is deprived of PRL with a dopamine
agonist for at least 24 h. This, and the observation that removal
of residues 322–333 within the cytoplasmic domain increases
PRL-induced activation of a milk protein promoter (16),
led us to propose that a PRL-inducible inhibitory protein was bound to
the receptor that suppressed JAK 2 activation (14).
Recently, a family of cytokine-inducible inhibitors of cytokine
signaling has been reported (17, 18, 19, 20, 21), although the
first member of the family, CIS (cytokine-inducible
SH2-containing protein), has been known for several years
(22). These SOCS (suppressor of cytokine signaling)
proteins appear to act as feedback inhibitors of signaling for a wide
range of cytokines that act through the JAK/STAT (signal transducer and
activator of transcription) pathway (20). They possess a
variable N-terminal region preceding an SH2 domain, followed by a
conserved C-terminal region of approximately 40 residues known as the
SOCS box, which is thought to regulate proteolysis (23).
This motif is found in 20 proteins in five structural classes
(24). SOCS-1 binds directly to JAK kinases, including JAK
2 and Tec kinases, inhibiting their catalytic activity (23, 25). CIS, on the other hand, binds to tyrosine-phosphorylated
ß-chain of IL-3, erythropoietin (22), and GH receptors
(26), inhibiting their proliferative actions and, in the
case of the GH receptor, increasing proteasomal degradation of the
receptor signaling complex (27). SOCS-3 was shown to be
induced by GH and leptin and to inhibit STAT 5-mediated transactivation
of the serine protease inhibitor 2.1 promoter by GH by binding to
membrane-proximal phosphotyrosine residues in the GH receptor
(26, 28, 29). These and other in vitro studies
of SOCS indicate that they act in a classic feedback loop manner to
inhibit the action of tissue-relevant cytokines. This is supported
in vivo by the recent SOCS-1 knockout in which the thymus
was smaller, a loss of mature B cells was observed, and the mice showed
stunted growth and died before weaning with fatty degeneration of the
liver (30). These symptoms could be explained by unopposed
action of interferon- overexpression, a cytokine that activates SOCS
expression (17, 30). On the other hand, CIS-overexpressing
transgenic mice exhibit impaired mammary gland development and growth
retardation, although CIS knockout mice have no obvious phenotype
(3, 31). SOCS-3 knockout and overexpressing transgenic
mice are both embryonic lethal, owing to the disruption of hepatic
hematopoiesis (31). Recently, SOCS-2 knockout mice were
found to exhibit a phenotype resembling GH-overexpressing transgenic
mice (32).
The aim of this study was to assess the involvement of SOCS genes in PRL action in vivo in major PRL target tissues. We have used PRL-suppressed lactating rats as our in vivo model, because we have shown that they respond to PRL injection with tyrosine phosphorylation of both the receptor and JAK 2 and with STAT 5 activation (33). We present evidence for the induction of expression of SOCS genes by PRL in vivo, for their involvement in PRL signaling, and for differential sensitivity of PRL target tissues to induction of these suppressors of signaling.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animals
Twenty-four or 48 h before experiments, pups were removed
from 5- to 10-day lactating Wistar rats, and the mothers were injected
sc with 1.5 mg/kg bromocriptine dissolved in saline-ethanol (7:3) to
block endogenous PRL secretion (33). Injections were
repeated every 12 h. Then, 24 or 48 h after the first
injection, rats were injected ip with 0.25 mg of oPRL in saline
containing 10 mM NaHCO3 or vehicle
alone. Rats were killed 0, 30, 60, 120, 240, or 440 min after PRL
injection. A separate group of animals was removed from the pups and
killed immediately.
For the initial suckling experiments (four per group), pups were removed for 16 h from lactating mothers, and then mothers were put back with the pups for 2 h before being killed. A control group was kept separate from the pups before being killed. To study the effect of suckling on subsequent response to PRL administration, four groups were used. Two groups ("suckled" groups) were put back to suckle their pups for 1 h and then the pups were removed for a further 1 h: one group was then given oPRL (0.25 mg, as above), while the other group was injected with saline vehicle. After a further 30 min, these rats were killed and their tissues were processed. With the other two groups ("nonsuckled" groups), after the 16-h pup deprivation period, one group was injected with oPRL (0.25 mg, as above) 30 min before being killed, while the other group was injected with saline vehicle.
In all cases, the mammary glands, ovaries, liver, and adrenal glands were quickly excised, snap frozen in liquid nitrogen, and stored at -80 C until processing for Northern, electrophoretic mobility shift assay (EMSA), or Western blot analysis. Tissues were fixed in 4% buffered paraformaldehyde for immunohistochemistry. These experiments were authorized by the University of Queensland institutional animal ethics committee according to National Health and Medical Research Council (Australia) guidelines.
Northern hybridization
Total RNA was isolated from tissues using TRIzol Reagent
(Life Technologies, Inc., Rockville, MD) according to the
manufacturer’s protocol. Aliquots of 20 µg of total RNA were
denatured and subjected to electrophoresis on a 1.2% agarose, 11%
formaldehyde gel and then transferred to a HyBond N membrane
(Amersham Pharmacia Biotech, Buckinghamshire, UK). After
UV cross-linking and then baking for 30 min at 80 C, the membranes were
prehybridized in NorthernMax prehyb/hyb buffer (Ambion, Inc., Austin, TX) for 4 h at 50 C.
32P-labeled cDNA probes for SOCS-1, -2, -3, or
CIS [prepared by random prime labeling using the Rediprime II system
(Amersham Pharmacia Biotech, Uppsala, Sweden)] were then
added, and the membranes were hybridized at 50 C for 16 h. After
this, the membrane was washed twice in 2x SSC (30 mM
sodium citrate, 300 mM sodium chloride, pH 7.0), 0.1% SDS
at 65 C for 30 min. After autoradiography at -70 C, all blots were
stripped and rehybridized with a 32P-end-labeled
oligomer probe specific for 18S rRNA to determine loading. SOCS mRNA
expression was then normalized to 18S RNA expression from densitometer
scans using the Bio-Rad Laboratories, Inc. (Hercules, CA)
GS-700 densitometer with Molecular Analyst software (Bio-Rad Laboratories, Inc.).
EMSA
For the preparation of nuclear extracts (35),
tissues were homogenized with a Polytron homogenizer and then incubated
for 15 min in 10 mM HEPES, 10 mM KCl, 0.1
mM EDTA, 0.1 mM EGTA, 1 mM
dithiothreitol (DTT), 20 mM NaF, 0.5 mM
phenylmethylsulfonyl fluoride (PMSF), 1 mM
Na4P2O7,
1 mM Na3VO4,
and 2 µg/ml aprotinin and leupeptin (pH 7.9) on ice. Cells were lysed
by the addition of Nonidet P-40 (Schleicher & Schuell, Inc., Keene, NH) to 0.6% and microfuged at 13,000 rpm for 30
sec. Nuclear pellets were resuspended in 20 mM HEPES, 0.4
M NaCl, 1 mM EDTA, 1 mM EGTA, 1
mM DTT, 1 mM
Na4P2O7,
1 mM Na3VO4, 20
mM NaF, 0.5 mM PMSF, 2 µg/ml aprotinin and
leupeptin, and 20% glycerol. Extracts were then divided into aliquots
and stored at -80 C.
The complementary strands of Stat 5 oligonucleotide probe were obtained from Pacific Oligos (Lismore, New South Wales, Australia) and contained the sequence 5'-AGATTTCTAGGAATTCAATCC-3' [binding sites are underlined (4)]. Annealed, double stranded oligonucleotides were 5'-end labeled with T4 polynucleotide kinase and purified on a 10% polyacrylamide gel.
To perform the gel shift assay, 5 µl of nuclear extract was added to a binding mixture (22 µl) containing 180 µg/ml BSA, 90 µg/ml poly(dI-dC), 12 mM HEPES (pH 7.9), 12% glycerol, 0.12 mM EDTA, 0.9 mM MgCl2, 0.6 mM DTT, 0.6 mM PMSF, 120 mM KCl, and 1.2 µg/ml aprotinin and leupeptin (35). This binding mixture was incubated with or without 3 µl of STAT 5b antiserum (Santa Cruz Biotechnology, Inc.; catalog no. sc 835), which recognizes both STAT 5a and 5b, for 1 h on ice and for a further 30 min after addition of 0.5–1 ng of labeled probe. Samples were then electrophoresed on 6% polyacrylamide, 1x TBE (90 mM Trisborate-20 mM EDTA) gels at 4 C. Gels were then dried and exposed to Fuji Photo Film Co., Ltd. (Tokyo, Japan) RX film at -70 C. Equal loading of the nuclear extracts on these gels was confirmed by a parallel run with an octamer-binding protein-1 probe as previously described by Clarkson et al. (35).
Immunohistochemistry
For localization of SOCS-3 and CIS immunoreactivity in ovary and
adrenal gland, 5-µm paraffin sections were dewaxed and rehydrated,
pretreated with 3% H2O2,
and then blocked with 10% normal rabbit serum. Sections were incubated
with goat anti-CIS (4 µg/ml; Santa Cruz Biotechnology, Inc.; cat sc 1529), goat anti-SOCS-1 (4 µg/ml; Santa Cruz Biotechnology, Inc.; cat sc 7006), or goat anti-SOCS-3 (4
µg/ml; Santa Cruz Biotechnology, Inc.; cat sc 7009)
overnight at 4 C. After washing, sections were incubated with
biotin-labeled rabbit antigoat IgG (Zymed Laboratories, Inc., San Francisco, CA; cat 50-232) and horseradish peroxidase
(HRP)-streptavidin (Zymed Laboratories, Inc.; cat
85–9943, with dilutions as recommended by manufacturer) for 1 h
at 20 C. Immunoreactivity was localized with 3-amino-9-ethylcarbazole
substrate (Zymed Laboratories, Inc.; cat 85-9943). For
negative controls, primary antibodies preincubated for 2 h at 20 C
with SOCS-3 and CIS blocking peptides (20 µg/ml; Santa Cruz Biotechnology, Inc.; cat sc 7009p and 1529p, respectively) were
used. Rabbit antibodies to CIS (a gift of Prof. A. Yoshimura; dilution,
1:500) and SOCS-3 (dilution, 1:50) were used to confirm the
immunoreactive distribution found with the goat antibody, with in this
case normal rabbit serum as the primary antibody control (see
above).
Mammary gland paraffin sections was dewaxed, rehydrated, and pretreated with H2O2 as described above. Sections were blocked with 10% normal horse serum followed by 1:20 rabbit anti-CIS antiserum (a gift from A. Yoshimura), 1:20 rabbit anti-SOCS antiserum, or 1:20 nonimmune rabbit serum (as negative control) overnight at 4 C. Thereafter, sections were incubated with 1:200 biotinylated donkey antirabbit IgG (cat RPN1004; Amersham Pharmacia Biotech, Buckinghamshire, UK) for 2 h at room temperature and followed by 1:200 streptavidin-biotinylated HRP complex (cat RPN 1051; Amersham Pharmacia Biotech, Buckinghamshire, UK) for 2 h at room temperature. The signal was visualized by incubation with DAB substrate for 5 min. Sections were counterstained with hematoxylin and dehydrated before mounting.
Western blotting
This was undertaken using the protocol described previously
(14) with minor modifications designed to allow the use of
cytosol directly. Frozen ovary was homogenized in 5 volumes of cytosol
buffer [40 mM Tris-HCl, 5 mM EGTA, with
complete protease inhibitor cocktail (Roche Molecular Biochemicals, Mannheim, Germany; cat 1697498) pH 7.6]
and then centrifuged at 15,000 rpm in a microfuge for 15 min at 4 C. To
20 µl of cytosol was added 10 µl of 3x Laemmli sample buffer, and
the solution was boiled for 3 min before loading on a 12%
(3.3%C) (cross-link) Laemmli gel (14). After
semidry transfer in 20% methanol in Laemmli electrophoresis buffer,
membranes were blocked with 2% BSA (ICN, Costa Mesa, CA; cat 160069)
in 0.1% Tween in Tris-buffered saline (14). The blots
were incubated overnight at 4 C with goat anti-SOCS-3 and anti-CIS at
1:250. HRP-labeled antirabbit and antigoat antisera were then used for
visualization by enhanced chemiluminescence.
For mammary gland, samples from 48-h bromocriptine-treated rats and from lactating mothers with or without pup withdrawal for 16 h were prepared as described above except that tissues were homogenized in RIPA buffer (150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris, pH 7.5) with complete protease inhibitor cocktail (Roche Molecular Biochemicals; cat 1697498). Lysates were then boiled in Laemmli sample buffer before being centrifuged. DNA in the supernatant was sheared by passing through a 25-gauge needle 10 times before storage. One hundred micrograms of lysate protein (approximately 2 µl of lysate) was then Western blotted as described above. After transfer, nitrocellulose membranes were blocked with Tris-buffered saline (pH 8) containing 5% teleostean gelatin (Sigma; cat G 7765), 0.1% Tween 20 and probed with goat anti-SOCS-3 antibody (1:250; Santa Cruz Biotechnology, Inc.) at 4 C overnight. Thereafter, membranes were incubated with HRP-conjugated rabbit antigoat antibody (Pierce Chemical Co., Rockford, IL) at 1:10,000 for 1 h at room temperature followed by development with Western blot Renaissance chemiluminescence reagent plus (New England Nuclear Life Science Products, Boston, MA).
Cell transfection
Monkey kidney COS-1 cells were maintained in DMEM supplemented
with 10% Serum Supreme and 5 µg/ml gentamycin. The cells were seeded
onto six-well plates, cultured to 40% confluence, and then serum
starved overnight in GC3 medium [Ham’s F12:DMEM (1:1) containing 1x
insulin-transferrin-sodium selenite media supplement
(Sigma), 1x nonessential amino acids (Life Technologies, Inc.), and 5 µg/ml gentamycin]. For each well,
90 µl of transfection reagent/plasmid mixture containing 2 µg of
rabbit PRL receptor expression plasmid, 1 µg of ß-lactoglobulin CAT
reporter, 0.4 µg of Stat 5a expression plasmid, 0.2 µg of
ß-galactosidase reporter, 0.5 µg of SOCS-1, SOCS-2, SOCS-3, or CIS
expression plasmid, and 20 µl of
(N-[1-(2,3-dioleoyl)propyl]-N,N,N-trimethylammonium
methyl-sulfate liposomal transfection reagent (Roche Molecular Biochemicals) in HBS (21 mM HEPES, 137
mM NaCl, 5 mM KCl, 0.7 mM
Na2HPO4, 5.5 mM
dextrose, pH 7.05) was used (6). The transfection mixture
was mixed gently and incubated for 15 min at 20 C before adding it to
each well in 2 ml of GC3 medium. After incubation for 7 h, the
medium was replaced and the cells were left for 16 h. Cells were
treated with 500 ng/ml oPRL for 24 h in GC3 medium. The cells were
then harvested for measurement of CAT and ß-galactosidase
activities.
CAT assays
CAT assay was performed as described by Downes et al.
(36). The cells were lysed by freezing and thawing three
times in 0.25 M Tris-HCl, pH 7.8, on a dry
ice/ethanol bath. Aliquots of cell supernatant were incubated with a
mixture of 0.2 mCi of [14C]chloramphenicol, 7
mM acetyl coenzyme A, and 0.25
M Tris-HCl, pH 7.8, at 37 C for 6 h. The
chloramphenicol and its acetylated forms were then extracted with 1 ml
of ethyl acetate and resolved by running the extract on Silica gel TLC
plates in chloroform/methanol (19:1). Plates were analyzed with an
AMBIS ß-scanner. Results were normalized to ß-galactosidase
expression as described (San Diego, CA) (6).
Statistical analysis
Multiple comparisons were undertaken by ANOVA using Tukey’s
post hoc analysis.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). In the case of the adrenal gland,
rapid induction of SOCS-3 and CIS mRNA was seen, with no significant
response for SOCS-2 and SOCS-1 mRNA. The induction of SOCS-3 mRNA was
even more transient than that seen with the ovary (Fig. 2). The liver did not respond
consistently to PRL administration. The mammary gland did not
show induction of any SOCS mRNAs in response to PRL
administration to these bromocriptine-treated, 24-h pup-deprived
mothers. In all tissues, no induction of SOCS mRNA was seen 60 min
after injection of saline, demonstrating that the induction we observed
is not stress related.
|
|
, it can be seen that an
immunoreactive protein band migrating just below FLAG-labeled CIS and
SOCS-3 (expressed in HEK293 cells) increased in intensity 2 h
after PRL injection.
|
), but we were unable to demonstrate
increased immunoreactivity in response to PRL administration. For both
tissues, immunoreactivity was abolished by pretreatment with the
peptide used for immunization. This could not be shown for SOCS-1
immunoreactivity (not shown). In the adrenal gland, immunoreactivity
was present in the cortex, particularly in the zona fasciculata, but
elements within medullary tissue were also immunoreactive. In the
ovary, immunoreactivity was strong in some corpora lutea cells but
absent from follicular granulosa cells. There was weak staining in
stroma and in the endothelium of blood vessels. Other ovarian
components were not immunoreactive. Localization of SOCS-3 and CIS
immunoreactivity in the ovary was confirmed separately with relevant
rabbit antibodies, and these gave clear cellular localization within
luteal cells and a stronger immunoreaction (Fig. 5, e and f). These rabbit antibodies were
able to detect cellular immunoreactive CIS and SOCS-3 in ducts and
alveolar epithelial cells of the lactating mammary gland (Fig. 5, a–c).
|
|
). Again, no induction
of any SOCS mRNA was apparent in the mammary gland, including that for
SOCS-3 (Fig. 6B).
|
shows
strong induction of STAT 5 element binding in response to PRL injection
in both the ovary and mammary gland. The identity of the bandshift
protein appears to be STAT 5, because the band was supershifted with a
STAT 5 antibody but not a STAT 4 antibody. In a previous study, we
showed that STAT 1 and STAT 3 oligonucleotides do not compete for this
band in rat mammary nuclear extracts (33).
|
). This was
associated with a decline in control (no stimulation) levels of SOCS-3
but not of SOCS-1, SOCS-2, or CIS mRNA expression (Fig. 9). Importantly, unstimulated SOCS-3 mRNA
was substantially increased in 24-h-deprived lactating mammary gland
compared with other tissues, whereas the level of SOCS-3 mRNA in
non-bromocriptine-treated rats kept with their litters was very low.
Thus, there is a correlation between unstimulated or basal SOCS-3 mRNA
and the ability of PRL to induce SOCS gene expression in the mammary
gland.
|
|
). Importantly, this figure also
shows that unstimulated/basal SOCS 3 protein is significantly increased
at 16 h compared with that in normal suckling animals.
|
).
|
).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is known that STAT 5a and STAT 5b both bind to multiple upstream regulatory elements in the CIS gene promoter (34, 38). These contribute to the induction of CIS mRNA by erythropoietin (34) and presumably by PRL, because, as we show here, PRL induces STAT 5a binding activity in both ovary and mammary gland in vivo, as also reported elsewhere (33, 39). Because in STAT 5b knockout mice GH is able to induce CIS or SOCS-2 mRNAs in virgin mammary tissue, it is likely that STAT 5a is the key STAT controlling CIS and SOCS-2 transactivation in the rodent mammary gland (40). In the case of the SOCS-3 promoter, which is likewise rapidly transactivated by a number of class 1 cytokines (20), its transcription is known to be regulated by a STAT 3 responsive element proximal to the transcription start site (41), and there is also good evidence that PRL induces STAT 3 activation in the rat ovary (42). SOCS-1 mRNA is also known to be induced by activated STAT 3 (19), which is consistent with SOCS-1 mRNA induction by PRL.
As indicated, the pattern of SOCS member response to PRL differs between rat tissues. In the ovary, all four SOCS mRNAs examined are induced by a high dose of PRL, although with the physiological suckling stimulus, only SOCS-1, SOCS-2, and CIS are induced. In the adrenal gland, where PRL influences steroid metabolism in the cortex, only CIS and SOCS-3 mRNAs are induced by PRL. In the liver, there was no consistent SOCS response to PRL, consistent with the difficulty in showing tyrosine phosphorylation and STAT activation in the rat liver (33). In the mouse liver, Pezet et al. (43) reported strong induction of SOCS-3 mRNA in response to PRL, which may relate to differences in the proportion of short and long form PRL receptors in the rat. In the 48-h PRL-deprived mammary gland (but not in the 24-h-deprived gland), SOCS-1, SOCS-3, plus CIS, but not SOCS-2 mRNAs, are induced. Thus, there is tissue specificity in the SOCS genes that are induced by PRL. In a survey of SOCS gene induction by cytokines, Aman and Leonard (20) concluded that CIS and SOCS-3 are induced by all 14 cytokines studied, whereas the induction of SOCS-1 and, to a lesser extent, SOCS-2 is more cytokine specific. Interestingly, the cytokine most closely related to PRL, GH, strongly induces SOCS-3 mRNA but induces other SOCS members only weakly. In mouse liver, as with PRL, SOCS-3 and CIS were the only SOCS members induced by GH (28). The selection of SOCS genes induced by PRL could be seen as an integrated response able to totally or selectively attenuate elements of the PRL signaling process. In the absence of significant SOCS-1 induction, it is likely that JAK 2 signaling would retain significant activity, whereas STAT 5 activation, and probably other signaling pathways dependent on PRL receptor association through SH2 domain interaction, would be selectively impaired by high CIS and SOCS-3 expression, because these presumably associate with the phosphorylated receptor in the same manner as with the IL-3, erythropoietin, and GH receptors (15, 22, 26).
Of particular interest in this study is the correlation between the unstimulated (control) level of SOCS-3 expression and the ability of the lactating mammary gland to respond to PRL at the level of gene induction. The gland was not able to respond to PRL (administered or suckling induced) 24 h after pup withdrawal, yet both the ovary and the adrenal gland were fully responsive, as indicated by the ability to induce SOCS genes. However, by 48 h after pup removal, the PRL response had returned, coincident with a decrease in unstimulated SOCS-3 level toward the baseline level seen in suckling mothers. We were also unable to obtain an induction of mammary acetyl coenzyme A carboxylase by PRL in 24-h pup-deprived mothers (not shown), although this is the most in vivo PRL-responsive gene in the rat mammary gland (44). Hormone resistance consequent to increased SOCS-3 has been reported for leukemia inhibitory factor-induced desensitization of the pituitary corticotroph (41), for hypothalamic resistance to leptin (29), for endotoxin-induced GH resistance in the liver (45), and for desensitization to GH-dependent induction of the insulin-like binding protein-3 acid-labile subunit after IL-1ß treatment (46). In the case of the mammary gland, the cause of the increased SOCS-3 could be an increase in active STAT-3, which is known to occur within 12 h of cessation of suckling, with a concomitant decrease in active STAT 5a (47). As indicated above, the SOCS-3 gene possesses a critical STAT 3 element proximal to the transcription start site (41). Up-regulation of SOCS-3 by STAT 3 as a result of milk stasis would prevent activation and action of STAT 5a by PRL, resulting in the cessation of transcription of new milk protein mRNAs. This would provide an effective feedback loop to prevent overstimulation of milk production in the absence of suckling.
Interestingly, mammary-specific knockout of STAT 3 using the cre-lox (cre recombinance-lox P sites) approach prevented the first phase of involution, during which expression of milk protein genes is down-regulated, active STAT 5a is decreased, active STAT 3 is increased, and apoptosis genes are induced in alveolar epithelium (48). We predict that the activation of SOCS-3 would be absent or much reduced in the mammary glands of such conditional STAT 3 knockout mice and that this would be associated with a sustained sensitivity to PRL. The mechanism for the activation of STAT 3 is not known, but it could be a result of the activation of a tyrosine kinase by hydrostatic pressure within the engorged gland, possibly mediated through extracellular matrix interactions (49, 50) or the accumulation of the feedback inhibitor of lactation, a small protein secreted with the milk and thought to act through receptors on the apical surface of the epithelial cell to inhibit lactation (51). We have found (P. Lau, S.P. Tam, and M.J. Waters, unpublished data) that a number of proteins show markedly increased tyrosine phosphorylation in 24-h pup-deprived, bromocriptine-treated mammary glands compared with those of suckling animals. These proteins are presumably targets of activated tyrosine kinase(s) that could also activate STAT 3. The identity of the responsible kinases is not yet known, but it does not include src or fyn kinases. Whatever the identity of the kinase(s), up-regulation of SOCS-3 provides a basis for the conclusion of Li et al. (47) that "the PRL signaling pathway is closed by local factors during the first stage of involution." Our finding that responsiveness to PRL returns 48 h after pup withdrawal, before second stage involution (lobulo-alveolar breakdown), could fit with either hypothesis, because feedback inhibition of lactation would be decreased after milk production ceases, or alternatively, opening of the tight junctions, which has occurred by this time, would decrease the milk hydrostatic pressure (52). This would allow suckling-induced systemic hormones to prevent second stage involution in teat-sealed mice, as has been observed (47).
One curious feature of our study is the observed ability of
administered PRL to increase activated STAT 5a in 24-h pup-deprived
mothers. This does not support the arguments described above if SOCS-3
is acting at the level of STAT 5a activation. However, we found
complete inhibition of the action of STAT 5a at the promoter level in
transient assays by SOCS-3 and blockade of SOCS mRNA induction in
vivo. One could reconcile these observations by postulating a
direct inhibitory action of SOCS-3 or another regulatory protein (such
as p53) at the promoter level. It is known that p53 is induced within 1
day of litter removal (53, 54) and that p53 is able to
inhibit transactivation by STAT 5 in reporter assays without affecting
the ability of STAT 5 to bind to its response element
(55). However, p53 expression remains increased for
several days after pup removal (53, 54), whereas SOCS 3
declines as PRL responsiveness returns by 48 h after weaning.
Currently, we are unaware of data supporting the hypothesis that SOCS 3
has a direct nuclear action, but neither are there data refuting this
view, and this question is being actively investigated in our
laboratory. The situation in the lactating rat evidently differs from
that seen in the rabbit, in which we previously showed PRL-dependent
refractoriness to PRL-induced tyrosine phosphorylation of its receptor
in vivo (14). This refractoriness could be a
result of the presence of PRL-induced SOCS, which would be appropriate
in a species that suckles at more infrequent intervals than the rat. A
model that accounts for the role of SOCS 3 in the regulation of mammary
gland sensitivity to PRL is shown in Fig. 13.
|
The role of SOCS-2 remains enigmatic, and it was only induced in the
ovary. Although it is induced by PRL, it does not inhibit PRL
activation of the lactoglobulin promoter, and its overexpression
actually increased GH transactivation of the serine protease inhibitor
2.1 gene (28). SOCS-2 does not inhibit
interferon--mediated antiviral and antiproliferative actions,
whereas SOCS-1 and SOCS-3 do (58), and it does not inhibit
leukemia inhibitory factor-mediated differentiation and growth arrest,
whereas SOCS-3 does (59). It has been proposed that the
delayed induction of SOCS-2 may attenuate the inhibition of signaling
by SOCS-1, because this SOCS is evidently able to overcome inhibition
by SOCS-1 (43). On the other hand, the SOCS-2 knockout
mouse shows increased growth postnatally, apparently with increased
activity of the GH axis (32). The relationship of this
finding to the report that SOCS-2 binds to the IGF-1 receptor in
vivo (60) remains to be established.
In summary, we have presented evidence that SOCS genes are induced by PRL in a tissue-specific manner and that SOCS-1, SOCS-3, and CIS, but not SOCS-2, are capable of abrogating PRL-dependent transactivation of a STAT 5-responsive gene in vivo. Although it appears that SOCS gene induction acts in a feedback manner to regulate the sensitivity of individual tissues to PRL secretory pulses, it is the chronic up-regulation of SOCS-3 by other means, as observed in the nonsuckling mammary gland, that may be of most physiological importance in the regulation of tissue sensitivity to PRL.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received April 11, 2001.
Accepted for publication July 10, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. C. do Amaral, E. E. Connor, S. Tao, J. Hayen, J. Bubolz, and G. E. Dahl Heat-stress abatement during the dry period: Does cooling improve transition into lactation? J Dairy Sci, December 1, 2009; 92(12): 5988 - 5999. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M Arendt, L. C Evans, D. E Rugowski, M. J. Garcia-Barchino, H. Rui, and L. A Schuler Ovarian hormones are not required for PRL-induced mammary tumorigenesis, but estrogen enhances neoplastic processes J. Endocrinol., October 1, 2009; 203(1): 99 - 110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. T Anderson, N. N M Isa, J. L Barclay, M. J Waters, and J. D Curlewis Maximal expression of suppressors of cytokine signaling in the rat ovary occurs in late pregnancy Reproduction, September 1, 2009; 138(3): 537 - 544. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L Hadsell, A. F Parlow, D. Torres, J. George, and W. Olea Enhancement of maternal lactation performance during prolonged lactation in the mouse by mouse GH and long-R3-IGF-I is linked to changes in mammary signaling and gene expression J. Endocrinol., July 1, 2008; 198(1): 61 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. M. Anderson, D. C. Kieser, F. J. Steyn, and D. R. Grattan Hypothalamic Prolactin Receptor Messenger Ribonucleic Acid Levels, Prolactin Signaling, and Hyperprolactinemic Inhibition of Pulsatile Luteinizing Hormone Secretion Are Dependent on Estradiol Endocrinology, April 1, 2008; 149(4): 1562 - 1570. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. H. Wall and T. B. McFadden Use it or lose it: Enhancing milk production efficiency by frequent milking of dairy cows J Anim Sci, March 1, 2008; 86(13_suppl): 27 - 36. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. E. Dahl Effects of short day photoperiod on prolactin signaling in dry cows: A common mechanism among tissues and environments? J Anim Sci, March 1, 2008; 86(13_suppl): 10 - 14. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Barclay, S. T. Anderson, M. J. Waters, and J. D. Curlewis Characterization of the SOCS3 Promoter Response to Prostaglandin E2 in T47D Cells Mol. Endocrinol., October 1, 2007; 21(10): 2516 - 2528. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. H. Wall, H. M. Crawford, S. E. Ellis, G. E. Dahl, and T. B. McFadden Mammary Response to Exogenous Prolactin or Frequent Milking During Early Lactation in Dairy Cows J Dairy Sci, December 1, 2006; 89(12): 4640 - 4648. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gentili, J. S. Schwartz, M. J. Waters, and I. C. McMillen Prolactin and the expression of suppressor of cytokine signaling-3 in the sheep adrenal gland before birth Am J Physiol Regulatory Integrative Comp Physiol, November 1, 2006; 291(5): R1399 - R1405. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Anderson, D. R. Grattan, W. van den Ancker, and R. S. Bridges Reproductive Experience Increases Prolactin Responsiveness in the Medial Preoptic Area and Arcuate Nucleus of Female Rats Endocrinology, October 1, 2006; 147(10): 4688 - 4694. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Anderson, P. Beijer, A. S. Bang, M. A. Fenwick, S. J. Bunn, and D. R. Grattan Suppression of Prolactin-Induced Signal Transducer and Activator of Transcription 5b Signaling and Induction of Suppressors of Cytokine Signaling Messenger Ribonucleic Acid in the Hypothalamic Arcuate Nucleus of the Rat during Late Pregnancy and Lactation Endocrinology, October 1, 2006; 147(10): 4996 - 5005. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Gentili, M. J. Waters, and I. C. McMillen Differential regulation of suppressor of cytokine signaling-3 in the liver and adipose tissue of the sheep fetus in late gestation Am J Physiol Regulatory Integrative Comp Physiol, April 1, 2006; 290(4): R1044 - R1051. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. E. Clarkson, M. P. Boland, E. A. Kritikou, J. M. Lee, T. C. Freeman, P. G. Tiffen, and C. J. Watson The Genes Induced by Signal Transducer and Activators of Transcription (STAT)3 and STAT5 in Mammary Epithelial Cells Define the Roles of these STATs in Mammary Development Mol. Endocrinol., March 1, 2006; 20(3): 675 - 685. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. T. Anderson, J. L. Barclay, K. J. Fanning, D. H. L. Kusters, M. J. Waters, and J. D. Curlewis Mechanisms Underlying the Diminished Sensitivity to Prolactin Negative Feedback during Lactation: Reduced STAT5 Signaling and Up-Regulation of Cytokine-Inducible SH2 Domain-Containing Protein (CIS) Expression in Tuberoinfundibular Dopaminergic Neurons Endocrinology, March 1, 2006; 147(3): 1195 - 1202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. H. Wall, T. L. Auchtung-Montgomery, G. E. Dahl, and T. B. McFadden Short Communication: Short-Day Photoperiod During the Dry Period Decreases Expression of Suppressors of Cytokine Signaling in Mammary Gland of Dairy Cows J Dairy Sci, September 1, 2005; 88(9): 3145 - 3148. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Curlewis, S. P. Tam, P. Lau, D. H. L. Kusters, J. L. Barclay, S. T. Anderson, and M. J. Waters A Prostaglandin F2{alpha} Analog Induces Suppressors of Cytokine Signaling-3 Expression in the Corpus Luteum of the Pregnant Rat: A Potential New Mechanism in Luteolysis Endocrinology, October 1, 2002; 143(10): 3984 - 3993. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.-Y. Yu-Lee Prolactin Modulation of Immune and Inflammatory Responses Recent Prog. Horm. Res., January 1, 2002; 57(1): 435 - 455. [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 |
