NEUROENDOCRINOLOGY
Runt-Related Transcription Factors Impair Activin
Induction of the Follicle-Stimulating Hormone
␤-Subunit Gene
Kellie M. Breen, Varykina G. Thackray, Djurdjica Coss, and Pamela L. Mellon
Department of Reproductive Medicine and Center for Reproductive Science and Medicine, University of
California, San Diego, La Jolla, California 92093
Synthesis of the FSH ␤-subunit (FSH␤) is critical for normal reproduction in mammals, and its
expression within the pituitary gonadotrope is tightly regulated by activin. Here we show that
Runt-related (RUNX) proteins, transcriptional regulators known to interact with TGF␤ signaling
pathways, suppress activin induction of FSH␤ gene expression. Runx2 is expressed within the
murine pituitary gland and dramatically represses activin-induced FSH␤ promoter activity, without
affecting basal expression in L␤T2 cells, an immortalized mouse gonadotrope cell line. This repressive effect is specific, because RUNX2 induces LH␤ transcription (with or without activin) and
does not interfere with GnRH induction of either gonadotropin ␤-subunit gene. Analysis of the
murine FSH␤ promoter by transfection and gel shift assays reveals that RUNX2 repression localizes
to a Runx-binding element at ⫺159/⫺153, which is adjacent to a previously recognized region
critical for activin induction. Mutation of this ⫺153 activin-response element or, indeed, any of the
five activin-responsive regions prevents activin induction and, in fact, RUNX2 suppression, instead
converting RUNX2 to an activator of the FSH␤ gene. Although the Runx-binding element is required for RUNX2-mediated repression of FSH␤ induction by either activin or Smad3, confirming
a functional role of this novel site, protein interactions in addition to those between RUNX2 and
Smads are necessary to account for full repression of activin induction. In summary, the present
study provides evidence for Runx2-mediated repression of activin-induced FSH␤ gene expression
and reveals the context dependence of Runx2 action in hormonal regulation of the gonadotropin
genes. (Endocrinology 151: 2669 –2680, 2010)
SH is essential for reproductive function in mammals.
Indeed, female mice lacking FSH are infertile due to
defects in proper development and maturation of ovarian
follicles (1). At the molecular level, FSH is a heterodimer
consisting of a common ␣-subunit complexed with a
unique ␤-subunit (2). The ␤-subunit confers biological
specificity, and its synthesis is the rate-limiting step in the
overall production of FSH (2, 3). Transcriptional regulation of FSH␤ occurs via endocrine, paracrine, and autocrine action from a variety of key players including hypothalamic GnRH, gonadal steroid hormones, and the
activin-inhibin-follistatin system (3, 4).
Activin, a member of the TGF␤ superfamily, is an important regulator of FSH synthesis and promotes expres-
F
sion of FSH␤ during the ovulatory cycle (5, 6). The activity
of activin is antagonized by a closely related TGF␤ family
member, inhibin, and both are critical for physiological
regulation of FSH synthesis. Mice lacking either activin or
inhibin exhibit abnormal FSH␤ expression and disrupted
fertility (6, 7), demonstrating the importance of their transcriptional regulation of FSH␤ for overall reproductive
fitness.
Although activin is considered an important regulator
of FSH␤-subunit gene expression, the mechanisms underlying transcriptional responsiveness are complex, requiring many individual elements that act in cooperation, and
exhibit species specificity (8 –11). Several of these activin
response elements have been found to harbor consensus
ISSN Print 0013-7227 ISSN Online 1945-7170
Printed in U.S.A.
Copyright © 2010 by The Endocrine Society
doi: 10.1210/en.2009-0949 Received August 12, 2009. Accepted March 4, 2010.
First Published Online March 31, 2010
Abbreviations: FoxL2, Forkhead transcription factor L2; RUNX, runt-related protein.
Endocrinology, June 2010, 151(6):2669 –2680
endo.endojournals.org
2669
2670
Breen et al.
Runx Repression of FSH␤ Gene Expression
Smad-binding sites, and a few of these sites have been
shown to bind Smad proteins, known mediators of TGF␤
signaling cascades, including activin (8, 11, 12). Of interest, the ⫺153 activin-response element in the mouse
(⫺167 in the ovine gene) is required to respond effectively
to activin (8, 11) but has not been shown to directly bind
Smad proteins or any other known activin-induced mediator. Interestingly, Su et al. (13) demonstrated that disruption of a sequence juxtaposed to this activin-responsive site within the context of the ovine FSH␤ promoter
causes severe dysregulation of basal expression and transcriptional regulation in L␤T2 cells. When this mutant
ovine FSH␤ reporter was introduced into transgenic mice,
the transgene revealed diminished basal expression, improper regulation by activin or follistatin, and failure to
exhibit the secondary FSH surge (13). In silico analysis
indicated that this sequence could represent a putative
binding site for the runt-related (RUNX) family of transcription factors.
The RUNX family of nuclear transcription factors
(RUNX, human; Runx, mouse) was originally identified
in Drosophila as the pair rule gene runt (14). In mammals,
the three Runx family members (Runx1, -2, and -3) all
play critical roles in cell differentiation, tissue development, and ultimately, human disease (15). The gene-regulatory actions of Runx factors are mediated not only by
binding specific promoter regions, using a conserved Runt
domain essential for DNA binding but also through the
formation of protein interactions that assist in the assembly of transcriptional complexes at specific subnuclear
sites (16, 17). These protein-protein interactions potently
influence transactivation or repression by Runx itself and
increase the complexity of the mechanisms of transcriptional regulation by Runx family members.
Given that Runx factors appear to function as scaffolding proteins involved in the integration of complex and
coordinated gene-regulatory mechanisms, including TGF␤
family signaling, we sought to investigate the role for Runx
proteins in the transcriptional regulation of FSH␤. In L␤T2
cells, a model of cultured gonadotrope cells that endogenously expresses FSH␤ and contains the machinery to respond to the activin-follistatin system, RUNX proteins abrogated activin induction of FSH␤ promoter activity. We
further identified a Runx cis-regulatory element at ⫺159 in
the murine FSH␤ promoter juxtaposed to the ⫺153 critical
region for activin responsiveness. This Runx-binding site not
only is necessary for the physical interaction of Runx2 with
the murine FSH␤ promoter. but it is also required for
RUNX2 repression of activin induction. These results clarify
the importance of Runt-related transcription factors in transcriptional regulation within the gonadotrope cell and pro-
Endocrinology, June 2010, 151(6):2669 –2680
vide an important role for Runx2 in feedback control of
reproductive hormone synthesis.
Materials and Methods
Plasmids
FSH␤ reporter plasmids have been previously described:
⫺1000 murine FSH␤luc (18, 19); ⫺985 ovine FSH␤luc (20, 21);
⫺1028 human FSH␤luc (12); and the activin-response element
cis mutations within the ⫺1000 murine FSH␤luc at ⫺267,
⫺153, ⫺120, and the 5X mutation (containing mutations in all
five activin-responsive elements at ⫺267, ⫺153, ⫺139, ⫺120,
and ⫺106) (11). The 1.8-kb rat LH␤luc was kindly provided by
Mark Lawson. The expression vectors for human RUNX1, -2,
and -3 were provided by Yoshiaki Ito (22), and murine Runx2
was provided by Jane Lian (16, 17). Smad3 was provided by Rik
Derynck (23).
Mutations were generated by PCR using specific primers
(Supplemental Table 1 published on The Endocrine Society’s
Journals Online web site at http://endo.endojournals.org) for the
⫺398 FSH␤ promoter (⫺156/⫺154 mutation) or murine Runx2
expression vector (HTY␮ and WRPY␮). Mutagenesis was performed using the QuikChange XL site-directed mutagenesis kit
(Stratagene, La Jolla, CA) and confirmed by sequencing.
Cell culture and transient transfection
L␤T2 cells, cultured as previously described (18), were seeded
into 12-well plates at 3 ⫻ 105 cells per well and incubated overnight at 37 C. Each well was transfected with 400 ng of the
luciferase-reporter plasmid or empty pGL3 vector, 200 ng of the
human RUNX1, -2, or -3 expression vector (vector pEF-BOS),
murine Runx2 (vector pHA) or empty vector, and 100 ng of a
␤-galactosidase reporter gene regulated by the thymidine kinase
promoter (TK-␤gal) as a control for transfection efficiency using
FuGENE 6 transfection reagent (Roche Applied Science, Indianapolis, IN). In experiments using Smad3 to induce FSH␤ promoter activity, cells were also transfected with 200 ng Smad3 or
empty pRK5 vector. Eighteen hours after transfection, cells were
transferred to serum-free DMEM supplemented with 0.1% BSA,
5 mg/liter transferrin, and 50 nM sodium selenite. After 6 h,
activin A (25 ng/ml; Calbiochem, San Diego, CA), follistatin (25
ng/ml; R&D Systems, Minneapolis, MN), or vehicle (0.1% BSA)
was administered for 24 h before harvest. Where indicated,
GnRH (10 ng/ml; Sigma-Aldrich, St. Louis, MO) treatment began 6 h before harvest. Cells were harvested and extracts prepared for assay of luciferase and ␤-galactosidase activity as previously described (24).
Quantitative real-time PCR
Preparation of cDNA from mouse pituitary or L␤T2 cells was
performed as previously described (25). Male and female
C57BL/6 mice (6 wk of age) were purchased from The Jackson
Laboratory (Bar Harbor, ME), and housed in a University of
California, San Diego (UCSD), animal facility under standard
conditions. At 8 wk of age, mice were decapitated and pituitaries
removed for immediate processing. All procedures were approved by the UCSD Institutional Animal Care and Use Committee. Briefly, RNA was extracted with Trizol reagent (Invitrogen/GIBCO, Carlsbad, CA) according to the manufacturer’s
Endocrinology, June 2010, 151(6):2669 –2680
RUNX proteins are present in the pituitary gland
and immortalized gonadotrope cells
Because Runx proteins are known regulators of TGF␤
signaling pathways, including activin, we first identified
the presence of these factors in the mouse pituitary gland
and in model gonadotrope cells. Quantitative RT-PCR
analysis detected total mRNA for the Runx family in both
male and female mouse pituitary tissue as well as in L␤T2
cells, an immortalized pituitary gonadotrope cell line (Fig.
1A). Total transcript level for the Runx family in L␤T2
cells is approximately 2-fold greater than in the pituitary
gland, a heterogeneous tissue that contains less than 10%
gonadotrope cells (28). Protein expression of Runx1, -2,
and -3 is readily detected by Western blotting analysis of
0.5
Total Runx
0.4
FSHβ
0.3
0.2
0.1
Statistics
All experiments were performed in triplicate and were repeated at least three times. To normalize for transfection efficiency, all luciferase values were divided by ␤-galactosidase, and
the triplicate values were averaged. To control for interexperimental variation, the empty pGL3 reporter plasmid was transfected with TK-␤gal and any relevant expression vectors, and the
average pGL3/␤gal value was calculated. Average luc/␤gal values were divided by the corresponding pGL3/␤gal value. Indi-
fe
le
ma
pit
T2
Lβ
αT3-1
LβT2
B
pit
LβT2
le
ma
EMSA
Mouse FSH␤ promoter oligonucleotides from ⫺164 to ⫺134
(Supplemental Table 1) or the corresponding sequences in the
sheep and human were annealed, end-labeled, and purified as
previously described (27). Binding reactions used 2 fmol 32Plabeled oligonucleotide and 8 ␮g L␤T2 nuclear extract. In competitor assays, 250-fold excess unlabeled oligonucleotide was
added to the binding reaction before addition of probe. For supershift assays, 1 ␮g rabbit antimouse Runx2 antibody (Santa
Cruz Biotechnology) or normal rabbit IgG control, was added to
the reaction. Reactions were electrophoresed on a 5% polyacrylamide gel at 250 V for 2 h and then dried under vacuum and
exposed to film.
0
αT3-1
A
LβT2
Nuclear extracts were prepared from ␣T3-1 and L␤T2 cells as
previously described (26). When experiments were conducted
with activin A (25 ng/ml), follistatin (25 ng/ml), or vehicle (0.1%
BSA), treatment began 24 h before harvest. Nuclear extract (30
␮g) was boiled for 5 min in 5⫻ Western loading buffer, fractionated on a 10% SDS-PAGE gel, and electroblotted for 90 min
at 300 mA onto polyvinylidene difluoride (Millipore, Billerica,
MA) in 1⫻ Tris-glycine-sodium dodecyl sulfate/20% methanol.
Blots were blocked overnight at 4 C in 3% BSA and then probed
for 1 h at room temperature with goat antihuman Runx1, rabbit
antimouse Runx2, or rabbit antihuman Runx3 antibody (Santa
Cruz Biotechnology, Santa Cruz, CA) diluted 1:500 in blocking
buffer. Blots were then incubated with a horseradish peroxidaselinked secondary antibody (Santa Cruz Biotechnology) and
bands visualized using the SuperSignal West Pico chemiluminescent substrate (Pierce Biotechnology Inc., Rockford, IL). BioRad prestained protein ladder plus serves as a size marker.
Results
αT3-1
Western blotting analysis
2671
vidual values obtained from each independent experiment were
then averaged, and statistics were performed using JMP 7.0
(SAS, Cary, NC). Significance was established as P ⬍ 0.05 by
two-way ANOVA followed by Tukey’s post hoc test.
Transcript Relative to GAPDH
instructions, treated to remove contaminating DNA (DNA-free;
Ambion, Austin, TX), and reverse transcribed using Superscript
III first-strand synthesis system (Invitrogen). Quantitative realtime PCR was performed in an iQ5 real-time PCR instrument
(Bio-Rad, Hercules, CA) and used iQ SYBR Green Supermix
(Bio-Rad) with specific primers for GAPDH, FSH␤, Runx2, and
total Runx (Supplemental Table 1).
The iQ5 real-time PCR program was as follows: 95 C for 15
min, followed by 40 cycles at 95 C for 15 sec, 55 C for 30 sec, and
72 C for 30 sec. Within each experiment, the amount of FSH␤,
Runx2 or total Runx, and GAPDH was calculated by comparing
a threshold cycle obtained for each sample with the standard
curve generated from serial dilutions of a plasmid containing
GAPDH, ranging from 1 ng to 1 fg. Values for FSH␤ and Runx2
or total Runx were determined from the same sample and are
expressed relative to GAPDH. All samples were assayed (in triplicate) within the same run, and the experiment was conducted
two times.
endo.endojournals.org
82 kD
64 kD
48 kD
Runx1
Runx2
Runx3
FIG. 1. Runx proteins are expressed in pituitary tissue and
immortalized gonadotrope cells. A, Quantitative RT-PCR analysis of
total Runx and FSH␤ mRNA from male and female mouse pituitary (pit)
tissue or L␤T2 cells. In each sample, the amount of total Runx or FSH␤
mRNA was compared with the amount of GAPDH mRNA and results
are expressed as relative transcript level. Results represent the mean ⫾
SEM of two independent experiments, each performed in triplicate. B,
Western blotting analysis of nuclear extracts from ␣T3-1 and L␤T2 cells
was performed using antibodies for Runx1, -2, and -3. Protein bands
were detected at the expected sizes of 53, 55, and 44 kDa for Runx1,
-2, and -3, respectively, in both ␣T3-1 and L␤T2 cells. An additional
higher molecular weight band in the Runx3 Western blot likely
represents posttranslationally modified Runx3 (15). The experiment
was repeated three times with similar results, and representative gels
are shown.
Runx Repression of FSH␤ Gene Expression
Differential regulation of FSH␤ and LH␤ by RUNX2
Appropriate expression of the gonadotropin subunit
genes is dependent on paracrine and autocrine actions
within the anterior pituitary. To determine whether the
effect of RUNX2 is dependent on hormonal milieu, L␤T2
cells were cultured in the presence of activin, follistatin, or
GnRH, and the effect of RUNX2 overexpression was assessed on FSH␤ promoter activity. Again, RUNX2 potently reduced activin-induced FSH␤ expression (Fig. 3A),
whereas it failed to alter FSH␤ promoter activity in cells
treated with follistatin, indicating that RUNX2 repression
requires a threshold level of promoter activation. Moreover, this effect does not occur with any hormonal induction because GnRH regulation, although it results in a
2.7-fold induction of FSH␤ expression, is not affected by
B
14
8
6
4
2
0
#
*
10
#
8
**
#
***
#
6
4
2
0
C
RUNX2
RUNX3
10
12
RUNX1
12
Empty Vector
Fold Induction by Activin
14
RUNX1
RUNX2
RUNX3
D
0.3
0.2
+ Follistatin
Runx2
FSHβ
0.4
+ Activin
Ab: α-Runx2
0.5
Control
RUNX proteins regulate FSH␤ gene expression in
gonadotrope cells
Because activin is a potent regulator of FSH␤ synthesis,
we sought to test whether Runx proteins are transcriptional effectors of FSH␤. The murine FSH␤ promoter
(⫺1000 bp) fused upstream of a luciferase reporter gene
(FSH␤luc) was transiently cotransfected along with human RUNX1, -2, or -3 expression vectors into L␤T2 cells.
Overexpression of RUNX1, -2, or -3 did not significantly
alter basal expression of the 1-kb murine FSH␤ promoter
(Fig. 2A). In contrast, all three RUNX proteins blunted the
robust induction by activin (Fig. 2B). Specifically, treatment with activin resulted in an 11-fold induction of
FSH␤, which was reduced by 42, 48, or 23%, byRUNX1,
RUNX2, or RUNX3, respectively. Although all three
mammalian Runx proteins are present in L␤T2 cells and
are capable of transcriptional repression of activin induction of the FSH␤ promoter, we focused on Runx2 based on
the intensity of its effect and evidence for its interaction
with members of TGF␤ signaling cascades (29, 30). We
first confirmed the presence of Runx2 within the murine
pituitary gland (Fig. 2C) and determined whether protein
expression is changed by treatment with activin or follistatin (Fig. 2D), a potent activin-binding protein that
neutralizes endogenous activin secreted by the L␤T2 cell
itself (31). Levels of Runx2 protein are similar in L␤T2
cells treated with vehicle, activin, and follistatin. Collectively, these data identify the Runx family as potential
regulators of gonadotrope function and focus our attention on the molecular mechanism whereby RUNX2 potently represses FSH␤ expression.
A
Empty Vector
nuclear extracts from both L␤T2 cells and ␣T3-1 cells, a
gonadotrope precursor cell line (Fig. 1B), revealing a cell
model for understanding the role of this novel family of transcription factors in gonadotrope subunit gene expression.
Endocrinology, June 2010, 151(6):2669 –2680
Fold Change
Breen et al.
Transcript Relative to GAPDH
2672
0.1
0
ma
le
pit
a
fem
le
pit
T2
Lβ
FIG. 2. RUNX proteins repress activin-induced FSH␤ gene expression.
A and B, Effect of overexpression of RUNX1, -2, or -3 on FSH␤
promoter activity. The 1-kb murine FSH␤luc reporter plasmid was
transiently cotransfected into L␤T2 cells with an expression vector for
human RUNX1, -2, or -3 or pEF-BOS as an empty vector control. Cells
were treated for 24 h with vehicle (A) or activin (B) and harvested for
luciferase activity as a measure of FSH␤ promoter activity. Results
represent the mean ⫾ SEM and are depicted as FSH␤ activity relative to
the vehicle empty vector control. #, Significant effect of activin vs.
vehicle empty vector control; *, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍
0.001, significant suppression by RUNX expression plasmid vs. empty
vector expression plasmid in the presence of activin as determined by
two-way ANOVA followed by Tukey’s post hoc test. C, Quantitative
RT-PCR analysis of Runx2 and FSH␤ mRNA extracted from male and
female mouse pituitary (pit) tissue or L␤T2 cells. D, Western blotting
analysis of Runx2 in nuclear extracts from L␤T2 cells treated with
activin, follistatin, or vehicle (control), using an antibody specific for
Runx2 (Ab: ␣-Runx2). The experiment was repeated three times with
similar results, and a representative gel is shown.
RUNX2, indicating that repression by RUNX2 is specific
for induction by activin.
Previously, we reported that, in addition to its affect on
FSH␤, activin signaling induces LH␤ expression, albeit to
a lesser extent (32). To further address the specificity of
RUNX2 action on gonadotropin subunit gene expression,
we tested the effect of RUNX2 on LH␤ promoter activity.
RUNX2 was transiently cotransfected along with ⫺1800
bp of the rat LH␤ promoter fused to a luciferase reporter
(LH␤luc) into L␤T2 cells. Upon treatment with activin
alone, LH␤ expression was induced 1.6-fold, whereas
Endocrinology, June 2010, 151(6):2669 –2680
endo.endojournals.org
A
FSHβ Fold Induction
12
FSHβluc
#
*
Empty Vector
10
RUNX2
8
#
6
4
#
#
2
0
Vehicle
Activin Follistatin GnRH
B
LHβ Fold Induction
4
*#
LHβluc
3
2
#
*
#
#
*
1
0
Vehicle
Activin Follistatin GnRH
FIG. 3. RUNX2 repression of activin induction is specific for FSH␤. The
1-kb murine FSH␤luc reporter gene (A) or 1.8-kb rat LH␤luc reporter
gene (B) was transfected into L␤T2 cells along with RUNX2 or its
empty vector. Cells were treated with vehicle, activin, follistatin, or
GnRH and harvested for luciferase activity to determine the effects of
RUNX2 on hormone induction of both LH␤ and FSH␤. Results are
depicted as fold induction by hormone treatment relative to the vehicle
empty vector control for each reporter plasmid. #, Significant induction
by hormone treatment vs. vehicle empty vector control; *, significant
effect of RUNX2 vs. empty vector on hormone-induced FSH␤/LH␤
expression.
GnRH induced LH␤ by 2.0-fold (Fig. 3B). In contrast to
its effect on FSH␤, RUNX2 induced LH␤ expression
nearly 2-fold, and this increase occurred in the presence of
vehicle, activin, or follistatin. Interestingly, RUNX2 does
not appear to further induce LH␤ in the presence of
GnRH. Taken together, these experiments demonstrate
that the repressive action of RUNX2 is specific to FSH␤
and dependent on elevated levels of circulating activin.
RUNX2 repression localizes to an
activin-responsive region of the FSH␤ promoter
We examined the proximal promoter region of the murine FSH␤ gene and identified three possible binding sites
for Runx transcription factors within 1 kb upstream of the
transcription start site [⫺652, ⫺455, and ⫺159; ⬎85%
identity each, by web-based software TFSEARCH (33)].
To identify regions of the FSH␤ gene that are functionally
involved in RUNX2 regulation, L␤T2 cells were transiently
transfected with a series of truncated FSH␤ reporter plasmids, ranging in length from ⫺1000 to ⫺95 bp of the 5⬘
2673
regulatory sequence. Figure 4 illustrates the effects of cotransfection of RUNX2 on the progressive 5⬘-promoter truncations on basal (Fig. 4A) or activin-induced FSH␤ promoter
activity (Fig. 4B). Although RUNX2 causes no change in
basal expression of the 1-kb FSH␤ promoter (Fig. 4A,
⫺1000), expression is increased by approximately 30%
when the reporter gene is truncated to ⫺230 or ⫺194. Further deletion of the region to ⫺127 results in a loss of
RUNX2 induction, indicating that elements important for
the effect of RUNX2 on basal FSH␤ expression reside within
the ⫺194- to ⫺127-bp region of the gene.
As observed previously (25), activin induction of the
FSH␤ gene declined incrementally as the promoter was
progressively truncated from ⫺1000 to ⫺95 (Fig. 4B).
RUNX2 repressed FSH␤ promoter activity by approximately 40% when the region contained at least ⫺304 bp
of the proximal promoter. Interestingly, further truncation of the region from ⫺304 to ⫺230, removing the most
5⬘ activin response element at ⫺267 (34), results in a loss
of activin induction and elimination of RUNX2 repression, indicating that activin responsiveness of the FSH␤
gene is required for RUNX2 repression.
The ⫺267 site is one of five important cis-regulatory
elements (Fig. 4C), each of which is critical for activin
responsiveness of the murine FSH␤ gene (8, 11, 34). These
bind proteins such as Smad family members (11, 34), Pbx1
and Prep1 homeodomain proteins (8, 11), or forkhead
transcription factor L2 (FoxL2) (35). We therefore analyzed the necessity of the known activin response elements
within the FSH␤ gene for RUNX2 repression (11). Remarkably, cis mutation of individual elements or combined mutation of all five sites in the 1-kb FSH␤ promoter
allowed RUNX2 to induce expression in the absence of
activin (Fig. 4D), similar to its effect on basal expression
of the promoter truncations (⫺230 and ⫺194; Fig. 4A). As
in the case of the ⫺230 truncation that lost activin responsiveness concurrent with RUNX2 repression, mutation of single elements in the FSH␤ gene prevents activin
induction and, thus, RUNX2 repression in the presence
of activin (Fig. 4E). In fact, mutation of all five activin
response elements actually converts the repression by
RUNX2 to strong induction. Taken together, these data
indicate that the direction of RUNX2 activation or repression is dependent on representation, availability, and coordinated binding of coregulatory factors, including those
involved in activin signaling.
Runx2 binds a novel Runx element within the
FSH␤ proximal promoter
Because we found that each mutation or truncation that
prevents activin induction also prevents RUNX2 repression, we focused on the region of the FSH␤ gene required
2674
Runx Repression of FSH␤ Gene Expression
Breen et al.
A
C
*
Empty Vector
3
Fold Change
murine FSHβ 5’ regulatory region
-267
4
Smad2/3/4
?
TGTGGCA
*
D 40
Fold Change
1
0
-139
-159 -153
-120 -106
Runx
Basal
RUNX2
2
Endocrinology, June 2010, 151(6):2669 –2680
?
Pbx1 FoxL2
Prep1
Smad4
*
Basal
*
30
*
20
*
10
8
6
4
2
*
E
*
Activin
*
Fold Induction by Activin
Fold Induction by Activin
10
12
9
20
5X
m
m
ut
ut
ut
m
53
m
-1
-1
-2
67
B
ut
W
T
0
*
Activin
*
*
6
*
competes complex i and iii but fails to compete complex ii (lane 3). Inclusion of control
IgG (lane 4) does not alter protein binding;
however, a Runx2-specific antibody results
in a supershift of complex ii (lane 5), identifying this complex as containing Runx2.
We next determined the nucleotides necessary for Runx2 interaction with DNA.
EMSA was performed with the wild-type
⫺164/⫺134 probe (Fig. 5B, WT) and scanning 2-bp mutant oligonucleotides as unlabeled competitors (Fig. 5B, A–M). The three
mutant oligonucleotides that were unable to
compete for Runx2 binding, mutations C–E
(Fig. 5C, lanes 5–7), together encompass six
of the seven nucleotides within the putative
Runx site identified at ⫺159/⫺153.
3
RUNX2 repression is conserved across
species
Elements conveying responsiveness to activin
are well conserved across species, and a
FIG. 4. RUNX2 suppression of the murine FSH␤ promoter maps to an activin-responsive
recent
report (13), combined with the curregion. A and B, L␤T2 cells were transfected with a series of 5⬘-truncated FSH␤luc
reporter plasmids along with RUNX2 or its vector control and treated with vehicle (A) or
rent study, implicates the Runx family of
activin (B) to determine regions of the FSH␤ promoter that are responsive to RUNX2.
transcription factors in the regulation of acResults are depicted as FSH␤ fold induction relative to the empty vector ⫺1000 FSH␤luc
tivin responsiveness in the sheep and mouse.
reporter plasmid. *, Significant effect of RUNX2 vs. empty vector on FSH␤ promoter
activity. C, Schematic of the murine FSH␤ 5⬘ regulatory region illustrating the known
We tested the hypothesis that RUNX regactivin-responsive elements (open circles) involved in expression of the FSH␤ gene.
ulation is a conserved mechanism of reProteins binding each site are indicated or, if unknown, are denoted with a question
pression among mouse, sheep, and humans
mark. Juxtaposed to the ⫺153 activin response element is a Runx-binding site ⫺159
by transiently transfecting RUNX2 into
TGTGGCA ⫺153. D and E, L␤T2 cells were transfected with FSH␤luc reporter plasmids
containing mutations (mut) in known activin response elements (11), along with RUNX2
L␤T2 cells along with either a ⫺985 ovine
or its vector control, treated with vehicle (D) or activin (E), and analyzed for FSH␤
FSH␤luc or ⫺1028 human FSH␤luc, and
promoter activity as described above. The FSH␤luc reporter plasmids either contained
compared repression with the ⫺1000 muelements mutated individually (⫺267 mut, ⫺153 mut, and ⫺120 mut) or all five
elements mutated in combination (5X mut; containing cis mutations at ⫺267, ⫺153,
rine FSH␤luc reporter (Fig. 6, A and B). The
⫺139, ⫺120, and ⫺106).
effect of RUNX2 on basal and activin-induced FSH␤ promoter activity is similar for
for RUNX2 activation in the absence of activin. In silico
all three species. Specifically, ovine, human, and murine
analysis of this region (⫺194/⫺127) identified a putative
basal FSH␤ promoter activity was not significantly altered
Runx-binding site based on homology to a consensus recby RUNX2 (Fig. 6A). In contrast, responsiveness to acognition sequence 5⬘-PyGPyGGTPy-3⬘ (36). This putative
tivin was reduced by 53% in the sheep, 39% in the human,
site, ⫺159 TGTGGCA ⫺153, is positioned immediately
and 47% in the mouse (Fig. 6B), demonstrating that reupstream of the activin response element at ⫺153 ATTTAGAC ⫺146 (Fig. 4C). EMSAs were performed to test pression by RUNX2 is functionally conserved in the ovine,
the hypothesis that Runx2 could physically interact at this human and murine FSH␤ genes.
DNA sequences of the ovine and human FSH␤ genes
site. When an oligonucleotide encompassing the ⫺164/
were
compared with that of the murine ⫺159 Runx-bind⫺134 region of the gene was used in EMSA with nuclear
ing
site
(Fig. 6C, murine, underlined) and investigated for
extracts from L␤T2 cells, specific complexes were observed (Fig. 5A, lane 1, labeled i, ii, and iii). Unlabeled conservation. Although this murine Runx site shows minwild-type oligonucleotide successfully competes with the imal conservation (⬍50% conserved) with the species inlabeled probe for protein binding of all three complexes vestigated, another site was recently reported as critical for
(lane 2), whereas an oligonucleotide containing a muta- activin responsiveness of the ovine FSH␤ gene in vivo and
tion of the 7-bp putative Runx2 consensus site (⫺159 proposed as a putative Runx-binding site based on its seAAAAAAA ⫺153; underlined in Fig. 5B) successfully quence (13). This Runx site sits 3 bp downstream of the
m
5X
m
20
-1
ut
ut
ut
53
-1
67
m
m
ut
W
T
0
-2
0
endo.endojournals.org
B
12
Empty Vector
RUNX2
Fold Change
10
Runx2 ss
i
8
6
4
2
ii
0
Runx2
ine
Ov
iii
Hu
ma
2675
*
12
Fold Induction by Activin
A
Runx2 Ab
IgG
Probe
250x WT
A
250x MUT
Endocrinology, June 2010, 151(6):2669 –2680
#
10
8
*
#
6
4
#
#
*
#
2
0
n ine
r
Mu
ine
Ov
Hu
ma
n ine
r
Mu
C
1
B
2
3
4
-164
WT:
A:
B:
C:
D:
E:
F:
G:
H:
I:
J:
K:
L:
M:
5
-134
TGCTCTGTGGCATTTAGACTGCTTTGGCGAG
TAATCTGTGGCATTTAGACTGCTTTGGCGAG
TGCAATGTGGCATTTAGACTGCTTTGGCGAG
TGCTCAATGGCATTTAGACTGCTTTGGCGAG
TGCTCTGAAGCATTTAGACTGCTTTGGCGAG
TGCTCTGTGAAATTTAGACTGCTTTGGCGAG
TGCTCTGTGGCGGTTAGACTGCTTTGGCGAG
TGCTCTGTGGCATGGAGACTGCTTTGGCGAG
TGCTCTGTGGCATTTCCACTGCTTTGGCGAG
TGCTCTGTGGCATTTAGGGTGCTTTGGCGAG
TGCTCTGTGGCATTTAGACCCCTTTGGCGAG
TGCTCTGTGGCATTTAGACTGGGTTGGCGAG
TGCTCTGTGGCATTTAGACTGCTGGGGCGAG
TGCTCTGTGGCATTTAGACTGCTTTAACGAG
C
WT A B C D E F G H I J K L M
i
Runx2
iii
1
2
3
4
5
6 7
8
9 10 11 12 13 14 15
* **
FIG. 5. Runx2 binds the ⫺164/⫺134 region of the murine FSH␤
promoter. A, EMSA was performed using L␤T2 nuclear extract and a
radiolabeled oligonucleotide probe containing the putative Runx2
binding site identified at ⫺159 bp of the FSH␤ promoter to test for
complex formation (lane 1, Probe; sequence in B, WT with Runx site
underlined). A 250-fold excess of unlabeled wild-type probe (lane 2,
250⫻ WT) or mutant probe (lane 3, 250⫻ MUT) or nonspecific IgG
(lane 4) were included in the binding reactions as indicated. The
addition of an antibody specific for Runx2 (lane 5, Runx2 Ab) resulted
in reduction of the Runx2 band and formed an antibody supershift as
indicated by Runx2 ss. B, An alignment of the wild-type murine FSH␤
promoter sequence (⫺164/⫺134; WT), and the oligonucleotides used
as competitors, labeled A–M, are shown. The scanning 2-bp mutations
introduced are underlined. C, EMSA was performed by using L␤T2
nuclear extracts incubated with labeled wild-type probe (lane 1) along
with 250-fold excess of the indicated wild-type (lane 2, WT) or mutant
competitor (lanes 3–15, A–M).
murine ⫺159 Runx site and is highly conserved between
human and ovine (Fig. 6C, ovine, dashed underline),
providing circumstantial evidence that the region encompassing the ⫺153 activin response element (Fig. 6C,
illustrated on murine promoter, gray shading) harbors
a potential element for Runx modulation of activin responsiveness across multiple species.
Ovine -178 TGATCTACTGCATTTAGACTGCTTTGGCGAG -148
Human -175 TAATCTACTGCGTTTAGACTACTTTAGTAAA -145
Murine -164 TGCTCTGTGGCATTTAGACTGCTTTGGCGAG -134
D
Ovine
Human Murine
LβT2 NE: + + + + + + + + +
IgG:
+
+
+
+
+
+
Runx2 Ab:
Runx2 ss
Runx2
1
2
3
4
5
6
7
8
9
FIG. 6. A and B, Effect of overexpression of RUNX2 on ovine, human,
and murine FSH␤ promoter activity. The ⫺985-bp ovine FSH␤luc,
⫺1028-bp human FSH␤luc, or ⫺1000 murine FSH␤luc reporter
plasmid was transfected into L␤T2 cells with RUNX2 or its vector
control, and FSH␤ promoter activity was assessed in cells treated with
vehicle (A) or activin (B). Results are depicted as FSH␤ fold induction
relative to the empty vector reporter control for each species. #, Significant
effect of activin; *, significant suppression by RUNX in the presence of
activin. C, Sequence comparison of the putative Runx element
identified in the murine FSH␤ gene (⫺159/⫺153, underline) with
corresponding gene regions in the ovine and human. The ⫺153 activin
responsive element (murine, gray shading) and putative ovine Runx site
recently identified (dashed underline) (13) are also highlighted. D,
Nuclear extracts from L␤T2 cells were incubated with the ovine,
human, or murine oligonucleotide probe, and complex formation was
assayed by EMSA. Inclusion of a nonspecific IgG antibody or an
antibody specific for Runx2 is indicated. Addition of an antibody
specific for Runx2 resulted in elimination of the Runx2 band and an
antibody supershift on the murine probe (lane 9, Runx2 ss).
EMSA was performed with species-specific oligonucleotide probes corresponding to the ⫺164/⫺134 FSH␤ region of the mouse to determine whether Runx2 binds this
region in the ovine and human FSH␤ genes as well. L␤T2
nuclear extracts were incubated with probe alone, nonspecific IgG, or a Runx2-specific antibody (Fig. 6D). Although an antibody specific for Runx2 causes a supershift
on the murine probe (Fig. 6D; lane 9, supershifted complex indicated as Runx2 ss), a complex was not detected
on the ovine or human probe that could be identified as
Runx2 or a supershift of Runx2. It is possible that we
could not visualize binding of Runx2 on the sheep and
human probes due to lower-affinity binding than the
mouse. Altogether, our results identify a conserved mechanism of repression of activin induction via RUNX2 but
raise the possibility that RUNX2 uses different sequences
2676
Breen et al.
Runx Repression of FSH␤ Gene Expression
from those identified in the mouse to mediate repression of
the sheep and human FSH␤ promoters.
Integrity of the Runx-binding element is critical for
RUNX2 repression
Once we had determined that Runx2 could bind the
Runx element at ⫺159/⫺153, transient transfection assays were used to determine whether this site plays a functional role in the suppression of activin-induced FSH␤
gene expression by RUNX2. For this experiment, L␤T2
cells were transfected with the ⫺398 FSH␤luc reporter
(wild type) or the same reporter plasmid containing a cis
mutation in which the ⫺156 GGC ⫺154 nucleotides,
known to be critical for high-affinity binding by Runx
proteins to DNA (37) (Fig. 5C), were mutated to AAA
(⫺156/⫺154 mutation). Basal FSH␤ promoter activity
was slightly, but not significantly, elevated by introduction of the Runx site cis mutation compared with the wildtype mouse FSH␤ promoter (Fig. 7A, WT); overexpression of RUNX2 did not alter promoter activity compared
with the empty vector control (Fig. 7A, ⫺156/⫺154 mutation). Activin induced FSH␤ expression of both the wildtype and ⫺156/⫺154 mutant FSH␤ promoters equally
(Fig. 7B). Although RUNX2 suppresses activin induction
of the wild-type promoter (Fig. 7B; WT, ⬃50% reduced
by RUNX2), mutation of the Runx element prevents
RUNX2 repression of activin-induced FSH␤ expression
(Fig. 7B; ⫺156/⫺154 mutation, not significantly suppressed as indicated by ns). Mutation of this binding site
does not alter the ability of RUNX1 or RUNX3 to repress
the induction of FSH␤ (P ⬎ 0.05, data not shown). Taken
B
5
C
5
Fold Change
RUNX2
3
2
1
0
WT
-156/-154
Mutation
Fold Induction by Activin
Empty Vector
4
25
*
ns
4
3
2
1
0
WT
-156/-154
Mutation
Fold Induction by Smad3
A
*
ns
20
15
10
5
0
WT
-156/-154
Mutation
FIG. 7. Cis mutation of the Runx element relieves RUNX2-mediated
repression of mouse FSH␤ transcription. A and B, L␤T2 cells were
transfected with RUNX2 or its control vector along with the ⫺398
mouse FSH␤luc wild-type (WT) reporter or the ⫺398 FSH␤luc reporter
with a cis mutation in the ⫺159 Runx site (⫺156/⫺154 Mutation).
FSH␤ promoter activity was assessed in cells treated with vehicle (A) or
activin (B) to determine the necessity of the ⫺159 site for RUNX2
repression. C, To test whether RUNX2 can interfere with Smad3induced FSH␤ expression and whether the ⫺159 Runx site is necessary
for this effect, cells were cotransfected with Smad3 and RUNX2.
*, Significant suppression of FSH␤ promoter activity by RUNX2 vs.
empty vector; ns, not significant.
Endocrinology, June 2010, 151(6):2669 –2680
together, these studies indicate that the ⫺159 Runx site
within the FSH␤ promoter is important and necessary for
the suppressive effect of RUNX2.
It is well established that Smad proteins are critical for
murine FSH␤ expression (9, 10, 34, 38). Indeed, mice
lacking Smad3 exhibit reduced FSH␤ expression (32), illustrating the importance of this Smad factor in vivo. We
hypothesized that if RUNX2 suppresses activin induction
via disrupting Smad signaling, then overexpression of
RUNX2 would repress the induction of FSH␤ by Smad3.
Alternatively, if RUNX2 were acting upstream of Smad3
in the activin signaling cascade, RUNX2 would not interfere with induction by Smad3. As expected, overexpression of Smad3 induces a robust increase in wild-type FSH␤
promoter activity; this response is potently repressed by
RUNX2 (Fig. 7C; WT, ⬃80% reduced by RUNX2). Similar to the case with activin, RUNX2 is unable to suppress
Smad3 induction of the FSH␤ promoter containing the
⫺156/⫺154 mutation (Fig. 7C; ⫺156/⫺154 Mutation).
Collectively, these data show that the ⫺159 Runx site is
necessary for RUNX2 to inhibit Smad3-induced transcriptional activity and support the hypothesis that
RUNX2 repression of FSH␤ expression acts on or downstream of Smad3 activation.
Mutation of the Smad interacting domain of
Runx2 fails to relieve Runx2 repression of activin
induction
The carboxy terminus of RUNX2 contains specific regions necessary for mediating functional interactions with
a number of coregulatory proteins involved in either transcriptional activation or repression (29, 39 – 43). For example, a three-amino-acid His-Thr-Tyr motif (HTY;
amino acids 426-428) within the Smad interacting domain
of Runx2 has been shown to mediate Smad protein interactions (29, 39), and a four-amino-acid Trp-Arg-Pro-Tyr
motif (WRPY; amino acids 525-528) is necessary for interactions with corepressors of the Groucho/TLE family
(40). We assessed the importance of these motifs for
Runx2-induced repression of activin- or Smad3-induced
FSH␤ promoter activity by cotransfecting a murine Runx2
plasmid containing either HTY (HTY␮) or WRPY
(WRPY␮) domain mutations (residues mutated to alanine; AAA or AAAA, respectively). Overexpression of
Runx2 or WRPY␮ results in a similar reduction in Smad3induced FSH␤ activity (Fig. 8A; Runx2 vs. WRPY␮). In
contrast, HTY␮ fails to significantly repress FSH␤ expression (Fig. 8A; Runx2 vs. HTY␮, P ⬍ 0.05), suggesting that
the Runx2 HTY motif is necessary to mediate repression
by Runx2. Interestingly, neither the WRPY nor the HTY
Runx2 mutation was sufficient to alleviate repression
when FSH␤ was induced by activin (Fig. 8B; Runx2 vs.
Endocrinology, June 2010, 151(6):2669 –2680
5
ns
3
2
1
0
HTYµ
1
4
Runx2
2
LHβLuc + Activin
WRPYµ
LHβ Fold Induction by Activin
3
0
C
Empty Vector
HTYµ
Runx2
0
WRPYµ
5
ns
HTYµ
ns
10
4
Runx2
15
5
FSHβLuc + Activin
WRPYµ
20
B
Empty Vector
*
Empty Vector
FSHβ Fold Induction by Smad3
25
FSHβLuc + Smad3
FSHβ Fold Induction by Activin
A
endo.endojournals.org
FIG. 8. Runx2 repression of activin induction is mediated, in part,
through an interaction with Smads. A and B, L␤T2 cells were
transfected with the ⫺398 FSH␤luc reporter and an expression
vector containing wild-type murine Runx2, Runx2 with a mutated
WRPY motif (WRPY␮), Runx2 with a mutated HTY motif (HTY␮), or
empty vector to assess whether Runx2 HTY or WRPY protein
interaction domains are necessary for repression of Smad3-induced
(A) or activin-induced (B) FSH␤ activity. C, To test whether Runx2
HTY or WRPY protein interaction domains are necessary for LH␤
induction, a 1.8-kb rat LH␤luc reporter gene was transfected into
L␤T2 cells along with Runx2, WRPY␮, HTY␮, or its empty vector,
and activin-induced promoter activity was assessed. *, Significant
relief of repression by Runx2 mutation vs. wild-type Runx2; ns, not
significant.
WRPY␮ or HTY␮, P ⬎ 0.05), although there was a trend
for reduced inhibition by HTY␮ (P ⫽ 0.08 vs. Runx2).
Furthermore, mutation of either domain did not alter the
ability of Runx2 to induce LH␤ (Fig. 8C), suggesting that
Runx2 coordinates induction of LH␤ via interaction with
proteins other than Smads or Groucho family members.
Taken together, these results demonstrate that the Runx2
HTY motif is critical for repression of Smad3-induced
FSH␤ activity but not solely responsible for mediating the
repressive actions of Runx2 on FSH␤ expression induced
by activin.
Discussion
In the present study, we identify a mechanism whereby the
Runx family of transcription factors potently regulates
activin induction of FSH␤ gene expression. Our investigation confirms the expression of Runx2 in the pituitary
gland of mice and focuses on the role of Runx2 as a potent
repressor of activin action using a gonadotrope cell model.
Promoter analyses show that RUNX2-mediated repression of activin induction is lost upon 5⬘ truncation or cis
mutation of any of the five previously characterized elements mutation of which causes a loss of activin induction,
converting RUNX2 repression to induction of FSH␤ gene
expression. Promoter truncations also reveal that in the
absence of activin signaling, RUNX2 induces FSH␤ ex-
2677
pression, an activity lost when region ⫺194/⫺127 is deleted, thereby focusing our attention on the ⫺159 Runx
consensus site within the FSH␤ promoter. With regard
to the mechanism of RUNX2-mediated transcriptional
activity, these findings demonstrate that the effect of
RUNX2 to induce or repress is closely tied to the highly
coordinated and complex mechanism of activin action
on the FSH␤ gene.
With regard to potential mechanisms of activin induction, Smad proteins are well-known signaling molecules
activated by TGF␤ family members. Smad2 and Smad3
are phosphorylated by activin receptors at the plasma
membrane and, together with Smad4, induce transcription of target genes (44), including FSH␤ (9, 11, 12, 34).
Of importance to our investigation, Smad3 and Runx2
can physically and functionally interact at Runx composite elements (29, 30) and induce repression of the osteocalcin promoter (29, 30), providing a potential mechanism
whereby Runx2 could inhibit activin induction of FSH␤
gene expression. Consistent with this possibility, RUNX2
not only suppresses activin induction of FSH␤ but also
represses induction by Smad3 as well. In actuality,
RUNX2 nearly abolishes FSH␤ induction by Smad3 in
comparison with the partial effect on activin induction
(⬃85% vs. ⬃50% reduced, Smad3 vs. activin, respectively; Fig. 7), and mutation of the Runx2/Smad interaction
domain eliminates Runx2-mediated repression of Smad3
induction. However, our finding that mutation of this domain only partially relieves repression of FSH␤ when it is
induced by activin suggests that the mechanism of Runx2mediated repression of activin induction likely involves
multiple factors in addition to Smad3.
With regard to other mediators, recent evidence suggests that the forkhead transcription factor, FoxL2, another transcriptional regulator known to bind Smads (45),
is a critical mediator of activin induction of FSH␤ gene
expression. Indeed, FoxL2 has been shown to bind the
murine ⫺106 activin response element (35) and has the
potential to physically interact at the ⫺153 site (46). Thus,
if FoxL2 is binding at ⫺153, Runx2 could interfere with
its action and interaction with Smads by binding at ⫺159.
Taken together, our results show that Runx2 action is
dependent on the availability of Smad3 and other coregulatory proteins that interact at the ⫺159 Runxbinding element and that the balance of these factors
contribute to the direction in which Runx2 regulates the
FSH␤ promoter.
Our findings reinforce the idea that Runx2 is a contextdependent transcription factor, functioning as either a
corepressor or coactivator depending on formation of specific coregulatory protein interactions at the DNA level.
These interactions are dictated by the carboxy terminus of
2678
Breen et al.
Runx Repression of FSH␤ Gene Expression
Runx2, which contains regions involved in either transcriptional activation or repression by Runx2, and motifs
necessary for mediating physical and functional interactions with a number of coregulatory proteins, including
the corepressors of the Groucho/TLE family (40), histone deacetylase 3 (41), homeodomain activators such
as Dlx3 (42), and Smad proteins (43). Although our
studies implicate the HTY Smad-interacting domain
motif as important for Runx2-mediated repression of
FSH␤, protein interactions via this domain are not necessary for the induction of LH␤ by Runx2. Furthermore, the idea that Runx2 can physically interact with
a host of transcriptional effectors, which collectively
dictate activation or repression, sheds light on our finding of induction of LH␤ as well as our observation that
basal expression of FSH was induced, rather than repressed, when the FSH␤ promoter was rendered unresponsive to activin via cis mutation or 5⬘ truncation.
Again, these findings underscore the complexity of gene
regulation by Runx2 and provide an intriguing and
highly exciting mechanism of hormonal regulation of
the gonadotropin genes.
Developmentally, the importance of Runx2 cannot be
underestimated because mice lacking Runx2 die immediately after birth due to the absence of mineralized bone
(47). In the reproductive axis, Runx2 is expressed within
the ovary during the follicular to luteal phase transition of
the ovulatory cycle and is important for luteinization of
ovarian follicles (48, 49). As for the pituitary gland, recent
evidence suggests that RUNX2 is involved in human pituitary tumor formation (37), and our findings confirm
the presence of Runx2 in the mouse pituitary and the L␤T2
gonadotrope cell model. Although we did not detect
changes in Runx2 mRNA across the ovulatory cycle in
mice (unpublished observations) or observe protein levels
of Runx2 within L␤T2 cells to change during hormone
treatment, we cannot exclude the possibility that this
tightly regulated transcription factor is altered by hormonal milieu within the gonadotrope cell itself. Another
intriguing scenario is the potential interplay between the
members of the Runx family. Indeed, all three Runx members are expressed in L␤T2 cells; however, we postulate
that the protein interactions unique to Runx2 enable this
regulatory factor to mediate specific regulatory effects
upon the gonadotropin subunit genes.
In summary, our results demonstrate a novel role for
Runx2 regulation of FSH␤ gene expression in gonadotrope cells. We identify a putative Runx binding site at
⫺159 in the mouse FSH␤ gene and show that this site is
necessary for RUNX2 repression of activin induction of
FSH␤. Although RUNX2 also inhibits activin induction of
ovine and human FSH␤, the mechanism of repression in
Endocrinology, June 2010, 151(6):2669 –2680
these species may be indirect or through an element at
another location. Collectively, this work has revealed new
insights regarding hormonal regulation of the FSH␤ promoter by identifying a mechanism whereby FSH␤ expression is dampened after induction by activin.
Acknowledgments
We thank Dr. Daniel Bernard (McGill University, Montreal,
Canada) for generously providing the human FSH␤-luciferase
plasmid. The rat LH␤-luciferase plasmid was kindly provided by
Dr. Mark Lawson (University of California, San Diego, La Jolla,
CA). We thank Dr. Yoshiaki Ito (National University of Singapore, Singapore) for providing the human RUNX1, -2, and -3
plasmids and Dr. Jane Lian (University of Massachusetts,
Worcester, MA) for providing the murine Runx2 plasmid. The
Smad3 expression plasmid was a kind gift of Dr. Rik Derynck
(University of California-San Francisco, San Francisco, CA).
Special thanks go to Dr. Rachel Larder for critical reading of the
manuscript and members of the Mellon laboratory for helpful
discussions throughout this work. DNA sequencing was performed by the UCSD Cancer Center DNA sequencing shared
resource (NCI P30 CA023100).
Address all correspondence and requests for reprints to: Pamela
L. Mellon, Department of Reproductive Medicine and Center for
Reproductive Science and Medicine, University of California,
San Diego, 9500 Gilman Drive, La Jolla, California 920930674. E-mail: pmellon@ucsd.edu.
This work was supported by National Institutes of Health
(NIH) Grant R01 HD020377 (to P.L.M.) and by the Eunice
Kennedy Shriver National Institute of Child Health and Human
Development/NIH through cooperative agreement (U54 HD012303)
as part of the Specialized Cooperative Centers Program in Reproduction and Infertility Research (P.L.M.). K.M.B. was partially supported by NIH Grant F32 HD051360. V.G.T. was
partially supported by NIH Grant K01 DK080467, and D.C.
was partially supported by NIH Grants R01 HD057549 and
R03 HD054595.
Disclosure Summary: The authors have nothing to disclose.
References
1. Kumar TR, Wang Y, Lu N, Matzuk MM 1997 Follicle stimulating
hormone is required for ovarian follicle maturation but not male
fertility. Nat Genet 15:201–204
2. Pierce JG, Parsons TF 1981 Glycoprotein hormones: structure and
function. Annu Rev Biochem 50:465– 495
3. Kaiser UB, Conn PM, Chin WW 1997 Studies of gonadotropinreleasing hormone (GnRH) action using GnRH receptor-expressing
pituitary cell lines. Endocr Rev 18:46 –70
4. Vale W, Rivier C, Brown M 1977 Regulatory peptides of the hypothalamus. Annu Rev Physiol 39:473–527
5. Matzuk MM, Kumar TR, Bradley A 1995 Different phenotypes for
mice deficient in either activins or activin receptor type II. Nature
374:356 –560
Endocrinology, June 2010, 151(6):2669 –2680
6. Vassalli A, Matzuk MM, Gardner HA, Lee KF, Jaenisch R 1994
Activin/inhibin ␤B subunit gene disruption leads to defects in eyelid
development and female reproduction. Genes Dev 8:414 – 427
7. Burns KH, Matzuk MM 2002 Genetic models for the study of gonadotropin actions. Endocrinology 143:2823–2835
8. Bailey JS, Rave-Harel N, McGillivray SM, Coss D, Mellon PL 2004
Activin regulation of the follicle-stimulating hormone ␤-subunit
gene involves Smads and the TALE homeodomain proteins Pbx1
and Prep1. Mol Endocrinol 18:1158 –1170
9. Gregory SJ, Lacza CT, Detz AA, Xu S, Petrillo LA, Kaiser UB 2005
Synergy between activin A and gonadotropin-releasing hormone in
transcriptional activation of the rat follicle-stimulating hormone-␤
gene. Mol Endocrinol 19:237–254
10. Suszko MI, Balkin DM, Chen Y, Woodruff TK 2005 Smad3 mediates activin-induced transcription of follicle-stimulating hormone
␤-subunit gene. Mol Endocrinol 19:1849 –1858
11. McGillivray SM, Thackray VG, Coss D, Mellon PL 2007 Activin
and glucocorticoids synergistically activate follicle-stimulating hormone ␤-subunit gene expression in the immortalized L␤T2 gonadotrope cell line. Endocrinology 148:762–773
12. Lamba P, Santos MM, Philips DP, Bernard DJ 2006 Acute regulation of murine follicle-stimulating hormone ␤-subunit transcription
by activin A. J Mol Endocrinol 36:201–220
13. Su P, Shafiee-Kermani F, Gore AJ, Jia J, Wu JC, Miller WL 2007
Expression and regulation of the ␤-subunit of ovine follicle-stimulating hormone relies heavily on a promoter sequence likely to bind
Smad-associated proteins. Endocrinology 148:4500 – 4508
14. Nüsslein-Volhard C, Wieschaus E 1980 Mutations affecting segment number and polarity in Drosophila. Nature 287:795– 801
15. Bae SC, Lee YH 2006 Phosphorylation, acetylation and ubiquitination: the molecular basis of RUNX regulation. Gene 366:58 – 66
16. Javed A, Guo B, Hiebert S, Choi JY, Green J, Zhao SC, Osborne MA,
Stifani S, Stein JL, Lian JB, van Wijnen AJ, Stein GS 2000 Groucho/
TLE/R-esp proteins associate with the nuclear matrix and repress
RUNX (CBF␣/AML/PEBP2␣) dependent activation of tissue-specific gene transcription. J Cell Sci 113:2221–2231
17. Zaidi SK, Javed A, Choi JY, van Wijnen AJ, Stein JL, Lian JB, Stein
GS 2001 A specific targeting signal directs Runx2/Cbfa1 to subnuclear domains and contributes to transactivation of the osteocalcin gene. J Cell Sci 114:3093–3102
18. Thackray VG, McGillivray SM, Mellon PL 2006 Androgens, progestins and glucocorticoids induce follicle-stimulating hormone
␤-subunit gene expression at the level of the gonadotrope. Mol Endocrinol 20:2062–2079
19. Coss D, Jacobs SB, Bender CE, Mellon PL 2004 A novel AP-1 site
is critical for maximal induction of the follicle-stimulating hormone-␤ gene by gonadotropin-releasing hormone. J Biol Chem 279:
152–162
20. Strahl BD, Huang HJ, Pedersen NR, Wu JC, Ghosh BR, Miller WL
1997 Two proximal activating protein-1-binding sites are sufficient
to stimulate transcription of the ovine follicle-stimulating hormone-␤ gene. Endocrinology 138:2621–2631
21. Vasilyev VV, Pernasetti F, Rosenberg SB, Barsoum MJ, Austin DA,
Webster NJ, Mellon PL 2002 Transcriptional activation of the ovine
follicle-stimulating hormone-␤ gene by gonadotropin-releasing hormone involves multiple signal transduction pathways. Endocrinology 143:1651–1659
22. Bae SC, Lee KS, Zhang YW, Ito Y 2001 Intimate relationship between TGF-␤/BMP signaling and runt domain transcription factor,
PEBP2/CBF. J Bone Joint Surg Am 83-A Suppl 1(Pt 1):S48 –S55
23. Zhang Y, Feng XH, Derynck R 1998 Smad3 and Smad4 cooperate
with c-Jun/c-Fos to mediate TGF-␤-induced transcription. Nature
394:909 –913
24. McGillivray SM, Bailey JS, Ramezani R, Kirkwood BJ, Mellon PL
2005 Mouse GnRH receptor gene expression is mediated by the
LHX3 homeodomain protein. Endocrinology 146:2180 –2185
25. Coss D, Hand CM, Yaphockun KK, Ely HA, Mellon PL 2007 p38
mitogen-activated kinase is critical for synergistic induction of the
endo.endojournals.org
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
2679
FSH␤ gene by gonadotropin-releasing hormone and activin through
augmentation of c-Fos induction and Smad phosphorylation. Mol
Endocrinol 21:3071–3086
Rosenberg SB, Mellon PL 2002 An Otx-related homeodomain protein binds an LH␤ promoter element important for activation during
gonadotrope maturation. Mol Endocrinol 16:1280 –1298
Cherrington BD, Bailey JS, Diaz AL, Mellon PL 2008 NeuroD1 and
Mash1 temporally regulate GnRH receptor gene expression in immortalized mouse gonadotrope cells. Mol Cell Endocrinol 295:
106 –114
Ooi GT, Tawadros N, Escalona RM 2004 Pituitary cell lines and
their endocrine applications. Mol Cell Endocrinol 228:1–21
Javed A, Bae JS, Afzal F, Gutierrez S, Pratap J, Zaidi SK, Lou Y,
van Wijnen AJ, Stein JL, Stein GS, Lian JB 2008 Structural coupling
of Smad and Runx2 for execution of the BMP2 osteogenic signal.
J Biol Chem 283:8412– 8422
Alliston T, Choy L, Ducy P, Karsenty G, Derynck R 2001 TGF-␤induced repression of CBFA1 by Smad3 decreases cbfa1 and osteocalcin expression and inhibits osteoblast differentiation. EMBO J
20:2254 –2272
Pernasetti F, Vasilyev VV, Rosenberg SB, Bailey JS, Huang HJ,
Miller WL, Mellon PL 2001 Cell-specific transcriptional regulation
of FSH␤ by activin and GnRH in the L␤T2 pituitary gonadotrope
cell model. Endocrinology 142:2284 –2295
Coss D, Thackray VG, Deng CX, Mellon PL 2005 Activin regulates
luteinizing hormone ␤-subunit gene expression through Smad-binding and homeobox elements. Mol Endocrinol 19:2610 –2623
Heinemeyer T WE, Reuter I, Hermjakob H, Kel AE, Kel OV,
Ignatieva EV, Ananko EA, Podkolodnaya OA, Kolpakov FA,
Podkolodny NL, Kolchanov NA 1998 Databases on transcriptional
regulation: TRANSFAC, TRRD, and COMPEL Nucleic Acids Res
26:362–367
Suszko MI, Lo DJ, Suh H, Camper SA, Woodruff TK 2003 Regulation of the rat follicle-stimulating hormone ␤-subunit promoter by
activin. Mol Endocrinol 17:318 –332
Lamba P, Fortin J, Tran S, Wang Y, Bernard DJ 2009 A novel role
for the forkhead transcription factor FOXL2 in activin A-regulated
follicle-stimulating hormone ␤-subunit transcription. Mol Endocrinol 23:1001–1013
Thornell A, Hallberg B, Grundström T 1991 Binding of SL3-3 enhancer factor 1 transcriptional activators to viral and chromosomal
enhancer sequences. J Virol 65:42–50
Zhang HY, Jin L, Stilling GA, Ruebel KH, Coonse K, Tanizaki Y,
Raz A, Lloyd RV 2009 RUNX1 and RUNX2 upregulate galectin-3
expression in human pituitary tumors. Endocrine 35:101–111
Bernard DJ 2004 Both SMAD2 and SMAD3 mediate activin-stimulated expression of the follicle-stimulating hormone beta subunit in
mouse gonadotrope cells. Mol Endocrinol 18:606 – 623
Afzal F, Pratap J, Ito K, Ito Y, Stein JL, van Wijnen AJ, Stein GS, Lian
JB, Javed A 2005 Smad function and intranuclear targeting share a
Runx2 motif required for osteogenic lineage induction and BMP2
responsive transcription. J Cell Physiol 204:63–72
Lutterbach B, Westendorf JJ, Linggi B, Isaac S, Seto E, Hiebert SW
2000 A mechanism of repression by acute myeloid leukemia-1, the
target of multiple chromosomal translocations in acute leukemia.
J Biol Chem 275:651– 656
Schroeder TM, Kahler RA, Li X, Westendorf JJ 2004 Histone
deacetylase 3 interacts with runx2 to repress the osteocalcin promoter and regulate osteoblast differentiation. J Biol Chem 279:
41998 – 42007
Hassan MQ, Javed A, Morasso MI, Karlin J, Montecino M, van
Wijnen AJ, Stein GS, Stein JL, Lian JB 2004 Dlx3 transcriptional
regulation of osteoblast differentiation: temporal recruitment of
Msx2, Dlx3, and Dlx5 homeodomain proteins to chromatin of the
osteocalcin gene. Mol Cell Biol 24:9248 –9261
Zaidi SK, Sullivan AJ, van Wijnen AJ, Stein JL, Stein GS, Lian JB
2680
Breen et al.
Runx Repression of FSH␤ Gene Expression
2002 Integration of Runx and Smad regulatory signals at transcriptionally active subnuclear sites. Proc Natl Acad Sci USA 99:8048 –
8053
44. Massagué J 1998 TGF-␤ signal transduction. Annu Rev Biochem
67:753–791
45. Blount AL, Schmidt K, Justice NJ, Vale WW, Fischer WH,
Bilezikjian LM 2009 FoxL2 and Smad3 coordinately regulate follistatin gene transcription. J Biol Chem 284:7631–7645
46. Corpuz PS, Lindaman LL, Mellon PL, Coss D 16 March 2010 FoxL2
is required for activin induction of the mouse and human folliclestimulating hormone ␤-subunit genes. Mol Endocrinol 10.1210/
me.2009-0425
Endocrinology, June 2010, 151(6):2669 –2680
47. Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell
IR, Stamp GW, Beddington RS, Mundlos S, Olsen BR, Selby PB,
Owen MJ 1997 Cbfa1, a candidate gene for cleidocranial dysplasia
syndrome, is essential for osteoblast differentiation and bone development. Cell 89:765–771
48. Park ES, Choi S, Muse KN, Curry Jr TE, Jo M 2008 Response gene
to complement 32 expression is induced by the luteinizing hormone
(LH) surge and regulated by LH-induced mediators in the rodent
ovary. Endocrinology 149:3025–3036
49. Jo M, Curry Jr TE 2006 Luteinizing hormone-induced RUNX1
regulates the expression of genes in granulosa cells of rat periovulatory follicles. Mol Endocrinol 20:2156 –2172
Refer a new active member and you could
receive a $10 Starbucks Card when they join.
www.endo-society.org/referral