|Home | About | Journals | Submit | Contact Us | Français|
Pin1 is a peptidyl-prolyl cis-trans isomerase which catalyzes the isomerization of phosphorylated Ser/Thr-Pro peptide bonds. Pin1 knockout mice have marked abnormalities in their reproductive development and function. However, the molecular mechanisms underlying their reproductive defects are poorly understood. Herein, we demonstrate that Pin1 is required for both basal and GnRH-induced gonadotropin β-subunit gene transcription, through interactions with the transcription factors SF-1, Pitx1, and Egr-1. Pin1 activates transcription of the gonadotropin β-subunit genes synergistically with these transcription factors, either by modulating their stability or by increasing their protein-protein interactions. Notably, we provide evidence that Pin1 is required for the Ser203 phosphorylation-dependent ubiquitination of SF-1, which facilitates SF-1-Pitx1 interactions and therefore results in an enhancement of SF-1 transcriptional activity. Furthermore, we demonstrate that in gonadotrope cells, sufficient levels of activated Pin1 are maintained through transcriptional and posttranslational regulation by GnRH-induced signaling cascades. Our results suggest that Pin1 functions as a novel player in GnRH-induced signal pathways and is involved in gonadotropin β-subunit gene transcription by modulating the activity of various specific transcription factors.
Pin1 is a unique peptidyl-prolyl isomerase (PPIase) that recognizes a phosphorylated serine or threonine residue preceding a proline (pSer/Thr-Pro) and isomerizes the cis-trans conformation of the prolyl-peptidyl bond. This sequence can be phosphorylated by various kinases, including the mitogen-activated protein kinases (MAPKs), which form an integral part of many signal transduction pathways. Phosphorylation is a common modification involved in transmission of intracellular signals, and recognition of this motif by Pin1 and the subsequent conformational change in the protein provide a partial explanation for one mechanism of phosphorylation-dependent protein regulation (34). The effects of the Pin1-induced isomerization on its target proteins are diverse and include altering the stability and localization of the target proteins, as well as modifying their interaction with other proteins (14, 27, 33, 41, 53). In this way, Pin1 regulates various cellular processes, including differentiation, progression through the cell cycle, gene transcription, and mRNA splicing, and its functions in cancer and Alzheimer's disease have been well established (32, 34, 53, 57, 64).
Pin1 is expressed ubiquitously and its transcription is stimulated by various growth factors through activation of E2F, and also by Notch1 in a positive feedback loop (48, 51, 65). Its activity is also regulated by phosphorylation of Ser16 in the center of the pSer/Thr-Pro binding pocket, which abolishes Pin1's capability to interact with its substrates and may facilitate it nuclear export (35). The kinases involved in this regulation have not been fully investigated, although protein kinase A (PKA) was implicated (35).
Some of the most marked abnormalities in knockout mice are in their reproductive development and function. The phenotype includes decreased fertility, testicular atrophy, reduced testis size, seminiferous tubule degeneration, and spermatogonial depletion, while the gradual degeneration of spermatogonia up to the age of 14 months suggests that Pin1 has multiple roles in regulating fertility in adults as well as during early stages of development (1). Notably, female Pin1 knockout mice exhibit severe reduction in mammary epithelial duct development (31). The molecular mechanisms underlying these reproductive defects have yet to be elucidated. However, given that spermatogonia are regulated by the gonadotropin hormones, it is likely that some of the noted effects of Pin1 knockout might involve the gonadotropins.
Reproduction is regulated by the hypothalamic hormone, gonadotropin-releasing hormone (GnRH), which binds a G-protein-coupled receptor on the pituitary gonadotrope to induce transcription of the three gonadotropin genes, a common α-subunit (αGSU) and the hormone-specific β-subunits (LHβ and FSHβ); synthesis of the β-subunit is the rate-limiting step in hormone synthesis (42). The gonadotropic hormones, luteinizing hormone (LH) and follicle-stimulating hormone (FSH), are then secreted into the circulation and reach the gonads to activate reproductive development and function.
GnRH-induced signaling to the gonadotropin genes is predominantly via the PKC pathway and downstream MAPK cascades, including extracellular signal-related kinase 1/2 (ERK1/2), Jun N-terminal kinase (JNK), p38 MAPK, and big MAPK (BMK or ERK5), while PKA is also activated following GnRH-induced increases in cyclic AMP (29, 42). Many of the transcription factors that regulate expression of the gonadotropin genes are activated through phosphorylation by these kinases and thus contain pSer/Thr-Pro motifs. The LHβ gene promoter is activated at a basal level by the orphan receptor steroidogenic factor 1 (SF-1) and the paired-like homeodomain transcription factor 1 (Pitx1) (19, 39); both factors are expressed in a cell-restricted manner (18, 24, 25, 44, 55). Also crucial in LHβ gene expression is the early growth factor 1 (Egr-1), whose expression and phosphorylation are induced by GnRH, allowing its synergistic interaction with SF-1 and Pitx1 and mediation of the GnRH effect (10, 15, 19, 21, 39, 59). The molecular mechanisms of the basal and GnRH-stimulated effects on the FSHβ gene promoter have been less well elucidated, but they appear to involve SF-1 and various Pitx proteins, including Pitx1, as well as the activator protein-1 (AP-1) transcriptional complex, comprising c-Jun and c-Fos, both of which are transcriptionally regulated and phosphorylated by GnRH (7, 23, 37, 39, 62, 70).
SF-1 and Pitx1 act synergistically on both the LHβ and Mullerian-inhibiting-substance gene promoters, and this was shown to be a result of Pitx1 binding directly to SF-1. Pitx1 enhances the transcriptional activity of SF-1 through an unmasking effect, in which the Pitx1 C terminus interacts with a region of SF-1 N-terminal to amino acid 279, possibly involving an activation domain located between amino acids 187 and 245, to activate SF-1 to a degree similar to that achieved by deletion of the ligand-binding domain (LBD) (58).
Both SF-1 and Pitx1 appear to be phosphorylated, either constitutively or possibly through GnRH-stimulated kinases (16, 38). Phosphorylation of SF-1 at Ser203, by ERK1/2 or CDK7, was shown to be required for maximal SF-1 target gene transactivation and recruitment of various cofactors (12, 16, 28). However, SF-1 is subject to additional posttranslational modifications which alter its activity, and it can be SUMOylated at two conserved lysines, Lys119 and Lys194, which are adjacent to the DNA-binding domain (DBD) and LBD, respectively. SUMOylation at Lys194 was seen to inhibit Ser203 phosphorylation and was also associated with reduced transactivation of various SF-1 target genes (5, 67). In gonadotropes, SF-1 was seen also to be ubiquitinated, and this form was associated with the LHβ gene promoter (60). While clearly affecting SF-1 transactivation, the relationship between these various modifications and their roles are not yet clear.
In this study we hypothesized that the MAPK-directed phosphorylation of some of these gene-specific transcription factors might cause them to be targeted by Pin1. We have demonstrated a role for Pin1 in transcription of the gonadotropin β-subunit genes and elucidated its function in the activation of SF-1, via a phosphorylation-regulated pathway which facilitates SF-1-Pitx1 interactions as a result of promoting SF-1 ubiquitination. We also show that Pin1 activity is regulated by GnRH, indicating that it constitutes a novel player in the GnRH-signaling pathway to gonadotropin gene expression.
The murine gonadotrope αT3-1 and LβT2 cells (gifts from P. Mellon, University of California, San Diego) were cultured in Dulbecco's modified Eagle's medium supplemented with 10% dialyzed fetal bovine serum (Gibco-Invitrogen), 10 mM HEPES [pH 7.4], 4 mM l-glutamine (HyClone), 100 U/ml penicillin (Gibco-Invitrogen), and 100 μg/ml streptomycin (Gibco-Invitrogen). COS-1 cells, HEK 293T control small interfering RNA (siRNA) cells, HEK 293T Pin1 siRNA cells (50), MEF wild-type (WT) cells, and MEF Pin1−/− cells (gifts from K. P. Lu, Harvard Medical School) were cultured in Dulbecco's modified Eagle's medium supplemented with 10% certified fetal bovine serum (HyClone) and the same antibiotics. As appropriate, cells were exposed to 10 to 100 nM concentrations of the GnRH agonist buserelin (Sigma), 100 nM phorbol myristate acetate (PMA) (Sigma), 1 μM forskolin (Sigma), 5 μM cyclosporine (Sigma), 10 μM MG132 (Sigma), 15 μM roscovitine (Sigma), or 1 μM U0126 (Promega), as indicated in the figure legends. All cells were maintained at 37°C under 5% CO2. Transfection was carried out at 50 to 60% confluence. LβT2 cells were transfected using Geneporter 2 (Gene Therapy Systems), HEK 293T cells were transfected using the calcium phosphate method, and the other cells were transfected using Lipofectamine 2000 (Invitrogen). After transfection, cells were incubated for a further 24 to 48 h before harvest.
Luciferase assay constructs were created by ligating 1,300 bp of the proximal murine LHβ gene promoter or 600 bp of the proximal murine FSHβ gene promoter into pGL2 Basic (Promega), as reported previously (11, 39). The expression vectors for Pitx1, Egr-1, and SF-1 were gifts from J. Drouin (Montreal, Canada), J. Milbrandt (St. Louis, MO), and K. Parker (Durham, NC), respectively. The coding sequence of Pin1 was synthesized by reverse transcriptase PCR (RT-PCR) from LβT2 cell mRNA and was ligated into the expression vector pCS2+. For FLAG- or hemagglutinin (HA)-tagged constructs, sequence-specific primers containing restriction enzyme recognition sites at the 5′ ends were used to PCR amplify the DNA fragments from the above expression vectors. The Pin1 coding sequence was inserted into the expression vector p3x-FLAG-CMV 10. Pitx1, SF-1, Egr-1, ubiquitin, and SUMO1 sequences were inserted into pxj40-HA, pxj40-FLAG, and pxj40-Myc (gifts from B. C. Low, National University of Singapore). pxj40-GFP was used to construct pxj40 GFP SF-1 and pxj40 GFP SF-1 K119R. pcDNA3 CDK7 HA (P#633) and pcDNA3 CDK7 D155A HA (P#812) were purchased from Addgene. For the two-hybrid assays, the reporter gene construct was created as described previously (36). Other constructs were prepared by inserting the sequences from Pin1, Pitx1, SF-1, and Egr-1 expression vectors into the Gal4 DBD or Gal4 activation domain (AD) vector pM or pVP16 (Clontech). Site-directed mutations were carried out using a QuikChange site-directed mutagenesis kit (Stratagene), according to the manufacturer's instructions, in order to introduce the specific amino acid substitutions. All constructs were verified by sequencing.
Antibodies used were as follows: anti-Pin1 (Santa Cruz; sc-15340), anti-HA (Santa Cruz; sc-805 and sc-7392), anti-c-Myc (Santa Cruz; sc-40) (Sigma; C3956), anti-FLAG (Santa Cruz; sc-807), anti-GAPDH (Santa Cruz; sc-47724), anti-ATF3 (Santa Cruz; sc-188), anti-Pitx1 (Santa Cruz; sc-18922), anti-Egr-1 (Santa Cruz; sc-110), anti-β-catenin (Santa Cruz; sc-7199), anti-Ub (P4D1; Santa Cruz; sc-8017), anti-ERK1 (Santa Cruz; sc-94), anti-ERK2 (Santa Cruz; sc-154), anti-phospho-ERK (Tyr 204; Santa Cruz; sc-7976), anticalcineurin (PP2B-Aa [D-9]; Santa Cruz; sc-17808), anti-green fluorescent protein (GFP) (Santa Cruz; sc-81045 and sc-8334), anti-SF-1 (Upstate Biotechnology; 07-618), anti-phospho-Pin1 (Ser 16) (Cell Signaling; 3721), anti-α-tubulin (Sigma; T9026), and anti-FLAG (Sigma; F3165).
Stealth small interfering RNAs (siRNAs) specific for Pin1 (sense, 5′-GCCGGGUGUACUACUUCAAdTdT-3′ ), and scrambled control (sense, 5′-GUGUUACAGCUCCAGAUGCdTdT-3′) were purchased from Invitrogen. The siRNAs were transfected into LβT2 cells using Oligofectamine (Invitrogen) according to the manufacturer's protocols.
For construction of calcineurin A short hairpin RNA (shRNA)-encoding plasmids, a pair of oligonucleotides including the specific 19-bp target sequence for calcineurin A (5′-TTACAATCTTCTCGGCACC-3′) was synthesized, annealed, and ligated into pSUPER Basic vector (Oligoengine, Seattle, WA). The construct was verified by sequencing. pSUPER Basic GFP shRNA was used as a control, as described previously (36). The vector-based shRNA constructs were transfected into LβT2 cells using Geneporter 2.
LβT2 cells were plated in 96-well white plates. At 12 to 24 h after plating, expression vectors (50 ng for Pin1, 80 ng for SF-1 and Egr-1, and 10 ng for Pitx1), the luciferase reporter vector (100 ng) and simian virus 40-Renilla luciferase (2 ng) were transfected after equilibration of the total amount of DNA with pCS2+ empty vector or pWhitescript. Cells were incubated for 24 to 48 h before harvest. Luciferase activity was measured using the Dual-Glo system (Promega) and a Veritas Microplate luminometer (Turner Biosystems) and normalized to levels of Renilla luciferase. Reporter gene activity was calculated as activity over basal levels (n-fold) generated from transfection of the luciferase reporter vector and pCS2+ empty vector or pWhitescript. Statistical analysis was performed using a simple unpaired t test to determine means that were statistically different. Differences were considered significant when P was <0.05. Synergy was determined by comparing the additive effects of expression vectors transfected separately (using SE(x + y) = √[SE(x)2 + SE(y)2]) with the effect when both were transfected together and was defined as a significantly (P < 0.05) greater-than-additive effect.
For the two-hybrid assays, COS-1 or LβT2 cells were plated in 96-well plates before transfection using 150 ng of the pM and pVP16 fusion constructs, 50 ng of the reporter gene, and 2 ng of simian virus 40-Renilla luciferase as an internal control. Cells were incubated for 24 to 48 h before harvest, and luciferase activity was measured as above. Reporter gene activity was calculated as activity over basal levels (n-fold) generated from transfection of the empty pM and pVP16 constructs, after normalization with Renilla luciferase levels. Statistical analysis to determine protein interaction compared the additive effect of the pM-fusion and pVP-fusion expression vectors transfected separately with their effect when transfected together, as above.
RNA was extracted using TRIzol (Invitrogen), and the total RNA (2 μg) was reverse transcribed using Superscript III reverse transcriptase (Invitrogen) and random oligo(dT) primers (New England Biolabs). Of the 20 μl cDNA obtained, 1 μl was used for a semiquantitative PCR with specific primers. Primers for the LHβ RT-PCR were as follows: forward, 5′-GCCTGTCAACGCAACTCTGG-3′; reverse, 5′-CAGGCCATTGGTTGAGTCCT-3′. Primers for the FSHβ RT-PCR were as follows: forward, 5′-AGCACTGACTGCACCGTGAG-3′; reverse, 5′-CCTCAGCCAGCTTCATCAGC-3′. Primers for the β-actin RT-PCR were as follows: forward, 5′-GCCATGTACGTAGCCATCCA-3′; reverse, 5′-ACGCTCGGTCAGGATCTTCA-3′. Primers for the Pin1 RT-PCR were as follows: forward, 5′-CCGGAATTCATGGCGGACGAGGAGAAG-3′; reverse, 5′-TGCTCTAGATCATTCTGTGCGCAGGAT-3′. Primers for the SF-1 RT-PCR were as follows: forward, 5′-ATGGACTATTCGTACGAC-3′; reverse, 5′-TCAAGTCTGCTTGGCCTG-3′. Primers for the Pitx1 RT-PCR were as follows: forward, 5′-ATGGACGCCTTCAAGGGAGGC-3′; reverse, 5′-TCAGCTGTTGTACTGGCAAG-3′.
Real-time PCR was carried out using SYBR green I dye with an ABI Prism 7900 sequence detector (Perkin-Elmer Applied Biosystems). PCRs were performed in a 5-μl volume, containing PCR Master Mix, 0.1 μg cDNA template, and forward and reverse primers. Primers for the LHβ real-time PCR were as follows: forward, 5′-CAGTCTGCATCACCTTCACC-3′; reverse, 5′-GCAGTACTCGGACCATGCTA-3′. Primers for the FSHβ real-time PCR were as follows: forward, 5′-TGCACAGGACGTAGCTGTTT-3′; reverse, 5′-TGAGATGGTGATGTTGGTCA-3′. Primers for the β-actin real-time PCR were as follows: forward, 5′-CCTTCCTTCTTGGGTATGGA-3′; reverse, 5′-ACGGATGTCAACGTCACACT-3′. The samples were heated to 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. The comparative cycle time (CT) method was used to compare mRNA levels in the various samples which were assayed in duplicate.
Cells were lysed in 1× lysis buffer (50 mM Tris-Cl [pH 8.0], 100 mM NaCl, 1% sodium dodecyl sulfate [SDS], 0.5% NP-40, 5% glycerol, 2 mM dithiothreitol, and protease inhibitors). After centrifugation at 13,400 × g for 40 min at 4°C, the supernatant was collected and protein content was measured with the Bradford protein assay reagent (Bio-Rad). The proteins were resolved by SDS-polyacrylamide gel electrophoresis (PAGE) and subsequently transferred to polyvinylidene difluoride membranes (Millipore). The membrane was blocked with 5% bovine serum albumin in TBST (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 0.1% Tween 20). After washing with TBST, the membrane was incubated with primary antibody (Ab) diluted in TBST for 2 h at room temperature or overnight at 4°C. After washing with TBST for 30 min, the membrane was then incubated with goat or bovine horseradish peroxidase-conjugated secondary Ab to mouse, rabbit, or goat immunoglobulin G (IgG) (Santa Cruz) for 1 h at room temperature, followed by extensive washes with TBST. The immunoreactive proteins were detected using the Super Signal Pico West chemiluminescent system (Pierce Chemical), followed by exposure to Fuji medical X-ray film (Fujifilm).
LβT2 cells were grown to approximately 80% confluence in 100-mm plates. The proteins were cross-linked to DNA with 1% formaldehyde. Cross-linking was arrested by addition of 0.125 M glycine. The cells were then rinsed, collected, and resuspended in 1,000 μl ChIP lysis buffer (50 mM Tris-Cl [pH 7.5], 1% NP-40, 0.5% Na deoxycholate, 0.05% SDS, 1 mM EDTA, 150 mM NaCl, and protease inhibitors). The samples were sonicated on ice with a Misonix XL2020 Sonifier to obtain DNA fragments of ~300 to 600 bp, at setting 3 for 10 s six times, with a 10-s rest between, and then pelleted by centrifuging at 13,000 rpm at 4°C for 10 min. 1 μg of Ab [anti-Pin1 (Santa Cruz; sc-15340) or anti-HA (Santa Cruz; sc-805)] and 20 μl protein A Sepharose CL-4B (GE Health) were added into 450 μl lysate and subsequently incubated overnight at 4°C with rotation. The Ab-bound complexes were washed extensively with each of the following buffers: low-salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8.0], 150 mM NaCl), high-salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8.0], 500 mM NaCl), and LiCl wash buffer (0.25 M LiCl, 1% NP-40, 1% deoxycholate, 1 mM EDTA, 10 mM Tris-HCl [pH 8.0]). This was followed by two washes with Tris-EDTA. The protein-bound DNA precipitated from the ChIP assay and input DNA were purified with phenol-chloroform and dissolved in 25 μl H2O. Precipitated and input DNA served as templates for promoter-specific PCR. Primers used to amplify the FSHβ and LHβ gene-proximal promoters were as described previously (11).
For endogenous IP, LβT2 cells were washed with cold 1× phosphate-buffered saline (PBS) and lysed in radioimmunoprecipitation assay (RIPA) buffer (20 mM Tris-Cl [pH 8.0], 125 mM NaCl, 0.5% NP-40, 5% glycerol, 20 mM NaF, 0.2 mM Na3VO4, 2 mM EDTA, protease inhibitors) for 30 min. After centrifugation at 13,400 × g for 40 min, 1 mg of total cell lysate was incubated with 1 μg of Ab and 20 μl protein A or G Sepharose CL-4B (GE Health) overnight at 4°C with gentle rotation. The protein-Ab-bead complexes were washed four times with the lysis buffer. For FLAG IP, 1 μg of FLAG-tagged construct and 1 μg of HA-tagged construct were cotransfected into cells in 60-mm plates. At 24 h after transfection, cells were washed with cold 1× PBS and lysed in RIPA buffer. After centrifugation at 13,400 × g for 40 min, 100 μg of total cell lysate was incubated with 4 μl of EZview red anti-flag M2 affinity gel (Sigma) for 2 h at 4°C with gentle rotation. The protein-Ab-bead complexes were washed at least four times with the lysis buffer to reduce nonspecific binding. SDS-PAGE coupled with Western blot analysis was used to resolve the complexes.
FLAG- or HA-tagged WT or mutant SF-1 and ubiquitin constructs were cotransfected, and after 24 h, the cells were washed with cold 1× PBS and lysed for 30 min in RIPA buffer. The supernatants of the cell lysates were incubated with EZview red anti-FLAG M2 affinity gel (Sigma) or mouse HA Ab (Santa Cruz; sc-7392)-conjugated protein G beads for 2 h at 4°C with gentle rotation, followed by four washes with lysis buffer, after which Western blot analysis was performed. The HA Ab was conjugated to the beads by incubating 1 μg of mouse HA Ab (Santa Cruz; sc-7392) with 20 μl protein G Sepharose CL-4B in RIPA buffer and for 1 h at 4°C, followed by extensive washes.
HA-tagged WT or mutant SF-1 and Myc-SUMO1 constructs were cotransfected, and after 24 h, cells were washed with cold 1× PBS and lysed in RIPA buffer containing 10 mM N-ethylmaleimide (NEM). The supernatants of the cell lysates were then incubated with mouse HA Ab-conjugated protein G beads for 2 h at 4°C with gentle rotation, followed by four washes with lysis buffer containing 10 mM NEM, after which Western blot analysis was performed.
LβT2 cells were seeded on ethanol-sterilized glass coverslips in 12-well plates and after 24 h were transfected with the pxj40 GFP empty vector, pxj40 GFP SF-1, or pxj40 GFP SF-1 K119R. Some of the cells were treated with 5 μM MG132 and/or 100 nM GnRH for 6 h, as indicated in the figure legends. The cells were fixed with cold-methanol at −20°C for 10 min and then permeabilized with PBS containing 0.1% Triton X for 20 min at room temperature. The fixed cells were washed twice with PBS and then incubated with DAPI (4′,6′-diamidino-2-phenylindole) for 20 min, followed by washing with PBS twice. The coverslips were mounted on glass slides using FluorSaveTM reagent (Calbiochem) and examined by LSM510 confocal fluorescence microscopy (Carl Zeiss).
We first assessed the effect of Pin1 overexpression on gonadotropin β-subunit gene promoter activity in mouse gonadotrope LβT2 cells using luciferase assays and saw that it increased both basal and GnRH-induced murine LHβ and FSHβ promoter activity (Fig. (Fig.1A;1A; levels of overexpression are shown in Fig. Fig.1E).1E). The ability of Pin1 to increase expression of the endogenous genes was shown by quantitative real-time RT-PCR using primers spanning an intron/exon border. As in the luciferase assays, Pin1 overexpression increased the basal and GnRH-induced LHβ and FSHβ gene transcription (Fig. (Fig.1B).1B). The finding that Pin1 increases LHβ and FSHβ gene transcription, as opposed to altering stability of the transcript, was verified by treating cells with actinomycin D and measuring rates of transcript degradation, which were similar for LHβ in cells with or without transfection of the Pin1 expression vector (data not shown).
To confirm the role of Pin1 in these effects, W34A and K63A mutants were overexpressed to test the requirement of the WW and PPIase domains. Real-time PCR confirmed that, while the WT Pin1 increases transcript levels of both genes, neither mutant was able to do the same, and the W34A mutant reduced LHβ transcript levels to below those of the control (Fig. (Fig.1C;1C; levels of overexpression are shown in Fig. Fig.1E1E).
The actual role of Pin1 in regulating expression of these genes was verified by knocking down Pin1 expression with siRNA (Fig. 1D and E). Transfection of the Pin1-targeting siRNA decreased the basal and, to a greater degree, the GnRH-increased LHβ and FSHβ mRNA levels (Fig. (Fig.1D).1D). This demonstrates that Pin1 plays a crucial role in GnRH-induced signaling pathways to induce mouse gonadotropin β-subunit expression.
Given that GnRH appeared to increase Pin1 protein levels (Fig. (Fig.1E),1E), we examined the ability of GnRH to affect Pin1 expression. LβT2 cells were exposed to 10 nM GnRH for 0, 6, 8, or 12 h, and RNA was extracted for RT-PCR. In addition to the Pin1 mRNA, LHβ and FSHβ mRNAs were also measured to confirm the effective GnRH treatment. Pin1 mRNA levels increased at 6 h but declined thereafter; a similar pattern was seen in the FSHβ mRNA levels, while those of LHβ remained elevated even after 12 h, and those of β-actin remained constant (Fig. (Fig.2A).2A). Western blot confirmed that Pin1 protein levels increased progressively with lengthening GnRH treatments, in a rapid but more moderate response than seen for calcineurin (CnA) and ATF3, while GAPDH levels remained virtually unchanged (Fig. (Fig.2B).2B). The cell specificity of the GnRH effect was shown by repeating the 2-h GnRH treatment in gonadotrope (αT3-1 and LβT2) and nongonadotrope (MEF and COS-1) cell lines; the increase in Pin1 protein levels was seen only in the gonadotrope cells (Fig. (Fig.2C2C).
To determine whether GnRH alters the phosphorylation status of Pin1, the level of Pin1 phosphorylated at Ser16 (pPin1) was examined by Western blotting. The specificity of the Ab was confirmed by overexpressing GFP-Pin1 or GFP-Pin1 S16A; IP of the GFP-tagged proteins was followed by Western analysis with the pPin1 Ab (Fig. (Fig.2D,2D, right). In GnRH-treated cells, pPin1 was easily detected after 15 min GnRH and remained at a high level after 30 min, after which it gradually decreased (Fig. (Fig.2D).2D). Both PKA and PKC are activated by the GnRH signaling cascade, and treatment of LβT2 cells with PMA or forskolin showed that both treatments strongly induced pPin1 levels, indicating involvement of either or both pathways in phosphorylating Pin1 (Fig. (Fig.2D2D).
Phosphorylation of Pin1 is reported to hinder its activity by preventing its interaction with the substrate and its nuclear localization (35), yet Pin1, although being temporarily phosphorylated by GnRH, appears required for LHβ and FSHβ gene transcription. We have shown that GnRH induces expression of the phosphatase calcineurin (30) (Fig. (Fig.2B),2B), and we therefore hypothesized that GnRH-activated calcineurin might dephosphorylate pPin1 to allow its activity. To test this, LβT2 cells were treated with a calcineurin inhibitor, cyclosporine (CsA), which was added 30 min prior to GnRH treatment for 2 h. There was a clear increase in pPin1 levels in the CsA-treated cells, which was more obvious in the GnRH-treated cells. This finding was confirmed by transfection of calcineurin siRNA, which effectively blocked calcineurin expression in untreated cells but, not surprisingly, was less effective after GnRH treatment, which likely explains the less obvious effect of the siRNA in the GnRH-treated cells (Fig. (Fig.2E2E).
To further confirm the actions of calcineurin on Pin1, co-IP was carried out to examine whether calcineurin is associated with pPin1. Pin1 was overexpressed in LβT2 cells, some of which were treated with GnRH for 0.5 h. The cells were lysed and precipitated with anti-CnA or IgG, followed by blotting with anti-CnA or pPin1 Ab. The pPin1 was coprecipitated by the calcineurin Ab but not rabbit IgG, and significantly more so in the GnRH-treated cell lysate (Fig. (Fig.2F).2F). Taken together, these results clearly demonstrate a role for GnRH-activated calcineurin in regulating Pin1 activity.
In order to elucidate the mechanisms through which Pin1 activates LHβ and FSHβ gene transcription, ChIP was carried out to examine whether Pin1 is present on the promoters of these genes. Association of Pin1 with the proximal regions of the LHβ and FSHβ gene promoters was detected in untreated LβT2 cells and after GnRH exposure for 4 h (Fig. (Fig.3A3A).
We hypothesized that Pin1 is recruited to these promoters through its interaction with gene-specific transcription factors. Therefore, co-IP was carried out using LβT2 cell lysates, and Pitx1, Egr-1, and β-catenin were detected in precipitates using the anti-Pin1 Ab but not in those using the control IgG (Fig. (Fig.3B).3B). Because the mobility of SF-1 is similar to that of the IgG heavy chain, we performed co-IP using Ab to SF-1 for the precipitation; Pin1 was detected in the lysates from untreated and GnRH-treated cells but not in those using the control IgG (Fig. (Fig.3C3C).
To confirm some of these interactions, mammalian two hybrid assays were carried out in COS-1 cells using Pin1 fused to the Gal4 DBD (in pM vector) and Pitx1, Egr-1, or SF-1 fused to VP16 AD (in pVP vector), together with Gal4-responsive luciferase reporter gene and pRL-SV40 Renilla luciferase as an internal control. Cotransfection of pM-Pin1, but not the empty pM vector, with pVP-Egr-1, pVP-Pitx1, or pVP-SF-1 revealed significant induction of the luciferase reporter gene activity (Fig. 3D and E), indicating that Pin1 interacts with Pitx1, Egr-1, and SF-1. The two-hybrid assay was also carried out using an expression vector for SF-1 with an S203A mutation. The interaction of this mutated protein with Pin1 was markedly reduced in comparison to that of the wild-type SF-1. The reduced interaction of SF-1 S203A with Pin1 was seen also in co-IP assays, in which the FLAG-tagged SF-1 proteins were expressed in LβT2 cells and precipitated before immunoblotting with Pin1 Ab (Fig. (Fig.3E3E).
Mutagenesis of the putative Pin1 target motifs in Pitx1 was also carried out to determine the residues involved in the interaction with Pin1. After expression of the FLAG-tagged Pin1 and HA-tagged Pitx1 proteins, the FLAG-tagged Pin1 was precipitated before immunoblotting with Ab to the HA tag on the Pitx1 protein or mutant proteins. The binding of FLAG-Pin1 with the Pitx1 mutants, HA-Pitx1 S207A, HA-Pitx1 S259A, or HA-Pitx1 T267AS270A, was clearly reduced in comparison with the wild-type HA-Pitx1 (Fig. (Fig.3F3F).
These results indicate that Pin1 interacts directly with Pitx1, Egr-1, and SF-1 and demonstrate a crucial role for S203 of SF-1 and for several residues of Pitx1 in these interactions.
In order to investigate whether Pin1 binding with these transcription factors plays a functional role in transcription of the gonadotropin β-subunit genes, the effect of Pin1 on transactivation of these genes by Pitx1, Egr-1, or SF-1 was examined in luciferase assays, in which their expression vectors were transfected alone or in combination. Pin1 and SF-1 had a synergistic effect on the LHβ but not the FSHβ gene promoter, while Pin1 interacted synergistically with Pitx1 on the FSHβ but not the LHβ gene promoter. Egr1, which targets only the LHβ gene, also had a synergistic effect with Pin1 on LHβ promoter activity (Fig. (Fig.4A4A).
Since we found that Pin1 binds directly to Pitx1, Egr-1, and SF-1 and enhances their transactivation, we speculated that Pin1 affects the stability of these factors. Western blotting showed that Pin1 overexpression lead to an increase in Pitx1 but a decrease in SF-1 levels, while Egr-1 and GAPDH protein levels remained unchanged over the 48 h tested (Fig. (Fig.4B).4B). Some of this could be due to Pin1 regulation of transcription of these factors, as RT-PCR after overexpression of Pin1 showed a clear increase in Pitx1 mRNA levels, while those of SF-1 decreased marginally (Fig. (Fig.4C).4C). In order to test for posttranscriptional effects of Pin1 on these proteins, constructs encoding HA-tagged Pitx1 or SF-1 were transfected, with or without Pin1 overexpression. Western blotting of the HA-tagged proteins showed the same trend of an increase in Pitx1 and a decrease in SF-1 protein levels, both of which were more pronounced following transfection of more of the Pin1 expression vector (Fig. (Fig.4D4D).
To explore the molecular mechanism underlying the Pin1-induced drop in SF-1 levels, we examined whether Pin1 is involved in ubiquitination of SF-1. Ubiquitination of wild-type HA-SF-1 and HA-SF-1 S203A was analyzed by immunoprecipitation of the tagged protein followed by Western blotting using Ab against ubiquitin. A monoubiquitinated form of SF-1 was evident (between the 50- and 75-kDa markers) following transfection of the WT but not the S203A mutant; in the latter samples, the SF-1 ubiquitination was reduced to an undetectable level (Fig. (Fig.5A).5A). Similarly, the SF-1 WT and S203A constructs were transfected with an expression vector for Myc-tagged ubiquitin, and IP was carried out using rabbit Ab to the Myc tag. The precipitated proteins were blotted with mouse Ab to HA, and again the mono- and polyubiquitinated SF-1 was visible only in the WT SF-1-transfected samples (Fig. (Fig.5B5B).
In order to assess the role of Pin1 in the SF-1 ubiquitination, in vivo ubiquitination assays were carried out in LβT2 after overexpression of FLAG-tagged SF-1, with or without Pin1 siRNA. SF-1 was clearly polyubiquitinated in LβT2 control cells, but in the Pin1 knockdown cells, SF-1 ubiquitination was reduced (Fig. (Fig.5C).5C). Further studies were carried out in HEK 293T control or Pin1 knockdown cells (expressing Pin1 but at lower levels) (Fig. (Fig.5D),5D), after overexpression of HA-tagged WT SF-1 or the S203A mutant. SF-1 was seen to be polyubiquitinated in HEK 293T control cells, but in the Pin1 knockdown cells SF-1 ubiquitination was clearly reduced, while in neither cell line did the SF-1 S203A appear to be ubiquitinated (Fig. (Fig.5D).5D). These results indicate that Pin1 increases SF-1 ubiquitination through a mechanism involving Ser203.
Ubiquitinated SF-1 is detectable in untreated LβT2 cells; however, GnRH treatment for 6 h greatly enhanced its levels (Fig. (Fig.6A).6A). Having demonstrated that Ser203 is crucial for Pin1-induced SF-1 ubiquitination, we hypothesized that phosphorylation of SF-1 by GnRH-activated ERK1/2 might lead to the increase in SF-1 ubiquitination. GnRH induced activation of ERK within 30 min, and an effect was still seen after 2 and 4 h, but treatment with the specific inhibitor U0126 virtually abolished this effect (Fig. (Fig.6B).6B). We went on to test the effect of ERK inhibition on SF-1 ubiquitination. In unstimulated cells, U0126 did not change the levels of poly-Ub SF-1; however, in GnRH-treated cells, U0126 reduced SF-1 polyubiquitination to a modest level compared to that in cells treated with GnRH alone (Fig. (Fig.6C).6C). These results suggest that phosphorylation of SF-1 at S203 by GnRH-activated ERK2 contributes to SF-1 ubiquitination.
SF-1 was reported to be constitutively phosphorylated under normal culture conditions (16), and CDK7, a component of the general transcription factor IIH, was shown to phosphorylate SF-1 on Ser203 (28). Therefore, we asked whether SF-1 ubiquitination is also affected by CDK7-mediated phosphorylation of SF-1. In vivo ubiquitination assays of WT and S203A SF-1 were performed in the absence or presence of the CDK7 inhibitor roscovitine (ROS), with or without GnRH. The treatment with ROS clearly inhibited SF-1 ubiquitination in both untreated and GnRH-treated cells (Fig. (Fig.6D).6D). The specificity of CDK7 effect was shown by overexpressing a WT or dominant negative (DN) CDK7. Overexpression of the WT protein clearly increased levels of SF-1 ubiquitination, while the DN form failed to do so and caused a minor decrease compared to levels in cells transfected only with FLAG-tagged SF-1 expression vector (seen in the SF-1 blot [Fig. [Fig.6E]).6E]). These results demonstrate that ubiquitination of SF-1 is Ser203 phosphorylation dependent and that GnRH treatment increases SF-1 ubiquitination at least partially via the activation of ERK1/2, although CDK7 likely also plays a role.
Given that SF-1 is SUMOylated at Lys119 and Lys194, and that a similar conjugation mechanism is shared for SUMOylation and ubiquitination, we examined whether SF-1 is ubiquitinated at these residues. Ubiquitination of WT HA-SF-1 and a series of mutants: HA-SF-1 S203A, HA-SF-1 K119R, HA-SF-1 K194R, and HA-SF-1 K119R and K194R (2KR) was examined. The in vivo ubiquitination assay showed that WT HA-SF-1 and HA-SF-1 K194R were efficiently polyubiquitinated in LβT2 cells, while mutation of Lys119 to Arg prevented SF-1 from being polyubiquitinated (Fig. (Fig.7A).7A). Thus, the ubiquitination site in SF-1 is Lys119.
Using an in vivo SUMOylation assay in which the same SF-1 mutant constructs were transfected, we found that SUMOylation of SF-1 was not affected by mutation of Ser203 to Ala (Fig. (Fig.7B).7B). However, SUMOylation at Lys119 does not occur when Lys194 is mutated to Arg (Fig. (Fig.7B),7B), indicating that SUMOylation at Lys194 is a prerequisite for Lys119 SUMOylation. Together, these results show that both ubiquitination and SUMOylation of SF-1 can occur at the same site, Lys119, and that this ubiquitination is Ser203 phosphorylation dependent, while SUMOylation of Lys119 is not, but requires Lys194 SUMOylation.
Since ubiquitination and SUMOylation of SF-1 can occur at the same Lys119, we sought to determine whether the Ser203 phosphorylation-dependent ubiquitination of SF-1 promotes gonadotropin β-subunit gene transcription by increasing SF-1-Pitx1 interactions and whether this is proteasome dependent. Initially, Western analysis was performed using FLAG-precipitated samples from LβT2 cell lysates after transfection with FLAG-SF-1 and HA-Pitx1 (Fig. (Fig.8A),8A), or on the endogenous proteins (Fig. (Fig.8B),8B), in the absence or presence of MG132 for 6 h. In both cases, treatment with the proteasome inhibitor increased the binding between SF-1 and Pitx1 (Fig. 8A and B).
Both SF-1 and Pitx1 are polyubiquitinated in LβT2 cells, and treatment of cells with MG132 substantially increased the levels of both polyubiquitinated proteins (Fig. (Fig.8C).8C). To exclude the possibility that their enhanced binding results from an effect of MG132 on Pitx1, we transfected LβT2 cells with FLAG-SF-1 and Myc-Ub or with FLAG-SF-1 and pxj40 Myc empty vector. The FLAG-SF-1 and associated proteins were precipitated and incubated with LβT2 cell lysates containing HA-Pitx1. Western blotting showed that Myc-Ub overexpression led to FLAG-SF-1 polyubiquitination, while the FLAG-SF-1-Ub-Myc bound more HA-Pitx1 than did the FLAG-SF-1 when transfected alone (Fig. (Fig.8D).8D). Similarly, Myc-SUMO1 was transfected with FLAG-SF-1, which revealed SUMOylated FLAG-SF-1 but led to a decrease in the interaction between FLAG-SF-1 and HA-Pitx1 (Fig. (Fig.8D).8D). These results suggest that ubiquitination of SF-1 increases, and SUMOylation decreases, the binding between SF-1 and Pitx1 and that the transcriptional activity of SF-1 results from the balance between these two modifications that compete for the same Lys119.
Polyubiquitin chains linked through Lys48 of ubiquitin reportedly target a protein for proteasomal degradation, while attachment through Lys63 is thought to indicate nonproteolytic functions, including transcriptional regulation (46, 56). Therefore, we performed ubiquitination assays to examine whether Lys48 or Lys63 is involved in the formation of polyubiquitin chains on SF-1. Lysates of LβT2 cells transfected with FLAG-SF-1 and HA-Ub K48R or HA-Ub K63R were precipitated by FLAG beads, and the precipitates were subject to Western analysis by HA Ab. Mutation of either residue failed to prevent SF-1 ubiquitination; however, mutation of Lys63 reduced polyubiquitin chain formation on SF-1 more than mutation of Lys48 (Fig. (Fig.8E).8E). This demonstrates that SF-1 can be ubiquitinated via Lys48- and Lys63-linked polyubiquitin chains and confirms the possibility that this modification is likely involved in several mechanisms of modulating SF-1 function.
Finally, to test whether the GnRH-induced ubiquitination of SF-1 causes a change in its localization, we examined the localization of GFP-SF-1 and GFP-SF-1 K119R in LβT2 cells treated with GnRH and/or MG132. In nontreated cells, GFP-SF-1 was exclusively localized in the nucleus; however, exposure to GnRH and/or MG132 resulted in both nuclear and cytoplasmic distribution of GFP-SF-1. GFP-SF-1 K119R was localized only in the nucleus, even after GnRH and/or MG132 treatment (Fig. (Fig.8F).8F). It appears, therefore, that nuclear export of SF-1 by GnRH is followed by either proteasome degradation or deubiquitination in the cytosol.
We have demonstrated that ubiquitination of SF-1 increases the binding between SF-1 and Pitx1 and that Pin1 upregulates this ubiquitination, indicating that Pin1 likely increases interaction of the two transcription factors. To test this possibility, MEF WT and Pin1−/− cells were transfected with FLAG-SF-1, HA-Pitx1, and/or HA-Egr-1 before carrying out co-IP using the cell lysates. Western blotting showed that the interaction between SF-1 and Pitx1 was markedly reduced in MEF Pin1−/− cells compared to that in MEF WT cells (Fig. (Fig.9A).9A). This contrasted with the interaction of SF-1 with Egr-1, which did not differ between the cell lines, while the interaction of Pix-1 with Egr-1 was only marginally reduced in the Pin1 knockout cells (Fig. (Fig.9A).9A). Notably, the reduced interaction of Pitx1 with SF-1 or Egr-1 was rescued by exogenous Pin1, but not to the same degree by the WW or PPIase mutants (Fig. (Fig.9B9B).
In order to confirm the factor targeted by Pin1 that is responsible for the increase in Pitx1-SF-1 interaction, we examined the binding between FLAG-SF-1 and wild-type HA-Pitx1 or mutant HA-Pitx1, in which various Pin1-binding sites were mutated. Although these residues were shown to bind Pin1 (Fig. (Fig.3E),3E), mutation in Pitx1 of Ser207, Ser259, or Thr267 and Ser270 did not lead to significant alteration of its interaction with SF-1 (Fig. (Fig.9C).9C). However, mutation of SF-1 at Ser203 markedly reduced the interaction with Pitx1 (Fig. (Fig.9D).9D). These results indicate that Pin1 increases the interaction of SF-1 with Pitx1 through targeting SF-1.
We hypothesized in this study that Pin1 plays a role in activating GnRH-stimulated gonadotropin subunit gene expression through interactions with phosphorylated transcription factors, involving changes in stability and/or protein-protein interactions. We further hypothesized that Pin1 activity might be regulated by GnRH. The data presented support this role for Pin. We show that Pin1 is required for the full GnRH response of the gonadotropin β-subunit genes, while its overexpression increased basal and GnRH-activated transcription, as well as that specifically activated by various transcription factors. One of the mechanisms of this Pin1 action is through targeting SF-1, a key factor regulating all three gonadotropin subunit genes (18, 71). We show that Pin1-mediated SF-1 isomerization coordinates phosphorylation and ubiquitination, which facilitates Pitx1 binding, to increase SF-1 transcriptional activity. Moreover, GnRH increases Pin1 expression levels and regulates its activity through both phosphorylation and dephosphorylation, indicating that Pin1 is an integral part of GnRH signaling (Fig. (Fig.1010).
Pin1 expression is reportedly mediated through the activation of E2F by various oncogenes and Ras-activated signaling, possibly involving Sp1, which can interact with E2F; notably, the Pin1 promoter has two GC boxes that might confer Sp1 binding (51). In the gonadotrope, the Ras-MEK-pathway is activated by GnRH, which stimulates E2F (4, 42), and Sp1 mediates some of the GnRH responsiveness of the LHβ gene (20), suggesting a possible mechanism of GnRH activation of Pin1 expression. However, GnRH also rapidly increased Pin1 phosphorylation on S16, which was shown to inactivate Pin1 by preventing its interaction with the substrate and its translocation into the nucleus (35). This apparent paradox is reconciled by the discovery that GnRH dephosphorylates Pin1 through calcineurin, which presumably reactives Pin1, providing an additional regulatable pathway to ensure sufficient levels of active nuclear Pin1. To our knowledge, this is the first report that links these two proteins and indicates that Pin1 activity can be regulated by the same extracellular signals that activate the phosphorylation cascade targeted by Pin1.
Pin1 was found at the promoters of both gonadotropin β-subunit genes, presumably recruited through SF-1, Pitx1, and/or Egr-1, with which it interacts and in some cases shows functional synergy, although this appears to be promoter context specific. Most striking was the functional interaction with SF-1 on the LHβ gene, despite a clear reduction by Pin1 of SF-1 protein levels, which contrasted with the opposite effect on Pitx1. SF-1 is ubiquitinated (6, 60), and we have shown that this is dependent on Pin1 and the phosphorylation of Ser203. Ser203, which is recognized and bound by Pin1, is found in the hinge region adjacent to helix 1, and its phosphorylation was shown previously to induce major conformational change in the protein, as indicated by a change in protease sensitivity. That study also showed that the activity of this hinge/helix 1 region was enhanced by phosphorylation and that the S203 phosphorylation increased cofactor recruitment, likely as a result of the change in conformation (9). It is quite possible that this conformational change is the Pin1-induced isomerization, which exposes SF-1 to the ubiquitin-conjugation machinery, leading to its interaction with Pitx1 and ultimately to its proteasome-mediated degradation.
The ability of GnRH to increase SF-1 ubiquitination was shown recently (60), and our use of an ERK-specific inhibitor indicates that it likely stems from activation of ERK1/2; the MAPK cascade was shown previously to phosphorylate SF-1 both in vivo and in vitro (9, 12, 16). However, SF-1 is also reportedly phosphorylated by CDK7 and TFIIH (28). Activation of CDK7 relies on its Thr170 phosphorylation (13), which is present even without GnRH stimulation (unpublished data), indicating that CDK7 may be constitutively active in gonadotrope cells. This low level of activation might maintain a small pool of phosphorylated SF-1 to ensure basal gene transcription, and it would explain an earlier report that SF-1 is constitutively phosphorylated following its expression in MCF7 and COS cells (16).
A number of recent works have highlighted the role for ubiquitination in transcriptional regulation, which likely involves monoubiquitination to increase certain protein-protein interactions, as well as polyubiquitination through various linkages, only some of which are recognized by the proteasome, and may facilitate promoter clearance (17, 40, 46, 54, 56, 61, 63, 66). We show here that SF-1 is modified by monoubiquitination and polyubiquitinated chains with K48 and K63 linkages, the K63 linkage being apparently more abundant. It should be noted, however, that polyubiquitination editing can occur, whereby ubiquitin first is K63 linked and then undergoes a change to K48 linkages to enable proteasomal degradation (43, 66).
We show that SF-1 ubiquitination increases SF-1 interaction with Pitx1 and requires phosphorylation of S203. Given that relatively large amounts of monoubiquitinated SF-1 were detected and that this was also increased by GnRH, we propose that this signals the increased interaction with Pitx1. The polyubiquitination and eventual promoter clearance of SF-1 are likely important for the next round of transcription initiation, as shown for other transcriptional activators (47, 63). This appears to be similar to the GnRH-induced ubiquitination of estrogen receptor α (ERα) in activation of the LHβ gene promoter, although it involves a very different mechanism, in which GnRH upregulates the ubiquitin-conjugating enzyme, Ubc4, that targets ERα (36). This pathway of SF-1-Pitx1 interaction, clearance, and subsequent degradation via Pin1-stimulated ubiquitination would explain the requirement for Pin1 and yet its reduction in SF-1 protein levels. However, this is clearly not a universal action of Pin1, as the same drop in protein levels of Pitx1 and Egr-1 were not noted, despite its interaction with both proteins at the LHβ promoter.
This model is consistent with ubiquitination of SRC-3, which is dependent on GSK3-induced phosphorylation and controls both its transactivation and its clearance from the promoter. It was proposed that the transition from mono- to polyubiquitination, which occurs during transcription, serves as a clock to regulate its lifetime (63). In a different study, Pin1 was seen to interact with phosphorylated SRC-3, regulate its cellular turnover, and enhance its functional interaction with CBP/p300 (69). These observations clearly suggest that the ubiquitin proteasome degradation of SRC-3 is phosphorylation and Pin1 dependent.
Pin1 has been linked to promoting ubiquitin-proteasome-mediated degradation of other proteins, such as Bcl2, c-Myc, and cyclin E (2, 45, 52, 68). Negative regulation by Pin1-induced ubiquitination was also shown for Daxx, which inhibits the apoptotic response as a result of its proteasomal degradation (49). A recent report showed that ubiquitination of SMAD proteins occurs after binding by Pin1, as a result of the phosphorylated SMAD's increased interaction with the E3 ligase Smurf (41). A change in protein-protein interactions in this pathway, as a result of Pin1 binding, was also shown for the RNA polymerase-binding protein Che-1, which is degraded in response to apoptotic stimuli after Pin1-induced conformational changes allow it to interact with the E3 ligase HDM2 (8). However, in these reports, the outcome of the Pin1-mediated ubiquitination is negative regulation. Our results, as well as those pertaining to the activity of SRC-3, indicate that the Pin1-induced isomerization allows access to E3 ligases which promote assorted ubiquitin linkages with diverse outcomes.
We have demonstrated that ubiquitin targets SF-1 on Lys119, which is also targeted by SUMO. SF-1 is SUMOylated at two conserved lysines, Lys119 and Lys194, which reside adjacent to the DBD and LBD, respectively, and this modification clearly represses SF-1 transactivation (5, 22, 26, 67). This was attributed to the SUMOylation on Lys194, which was shown to reduce Ser203 phosphorylation, decrease coregulator binding, and cause a selective loss of binding to certain target genes (5, 67). In contrast, we found that SF-1 SUMOylation was not affected by S203 phosphorylation and that SUMOylation on Lys194 is a prerequisite for SUMOylation on Lys119, which has been confirmed in other reports (22, 69). Our results indicate that ubiquitination of SF-1 on Lys119 enhances, while SUMOylation inhibits, its interaction with Pitx1, which functions as a crucial coactivator in gonadotropin gene expression (58).
The transcriptional activity of SF-1 is thus dynamically modulated by a pathway of posttranslation modifications. The de-SUMOylation on Lys194 which facilitates Ser203 phosphorylation allows Pin1-mediated isomerization, enabling the subsequent Lys119 ubiquitination, to facilitate the interaction with its specific coactivator, Pitx1, and promoter activation. We have demonstrated a central role for Pin1 in controlling the activity of SF-1, a key factor regulating all three subunit genes. Given that Pin1 expression and activity are also regulated by GnRH, we propose Pin1 as a novel mediator in GnRH-activated signal transduction pathways. Transcription is often stimulated by extracellular signaling involving activation of MAPK kinases which phosphorylate gene-specific transcription factors, but this is just one of many types of posttranslational modifications that determine the rapid activation and tight regulation of gene-specific transcription factors (3). It seems clear that in many cases Pin1 may have a crucial role in recognizing this phosphorylation signal and in coordinating additional modifications through isomerization to facilitate access to various modifying enzymes, which lead to diverse outcomes.
This research was supported by funding from the Ministry of Education Academic Research Fund, grant number R-154-000-410-112.
We thank Lilach Pnueli and Ye Fan for technical help. Thanks go to P. Mellon for the αT3-1 and LβT2 cells, K. P. Lu for MEF WT and MEF Pin1−/− cells, and Q. Y. Yang for HEK 293T Ctrl siRNA and Pin1 siRNA cells. We also thank J. Drouin, J. Milbrandt, and K. Parker for Pitx1, SF-1, and Egr-1 expression vectors and B. C. Low for pxj40 HA, FLAG, and Myc vectors.
Published ahead of print on 7 December 2009.