| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Gene Center and Institute of Biochemistry, University of Munich,
GSF National Research Center for Environment and Health, and
Max-Planck-Institute of Psychiatry, Munich, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and Wnt agonists revealed a modest reduction in the activation of endogenous and transfected target genes. Two-dimensional analysis of total proteins and SUMO-modified proteins in transformed pre-B cells showed no significant differences between wild-type mice and homozygous mutant mice. Taken together, our data indicate that PIASy has a modest effect on cytokine and Wnt signaling, suggesting a redundancy with other members of the family of PIAS proteins. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IFN- are the main cytokines for innate immune responses against viral infections. IFN- is produced by many cell types, whereas IFN- is produced predominately by hemopoietic cells (reviewed in Refs.6 and 7). The signaling pathways for both IFNs are similar and involve the Jak tyrosine kinase-mediated activation of STAT proteins that stimulate transcription of target genes alone or in concert with IFN regulatory factors (reviewed in Refs.8 and 9). The response of cells to Wnt signals involves the stabilization and nuclear translocation of 
-catenin, which associates with LEF/TCF proteins and activates Wnt-responsive genes (reviewed in Ref.10).
In addition to STATs and LEF1/TCF proteins, many other transcription factors interact with PIAS proteins. These include nuclear hormone receptors, such as the androgen receptor (AR), p53, Smad4, Sp3, HMGI-C, Gfi-1, IRF-1, TFII-I and yeast septins (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25). In mouse and man, four Pias genes (Pias1, Pias3, Piasx, and Piasy) have been identified, which encode proteins that share a similar domain structure but differ in their specificity of interaction with other proteins. PIAS1 and PIAS3 inhibit DNA binding of STAT1 and STAT3, respectively (1, 2). In contrast, PIASy represses STAT1, LEF1, Smad4 and the AR without interfering with DNA binding by these proteins (3, 24, 26, 27). The Piasx gene encodes two splice variants, x (ARIP3) and x (Miz1), which interact with the AR and with the homeodomain protein Msx2, respectively (28, 29). In addition, PIASx interacts with STAT4, following IL-12 stimulation of T cells (5). In most cases, the association of PIAS proteins with transcription factors results in repression (reviewed in Ref.30); however, PIAS proteins can also augment gene expression by synergizing with other transcriptional coactivators (4, 18).
PIAS proteins share a similar domain structure, including an N-terminal domain, termed SAP domain, which mediates, in part, the interaction with partner proteins (3, 31) and a central RING domain that resembles the catalytic domain of ubiquitin ligases (32, 33). Recent experiments aimed at the mechanisms by which PIAS proteins modulate the activity of their interaction partners, have shown that PIAS proteins can augment the covalent modification of proteins with the small ubiquitin-related modifier (SUMO) (3, 14, 15, 16, 22, 34, 35, 36). Two families of SUMO proteins, SUMO-1 and SUMO-2/SUMO-3 resemble ubiquitin proteins and are also attached to the 
-amino groups of lysines in target proteins (37). The sumoylation pathway resembles that of ubiquitination, with the activation of SUMO by an E1 protein and its transfer to an E2 enzyme (37, 38, 39). Multiple lines of evidence suggest that PIAS proteins act as E3 ligases for SUMO. First, PIAS proteins can cooperate with E1 and E2 enzymes in reconstituted systems to augment SUMO modification of target proteins (3, 14, 22, 35). Second, PIAS proteins contain a cysteine-rich RING domain, which is a hallmark for ubiquitin ligases and is functionally important for the sumoylation activity of PIAS proteins (3, 22, 33, 34, 35). Finally, PIAS proteins display some specificity of protein substrates, although the specificity is not as pronounced as for ubiquitin ligases (5, 22, 40).
The function of PIAS proteins as SUMO E3 ligases appears to involve multiple mechanisms. The repression of transcription factors by PIAS proteins often correlates with the subnuclear sequestration into speckles, such as promyelocytic leukemia nuclear bodies (3, 20, 21, 23, 34, 41). This observation could account for the repression of transcription factor activity without changes in DNA binding. In addition, PIASx has been found to interact with histone deacetylase-3, which is involved in repression of TFII-I family proteins (25). The activation function of PIAS proteins could also be accounted for by the sumoylation-dependent augmentation of protein-protein interactions. In nuclear hormone receptors, the SUMO modification sites coincide with the protein motifs that mediate synergy in the activation of reporters containing multimerized binding sites for the hormone receptors (42, 43, 44).
Genetic experiments in Drosophila indicated that the PIAS orthologue dPIAS, also known as Zimp, interacts functionally with the STAT orthologue stat92E to regulate blood cell and eye development (45). An additional role of PIAS in the regulation of chromosome structure and function was inferred from the identification and characterization of dPIAS as a suppressor of position-effect-variegation, Su(var)2-10 (46). Likewise, a yeast orthologue of PIAS proteins, termed Siz1 was identified through its genetic interaction with the condensing complex (47) and shown to act as a SUMO E3 ligase for septin proteins (14, 16). However, the function of PIAS proteins in the mouse has not yet been elucidated. We report the generation and analysis of a mouse carrying a targeted mutation of the Piasy gene. We find no gross phenotypic abnormalities, although the Mendelian ratio of viable homozygous mutant offspring is significantly reduced relative to that of homozygous mutant embryos at E18.5. Molecular analysis of signaling by IFN- and Wnt proteins in PIASy-deficient mouse splenocytes and mouse embryonic fibroblasts (MEFs) revealed a modest but reproducible decrease in the nuclear responses, suggesting that PIASy acts redundantly with other PIAS proteins.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A mouse genomic library from 129/Sv mice (
dashII) was screened with a Piasy cDNA probe encompassing the full-length coding sequence of mouse Piasy. A phage clone was isolated that included exons 2 through 9 of the mouse Piasy gene. A 10-kb genomic Piasy gene fragment (48), including exons 2 through 9, was excised by EcoRI and subcloned into a Bluescript vector (Stratagene, La Jolla CA). For the generation of a targeting construct a 2.1-kb genomic EcoRI/BamHI fragment, encompassing parts of intron 1, exon 1, and 217 bp of exon 2 was used in combination with a 4.2-kb genomic, PCR-amplified fragment encompassing the 3' end of intron 2, exons 3 through 9 and the 5' end of intron 9. A LacZ-PGKneo gene cassette was inserted in frame into a BamHI site located in exon 2 giving rise to the final targeting construct, where the 3' 188 bp of exon 2 and 5150 bp of intron 2 were deleted (see Fig. 1). To allow positive-negative selection, a HSV-tk expression cassette was inserted into the EagI site of the Bluescript vector. For electroporation of D3 embryonic stem cells the targeting vector was linearized with SalI. Electroporation and selection of embryonic stem (ES) cells and the processing of G418-resistant ES cell clones were performed as described (49). Homologous recombination was identified by DNA blot analysis with internal, 5' and 3' external probes (see Fig. 1 and data not shown).
|
The animals used in all studies were maintained and bred according to institutional guidelines in the mouse facility of the Institute of Biochemistry (University of Munich, Munich, Germany). A PCR-based screening method was used to distinguish wild-type, heterozygous, and homozygous mutant mice using oligonucleotide primers as follows: 60–12c, 5'-GATGCTGCTGGGCTTTGTAG-3'; 60–10nc, 5'-CGTCCAGCATGTTGAAGAAG-3'; lac1–2, 5'-GTCGGATTCTCCGTGGGAA-3'.
DNA blot analysis
Genomic DNA was isolated from ES cells and tail biopsies. Then 15 µg of each sample was digested by restriction enzymes (see Fig. 1). The fragments were separated on a 0.8% agarose gel and transferred to nylon membranes. The internal, the 5' and 3' external probes were labeled with [-32P]dCTP by Random priming (Roche Diagnostics, Indianapolis, IN) and hybridized overnight at 42°C. Blots were washed to final stringency of 0.2% SDS, 0.1% SDS at 65°C, and exposed to film.
Immunoblot analysis
Total protein extracts were separated by 10% SDS-PAGE and transferred to nitrocellulose. Immunoblots were conducted by the ECL method according to the manufacturer’s instructions (Amersham Pharmacia Biotech, Piscataway, NJ), with the Abs rat anti-PIASy (mouse) mAb.
Flow cytometry
Thymus, spleen, and bone marrow were isolated and dispersed in RPMI 1640, containing antibiotics and glutamine with 10% heat-inactivated FCS. Cells (1 x 106) were resuspended in 100 µl of FACS buffer (PBS, 1% FCS) and incubated with Ab for 20 min. Cells were washed three times with FACS buffer, and the process was then repeated with secondary Abs as necessary. After staining, cells were resuspended in 500 µl of FACS buffer containing propidium iodine. Cells were then analyzed using a FACSCalibur (BD Biosciences, San Jose, CA). Data generated was interpreted using CellQuest Software (BD Biosciences). Data shown in the figures are gated for lymphocytes by size and complexity and for living cells by excluding cells that uptake propidium iodine. Directly conjugated Abs specific for mouse CD4, CD8, Mac1, and streptavidin-conjugated PE, FITC, and allophycocyanin were obtained from Caltag Laboratories (San Francisco, CA). Directly conjugated Abs specific for mouse B220, CD43, GR1, and IgM were obtained from BD Biosciences.
-galactosidase (-Gal) detection of Piasy--Gal expressing lymphocytes
Thymus, spleen, and bone marrow cell suspensions were prepared as described above. After cell surface staining for flow cytometry, cells were resuspended in growth medium. Cells were mixed with an equal volume of 2 mM fluorescein-digalactoside (Molecular Probes, Eugene, OR), and incubated for 1 min at 37°C. Afterwards the cell suspensions were diluted 10-fold with ice-cold growth medium and incubated on ice for 30–60 min. Cells were centrifuged and resuspended in 500 µl of FACS buffer and run on a FACSCalibur (BD Biosciences).
X-Gal staining of whole mount embryos
Dissected E16.5 embryos were fixed with 4% paraformaldehyde in PBS for 10 min on ice and stained for -Gal activity by incubation in PBS containing 2 mM MgCl2, 5 mM K4Fe(CN)6, 5 mM K3Fe(CN)6, and 0.4 mg/ml 5-bromo-4-chloro-3-indolyl -D-galactoside (X-Gal) at 37°C for 2–24 h. After staining, the embryos were washed in PBS, further fixed in 4% paraformaldehyde in PBS at 4°C and cleared using a benzyl alcohol to benzyl benzoate ratio (1:2) according to standard procedures.
X-Gal staining of cryosections of mouse embryos
E14.5 and E16.5 embryos were dissected in cold PBS and dry-frozen on dry ice. Prior cryostat sectioning, embryos were embedded and oriented in tissue freezing medium (Jung; Leica, Deerfield IL). Twenty-micrometer cryosections were fixed for 15 min in 0.2% glutaraldehyde in fixing solution (PBS, pH 7.4, containing 5 mM EGTA, pH 7.3, 2 mM MgCl2) at room temperature. After washing three times for 10 min in washing solution (PBS, pH 7.4, containing 2 mM MgCl2, 0.02% Nonidet P-40), sections were stained overnight in X-Gal staining solution (washing solution, containing 4 mM potassium ferrocyanide, 4 mM potassium ferricyanide, 1 mg/ml X-Gal) at 37°C. Sections were counterstained for 75 s as indicated using Nuclear Fast Red (Vector Laboratories, Burlingame, CA) and mounted in DPX (Fluka, Buchs, Switzerland).
Plasmid constructs
Plasmids encoding for full-length LEF1 have been previously described (50, 51). The pLEF7-fos-luciferase vector and the pCMV--catenin expression vector have been previously described (52). The pRL-TK Renilla control vector (Promega, Madison, WI) was used for normalization in luciferase reporter assays.
Cell culture, transient transfections, and reporter assays
MEFs were harvested from E12.5 embryos according to standard procedures (53). Genotyping was performed as described. MEFs were cultured in high glucose DMEM supplemented with penicillin-streptomycin-glutamine and 10% FCS (Invitrogen Life Technologies, Grand Island, NY). MEFs were transfected using the Amaxa Nucleofactor (Amaxa Biosystems, Cologne, Germany) and the MEF2 Nucleofactor kit (programme A23). The total DNA concentration was kept constant by adding empty vector plasmid DNA. Cells were harvested 26–28 h posttransfection in 200 µl of reporter lysis buffer, and luciferase assays were conducted according to the manufacturer’s instructions (Promega). Firefly luciferase activities were normalized against simultaneously measured Renilla luciferase activity (Promega Dual Luciferase Reporter Assay System).
Establishment of Abelson murine leukemia virus (MuLV) transformed pre-B cell lines
Fetal livers were isolated from E16.5 embryos and dispersed in RPMI 1640 containing antibiotics and glutamine with 10% heat-inactivated FCS by standard techniques. The Abelson MuLV was used to immortalize pools of pre-B cell populations (54).
Splenocyte cultures
Spleens were isolated from 3-mo-old mice and dispersed in RPMI 1640 containing antibiotics and glutamine with 10% heat-inactivated FCS by standard techniques. Cells were cultured overnight, counted, and 1.8 x 107 cells each were replated at a density of 2.25 x 106/ml. After 24 h, splenocytes were stimulated for 2 h with 500 U/ml murine IFN- (R&D Systems, Minneapolis, MN) and 30 ng/ml murine IFN- (R&D Systems). Splenocytes were harvested by centrifugation followed by total RNA preparation using TRIzol (Invitrogen Life Technologies).
Bone marrow macrophage cultures
Bone marrow was harvested from femurs of 3-mo-old animals by rinsing marrow cavities with PBS. Cells were cultured overnight in RPMI 1640 containing antibiotics and glutamine with 10% heat-inactivated FCS, and 10 ng/ml murine M-CSF (R&D Systems). Cells were counted and 
2.5 x 106 cells per plate were reseeded on new tissue culture dishes, and macrophages were expanded for 6 days using 10 ng/ml M-CSF. Expanded macrophages were stimulated for 2 h with 500 U/ml murine IFN- (R&D Systems) and 30 ng/ml murine IFN- (R&D Systems). Cells were harvested by centrifugation followed by total RNA preparation using TRIzol (Invitrogen Life Technologies).
SYBR-green based real-time PCR
The mRNA expression level of Wnt and STAT signaling target genes was measured using SYBR Green based real-time PCR. Total RNA was isolated with TRIzol (Invitrogen Life Technologies) and 2 µg of RNA were transcribed to cDNA using 200 U SuperscriptII (Invitrogen Life Technologies) and 0.1 nM oligo(dT)12–18 in a total volume of 20 µl. For one real-time reaction, 2x SYBR Green Mix (Applied Biosystems) was supplemented with 1/40 of synthesized cDNA in addition to the appropriate oligonucleotide primer pair and run on the AbiPrism 7000 sequence detection system (Applied Biosystems, Foster City, CA). Reverse transcriptase controls were done in parallel without adding enzyme. Obtained cycle numbers were normalized to -actin. All primer sequences were designed with the Primer Express software (Applied Biosystems). The primer sequences and their final concentrations used were as follows: Axin2 (forward, 600 nM), 5'-GAGGTGGTACCTTGCCAAAA-3', Axin2 (reverse, 50 nM), 5'-TTCCTGTCCCTCTGCTGACT-3', IFN regulatory factor (Irf)1 (forward, 600 nM), 5'-CCCTGGCTAGAGATGCAGATTAAT-3', Irf1 (reverse, 600 nM) 5'-CATGGAATCTGGAAGATCATCTCTT-3', -actin (forward, 40 nM), 5'-TGTGGTGGTGAAGCTGTAGC-3', -actin (reverse, 40 nM), 5'-GACGACATGGAGA AGATCTGG-3'.
Two-dimensional gel electrophoresis of total cell extracts
Wild-type and Piasy-lacZ homozygous mutant Abelson MuLV-transformed pre-B cell lines were cultured under standard conditions. Total cell extracts were generated in two-dimentional gel sample buffer (7 M urea, 2 M thiourea, 2% CHAPS, 100 mM DTT, 0.2% biolyte 3-10, (Bio-Rad, Hercules, CA), 10 mM iodacetamide, 0.5 mM PMSF in the presence of standard protease inhibitors) and the total protein concentration was determined using PlusOne 2-D Quant kit (Amersham Biosciences, Piscataway, NJ). A total of 300 µg of protein was separated in the first dimension using ReadyStrip IPG strip (Bio-Rad), pH 3–10, nonlinear, according to suppliers instructions (Bio-Rad). Focusing proceeded from 0 up to 60 kVh over 9 h followed by 12% SDS-PAGE in the second dimension. After fixation (5% acetic acid, 5% methanol) protein spots were visualized by colloidal Coomassie staining. For anti-SUMO1 immunoblot analysis (anti-GMP-1 mouse monoclonal, 1:1000; Zytomed, Berlin, Germany), two-dimensional gel electrophoresis was performed as described followed by standard blocking and immunoblotting techniques.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
With the aim of inactivating the murine Piasy gene in vivo, we used a cloned 10-kb genomic Piasy DNA fragment to generate an insertion-type targeting construct, in which a LacZ-PGKneo gene cassette has been inserted in-frame into the second exon of the Piasy gene (Fig. 1A). Homologous recombination of the targeting construct in murine ES cells should yield cell clones in which amino acid 81 of PIASy has been fused with the N terminus of the bacterial -Gal protein. This targeting strategy should allow for the visualization of the expression pattern of the Piasy gene in heterozygous mutant mice and the inactivation of the gene in homozygous mutant mice because the fusion protein lacks most of the PIASy protein domain, including the functionally important RING domain. Mutations in the RING domain have been found to abrogate the function of PIAS proteins in vitro and in transfection assays (3, 22, 33, 34, 35), and therefore we will refer to the homozygous Piasy-lacZ mutation also as Piasy–/– mutation.
The linearized targeting construct was transfected into D3 ES cells and clones that had been subjected to positive and negative selection were screened for homologous recombination by digesting genomic DNA with EcoRV and hybridizing DNA blots with an internal probe, derived from genomic sequences at the end of intron 2 of the Piasy gene (Fig. 1A, data not shown). Ten targeted ES cell clones were identified among 108 analyzed colonies and the homologous recombination events were confirmed by hybridization of XhoI/HindIII- and BamHI-digested genomic DNA with 5' and 3' external probes, respectively (Fig. 1, A, C, and D). Six targeted ES cell clones were microinjected into C57BL/6 blastocysts and chimeric males were crossed with C57BL/6 wild-type females. One chimera, derived from ES cell clone 198, transmitted the targeted allele through the germline. The genotypes of the F1 offspring were determined by PCR (data not shown) and DNA blot analysis of BamHI-digested genomic DNA that had been hybridized with a 3' external probe (Fig. 1E). Heterozygous animals did not show any obvious differences in phenotype or fertility as compared with wild-type littermates (data not shown). However, genotyping of 209 offspring from heterozygous intercrosses revealed a significantly reduced Mendelian frequency of 14%, instead of 25%, homozygous mutant (Piasy–/–) mice. The Mendelian frequency of homozygous mutant Piasy–/– embryos up to E18.5 was normal, suggesting a decreased perinatal viability. The surviving PIASy-deficient mice showed no obvious phenotype and were fertile.
To confirm that the targeted Piasy allele does not produce functional PIASy protein (Fig. 1B), we prepared total protein extracts from testis of wild-type and homozygous mutant (Piasy–/–) mice and assayed the extracts for the presence of PIASy protein by immunoblot analysis. Using a rat anti-PIASy mAb directed against amino acids 1 to 97 of the mouse PIASy protein, we failed to detect wild-type PIASy protein in the mutant tissue extract, although the slower-migrating PIASy--Gal fusion protein was readily detectable (Fig. 1, B and F).
Analysis of Piasy expression during embryonic development and morphology of PIASy-deficient mice
With the aim of visualizing individual cells and tissues that express the Piasy gene in the embryo, we performed whole mount -Gal staining of heterozygous Piasy-lacZ embryos at E12.5 and E14.5 (Fig. 2B and data not shown). Abundant expression of PIASy--Gal was detected in multiple tissues, including brain, skin, limb buds, tail bud, whisker follicles, and pelage hair follicle primordia (Fig. 2B). In sagittal cryosections of heterozygous Piasy-lacZ embryos at E16.5, an almost ubiquitous expression pattern could be observed with stronger expression in the neopallial cortex and the ventricular zone of the brain, the midbrain, the developing cerebellum, throughout the spinal cord and in the olfactory epithelium (Fig. 2E). Strong PIASy expression was also detected in the epidermis, the dorsal surface of the tongue, in the trapezius muscle and the cartilage primordium of the ribs and bones, the thyroid cartilage, tracheas, and lung (Fig. 2E). In wild-type embryos, no -Gal staining was detected, except the duodenum, indicating the observed expression in heterozygous mice is specific (Fig. 2D). These results correspond to previous in situ hybridization studies of E8.5 up to E16.5 mouse embryos, which detected Piasy transcripts in the limbs, the neuroepithelium, and hair follicles (55). Taken together, these studies suggest that PIASy may play a modulatory role during the development of multiple organs.
|
-Gal fusion protein is detected in Piasy-lacZ homozygous mutant mice (Fig. 2F). Although adult Piasy–/– mice also show no obvious phenotype, the basis for the reduced viability of these mice remains to be established. PIASy expression in hemopoietic cells and analysis of cell populations in PIASy-deficient mice
To address the question of whether PIASy is specifically expressed in distinct hemopoietic cell populations, we incubated cells of lymphoid organs with fluoresceine-digalactoside, a fluorogenic substrate for -Gal to detect Lac-Z-positive cells by flow cytometry. Using this approach, we found that PIASy is abundantly expressed in developing and mature T cells of the thymus (Fig. 3A). Histogram overlays of cells gated for PIASy--Gal expression indicated that 98–99% of CD4+CD8+ double positive cells, CD4+ single positive and CD8+ single positive cells express PIASy at abundant levels (Fig. 3A). Analysis of the T cell compartments in thymus and spleen of wild-type mice and Piasy-lacZ homozygous mutant mice by flow cytometry indicated that the numbers of CD4+CD8+ double positive cells, as well as the numbers of CD4+ and CD8+ single positive cells, are equivalent in both wild-type and mutant mice (Fig. 3, B and C).
|
-Gal (Fig. 3, D and F). However, the analysis of the B cell compartments in bone marrow and spleen of homozygous mutant mice showed a normal B220/CD43 profile and B220/IgM profile, respectively (Fig. 3, E and G).
Finally, PIASy--Gal expression was detected at lower levels in myeloid cells of bone marrow, using the markers Mac1 and GR1 (Fig. 3H). Analysis of the myeloid compartment in the bone marrow of Piasy–/– mice showed modest but reproducible 20% increase in the numbers of Mac1+GR1+ double positive cells (Fig. 3I). Thus, PIASy is abundantly expressed in multiple cell types of the lymphoid and myeloid lineages. However, PIASy is not essential for normal B and T cell development and has a modest effect on myeloid cell differentiation.
Cytokine signaling in Piasy-lacZ homozygous mutant mice
IFN- and IFN- signal to the nucleus by activating the latent cytoplasmic transcription factor STAT1, which is translocated to the nucleus and activates IFN-responsive target genes (reviewed in Refs.6 and 7). PIASy was shown to act as a transcriptional corepressor of STAT1 and the association of these proteins was found to be augmented by IFN signals (27). This observation has been interpreted to suggest that PIASy plays a role in the attenuation of the IFN response. STAT1 is indispensable for the IFN pathways (56, 57) and therefore, we examined the ability of IFN- and IFN- to induce target gene transcription in splenocytes and bone marrow-derived macrophages from wild-type or Piasy-lacZ homozygous mutant mice. Using quantitative real-time PCR analysis to detect transcripts of Irf1 gene, a direct target of STAT1 (58, 59), we found that Irf1 gene expression was induced 5- to 9-fold in IFN--treated wild-type splenocytes (Fig. 4A) and 7- to 17-fold in bone marrow-derived wild-type macrophages (Fig. 4B). For these experiments, we analyzed duplicate samples of two mice to assess the variabilities of the nuclear responses to cytokine signaling. IFN--stimulation resulted in a 14- to 16-fold augmentation of Irf1 expression in wild-type splenocytes (Fig. 4A, PIASy+/+) and 26-fold stimulation in wild-type macrophages (Fig. 4B, PIASy+/+).In splenocytes and macrophages from PIASy-deficient mice, the basal and IFN--induced levels of Irf1 mRNA expression were both similar to those of wild-type mice (Fig. 4), suggesting that the response to IFN- is normal in the absence of PIASy. However, the response of the mutant splenocytes and macrophages to IFN- was reproducibly reduced by a factor of 1.5 and 2, respectively (Fig. 4). In particular, the stimulation of wild-type macrophages by IFN- resulted in an 26-fold induction of Irf1, whereas only a 9- to 16-fold induction was observed in the PIASy-deficient cells (Fig. 4B).
|
LEF1/TCF proteins are nuclear mediators of Wnt signals that activate transcription of Wnt-responsive target genes in conjunction with -catenin (reviewed in Ref.10). LEF1 and TCF4 were found to interact specifically with PIASy, whereby the activity of LEF1 is repressed and the activity of TCF4 is enhanced (3, 4). Biochemical and genetic analyses identified Axin2, which acts as a negative regulator of soluble -catenin, as a target of Wnt signaling, thus establishing a negative feedback loop for Wnt signaling (60). To examine the Wnt responsiveness of the endogenous Axin2 gene in wild-type and Piasy-lacZ homozygous mutant mice, we determined the levels of Axin2 mRNA in MEF cell lines that have been treated with LiCl, which inhibits glycogen-synthase-kinase-3 and acts as a surrogate Wnt signal (61). For this experiment, we used two wild-type and four mutant mice to derive primary MEFs, and we performed the expression analysis with duplicate samples in three independent experiments. Determination of basal Axin2 levels in untreated MEFs by quantitative real-time PCR analysis did not reveal any significant differences between wild-type and Piasy-lacZ homozygous mutant cells (Fig. 5A). Stimulation of the MEF cells derived with LiCl resulted in quite variable induction efficiencies, ranging from 5.6-fold (Piasy–/– MEF 17) up to 24.2-fold (Piasy–/– MEF 12) (Fig. 5A). The induction efficiencies of the wild-type MEF cell lines ranged between 7.8-fold (wild-type MEF 5) and 13.1-fold (wild-type MEF no. 18). Therefore, the Wnt response of Axin2 in PIASy-deficient MEFs is comparable to that in wild-type MEFs.
|
-catenin-dependent activation of this reporter construct in transiently transfected 293 cells (3). In a first experiment, we compared the expression levels of the transfected LEF1 reporter in wild-type with PIASy-deficient MEF cells in the absence or presence of the surrogate Wnt signal, LiCl, but without the addition of exogenous LEF1 or -catenin expression constructs (Fig. 5B). The basal levels of reporter activity were similar in NaCl-treated wild-type and mutant MEF cells (Fig. 5B). LiCl treatment of these cells induced a 2- to 3-fold increase in reporter gene expression in both wild-type and mutant MEFs, suggesting that the loss of functional PIASy had no effect on basal or Wnt-induced reporter gene activity. Because the effects of LiCl in MEFs were limited, possibly due to the presence of low levels of endogenous LEF1/TCF proteins, we compared the activity of the LEF7-fos-luciferase reporter in wild-type and Piasy-lacZ homozygous mutant MEFs in the presence of exogenous LEF1 and -catenin. Cotransfection of LEF1 and -catenin expression plasmids into wild-type MEFs resulted in a robust 11.2-fold activation of the LEF7-fos-luciferase reporter (Fig. 5C, wt MEF 18). However, in Piasy–/– MEFs, the transcriptional activation of the reporter construct was reduced almost 3-fold (Fig. 5C, MEF 11), whereas no effect was observed in cells that have been transfected with only LEF1 or -catenin expression plasmids. This result suggests that PIASy may modulate the Wnt responsiveness of genes in vivo. The modest effects of the deficiency of PIASy could be, in principle, accounted for by a redundancy with another member of the PIAS family and/or an up-regulation of the expression of another PIAS gene. Therefore, we examined the expression of the PIAS genes in wild-type and Piasy–/– MEFs. To this end, we performed a quantitative real-time PCR analysis to detect transcripts of the Pias1, Pias3, and Piasx genes. The levels of expression of these Pias genes were found to be similar in wild-type and PIASy-deficient MEFs, suggesting that the absence of PIASy does not result in a compensatory up-regulation of the expression of another Pias gene (data not shown).
Two-dimensional gel electrophoresis of wild-type and Piasy–/– cell extracts
To investigate whether PIASy modulates the expression of Wnt or STAT target genes and to assess whether a deficiency of PIASy alters the SUMO modification of proteins, we adopted a protein expression profiling approach. Total cell extracts of wild-type and Piasy–/– Abelson-MuLV-transformed pre-B cells were separated in two dimensions by isoelectric focusing followed by SDS-PAGE. Protein spots were visualized by colloidal Coomassie staining and the gels were scanned for PDQUEST (Bio-Rad) software analysis (Fig. 6, A and B). We prepared pre-B cell extracts from two wild-type and two PIASy-deficient mice and assayed the extracts in duplicates by two-dimensional electrophoresis (Fig. 6, A and B). The comparison of the pattern of Coomassie-stained protein spots in wild-type and mutant cell extracts by the PDQUEST software (Bio-Rad) showed no differential expression pattern of proteins or shifts in protein migration due to potential secondary modifications of proteins comparing total cell extracts from wild-type (Fig. 6A) and Piasy-lacZ homozygous mutant Abelson MuLV-transformed pre-B cells (Fig. 6B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. In addition, we observed a reduced viability of the homozygous mutant mice and found that the activation potential of a Wnt-responsive reporter by LEF1 and -catenin is decreased. Together, these data suggest that PIASy may play a regulatory role in signaling by IFN- and Wnt proteins in vivo.
The subtle effects of the Piasy-lacZ mutation raise the question of whether the insertion of the lacZ gene into the second exon of Piasy generates a null allele or whether homozygous Piasy-lacZ mice retain a residual activity of PIASy protein. The immunoblot analysis of testis tissue extracts from mutant mice indicates that neither wild-type PIASy nor smaller forms of PIASy protein can be detected. This observation rules out the principle possibility that exon skipping generates a PIASy protein, lacking only the exon with the lacZ insertion. Although the N-terminal SAP domain is retained in the PIASy--Gal fusion protein, this protein lacks the central RING domain, which contributes to the interaction with partner proteins and is essential for the functional activity of PIASy. Mutations in the RING domain of PIASy and other PIAS proteins have been shown to abrogate the ability of these proteins to modulate the activity of partner proteins and act as SUMO E3 ligases (3, 22, 34, 35).
Several lines of evidence suggest that other PIAS proteins may compensate, at least in part, for the deficiency of PIASy in the mouse. First, PIAS proteins share a similar domain structure and show overlapping specificities in the association and sumoylation of partner proteins. For example, transient overexpression of PIASy and PIAS1 both result in the repression of a STAT1-responsive reporter construct in transfection experiments (2, 27). Moreover, PIAS1 and PIASx proteins both act as SUMO E3 ligases and corepressors of AR and p53 (12, 22, 34, 35). However, some specificity of PIAS proteins can be demonstrated in transfection assays. PIASx, but not PIAS3, antagonizes the IL-12-dependent activation of a reporter gene by STAT4 (5). Likewise, the repression and sumoylation of the nuclear matrix protein SATB2 is mediated specifically by PIAS1, but not by PIASx and PIAS3 (62). Second, PIAS proteins are widely expressed and show overlapping pattern of expression in the developing mouse (55). Our analysis of heterozygous and homozygous Piasy-lacZ embryos shows an almost ubiquitous expression at a low level but a more specific expression in multiple tissues. Third, a redundancy of PIAS proteins has been found in yeast, in which two PIAS-related proteins, termed Siz1 and Siz2, act redundantly in promoting SUMO modification of septins (14). Both Siz1
and Siz2 strains show no obvious phenotypes, but the double mutation resulted in poor cell growth and accumulation in G2/M of the cell cycle. Moreover, the overall sumoylation of proteins was found to be diminished in Siz1 cells but was virtually abrogated in Siz1Siz2 cells (14).
One obvious effect of the Piasy-lacZ mutation was the reduced viability of newborn mice. In contrast to the normal Mendelian frequency of homozygous mutant mice before birth, we observed only 14% homozygous mutants at weaning age. As we could not detect any obvious abnormalities, the basis for the reduction of the viability is unknown. The reduced viability may also reflect some heterogeneity in the genetic background of the mice, which were generated in a 129 strain but were backcrossed to a C57BL/6 strain. Strain-specific variations in the penetrance of mutations have been observed (63).
Although PIASy is expressed in multiple hemopoietic cell lineages, the only effect of the Piasy-lacZ mutation that we observed was a 25% increase in the number of Mac1- and Gr1-positive myeloid cells in the bone marrow. This modest increase in myeloid cells is reproducible in multiple mice and is consistent with the finding that the expression of Pias3 is lost in leukemic cells in which STAT3 is persistently activated (64, 65). Although PIASy has not been implicated in the regulation of STAT3, it is possible that PIASy and PIAS3 act redundantly in myeloid cells to modulate cytokine signaling.
In macrophages of the bone marrow and to a lesser extent in splenocytes, we detected a reproducible 2- to 3-fold decrease in the IFN- responsiveness of the endogenous Irf1 gene, a member of the IRF family of transcription factors, that is a direct target gene of IFN signaling (58, 59). IRF1 is involved in the regulation of the expression of both IFN-stimulated genes and IFN genes themselves. The induction of Irf1 transcription by IFN- in wild-type macrophages appears to be governed specifically by STAT1 because no Irf1 induction is observed in IFN--treated macrophages of Stat1–/– mice (56, 57). The decrease in the IFN- response of Irf1 is surprising, given the proposed function of PIAS proteins in the down-regulation of the activities of STAT proteins. Several possibilities could account for this difference. First, in our analysis we examine the effects of the PIASy deficiency in the context of a natural gene. In contrast, the effects of PIASy on STAT1 activity were determined with synthetic reporter constructs containing multimerized STAT1 binding sites, and in conjunction with overexpressed PIASy protein (27). The context of transcription factor binding sites has been shown to be important for the proper regulation of the IFN- gene by IRF, NF-
B, ATF2, and HMG-I/Y, which act in concert to mediate a proper transcriptional response (66, 67). For example, replacement of the ATF2-c-jun site with an AP-1 site in the IFN- promoter, or reversal of the orientation of the asymmetric ATF2-c-jun site impairs its virus induction (68). Therefore, changes in the context of transcription factor binding sites can result in different regulatory contributions of proteins. Second, we examined the effects of PIASy in primary cells and without the overexpression of proteins, which occurs typically in transfection assays and may squelch a regulatory response by titrating regulatory cofactors. Although the relatively modest effect of the PIASy deficiency on Irf1 expression may reflect a redundancy of PIASy with other PIAS proteins, such as PIAS1, we cannot rule out the possibility that the contribution of PIAS proteins to the regulation of signaling by STAT proteins is limited. The overexpression of PIASx augments the activation of a synthetic reporter construct by STAT1 in IFN--treated U3a cells, which lack endogenous STAT1 only 2-fold (36).
In contrast to the modest reduction of the IFN- response in PIASy-deficient macrophages and splenocytes, no changes could be observed in the IFN- response. Both IFNs use STAT1 as a nuclear mediator but the responses differ in the specificity of transcription factor activation. IFN- activates STAT1 as a dedicated transcription factor, whereas IFN- activates both STAT1 and STAT2 (69, 70). In addition, other signaling molecules have been found to activate STAT1 in vitro. These have been shown to include epidermal growth factor, M-CSF, IL-6, IL-10, and others (71, 72, 73, 74). STAT1-deficient mice show a normal responsiveness to these other cytokines, whereas both IFN-- and IFN--signaling was impaired, indicating that STAT1 is a transcription factor dedicated to IFN signaling (56, 57, 75). The basis for the differential effect of PIASy on signaling by IFN- and IFN- is unclear but it may be related to the cross-talk between these signaling pathways (6).
Based on the association of LEF1/TCF proteins with PIASy in vitro, we also anticipated an alteration in the expression of a Wnt-responsive reporter construct by LEF1 and -catenin in PIASy-deficient MEFs. We have previously shown that PIASy represses the activity of LEF1 in transiently transfected HeLa or 293 cells (3). In addition, PIASy was found to augment the activation of a LEF1/TCF reporter by TCF4 and -catenin (4), which is consistent with our observation that the activation potential of LEF1 in PIASy-deficient MEFs is reduced relative to that in wild-type MEFs. In our previous transfection experiments, in which we observed repression of the LEF1 activity by PIASy, we have used 293 and HeLa cells and noted that the repression was not observed in fibroblastic cells, suggesting that the regulation of transcription factor activity by PIASy is cell type-specific (3). In contrast to the changes of LEF1- and -catenin-mediated activation of a reporter construct in PIASy-deficient MEFs, we did not detect any significant change in the Wnt-response of the endogenous Axin2 gene, a known Wnt target gene (60). In MEFs, the amount of LEF/TCF proteins is possibly too low to elicit a strong Wnt response and PIASy may have an effect only under conditions of strong Wnt signals.
A common hallmark of PIAS proteins is their ability to augment the modification of proteins with SUMO (38). Therefore, we examined whether the sumoylation pattern of proteins is altered in PIASy-deficient cells. In transformed pre-B cells from mutant mice, we did not detect a significant change in the pattern of protein expression relative to wild-type mice. Notably, the sumoylation of proteins by PIAS proteins does not necessarily correlate with changes in their activity. For example, PIAS1 and PIASx both modulate the activity of STAT1 in transfection assays (36, 40). However, the sumoylation of STAT1 by PIASx does not correlate with changes in transcriptional activation as a mutation of the SUMO acceptor site at lysine 703 of STAT1 has no effect on STAT1 activity (36). Likewise, mutations of two SUMO acceptor sites of LEF1 have no effect on the activity of LEF1, although the association with PIASy and the presence of an intact RING domain is required for the repression of LEF1 in 293 cells (3). In contrast, sumoylation of some proteins, such as AR and SATB2 alters their activities (35, 62). Thus, PIAS proteins may use multiple mechanisms to modulate the activities of partner proteins.
In conclusion, the analysis of the PIASy-deficient mice has revealed a functional role for PIASy in the viability of mice and IFN- signaling, although the effects are modest. Therefore, it will be of interest to examine the effects of mutations of multiple Pias genes in the mouse to determine the role of PIAS proteins in sumoylation and regulation of transcription factor activity. While this study was under review, an independent study analyzing the effects of a targeted mutation of Piasy came to similar conclusions (76).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 This work was supported by a Grant of the German Research Foundation TR-5. 
2 Address correspondence and reprint requests to Dr. Rudolf Grosschedl at the current address: Max-Planck-Institute of Immunobiology, Stuebeweg 51, 79108 Freiburg, Germany. E-mail address: grosschedl{at}immunbio.mpg.de
3 Abbreviations used in this paper: PIAS, protein inhibitor of activated STAT; LEF, lymphoid enhancer binding factor; TCF, T cell-specific factor; SUMO, small ubiquitin-like modifier; ES, embryonic stem; IRF, IFN regulatory factor; AR, androgen receptor; -Gal, -galactosidase; X-Gal, 5-bromo-4-chloro-3-indolyl -D-galactoside; MuLV, murine leukemia virus; MEF, mouse embryonic fibroblast.
Received for publication May 21, 2004. Accepted for publication August 20, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-catenin-dependent activation of Tcf-4. EMBO J. 22:2047.[Medline]
/ revisited. Nat. Rev. Mol. Cell Biol. 2:378.[Medline]
-subunit of epithelial Na+ channel (-ENaC) transcription by Ras activation in salivary epithelial cells. J. Biol. Chem. 276:29805. family signaling. J. Biol. Chem. 278:27853.. Proc. Natl. Acad. Sci. USA 99:12807. function as SUMO-E3 ligases toward androgen receptor and repress androgen receptor-dependent transcription.. J. Biol. Chem. 277:41311. enhancer function. Genes Dev. 11:640.-catenin. Mol. Cell. Biol. 18:4807., in developing skin and neurons. J. Mol. Neurosci. 14:107.[Medline]
, , or using oligonucleotide arrays. Proc. Natl. Acad. Sci. USA 95:15623.B dependent virus induction of the human IFN- gene. Cell 71:777.[Medline]
gene enhancer by transcription factors and the architectural protein HMG I(Y). Cell 83:1101.[Medline]
interferon enhanceosome is dependent on ATF-2-c-jun heterodimer orientation. Mol. Cell. Biol. 20:4814. triggers transcription through cytoplasmic tyrosine phosphorylation of a 91 kDa DNA-binding protein. Science 258:1808. and STAT3 complexes in human T cells and monocytes. J. Immunol. 155:1079.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  T. Kubota, M. Matsuoka, T.-H. Chang, P. Tailor, T. Sasaki, M. Tashiro, A. Kato, and K. Ozato Virus Infection Triggers SUMOylation of IRF3 and IRF7, Leading to the Negative Regulation of Type I Interferon Gene Expression J. Biol. Chem., September 12, 2008; 283(37): 25660 - 25670. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Martin, K. Schwamborn, H. Urlaub, B. Gan, J.-L. Guan, and A. Dejean Spatial Interplay between PIASy and FIP200 in the Regulation of Signal Transduction and Transcriptional Activity Mol. Cell. Biol., April 15, 2008; 28(8): 2771 - 2781. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Tahk, B. Liu, V. Chernishof, K. A. Wong, H. Wu, and K. Shuai Control of specificity and magnitude of NF-{kappa}B and STAT1-mediated gene activation through PIASy and PIAS1 cooperation PNAS, July 10, 2007; 104(28): 11643 - 11648. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Reindle, I. Belichenko, G. R. Bylebyl, X. L. Chen, N. Gandhi, and E. S. Johnson Multiple domains in Siz SUMO ligases contribute to substrate selectivity J. Cell Sci., November 15, 2006; 119(22): 4749 - 4757. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Song, S. Bhattacharya, A. A. Yunus, C. D. Lima, and C. Schindler Stat1 and SUMO modification Blood, November 15, 2006; 108(10): 3237 - 3244. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Albor, S. El-Hizawi, E. J. Horn, M. Laederich, P. Frosk, K. Wrogemann, and M. Kulesz-Martin The Interaction of Piasy with Trim32, an E3-Ubiquitin Ligase Mutated in Limb-girdle Muscular Dystrophy Type 2H, Promotes Piasy Degradation and Regulates UVB-induced Keratinocyte Apoptosis through NF{kappa}B J. Biol. Chem., September 1, 2006; 281(35): 25850 - 25866. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Motohashi, F. Katsuoka, C. Miyoshi, Y. Uchimura, H. Saitoh, C. Francastel, J. D. Engel, and M. Yamamoto MafG Sumoylation Is Required for Active Transcriptional Repression Mol. Cell. Biol., June 15, 2006; 26(12): 4652 - 4663. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Hattori, H. Eberspaecher, J. Lu, R. Zhang, T. Nishida, T. Kahyo, H. Yasuda, and B. de Crombrugghe Interactions between PIAS Proteins and SOX9 Result in an Increase in the Cellular Concentrations of SOX9 J. Biol. Chem., May 19, 2006; 281(20): 14417 - 14428. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. Sharrocks PIAS proteins and transcriptional regulation--more than just SUMO E3 ligases? Genes & Dev., April 1, 2006; 20(7): 754 - 758. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Yueh, J. Leung, S. Bhattacharyya, L. A. Perrone, K. de los Santos, S.-y. Pu, and S. P. Goff Interaction of Moloney Murine Leukemia Virus Capsid with Ubc9 and PIASy Mediates SUMO-1 Addition Required Early in Infection J. Virol., January 1, 2006; 80(1): 342 - 352. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. van den Akker, S. Ano, H.-M. Shih, L.-C. Wang, M. Pironin, J. J. Palvimo, N. Kotaja, O. Kirsh, A. Dejean, and J. Ghysdael FLI-1 Functionally Interacts with PIASx{alpha}, a Member of the PIAS E3 SUMO Ligase Family J. Biol. Chem., November 11, 2005; 280(45): 38035 - 38046. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H Santti, L Mikkonen, A Anand, S Hirvonen-Santti, J Toppari, M Panhuysen, F Vauti, M Perera, G Corte, W Wurst, et al. Disruption of the murine PIASx gene results in reduced testis weight J. Mol. Endocrinol., June 1, 2005; 34(3): 645 - 654. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ihara, H. Yamamoto, and A. Kikuchi SUMO-1 Modification of PIASy, an E3 Ligase, Is Necessary for PIASy-Dependent Activation of Tcf-4 Mol. Cell. Biol., May 1, 2005; 25(9): 3506 - 3518. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) indicated. The 5', 3', and internal probes are shown as bold bars. The targeting construct includes the lacZ coding sequence, fused in-frame with amino acid 81 of PIASy encoded by exon 2, in combination with a PGK-neor gene cassette (
). The length of the fragments generated by XhoI/HindIII (5' external probe), EcoRV (internal probe), and BamHI (3' external probe) restriction digests of the wild-type and mutant allele are indicated below. Representative restriction sites are indicated above the lines as: B, BamHI; H, HindIII; R, EcoRI; X, XhoI. B, Wild-type PIASy protein is shown schematically on the top. PIASy contains a putative chromatin-binding SAP (SAP) domain, a C2HC3 RING (RING), a C-terminal serine-rich and acidic (Ser/Ac) domain. The nuclear localization signal (NLS, 
) is shown. IFN-
