|Home | About | Journals | Submit | Contact Us | Français|
The Wip1 gene is a serine/threonine phosphatase that is induced in a p53-dependent manner by DNA-damaging agents. We show here that Wip1 message is expressed in moderate levels in all organs, but is present at very high levels in the testes, particularly in the postmeiotic round spermatid compartment of the seminiferous tubules. We have confirmed that Wip1 mRNA is induced by ionizing radiation in mouse tissues in a p53-dependent manner. To further determine the normal biological function of Wip1 in mammalian organisms, we have generated Wip1-deficient mice. Wip1 null mice are viable but show a variety of postnatal abnormalities, including variable male runting, male reproductive organ atrophy, reduced male fertility, and reduced male longevity. Mice lacking Wip1 show increased susceptibility to pathogens and diminished T- and B-cell function. Fibroblasts derived from Wip1 null embryos have decreased proliferation rates and appear to be compromised in entering mitosis. The data are consistent with an important role for Wip1 in spermatogenesis, lymphoid cell function, and cell cycle regulation.
The p53 tumor suppressor protein plays a critical role in integration of signals that regulate cell growth and death in response to a variety of stressors (11, 12, 26). Loss or mutation of the gene encoding the p53 protein predisposes a cell to neoplastic transformation (8, 21, 23). Mutation of the p53 gene has been observed in half of all human cancers, and loss of p53 signaling may play a role in over 80% of all tumors (8, 12, 13).
In response to a variety of cell stresses, including DNA damage, aberrantly activated oncogenes, and hypoxia, p53 is phosphorylated on multiple serine and threonine residues, becomes stabilized, and accumulates in the nucleus (11, 12, 26). The activated p53 protein is a transcription factor that can upregulate as well as downregulate a battery of genes which participate in the biological response to the cell stress. This response can be in the form of cell cycle arrest, apoptosis, differentiation, or senescence. Any of these responses are likely to prevent the outgrowth of a nascent cancer cell.
Among the many p53 transcriptional targets that have recently been identified is the wild-type p53-induced phosphatase (Wip1) (5). The Wip1 gene exhibits a close similarity to type 2C phosphatases (2, 5). The type 2C phosphatases are distinguished from the PP1, PP2A, and PP2B phosphatases by their insensitivity to phosphatase inhibitors such as okadaic acid and requirement for divalent cations (Mg2+ or Mn2+) (3, 15). The type 2C phosphatases have been identified in a wide range of organisms and have often been associated with stress response, sexual differentiation, and cell cycle control (3, 15).
The Wip1 gene was originally identified in a screen for p53 target genes (5). Wip1 was shown to be upregulated in a p53-dependent manner by ionizing radiation (IR) and by UV radiation (UV) (5, 25). Following IR, Wip1 mRNA was rapidly induced and the protein was localized to the nucleus. The first specific target of the Wip1 phosphatase has recently been shown by Takekawa et al. (25) to be the p38 mitogen-activated protein (MAP) kinase protein. Phosphorylation of p38 is directly inhibited by Wip1. UV irradiation of cells causes rapid phosphorylation of a serine and threonine residue on p38, which activates its kinase activity. In turn, activated p38 MAP kinase can phosphorylate p53 on residues Ser33 and Ser46 in response to UV damage. UV-induced phosphorylation of p53 by p38 is associated with increased p53 transcriptional activity and apoptosis induction. Takekawa et al. (25) have shown that Wip1 selectively inhibits p38 through dephosphorylation of an activation-associated threonine residue. This results in downregulation of phosphorylation on UV-responsive p53 Ser33 and Ser46 residues. These data suggest that Wip1 may be a mediator of a negative feedback loop for p53 following its UV-induced activation.
Given the emerging relevance of Wip1 in p53-mediated cell cycle control and DNA damage response, we have begun to investigate the biological functions of Wip1 through studies of its activity in the mouse. In this paper, we confirm that Wip1 is activated by IR in a p53-dependent manner in multiple tissues of the mouse. We demonstrate that Wip1 mRNA is ubiquitously expressed, though at particularly high levels in the postmeiotic round spermatid compartment of the testes. We have also generated Wip1 null mice through gene targeting techniques and show that while such mice appear to exhibit normal embryonic development, they show defects in growth, organ structure, fertility, and lymphoid function. Embryo fibroblasts deficient for Wip1 show multiple defects in cell cycle control. All of these organismal and cellular phenotypes indicate that Wip1 does play an important role in the regulation of cell growth and in the function of a number of adult organ systems.
For Northern blot analyses of Wip1 mRNA, tissues were harvested and homogenized with a Polytron homogenizer in Trizol buffer reagent (Life Technologies), followed by purification of total RNA with the Trizol RNA purification kit according to the manufacturer's specifications. Ten micrograms of total RNA from each tissue was heated to 65°C and then loaded onto a formaldehyde-agarose gel and subjected to electrophoresis as previously described (20). Northern blot hybridization was done with the Ambion NorthernMax kit according to the manufacturer's directions with a Wip1 cDNA probe labeled with 32P by random primed oligonucleotides (Roche High Prime kit). After rinsing of the blot, hybridizing bands were visualized following autoradiography on Kodak AR-5 X-ray film for 1 to 5 days.
For reverse transcription (RT)-PCR of Wip1 RNA, Wip1+/+ and Wip1−/− spleens, kidneys, and testes from irradiated and unirradiated mice were isolated and total RNA was prepared from them as described above. RT-PCR on the isolated total RNAs was performed with the SuperScript preamplification system (Gibco-BRL) according to the manufacturer's protocol. Primers used for RT-PCR were primer 1 (Wip1-exon 1 sense; 5"-GCCCGCCAAGGTGTGCGCTG-3"), primer 2 (Wip1-exon 3 antisense; 5"-CAGGTCCACTGTGAGTGAGC-3"), primer 3 (Wip1-exon 3 sense; 5"-TCACTCACAGTGGACCTG-3"), primer 4 (Wip1-exon 6 antisense; 5"-AGGGCAGGTATATGGTCCTTAGA-3"), and primer 5 (Wip1-exon 4 sense; 5"-AGTGATGGACTTTGGAATAT-3").
In situ hybridization of testes sections from Wip1+/+ and Wip1−/− mice was performed essentially as described previously by Choi et al. (2).
The targeting construct was generated from a murine genomic Wip1 clone derived from a 129/Sv phage genomic library as previously described (2). The 5" homology arm of the construct was derived from a 3.2-kb NheI fragment located between exons 3 and 4 of the Wip1 gene. The 3" homology arm of the targeting construct was derived from an intron 5 2.2-kb BamHI segment. Between these two homology arms was placed a PGK-Puro selectable marker cassette. Successful homologous recombination of the targeting construct and the endogenous Wip1 allele results in deletion of all of exons 4 and 5, resulting in a severely truncated Wip1 protein. A negative selectable marker, MC1-TK, was located adjacent to the 5" homology region.
The targeting construct was electroporated into AB1 embryonic stem (ES) cells, and puromycin-resistant, 1-(2-deoxy,2-fluoro-β-d-arabinofuranosyl)-5-iodouracil (FIAU)-resistant colonies were selected. Resistant colonies were amplified, and DNA was prepared from them and screened by Southern blot procedures as previously described (16, 17). Probes derived from genomic Wip1 sequences upstream and downstream of the original targeting construct were used to screen for correctly targeted clones. One clone generated fragments of the expected sizes for successful homologous recombination, and this clone was further amplified and injected into 3.5-day blastocysts to generate chimeric offspring as previously described (17).
Chimeric offspring were then mated to C57BL/6 mice homozygous for a spontaneous tyrosinase gene mutation. These C57BL/6 mice were albino, so that any agouti offspring of the chimeric crosses were then analyzed for transmission of the mutant Wip1 allele. Genotyping for mutant and wild-type Wip1 alleles was performed by removing a 1-cm segment of the tail of weanling mice, extracting tail DNA, cleaving it with NcoRI, and performing agarose gel electrophoresis and Southern blot hybridization as previously described (16). The genotyping probe was a 2.2-kb BamHI fragment derived from intron 5. Mutant Wip1 alleles generated a 10-kb fragment, while the wild-type Wip1 allele generated a 7-kb fragment.
A Wip1 peptide corresponding to the mouse sequence containing amino acids 20 through 60 (APVPRRALGLPATPTLAGVGP) was synthesized by the solid-phase method with Fmol chemistry using an Applied Biosystems 430A peptide synthesizer (Foster City, Calif.). The peptide after cleavage from the resin and removal of the side chain protecting groups was incubated in reagent K (trifluoroacetic acid-phenol-thioanisole-H2O-EDTA, 82.5:5:5:5:2.5) for 3 h at room temperature and purified by high-pressure liquid chromatography (HPLC) on a pH-stable Vydac C-8 column (Hesperia, Calif.) with 0.05% trifluoroacetic acid-water-acetonitrile. The mass of the peptide was confirmed by electrospray ionization mass spectrometry on a Finnigan MAT SSQ 7000 (Finnigan MAT, San Jose, Calif.). Rabbit polyclonal antibodies were raised against the mouse Wip1 peptide coupled to keyhole limpet hemocyanin. The antibodies were affinity purified from the resulting serum by use of the peptide coupled with Sulfolink (Pierce), and the specificity of the antibody was confirmed by enzyme-linked immunosorbent assay (ELISA) and immunoblot assays.
BALB/c and Wip1−/− knockout mice were irradiated with 10 Gy generated from a 137Cs Shepherd Mark II irradiator. One day after irradiation, they were euthanized, and the whole testis was homogenized with a glass homogenizer. Homogenized testes or mouse embryo fibroblasts were extracted in 20 mM Tris (pH 7.5)-150 mM NaCl-0.1% Triton X-100-1 mM EDTA and a proteinase inhibitor cocktail (Roche). The extracts were then precipitated with acetone, pelleted, redissolved in sodium dodecyl sulfate (SDS) sample buffer, boiled, separated on an SDS-10% polyacrylamide gel electrophoresis (PAGE) gel and transferred to a polyvinylidene difluoride membrane. The membrane was blocked with 5% dry milk and probed either with a purified polyclonal antibody against Wip1 peptide, polyclonal antibody to p53 (M-19, Santa Cruz), polyclonal antibody to p21 (C-19, Santa Cruz), or a polyclonal antibody specific for a p53 phosphorylated serine 15 (Cell Signaling). The filter was subsequently incubated with horseradish peroxide-conjugated secondary antibody, developed with ECL Plus, and exposed to X-ray film. To standardize the amount of protein in each lane, the filter was stripped and probed with a polyclonal antibody against actin (Santa Cruz) and developed with ECL Plus.
After Wip1 germ line heterozygotes were obtained, these mixed 129/Sv-C57BL/6 mice were intercrossed to obtain Wip1−/− offspring. Wip1−/−, Wip1+/−, and Wip1+/+ offspring were monitored for disease (including cancer), body weight, fertility, and longevity over a period of 2 years in the Baylor College of Medicine vivarium. Seriously morbid mice were sacrificed and subjected to a complete necropsy. Organs were examined, weighed, and fixed in formalin in preparation for histopathology as previously described (24). Spontaneous deaths were recorded and necropsies were performed if death was recent. The experiments with mice described here complied with all relevant federal guidelines and Baylor College of Medicine policies.
Histopathology on organs from Wip1+/+, Wip1+/−, and Wip1−/− mice was performed by standard methods as previously described (18, 24). Hematoxylin- and eosin-stained slides were examined for histopathology by standard methods (18, 24).
For analysis of B cells, freshly isolated splenic cells were placed in phosphate-buffered saline (PBS) and stained with a mixture of fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse immunoglobulin G (IgG) polyclonal antibody (Sigma Chemical Company, St. Louis, Mo.) and a phycoerythrin (PE)-conjugated monoclonal rat anti-mouse CD19 (BD Pharmingen, San Diego, Calif.). For T cells, the lymphocytes were stained with a mixture of rat anti-mouse CD4-FITC, CD8-PE, and CD3-peridinin chlorophyll protein (PerCP) (BD Pharmingen). Appropriate preimmune FITC-conjugated goat IgG and rat IgG isotype controls were also used. After incubation at room temperature for 30 min, the cells were washed and suspended in 1% paraformaldehyde. Stained cells were stored at 4°C in the dark and analyzed within 24 h using two-color or three-color flow cytometry (Epics XL MCL; Beckman Coulter, Miami, Fla.).
Mice were challenged with 1 50% lethal dose (LD50) of mouse-adapted influenza A/Taiwan/1/86 (H1N1) virus using small particle aerosols as previously described (14). Briefly, mouse lung virus pools were diluted in minimal essential medium (stabilized with 0.05% gelatin) and placed in a nebulizer. Male and female mice were placed in microisolator cages covered with a Plexiglas with two vents (1 mm in diameter) and sealed on the edges with masking tape. They were then exposed to virus aerosols generated by compressed N2 and driven via an aerosol tube into the cages. Aerosolization lasted approximately 60 min. Following exposure, the infected mice were monitored daily for body weight and survival up to 21 days.
Splenic lymphocytes were cultured at a concentration of 105 cells per well (96-well plate) in RPMI 1640 medium supplemented with 2 mM l-glutamine, 100 U of penicillin, 100 μg of streptomycin (JRH Biosciences, Lenexa, Kans.), 5 × 10−5 M 2-mercaptoethanol (2-ME), 10 mM HEPES buffer, and 10% heat-inactivated fetal bovine serum (FBS) per ml. T cells were stimulated with monoclonal anti-CD3 and anti-CD28 (1 μg/ml) (BD Pharmingen) or phytohemagglutinin (PHA; 2 μg/ml; Wellcome Diagnostics, Dartford, England), and B cells with lipopolysaccharide (LPS) (Escherichia coli O55:B5, 10 μg/ml; Sigma). After incubation for 48 h at 37°C in a 5% CO2 humidified air incubator, the cells were pulsed with 0.5 μCi of [3H]thymidine ([3H]TdR, 6.7 mCi/ml; ICN Radiochemicals, Irvine, Calif.) for 18 to 24 h and harvested on glass fiber filter mats using a semiautomatic harvester (Skatron). Radioactivity incorporation was determined using a beta scintillation counter. The stimulation index (SI) for each mitogen was obtained by dividing the mean cpm of incorporation by mean control (medium alone) cpm. Differences between mean SI among groups were determined by Student's t test in analysis of variance.
Wip1+/− adult males and females were mated and examined for vaginal plugs each morning. Twelve days after visualization of the vaginal plug, the pregnant female was sacrificed and embryo fibroblasts were prepared from each embryo as previously described (6). DNA was prepared from a small portion of each embryo, and the Wip1 genotype was determined as described above. Several vials of passage 1 mouse embryo fibroblasts (MEFs) were frozen for each embryo.
For growth curve experiments, three lines each of Wip1+/+ and Wip1−/− MEFs were thawed, plated, and counted. Procedures were those described previously (9). A total of 7.5 × 105 MEFs were plated on multiple 100-mm dishes for each line in 10% fetal calf serum and Dulbecco's modified Eagle's medium (DMEM). Two plates were then harvested, and cells were counted by hemacytometer at intervals of 2, 4, 7, 8, 10, and 13 days. Fluorescence-activated cell sorting (FACS) analysis of early-passage primary MEFs was performed by standard procedures (9).
The number of cells in S phase was counted after pulse labeling with bromodeoxyuridine (BrdU) as previously described (1). The percentage of cells in mitosis was determined by microscopic visualization after fixing and staining the cells with antiphosphohistone H3 antibody (Upstate Biotechnology). For the colony plating assay, early-passage primary MEFs were plated at low density (10,000 cells per 10-cm plate) and incubated for 2 weeks with regular changes of medium. Cell colonies (>30 cells) were then fixed, stained, and counted. Experiments were performed in triplicate.
For long-term passaging experiments, three Wip1+/+ and Wip1−/− early-passage lines were used and conditions were similar to those described previously (9). One million cells of each line were plated on each of three 60-mm dishes, and every 3 days the cells were trypsinized, counted, and replated at 106 cells per 60-mm dish. When there were fewer than 106 Wip1−/− MEFs per plate, all of the cells were replated.
For G1 arrest assays, cells were normally plated, grown to confluency, and allowed to remain confluent for 4 to 5 days. Virtually all of these cells were arrested in G0/G1 by flow cytometry analysis. These synchronized cells were split into two groups, radiated and unirradiated. One group was treated with 5 Gy of ionizing radiation, and one group remained unirradiated. BrdU (10 μM) was added, and the cells were plated at lower density. The cells were then harvested at 24, 48, and 72 h, fixed in 70% ethanol, and subjected to double staining for DNA content (propidium iodide staining) and BrdU incorporation (staining with FITC-conjugated anti-BrdU antibody) prior to flow cytometry. Levels of relative G1 arrest were determined by the ratio of BrdU incorporation in the irradiated cells to the level of BrdU incorporation in the unirradiated cells for each time point.
Previously, we had shown that Wip1 mRNA is expressed in all mouse tissues, but was present at very high levels in the testes (2). To ascertain whether a specific compartment of the testes shows high expression of Wip1 mRNA, we performed in situ hybridizations with a Wip1 probe on sections from normal adult mouse testes. Wip1 is expressed within the cells of the seminiferous tubules (Fig. 1A and B).Wip1 is expressed in every tubule,but its level of expression is clearly stage dependent, with highest levels of expression detected in stage 6 to 8 tubules. Wip1 is expressed in the spermatogenic cells, with no expression detected in spermatogonia or early to mid-stage spermatocytes.
Expression of Wip1 is first detected in the late-stage pachytene spermatocytes just prior to the completion of meiosis. Expression of the gene is maintained postmeiotically in round spermatids, with highest levels of expression in stage 6 to 8 round spermatids. The level of Wip1 expression decreases progressively through the elongating spermatid stage. Thus, Wip1 is expressed in spermatocytes just prior to the completion of the meiotic divisions and during round spermatid development.
To determine the onset of the high testicular expression of the Wip1 mRNA, we performed Northern blot analyses for Wip1 RNA levels on testes from 15-, 20-, 25-, 30-, 35-, and 40-day-old mice (Fig. (Fig.1C).1C). Quantitation of Wip1 mRNA normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression showed that at 15 days of age, the testes expressed very low levels of Wip1, but by 25 days they were expressing very high levels and these remained very high thereafter, though there was a slight dip at 30 days (Fig. (Fig.1D1D).
The Northern analysis shows a correlation between the onset of puberty and spermatogenesis in the mouse and the appearance of the first wave of spermatogenesis. The lack of expression of Wip1 at day 15 is consistent with the in situ hybridization analysis in that Wip1 is not expressed in somatic cells or early-stage spermatogenic cells, spermatogonia, and early- to mid-stage pachytene spermatocytes, which would be present at this time point. Expression of Wip1 is only detected by day 20, when late-stage pachytene spermatocytes would be present and the first few round spermatids would populate the tubules. By day 25 most of the tubules will have undergone meiosis, leading to a large population of round spermatids, which as they elongate through day 30 leads to a decrease in Wip1 expression. After day 30 and the completion of the first round of spermatogenesis, the next rounds initiate, leading to randomization of the spermatogenic cell populations and the steady-state level of Wip1 expression in the testis.
The Wip1 gene was identified in a cell-based screen for p53 target genes by its p53-dependent response to ionizing radiation (5). To confirm that Wip1 is indeed a direct target of p53 in an organismal context, we investigated Wip1 mRNA induction in tissues of irradiated mice. p53+/+ and p53−/− mice were treated with 10 Gy of ionizing radiation, and tissues were harvested 6 h later. Total RNA was prepared from spleen, kidney, and testes of the irradiated mice and subjected to semiquantitative RT-PCR with Wip1, p21WAF/1/CIP1 (positive control), and actin (loading control). Spleen and kidney RNA from p53+/+ mice showed a large induction in Wip1 and p21 levels 6 h after ionizing radiation treatment (Fig. (Fig.22).Wip1 levels in irradiated p53+/+ testes were only marginally increased, perhaps due to high endogenous levels of this message already present. In contrast to p53+/+ tissues, p53−/− tissues showed little or no induction of Wip1 mRNA following radiation. These data are consistent with previous cell culture studies (5, 25) and indicate that Wip1 is induced by ionizing radiation in a p53-dependent manner both in vitro and in vivo.
To understand some of the biological functions of Wip1, we initiated gene targeting of the Wip1 gene in embryonic stem (ES) cells. To inactivate the Wip1 gene, we designed a targeting construct in which a puromycin expression cassette replaced genomic Wip1 sequences containing exons 4 and 5 (Fig. (Fig.3A).Exons3A).Exons 4 and 5 contain highly conserved sequences, and the expected targeting event would generate an allele expressing a severely truncated protein with no Wip1 coding sequences beyond exon 3. Introduction of the targeting construct into ES cells followed by selection in puromycin (positive selection) and FIAU (negative selection) led to the identification of several clones. Restriction mapping of the Wip1 allele in these selected clones identified one clone with the map expected after successful homologous recombination of the targeting construct and the endogenous Wip1 allele (Fig. (Fig.3B).3B). This clone was amplified, and standard procedures were used to reconstitute mice with a germ line mutation in the Wip1 allele.
Germ line Wip1 heterozygotes were crossed with each other to determine the effects of Wip1 nullizygosity on embryonic development. In fact, while Wip1−/− offspring of Wip1+/− parents were viable at birth, there was a reduction from the expected Mendelian ratio for null pups only. Of 268 offspring of Wip1+/− parents, 70 (26%) were Wip1+/+, 155 (58%) were Wip1+/−, and only 43 (16%) were Wip1−/−. Of the 43 Wip1−/− offspring, 26 were female and 17 were male, indicating a possible selection during development against male null embryos. Analysis of midgestation embryos failed to identify any obvious developmental abnormalities associated with Wip1 nullizygosity, suggesting the possibility that a portion of Wip1−/− males may have early embryonic defects. However, as described below, the Wip1−/− males also displayed a number of postnatal defects.
To confirm the absence of intact Wip1 expression in the Wip1 null mice, we performed Northern blot hybridization and RT-PCR analysis on RNAs from various Wip1−/− tissues. Northern blot hybridization showed that the Wip1 message was present in all of the Wip1+/+ tissues, but was not observed at all in the Wip1−/− tissues (Fig. (Fig.3C).3C). RT-PCR analysis did detect a Wip1 message with a deletion of exons 4 and 5, but presumably this message was unstable, because even reduced-size Wip1 mRNAs were not observed in the Wip1−/− lanes of the Northern blots (data not shown).
Confirmation that the Wip1−/− animals produce no viable Wip1 protein was provided by immunoblot analyses. Protein lysates were prepared from unirradiated and irradiated (10 Gy of IR) Wip1+/+ testes and irradiated Wip1−/− testes, subjected to SDS-polyacrylamide gel electrophoresis, and immunoblotted with a Wip1 antibody specific for mouse Wip1. The irradiated Wip1+/+ testes showed high levels of the expected 64-kDa Wip1 protein, while Wip1 of the same size was not discernible in the irradiated Wip1−/− testes (Fig. (Fig.3D,3D, left panel). The same mouse antibody does not react with human Wip1, but a human-specific Wip1 antibody does react with intact 64-kDa Wip1 protein, particularly in irradiated human cell lines (Fig. (Fig.3D,3D, right panel).
While Wip1 nullizygosity may have had some effect on embryonic development in males, a number of postnatal defects were observed in these males. Wip1−/− males were sometimes runted as pups and were significantly smaller than their Wip1+/− and Wip1+/+ siblings, which invariably appeared normal (Fig. (Fig.4A4A).Wip1−/− females did not exhibit this reduced body size. Mean body mass for 2- to 6-month-old Wip1−/− males was 21.6 g, whereas age-matched Wip1+/+ males averaged 28.1 g. The difference between the Wip1−/− and Wip1+/+ males was significant by t test (P < 0.0001). Mean body masses of 2- to 6-month-old Wip1−/− females and their Wip1+/+ counterparts were 22.7 and 21.9 g, respectively, indicating no significant differences among the female genotypes (P = 0.54).
Gross examination of the organ systems of Wip1−/− animals did not reveal overt morphological differences in structure and size compared to Wip1+/− and Wip1+/+ animals except that the male reproductive organs of the Wip1−/− males were often dramatically smaller than those of their wild-type and heterozygous counterparts (Fig. (Fig.4B).4B). This was particularly true of the testes, which in the adult Wip1−/− males had an average mass of 64 mg, versus 110 mg for the Wip1+/− and Wip1+/+ males. This difference was highly significant by t test (P < 0.0001). The reduced testis masses were not due solely to the reduced body masses of the Wip1−/− males, as even Wip1−/− males of normal weight often had testes of reduced size.
As might be expected, fertility in the Wip1−/− males was reduced. Four Wip1+/+ and 16 Wip1−/− males mated to wild-type females produced an average of 20.0 and 6.0 offspring, respectively, over a period of 5 months. Seven of 16 Wip1−/− males produced no offspring but did generate postcoital vaginal plugs in the females, indicating that the infertile Wip1 males were capable of mating. Other Wip1−/− males produced infrequent small litters, and a few produced litters at the same size and frequency as wild-type males. Wip1+/− males and Wip1−/− females demonstrated normal levels of fertility.
Another occasional observation was the presence of ulcerated skin lesions in approximately 5% of the Wip1−/− mice (Fig. (Fig.4C).4C). These particular lesions were never observed in the Wip1+/− and Wip1+/+ littermates. In addition, over 25% of the older null animals exhibited fluid-filled neck abscesses. The persistence of these skin lesions and abscesses was suggestive of an immune defect in the Wip1−/− mice.
Cohorts of Wip1−/−, Wip1+/−, and Wip1+/+ mice were monitored for longevity over a period of 24 months. Wip1−/− male animals tended to appear less robust over their life span and showed a much higher incidence of spontaneous deaths compared to their Wip1−/− female and Wip1+/− and Wip1+/+ littermates (Fig. (Fig.5).Only5).Only 3 of 20 Wip1−/− males survived to 2 years of age, whereas 15 of 17 of their Wip1−/− female littermates survived to 2 years of age. A log rank nonparametric survival test indicated that this difference was highly significant (P < 0.0001). In contrast to the Wip1−/− males, the Wip1−/− females had a survival rate similar to that of wild-type mice of the same background. Moreover, Wip1+/− males and females did not differ significantly in longevity from wild-type mice, as confirmed by the log rank test (P > 0.5). The cause of early deaths for the Wip1−/− males could not be determined directly, but cancers were virtually never observed in these animals, suggesting that nonneoplastic mechanisms might have been the more proximal causes of the deaths.
The reduced size of the Wip1−/− male reproductive organs prompted a closer examination of the histopathology of these organs. Wip1 null testes often exhibited moderate or extensive degeneration of the seminiferous tubules, with extensive vacuolization, loss of normal cellular architecture, and, in the worst examples, absence of mature spermatozoa (Fig. (Fig.6A).However,6A).However, in all cases examined there were clearly round and elongating spermatids, confirming that they had undergone meiosis and initiated terminal differentiation. Extensive cell death appears to be occurring within the germ cells, as marked by the presence of pycnotic nuclei.
The degeneration of the seminiferous epithelium suggests an inability to maintain spermatogenesis over time in the Wip1 null animals. The epididymis of the Wip1−/− male often exhibited atrophy as well, showing irregular tubules with abnormal architecture and an absence of spermatozoa, in contrast to the epididymi from wild-type mice, which were usually packed with spermatozoa (Fig. (Fig.6B).6B). Thus, the diminished fertility of the Wip1−/− males is likely to be a result of the reduced number of spermatozoa produced by the null testes and defects in the epididymi.
Earlier gross examination revealed evidence of splenic enlargement in some of the 8- to 11-month-old Wip1 null male and female animals, with spleen weights of 300 to 700 mg compared to the normal 100-mg spleens usually observed in age-matched wild-type animals. Histological comparison of the lymphoid organs of the Wip1−/− and Wip1+/+ mice revealed that the lymph nodes, spleen, and bone marrow of the null animals often exhibited a pronounced myeloid and plasmacytic hyperplasia. In some cases the splenic hyperplasia was accompanied by loss of normal splenic architecture (Fig. (Fig.6C).6C). In wild-type spleens, the white pulp regions were clearly demarcated from the surrounding red pulp, whereas in the Wip1−/− sections, the white pulp regions were reduced in size and poorly demarcated from the red pulp regions due to infiltration by other cell types (Fig. (Fig.6C).6C). No germinal centers could be identified in the white pulp, though prominent hematopoiesis was seen in the red pulp. The lymph nodes in some Wip1−/− mice also exhibited hyperplasia, accompanied by cellular infiltration with macrophages, neutrophils, and eosinophils. Bone marrow also showed evidence of myeloid hyperplasia in some of the older Wip1−/− mice. The thymus of some older Wip1 null mice also showed a reduction in size and a loss of normal medullary-cortical architecture.
A number of organs in the Wip1−/− mice exhibited increased levels of inflammation, particularly in those animals exhibiting skin ulcerations. These lesions showed extensive pyogranulomatous inflammation, infiltrating mast cells, epidermal proliferation, acanthosis, and hyperkeratosis. Inflammation in organs of Wip1+/− and Wip1+/+ animals was observed at relatively minimal levels.
The increased presence of inflammation, skin ulcerations, and abnormal lymphoid histopathology suggested defects in immunocompetence in the Wip1−/− mice. To more directly test this possibility, we challenged four Wip1−/− male mice and five Wip1+/− male mice with 1 LD50 of a mouse-adapted pathogenic influenza virus according to previous protocols (14). The virus-infected mice were then monitored for 21 days. All of the infected mice of both genotypes showed significant morbidity and body weight loss by 7 days postinfection. All four Wip1−/− males died by 11 days postinfection, while four of the five Wip1+/− males recovered from the virus and appeared healthy at 12 to 15 days postinfection (Fig. (Fig.7A).This7A).This difference in mortality was statistically significant by log rank analysis (P = 0.013). Thus, Wip1−/− mice do appear to be immunologically deficient, as measured by the ability to resolve an infection by a lethal pathogen.
The increased susceptibility of the Wip1 null animals to a lethal pathogen prompted an examination of their T- and B-cell populations and functions. Fresh, uncultured spleen cells from 3- to 6-month-old Wip1−/− and Wip1+/+ male mice were examined by flow cytometry for the presence of a variety of lymphoid cell surface antigens, including CD19+ and CD19+/IgG+ (for B cells) and CD3+, CD4+, and CD8+ (for T cells). CD19+ and CD19+/IgG+ B cells were not significantly different in number between Wip1−/− and Wip1+/+ animals (Fig. (Fig.7B).7B). Moreover, CD3+ T cell numbers were similar in both genotypes. However, CD4+ T cells were increased in number in Wip1−/− mice, and CD8+ T cells were decreased in number in Wip1−/− mice (Fig. (Fig.7C7C).
The T- and B-cell proliferative responses to mitogenic stimulation were further examined. Stimulation of splenic T cells with anti-CD3 and -CD28 elicited a reduced proliferative response in the Wip1−/− cells compared to the Wip1+/+ cells at both 2 and 3 days poststimulation, indicating defects in normal T-cell receptor-mediated signaling (Fig. (Fig.7D).7D). Addition of the polyclonal T-cell mitogen PHA to splenocytes also resulted in a dramatic reduction in response by the Wip1−/− T cells compared to their Wip1+/+ counterparts (Fig. (Fig.7E).7E). Finally, addition of the B-cell mitogen LPS evoked a less robust proliferative response by Wip1−/− B cells, suggesting that both B and T cells in the Wip1 null mice have partially compromised function (Fig. (Fig.7F).7F). When these T- and B-cell functional assays were performed on splenocytes from older Wip1−/− and Wip1+/+ mice, the results were similar (data not shown).
To assess whether the absence of Wip1 had any effect on other cell growth phenotypes, we isolated fibroblasts from midgestation Wip1−/− and Wip1+/+ embryos and cultured them under standard conditions. Soon after plating the Wip1−/− MEFs, it became obvious that they grow much more slowly than their Wip1+/+ counterparts. Growth curve analyses of Wip1−/− and Wip1+/+ MEFs confirmed that the proliferation rates of early-passage Wip1−/− MEFs were substantially retarded compared to Wip1+/+ MEFs (Fig. (Fig.8A).Thirteen8A).Thirteen days after plating three lines of 7.5 × 105 Wip1−/− early-passage MEFs, they grew to a mean of 2.5 × 106 cells per plate, versus a mean of 5.5 × 106 cells per plate for three lines of Wip1+/+ cells.
Flow cytometric analysis of dividing Wip1−/− and Wip1+/+ MEFs for DNA content and BrdU labeling indicated a decreased percentage of cells in S phase and mitosis (Table (Table1).1). The G2/M ratio of Wip1−/− cells was over threefold greater than that of the Wip1+/+ MEFs, suggesting a block in entering mitosis in the Wip1 null MEFs. Colony formation following plating at low density was also significantly reduced in the Wip1−/− MEFs (Fig. (Fig.8B8B).
Continued passaging of the Wip1−/− MEFs led to a rapid cessation of division, suggestive of a premature senescence. By passage 13, the Wip1−/− MEFs had essentially stopped dividing, while their Wip1+/+ counterparts were still dividing at a robust rate (Fig. (Fig.8C).8C). The morphology of these proliferatively inactivated Wip1−/− fibroblasts was similar to that of senescent normal fibroblasts, with a flattened, vacuolated, enlarged appearance. Whether or not the Wip1−/− MEFs are truly senescent remains uncertain, but it is clear that they have a greatly reduced number of population doublings in culture compared to their Wip1+/+ counterparts.
To determine whether Wip1 null MEFs maintained a normal G1 DNA damage checkpoint, we treated Wip1+/+, Wip1−/−, and p53−/− MEFs with 5 Gy of ionizing radiation (IR) and monitored their G1 arrest checkpoints at 24 h after irradiation. Not surprisingly, p53−/− MEFs showed no defect in S-phase entry (as measured by BrdU incorporation) compared to unirradiated p53−/− MEFs 24 h after IR treatment (Fig. (Fig.8D).8D). At 24 h post-IR, both Wip1−/− and Wip1+/+ MEFs showed a strong G1 arrest phenotype, though the Wip1−/− MEFs consistently showed a more robust G1 arrest than their Wip1+/+ counterparts (Fig. (Fig.8D).8D). These experiments indicate that Wip1 null cells do not have a defect in initiating G1 arrest in response to DNA damage.
The decreased cell cycle progression and more robust DNA damage-induced G1 arrest phenotype of the Wip1−/− MEFs suggested the possibility that they may have higher levels of activated p53. This would be consistent with the inhibitory effects of Wip1 on the p53-activating p38 MAP kinase (25). To assess this possibility, we determined p53 levels in dividing Wip1+/+ and Wip1−/− MEFs by Western blot analysis using a p53-specific antibody. p53 levels were similar in both types of MEFs, indicating that p53 is not produced at higher levels and does not have higher stability in Wip1−/− MEFs (Fig. (Fig.8E8E).
The relative activity of p53 in Wip1+/+ and Wip1−/− MEFs was measured by determining expression levels of a p53 target gene, p21 WAF1/CIP1, and phosphorylation of serine 15, markers for activated p53 (4, 22). Wip1−/− MEFs consistently showed higher levels of p21 protein and p53 phosphorylated serine 15 than Wip1+/+ MEFs, suggesting that p53 functional activity was higher in these cells (Fig. 8E and F).
The Wip1 gene has been shown to be a serine/threonine phosphatase inducible by UV and ionizing radiation in a p53-dependent manner in cells in culture (5, 25). While the function of Wip1 is presently unclear, its p53 responsiveness (5) and association with p38 MAP kinase (25) suggest that it may play an important role in cell signaling in response to cellular stress. Our finding that Wip1 mRNA is expressed in all examined tissues of the mouse is consistent with the hypothesis that Wip1 plays a global role in signaling in all cell types. Moreover, the demonstration that Wip1 is induced by ionizing radiation in multiple mouse tissues only when p53 is present confirms that it is a bona fide p53 target gene. The mechanism through which p53 induces Wip1 RNA expression is not known, since a consensus p53 response element has not been identified in the Wip1 promoter region (2), but we have observed that the Wip1 promoter does confer p53 inducibility in cell transfection assays (data not shown).
While Wip1 mRNA expression appears to be ubiquitous in all mouse tissues, the testes contain extremely high RNA levels. In situ hybridization experiments on testis sections revealed that Wip1 was expressed in spermatogenic cells, with its expression first detected in late-stage pachytene spermatocytes just prior to the completion of meiosis. The highest expression levels of Wip1 are in the postmeiotic round spermatid compartment. Thus, the expression of Wip1 correlates with the final stages of meiosis, and it may play a role in regulating the meiosis I and II divisions.
The high level of Wip1 expression in round spermatids suggests that it plays a role in inhibiting further cell cycles, thus maintaining the haploid state. Since there are round and elongating spermatids in the seminiferous tubules of Wip1 null animals, this suggests that Wip1 is not essential for the completion of meiosis. However, a more detailed analysis of the cell cycle state of the haploid spermatids in these animals might highlight a subtle meiotic defect. Alternatively, there could be a closely related family member that is compensating for its function in male germ cells.
The defects in male fertility of the Wip1 null mice are probably due to the dual effects on the spermatogenic cells and the epididymi, similar to what was observed with the estrogen receptor alpha knockout mouse, whose mutation affected both the somatic cells of the testis and the epididymi (10). The fact that many of the Wip1−/− male mice have smaller reproductive organs and reduced fertility suggests that this gene plays a major role in development of the male reproductive structures. However, the precise mechanisms by which Wip1 could influence this process remain to be elucidated.
The viability of Wip1−/− males and females indicates that Wip1 is not essential for embryonic development, yet the reduced ratio of Wip1−/− male offspring suggests that Wip1 may play a subtle role in influencing development of the males. Moreover, the Wip1−/− males exhibit a number of pleiotropic postnatal defects, such as runting, reproductive defects, and reduced longevity, that are not exhibited by the Wip1-deficient females. The reproductive deficiencies of the Wip1−/− males are presumably related to the aberrant architecture seen in the seminiferous tubules and epididymi, leading to the paucity of mature spermatozoa seen in many of the seminiferous tubules and epididymi. These defects in spermatogenesis contrast to those observed in p53 null mice, whose testes exhibited multinucleated giant cells suggestive of a degenerative syndrome resulting from the inability of the tetraploid primary spermatocytes to complete meiotic division (19).
The mechanisms of runting and reduced longevity seen in the null males are less clear, but may be a result of male hormone deficiencies accompanying the reduction in reproductive organ development. A complete age-related analysis of the hormonal status of these animals should shed light on the contribution of endocrinological defects to these phenotypes. The normal phenotypic appearance and longevity of the female Wip1−/− mice indicate that Wip1 is relatively dispensable for normal prenatal and postnatal function, though there may be changes in immune function which do not affect longevity.
Several observations indicate that the Wip1−/− animals have immune defects. These include the presence of ulcerated skin lesions in some animals, the disorganized and hyperplastic architecture in the lymphoid organs, the increased inflammation observed in normal organs, and the defects in the proliferative responses of T and B cells to antigenic and mitogenic stimulation. The most compelling evidence for an immunological defect was provided by the increased mortality of influenza virus-challenged Wip1−/− mice. The reduced T- and B-cell responses may partially explain the increase in the inflammatory response, since a defect in one part of the adaptive immune system may result in compensation by innate immune system components, including the neutrophils and macrophages. Whether these potential defects in immune function are related to the decreased longevity of the null males is unclear, since some of the older null females also show evidence of inflammatory lesions and hyperplastic disorganization of lymphoid structures. The hyperplasias and atrophic disorganization observed in the Wip1 null spleens were not preneoplastic, in contrast to the hyperplasias observed in p53 null spleens (27).
While T and B cells isolated from Wip1−/− animals have defects in proliferation, Wip1−/− embryo fibroblasts show even more profound deficits in growth in vitro . Wip1 null MEFs show low proliferation rates and enter S phase and especially mitosis less frequently than normal MEFs. Not only do Wip1−/− MEFs grow more slowly, but they seem to exhibit relatively few divisions prior to entering a senescence-like state. Thus, Wip1 is clearly an important component of cell cycle progression, at least in vitro. Why it does not seem to be equally important for cell division in the intact animal is unclear, but there may be compensatory pathways in vivo which mask the effects of Wip1 absence.
Based on early studies, Wip1 is likely to be an important component in the response to cell stress induced by radiation. Fiscella et al. (5) have shown that radiation induces increased Wip1 expression in a p53-dependent manner, and Takekawa et al. (25) have shown that Wip1 can inhibit p38 MAP kinase through dephosphorylation of a UV-activated threonine residue. We have demonstrated here that ionizing radiation induces a robust G1 arrest in Wip1−/− MEFs. The slight enhancement of the G1 arrest in Wip1 null cells compared to normal cells is consistent with the data of Takekawa et al. (25) indicating that the presence of Wip1 may inhibit the activated status of p53 through p38 MAP kinase deactivation.
Unstressed Wip1−/− MEFs divide much more slowly than their normal counterparts, as indicated by growth curves, long-term passaging, and flow cytometric analyses. These results indicate that Wip1 plays an important cell cycle regulatory role. Wip1−/− MEFs appear to be particularly compromised in the mitotic transition, as Wip1−/− cells have a much higher G2/M ratio than normal cells. This suggests that the phosphatase activity of Wip1 may target proteins involved in mediating mitosis. However, we have provided evidence that at least some of the decreased cell cycle progression in Wip1 null MEFs may be a result of a more activated p53 in these cells. This may be due to the effects of Wip1 in inactivating the p53-specific p38 MAP kinase (13), although action of Wip1 on other p53 effectors cannot be ruled out. The elucidation of such Wip1 targets will be of considerable importance for our understanding of its role in stress response and cell cycle control. Moreover, the availability of viable Wip1 null cells and animals should provide additional tools for understanding the function of Wip1 at both the cellular and the organismal level.
We thank Jerrold Ward for histopathological analyses of some of the Wip1−/− spleens and thymic tissues. We thank Eunice Varghese and Chao-Ling Wu for technical assistance and Marco Schito for helpful discussions.
This work was supported by a grant from the National Cancer Institute (CA54897) to L.A.D.