Wip1 and p53 contribute to HTLV-1 Tax-induced tumorigenesis
© Zane et al.; licensee BioMed Central Ltd. 2012
Received: 16 October 2012
Accepted: 15 December 2012
Published: 21 December 2012
Skip to main content
© Zane et al.; licensee BioMed Central Ltd. 2012
Received: 16 October 2012
Accepted: 15 December 2012
Published: 21 December 2012
Human T-cell Leukemia Virus type 1 (HTLV-1) infects 20 million individuals world-wide and causes Adult T-cell Leukemia/Lymphoma (ATLL), a highly aggressive T-cell cancer. ATLL is refractory to treatment with conventional chemotherapy and fewer than 10% of afflicted individuals survive more than 5 years after diagnosis. HTLV-1 encodes a viral oncoprotein, Tax, that functions in transforming virus-infected T-cells into leukemic cells. All ATLL cases are believed to have reduced p53 activity although only a minority of ATLLs have genetic mutations in their p53 gene. It has been suggested that p53 function is inactivated by the Tax protein.
Using genetically altered mice, we report here that Tax expression does not achieve a functional equivalence of p53 inactivation as that seen with genetic mutation of p53 (i.e. a p53 −/− genotype). Thus, we find statistically significant differences in tumorigenesis between Tax + p53 +/+ versus Tax + p53 −/− mice. We also find a role contributed by the cellular Wip1 phosphatase protein in tumor formation in Tax transgenic mice. Notably, Tax + Wip1 −/− mice show statistically significant reduced prevalence of tumorigenesis compared to Tax + Wip1 +/+ counterparts.
Our findings provide new insights into contributions by p53 and Wip1 in the in vivo oncogenesis of Tax-induced tumors in mice.
Human T-cell Leukemia Virus type 1 (HTLV-1) is the first identified human retrovirus. The virus belongs to the deltaretrovirus family and is the etiological agent of a highly aggressive neoplastic disease, Adult T-cell Leukemia/Lymphoma (ATLL), and inflammatory diseases including HTLV-1 Associated Myelopathy (HAM)/Tropical Spastic Paraparesis (TSP), uveitis, infective dermatitis and myositis [1–9]. HTLV-1 infects approximately 20 million individuals world-wide, and 1-5% of infected individuals will develop ATLL after a long latency period of 20 to 60 years .
HTLV-1 encodes a viral Tax oncoprotein. The singular expression of Tax is sufficient to transform primary rodent cells  and potentially human embryonic stem cells , immortalize human primary T lymphocytes [12, 13], and induce tumors in transgenic mice [14–17]. Tax confers pro-proliferative and pro-survival properties to HTLV-1 infected cells [18–20] by pleiotropically activating effector proteins including the Cyclic AMP Responsive Binding Protein (CREB) and CBP/p300 [21–24], Nuclear Factor kappa-B (NF-κB) [25–29], Cyclin-Dependant Kinases (CDKs) [30–33], and Akt [34–36] amongst others. Tax also triggers DNA damage [37–42]. In transforming a normal T-cell into a leukemic cell, it is believed that Tax must also neutralize cellular checkpoints (e.g. p53 and mitotic spindle assembly checkpoint) that act to censor DNA damage [43, 44] and aneuploidy [45, 46].
p53 is a DNA-binding transcription factor that plays a key role in cell cycle regulation, apoptosis, and DNA repair . The p53 gene is recognized as one of the most important tumor suppressor genes and is frequently mutated in human tumors including hematologic malignancies [48–50]. In many human malignancies, the frequency of p53 genetic mutation is ≥50% [51, 52]; however, the frequency of mutated p53 in ATL patients is reported to be around 15% [53–58], suggesting that loss of p53 activity in ATL may largely arise through a mechanism other than genetic mutation. Several in vitro studies in different cell types have shown that Tax represses p53 activity [59–65]. Various mechanisms have been proposed for Tax-inactivation of p53. Indeed, it has been suggested that Tax inactivates p53 by acting through either the CREB  or the NF-κB [66, 67] pathway; however, it has also been noted that neither mechanism satisfactorily explains Tax-p53 interaction , leaving the question of how Tax effectively disables p53 function incompletely answered.
Here, we have conducted in vivo experiments in mice to address two questions. First, we have assessed the effectiveness of Tax mediated inactivation of p53 versus inactivation of p53 by genetic mutations. Second, we have characterized Wip1 as a cooperating in vivo Tax co-factor in p53 inactivation. Using various genetically altered mice, we show that Tax inactivation of p53 is functionally less stringent than p53 inactivation by genetic mutation, and we report that the cellular Wip1 phosphatase protein collaborates functionally with Tax in inhibiting p53 activity.
Transient over-expression assays generally are imperfect reflections of physiological regulation. To ask in a more physiological manner how endogenous Wip1 expression regulates p53 activity, we independently isolated several primary MEF clones from Wip1 −/− knock-out mice  and their Wip1 +/+ wild type siblings (genotyping examples of MEFs are shown in Figure 3D, top). We then compared cell endogenous p53 activity in several independently isolated Wip1 −/− MEFs to other independently isolated control Wip1 +/+ MEFs employing either the pG13-Luc reporter assay (Figure 3D, bottom) or by determining the mRNA expression levels of a known p53-responsive target gene, p21WAF1/CIP1 (Figure 3E). Notably, the Wip1 −/− MEFs showed statistically significant higher levels of pG13-Luc expression (p=0.0076; t-test) and higher levels of p21 mRNA (p=0.0425; t-test) than the Wip1 +/+ MEFs, suggesting that cell endogenous Wip1 does physiologically reduce p53 function in primary cells (Figures 3Dand E). This regulation of p53 by Wip1, however, does not occur at the level of transcription because there was no statistically significant difference in the amounts of p53 mRNA in Wip1 +/+ versus Wip1 −/− MEFs (Figure 3F).
In our immunostainings, we did note that Tax and Wip1 colocalize in the nucleus (Figure 6A and Additional file 2: Figure S2A). Moreover, additional immunostainings also show that Wip1 and p53 colocalize in the nucleus (Figure 6B and Additional file 2: Figure S2B). Thus, conceivably, Tax, p53, and Wip1 interaction occurs through intranuclear contacts. Currently, we do not have sufficient data to fully understand whether the colocalization of Tax, Wip1, and p53 manifests in direct protein-protein interactions or the proteins interact through bridging by additional factors. Experiments are in progress to define better these mechanistic interactions.
Colloquially known as the guardian of the genome, p53 is an important player in cancer biology, as exemplified by its ubiquitous loss of function in cancers. Thus, approximately 50% of human cancers are genetically mutated in p53 [29, 82–85], and the other 50% show attenuated or abrogated p53 activity through means other than mutation . In the case of ATLL, the frequency of p53 gene deletion and mutation is lower than in many other types of cancers and has been reported to approximate 15% . Indeed, our own anecdotal findings are consistent with this low prevalence; in a recent survey of 7 primary ATLL cells, we found no evidence for any of the 11 most frequent p53 somatic gene mutations that have been described for lymphoid neoplasms (Zane, data not shown).
Cancers that retain wild-type p53 gene, nevertheless, can have attenuated p53 activity via other mechanisms. For example, Mdm2, an E3 ubiquitin ligase that promotes p53 degradation, is a major negative regulator of p53 [87–89]. Another example of negative regulation arises from the Twist1 protein. Twist1 accumulates in sarcomas that are genotypically p53 wild-type; it dysregulates p53 phosphorylation promoting its degradation . Additional examples come from DNA tumor viruses; some encode proteins that repress p53 activity. Hence, SV40 large T-antigen stabilizes, but inactivates, p53; adenovirus E1B-55-kDa protein, and the E6 oncoprotein of human papilloma virus (HPV) types 16 and 18 target p53 for ubiquitinylation and degradation [91–93]. In the case of HTLV-1, our work here reaffirms previous findings that Tax indeed attenuates p53’s transcriptional activity in cultured cells (Figure 3). However, a perhaps more important implication to arise from our study is that we compare for the first time the impact of Tax inactivation of p53 versus p53 inactivation by genetic mutation for their relative contributions to in vivo tumorigenesis in mice. To date, it generally has been believed that Tax stringently inactivates p53 activity reducing the need for ATL cells to acquire p53 inactivating mutations. Our results are, however, incongruent with this notion. Thus, we found that Tax induces tumorigenesis in mice much more robustly in a p53 −/− setting than in a p53 +/+ context (Figure 2A), suggesting that Tax inhibition of p53 in the latter context is significantly less complete than p53 inactivation via gene mutation. Our findings differ somewhat from those reported by Portis et al.. The differences may be due to variances in the mouse numbers, the mouse strains, and the criteria used to determine tumor-free survival and when euthanasias of mice are performed. To date, in the published literature, only cross-sectional findings are associated between p53 genetic mutations and human ATLLs . These findings do not offer clarity on when p53 mutations occurred relative to HTLV-1 infection, Tax expression, and the onset of transformation of ATLL cells. Our results in mice provide prospective analyses of the contribution of a p53 −/− genotype to the initiation of in vivo tumorigenesis by Tax. Accordingly, extrapolating our mouse findings to humans suggests that early loss of p53 through a p53 −/− genetic mutation in cells infected by HTLV-1 foretells a worse prognosis compared to a corresponding infection in a counterpart p53 +/+ setting.
In our investigation of p53 inactivation, we report for the first time a contributory role by Wip1 in Tax-tumorigenesis. Our insight into the role of Wip1 arose from the observation that loss of Wip1 (i.e. Wip1 −/− ) significantly reduced the frequency of tumor development in Tax transgenic mice (Figure 4B). We linked this observation to a Wip1-mediated p53 effect because we found that Wip1 −/− MEFs have significantly increased p53 activity over their Wip1 +/+ counterparts. Thus, a parsimonious interpretation of the collective findings is that loss of Wip1 phosphatase (i.e. Wip1 −/− ) increases cell endogenous p53 activity (Figures 3Dand E), and this increase in p53 function reduces Tax-tumorigenicity in Tax + Wip1 −/− mice (Figure 4B). Hence, the magnitude of p53 activity is important in regulating the extent of in vivo Tax tumorigenesis, and this view is further consistent with the tumor-free survival results comparing Tax + p53 +/+ and Tax + p53 −/− mice (Figure 1).
The potential value of inhibiting Wip1 in moderating cancer progression is not only limited to Tax–induced tumors because a Wip1 effect has also been suggested in mammary gland tumors , lymphomas , colorectal cancers , and other spontaneous tumors . Going forward further clarification is needed to understand whether Wip1’s effect on many cancers and its impact on Tax-driven tumor formation are primarily due to its effect on p53 signaling or may also arise from its known effects on other pathways, such as ARF, ATM, and p38 MAPK signaling [96, 99]. Studies that compare the in vivo tumorigenesis frequencies seen in Tax + Wip1 −/− p53 −/− versus Tax + Wip1 +/+ p53 −/− mice (two genotypes currently being bred in our laboratory) may help to address whether Wip1 has important substrates other than p53 that contribute to Tax-mediated transformation. In other models of carcinogenesis, it has been shown that the singular over-expression of Wip1 is insufficient to initiate oncogenesis  and that Wip1 mostly promotes tumors by cooperating with known oncogenes . Nevertheless, amplification of the Wip1 gene has been described for numerous human primary tumors [101–112], with virtually all such tumors being genetically p53 wild-type [71, 72, 113]. Based on this observation, one wonders if the low selective pressure for p53 mutations in ATLL could be due to Wip1 gene amplification in these cells. To our knowledge, this important question has not yet been investigated in ATLLs.
In summary, despite much progress in HTLV-1 research over the past three decades , a salient finding to emerge from this work is the new identification of Wip1 as a cooperating cellular co-factor of Tax in p53-inactivation and in vivo tumorigenesis. Currently, our confocal imaging results suggest a colocalization between Tax, Wip1, and p53 within the nucleus (Figure 6 and Additional file 2: Figure S2), but we still lack sufficient data to decipher mechanistically how Tax and Wip1 cooperate to inactivate p53. Amongst several plausible mechanisms, we remain unable to conclude whether Tax can increase Wip1 dephosphorylation of p53 and/or MDM2, a major inhibitor of p53 that has been reported to also be a target of Wip1 . Nonetheless, the functional delineation here of a contribution by Wip1 to Tax tumorigenesis (Figure 4B) does raise the possibility that future uses of small molecule Wip1 phosphatase-inhibitors  may benefit ATLL treatment.
The Tax and Wip1+/− transgenic mice were previously described [15, 74]. The p53-mutant mice were purchased from the Jackson lab (strain:B6.129S2-Trp53tm1Tyj/J) . The Wip1 and p53 knockout and Tax transgenic mice were all generated in C57BL/6 × 129/sv backgrounds [15, 68, 74]. Genotypes of the mice were determined by polymerase chain reactions (PCRs) using primers: Tax (Tax-F-7511-7530: 5′-tcggctcagctctacagttc-3′; Tax-R-8044-8025: 5′-tgagggttgagtggaacgga-3′), p53 (wt: 5′-acagcgtggtggtaccttat-3′, mutant: 5′-ctatcaggacatagcgttgg-3′ and common: 5′-tatactcagagccggcct-3′) and Wip1 (Wip1 Exon4 F: 5′-gtggagctatgatttcttcagtgg-3′; Wip1 Exon4 R: 5′-gatacgacacaagacaaacctcc-3′; Wip1 intron 3: 5′-acaagcttgcagggctgtttgtgg-3′; PGK promoter: 5′-cttcccagcctctgagcccagaaagc-3′). Experimental research on mice follows NIH approved animal study protocols and guidelines.
Mice were necropsied and examined by mouse pathologists. All of the internal organs (spleen, liver, pancreas, kidney, stomach, intestine, lung, heart, brain, lymph node, thyroid gland) were fixed, paraffin embedded, sectioned and stained with H&E for analyses. Tissues that were found to be grossly abnormal at time of necropsy were multiply sectioned and stained by H&E (hematoxylin and eosin) for microscopic histological analyses.
Human cervical cancer cell line HeLa and human colorectal carcinoma cell lines p53+/+HCT116 and p53−/−HCT116  were cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum (FBS) and antibiotics. Human T cell lines MT2, MT4, C8166, Jurkat, A301, CEM, and H9 were maintained in RPMI 1640 with 10% FBS.
Mouse monoclonal anti-Tax (NIH AIDS Research and Reference Reagent Program) was used to detect Tax protein in immunoblotting and by confocal microscopy. Anti-Flag monoclonal antibody (M2; mouse; Sigma), anti-Wip1 polyclonal antibody (rabbit; Santa Cruz), anti-p53 monoclonal antibody (mouse; Cell Signaling) and anti-tubulin monoclonal antibody (DM1A; mouse; Sigma) were purchased.
pG13-Luc, p53 (human wild type) (gifts from B. Vogelstein) and Wip1 (gift from L.A. Donehower) expression plasmids were previously described [73, 116, 117]. HeLa or p53+/+ HCT116 or p53−/−HCT116 cells were seeded into twelve-well tissue culture plates for the luciferase assays and into 10 cm-dishes for Tax transfections. Transfections were performed 24 h later, using Lipofectamine and Plus reagent (Invitrogen) as described by the manufacturer. At 24 h after transfection of the reporters, cell lysates were subjected to luciferase assay. Total amounts of DNA to be transfected were adjusted by the addition of empty vectors. To detect luciferase and β-Gal activity, luciferase substrate (Promega) and the Galacto-Star assay system (Applied Biosystems) were used. Relative values of luciferase activity were calculated using β-Gal activity as an internal control for transfection.
For real-time quantitative reverse transcriptase–polymerase chain reaction (qRTPCR), total cellular RNA from samples was isolated using TriZol reagent according to the manufacturer’s instructions (Invitrogen Life technologies). Before reverse transcription, RNA was treated by DNase (Invitrogen) to prevent DNA contamination. First-strand cDNA was synthesized from 1 μg RNA using oligodT and Superscript III reverse transcriptase (Invitrogen). RNA concentration and purity were determined by UV spectrophotometry (nanodrop). The primer pairs were designed using the Universal Probe Library website (Roche diagnostics) (Wip1-L hs: 5′-cccatgttctacaccaccagt-3′; Wip1-R hs: 5′-tggtccttagaattcacccttg-3′; p53-L hs: 5′-ccccagccaaagaagaaac-3′; p53-R hs: 5′-aacatctcgaagcgctcac-3′; p21-L hs: 5′-cgaagtcagttccttgtggag-3′; p21-R hs: 5′-catgggttctgacggacat-3′). The primers of each pair were located in different exons to avoid genomic amplification. Primer and probe sequences to detect Tax in human T-cells  and Tax-SK43: 5′-cggatacccagtctacgtgt-3′ and Tax-SK44: 5′-gagccgataacgcgtccatcg-3′ to detect Tax in mouse spleens. GAPDH was used as the reference gene for he normalization of results (GAPDH-R: 5′-agtgggtgtcgctgttgaag-3′; GAPDH-F: 5′- tggtatcgtggaaggactca-3′). PCRs were performed using iQSupermix (Bio-Rad) (for quantification of Tax cDNAs in human T-cells) and iQSYBR Green Supermix (for quantification of other cDNAs) on a CFX96 system (Bio-Rad). A large amount of cDNA was prepared from the MT2, C8166, and MT4 cell lines prior to the experiment. This cDNA was 10 fold-diluted, aliquoted and used as a calibrator for Tax and other RT-PCR runs, respectively. For relative quantification and normalization, the comparative Ct (or Eff-DDC) method was used .
Cells were cultured on glass coverslips, and fixed in 4% paraformaldehyde at 24 h after transfection. After blocking of nonspecific reactions with 1% bovine serum albumin (BSA), cells were then incubated with the indicated primary antibodies, followed by a subsequent incubation with the secondary antibodies conjugated with Alexa Fluor 488 or 594 (Molecular Probes). DNA was counterstained with 0.1 μg/ml Hoechst 33342. Coverslips were mounted in Prolong Antifade (Molecular Probes), and cells were visualized with a Leica TCS SP2 confocal microscope.
The statistical analysis of tumor numbers, survival curves, and spleen weights were computed using the PRISM software (version 5.03).
This work was supported by intramural NIAID funding. XL was supported by NIH grant R01CA136549. We thank members of the Jeang laboratory for critical readings of the manuscript; Qingpin Liu for her help with mouse tissues and Melissa Foster and Ryan Hamilton for caring of the mice.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.