Molecular Mechanism of Tumorigenesis in Mice Transgenic for the Human T Cell Leukemia VirusTaxGene

VIROLOGY

248, 332±341 (1998) VY989298

ARTICLE NO.

Molecular Mechanism of Tumorigenesis in Mice Transgenic for the Human T Cell Leukemia Virus Tax Gene Laurent Coscoy, Daniel Gonzalez-Dunia, FreÂdeÂric Tangy, Sylvie Syan, Michel Brahic, and Simona Ozden1 Unite des Virus Lents, ERS CNRS 572, Institut Pasteur, 28 rue du Dr. Roux, 75724 Paris Cedex 15, France Received November 3, 1997; returned to author for revision March 26, 1998; accepted June 19, 1998 The infection by human T lymphotropic virus type I is associated with adult T cell leukemia and several inflammatory degenerative disorders, including tropical spastic paraparesis. To investigate the role of the Tax protein in the development of diseases linked to human T lymphotropic virus type I infection, we generated two lines of transgenic mice carrying the tax gene under the control of the viral promoter. The expression of the transgene was low in these mice and was restricted to the central nervous system and testis. Mice from both lines developed various types of tumors, including fibrosarcomas and adenocarcinomas. Tax was expressed at a high level in fibrosarcomas and in cell lines derived from these tumors. In tumor-derived cells, the expression of Tax led to an increased degradation of IkBa and IkBb and caused stable nuclear translocation of nuclear factor-kB. This translocation was essential for cell proliferation, as shown by expressing a nondegradable form of IkBb in these cells. Therefore, Tax-induced cell transformation in mice correlates with the degradation of IkBa and IkBb and with the constitutive activation of NF-kB. © 1998 Academic Press

Ras protein (Pozzati et al., 1990) and immortalizes primary T cells in the presence of interleukin 2 (Grassmann et al., 1989). In addition, transgenic mice carrying the tax gene develop different types of tumors (Grossman et al., 1995; Hinrichs et al., 1987; Nerenberg et al., 1987). Tax binds directly to DNA but acts in cooperation with several cellular transcription factors (Yoshida, 1996); among these are the serum responsive factor (SRF) (Fujii et al., 1992), the activating transcription factor/cyclic AMP response element binding protein (CREB/ATF) (Bantignies et al., 1996; Zhao and Giam, 1991), and other members of the ATF family (Baranger et al., 1995; Perini et al., 1995), NF-kB-like factors (Ballard et al., 1988; BeÂraud et al., 1994; Munoz et al., 1994; Rousset et al., 1996), and p16INK, a cell cycle inhibitor (Suzuki et al., 1996). The role of these different interactions in the cell transformation mediated by Tax is still unclear. Contradictory findings have been reported concerning the molecular mechanism of rodent fibroblast transformation. The results of Smith and Greene (1991) suggest that the CREB/ATF pathway may play an important role. However, Kitajima et al. (1992) and Yamaoka et al. (1996) showed that the activation of NF-kB is essential for Tax-induced transformation. In a recent report, Matsumoto et al. (1997) suggested that either the NF-kB or the SRF pathways may operate, depending on the cell type. The aim of the present work was to study the effects of the expression of Tax in mice transgenic for the tax gene driven by an HAM/TSP long terminal repeat (LTR). The mice developed fibrosarcomas. Because transformation by Tax has been studied mainly in vitro in cells expressing Tax under the control of

INTRODUCTION Human T cell leukemia virus type I (HTLV-I) causes a variety of diseases, including adult T cell leukemia/lymphoma (ATLL) (Hinuma et al., 1981; Poiesz et al., 1980; Yoshida et al., 1982) and a non-neoplasic inflammatory neurological syndrome called human T lymphotropic virus type I (HTLV-I)-associated myelopathy or tropical spastic paraparesis (HAM/TSP) (Gessain et al., 1985; Osame et al., 1986). Several other clinical conditions have been linked to HTLV-I infection on the basis of seroepidemiological studies; these include uveitis (Mochizuki et al., 1992), SjoÈgren's syndrome (Terada et al., 1994; Vernant et al., 1988), inflammatory arthropathies (Nishioka et al., 1989), polymyositis (Morgan et al., 1989), and pneumopathies (Vernant et al., 1988). The viral protein Tax seems to play a major role in the pathogenesis of HTLV-I-associated diseases. Tax is known to stimulate the transcription of viral and cellular genes such as the genes coding for interleukin 2, interleukin 2 receptor, proto-oncogenes, c-jun and c-fos, several cytokines, and MHC molecules (Inoue et al., 1986; Fujii et al., 1988; Yoshida, 1993). The transforming potential of Tax has been demonstrated in different experimental systems. It has been shown that rodent fibroblastic cell lines expressing Tax form colonies in soft agar and tumors in nude mice (Tanaka et al., 1990). Also, Tax transforms primary fibroblasts in cooperation with the

1 To whom reprint requests should be addressed. Fax: 33(1) 40.61.31.67. E-mail: [email protected].

0042-6822/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

332

MECHANISM OF TAX TUMORIGENESIS IN MICE

333

SpeI, purified, and injected into one-cell embryos of C3H mice. Two founder animals (Tax I and Tax II) were identified, and C3H transgenic lines were obtained. In both lines, PCR analysis of the LTR, tax, and splice site regions (Table 1) gave the expected amplified products and did not reveal any recombination in the transgene. The integrity of the transgene was also verified by Southern blot analysis after BamHI digestion (data not shown). The copy number of the transgene in Tax I and Tax II mice was estimated after hybridization with a tax probe by comparing the signals obtained, on the same blot, for DNA from transgenic mice with those from the MT2 cell line (a chronically infected T cell line containing two proviral copies per cell). Densitometric analysis showed that Tax I transgenic animals had 4 or 5 copies of the transgene, whereas Tax II animals had 8±10 copies (data not shown). Although homozygous transgenic mice were obtained recently, the experiments presented in this article were performed on heterozygous animals. FIG. 1. Structure of the TSP LTR-tax transgene. Expression of the tax RNA in transgenic mice. (A) Structure of the TSP LTR-tax transgene: tax id driven by an HTLV-I LTR (U3, R, U5) isolated from a TSP patient. The SV40 splice site and polyadenylation signal are derived from the BamHI fragment of pSVK3 vector. (B) Total RNA from tissues of a Tax I mouse was subjected to tax specific RT PCR and analyzed by Southern blot hybridization (the same results were obtained with mice from the Tax II line). (1) Positive control (DNA extracted from tax transgenic mouse. (2) Negative control (DNA extracted from a non-transgenic mouse). (C) As a control for DNA contamination, the PCR was performed on total RNA without RT.

nonviral promoters, we derived cell lines from these tumors and investigated the mechanism of transformation of these lines. We show that the cellular proliferation of these lines is mediated by the NF-kB/IkB pathway. RESULTS

Analysis of transgene expression Total RNA was extracted from the brain, spinal cord, lung, kidney, heart, gonads, salivary glands, liver, lymph nodes, Peyer's patches, spleen, and thymus of mice from both transgenic line. Reverse transcription-polymerase chain reaction (RT-PCR) were performed using total RNA with Taxand RT-TG primers (Table 1). Tax mRNA was expressed in the central nervous system (brain and spinal cord), as well as in testis (Fig. 1B). RT-PCR analysis for glyceraldehyde3-phosphate dehydrogenase (GAPDH) expression was always positive, demonstrating the integrity of the RNA samples and the efficacy of RT. Although the transgene was detected reproducibly by RT-PCR, itcould not be shown in in situ hybridization or immunocytochemistry, indicating that the level of expressionwas low.

Transgenic mice

Development of tumors in transgenic mice

The tax construct, with minimal flanking plasmid sequences (Fig. 1A), was excised by digestion with KpnI/

At the age of 6±10 months, various types of tumors developed in both transgenic lines (Fig. 2A). Most of them

TABLE 1 Primers Used in This Study Region LTR tax

Splice site GAPDH

Name of primers

Sequence

LTR1 LTR2 tax 1 HTL171 targtax RTTG1 RTTG2 18 SL 20 SL

59-TCCGGTCGACCATGAGCCCCAAATATCCCCC-39 59-TCCCAAGCTTAATTTCTCTCCTGAGAGTGCTATAG-39 59-CTGGCCACCTGTCCAGAGCA-39 59-TCTGGAAAAGACAGGGTTGGGA-39 59-CGGCTCAGCTCTACAGTTCC-39 59-CTGTGGTGTGACATAATTGGAC-39 59-CCACCACTGCTCCCATTCAT-39 59-CCATGGAGAAGGCTGGGG-39 59-CAAAGTTGTCATGGATGACC-39

Size of amplimer 732 bp 268 bp

168 bp (DNA) 100 bp (RNA) 192 bp

Note. Primers used for screening transgenic mice and for RT-PCR analysis. The expected size of the fragments obtained after amplification with RT-TG1 and RT-TG2 primers, which flank the splice site present in the transgene, is 168 bp for the amplification of genomic DNA and 100 bp for the amplification of cDNA.

334

COSCOY ET AL.

FIG. 2. Characterization of tumors in tax transgenic mice. (A) Macroscopic aspect of mesenchymal tail tumors observed in a Tax II mouse (the same tumors were observed in mice from the Tax I line). (B) Histological aspect of a tail tumor (T); paraffin-embedded section. (C) Detection of vimentin in SO/CR tumoral cells by immunoperoxydase. (D) Detection of Po, a Schwann cell-specific marker, by in situ hybridization (arrows). Po is not expressed in tumor cells (T), indicating that they are not of glial origin. (E) Detection of tax mRNA by in situ hybridization on a paraffin section of a tail fibrosarcoma, showing that all tumoral cells are positive. (F) Detection of the Tax protein in a tail tumor using immunoperoxidase.

were solid tumors of the tail; some were mammary adenocarcinomas, hepatocarcinomas, and lung carcinomas. The incidence of tumors was higher in Tax II mice, suggesting an effect of the number of copies of the transgene. Approximately 13% (6 of 47) of Tax II mice presented with tumors of the tail, and ;30% showed weight loss and became cachectic after 10 months of age. The incidence of mammary carcinomas, hepatocarcinomas, and lung carcinomas

was 6.4% (3 of 47), 2.1% (1 of 47), and 2.1% (1 of 47), respectively. In comparison, ;5% of the Tax I animals developed tail tumors, and weight loss was not observed. Histological examination of the tail tumors revealed elongated, spindle-shaped cells characteristic of a fibrosarcoma (Fig. 2B). The expression of vimentin in these cells (Fig. 2C) suggested that they were of mesenchymal origin. Given their perineural localization, we looked for the ex-

MECHANISM OF TAX TUMORIGENESIS IN MICE

335

Tax expression in cells isolated from tail fibrosarcomas

FIG. 3. Western blot analysis of Tax protein in tumoral cell lines. Total cell extracts were performed to analyze Tax expression in cell lines. (Lane 1) Primary fibroblasts from a nontransgenic C3H mouse. (Lane 2) Primary nontumoral fibroblasts from a Tax transgenic mouse. (Lane 3) L929 mouse fibroblastic cells. (Lanes 4 and 5) SO/CR 2 and SO/CR4 cells.

pression of Po-mRNA, a specific marker of Schwann cells. Tumoral cells did not express this mRNA, indicating that the tumors observed in transgenic animals are not schwannoma (Fig. 2D). In situ hybridization and immunocytochemistry showed a high level of expression of tax mRNA (Fig. 2E) and Tax protein (Fig. 2F) in tail tumors. In contrast, Tax was not expressed in the other types of tumors found in the transgenic mice.

We derived cell clones from the tail fibrosarcoma of Tax II mice to study the role of Tax in tumorigenesis. Five clones (named SO/CR1 to SO/CR5) were obtained by limiting dilution and routinely propagating for up to 100 passages. Compared with control untransformed fibroblasts, these cells were much more refractile, grew to higher density, and did not undergo contact inhibition at confluence. Instead, they formed a large number of foci after 1 week in culture. All SO/CR clones were vimentin1, Po2, Mac12, F4/802, Class I1, and Class II2 as indicated by immunocytochemistry and FACS analysis (data not shown). They all expressed high levels of Tax protein as shown by Western blotting (Fig. 3). In situ hybridization and immunocytochemistry confirmed the presence of high levels of tax mRNA and protein expression, respectively (Figs. 4A and 4B). Primary fibroblasts derived from nontumoral tail tissue of transgenic mice were negative for the expression of Tax, including by RT-PCR. The tumorigenicity of one clone (SO/CR2) was tested in C3H and BALB/c nude mice. The subcutaneous inoculation of 106 cells in BALB/c nude mice and 5 3 106 cells in

FIG. 4. Characterization and tumorigenic capacity of SO/CR2 cells. (A) tax mRNA detection by in situ hybridization in SO/CR cells. (B) Tax protein detection by immunoperoxidase in SO/CR cells. (C) Fibroblastic tumors in a 2-month-old nude Balb/c mouse, 10 days after subcutaneous injection of 106 SO/CR2 cells. (D) Tumor in a 2-month-old C3H mouse, 10 days after subcutaneous injection of 5 3 106 SO/CR2 cells.

336

COSCOY ET AL.

6-week-old C3H mice induced fibroblastic tumors (1 cm diameter) as early as 1 week after inoculation (Figs. 4C and 4D). To examine the possibility that mutations in the transgene were responsible for the tumors observed in the Tax II and Tax I cell lines, we used PCR for amplification and sequenced the transgene from a tail biopsy of a transgenic mouse and from the SO/CR2 cell line. The sequences were identical to that of the initial tax transgene. p53 and p16 status in SO/CR clones The level of Tax expression was similar in all SO/CR clones. Therefore, all experiments described in the rest of this article were performed in parallel on two clones taken at random, SO/CR2 and SO/CR4. It has been shown that a mutated form of the tumor suppressor p53 accumulates in the nucleus of PBMC transformed by HTLV-I (Reid et al., 1993; Yamato et al., 1993). However, we did not detect an accumulation of the p53 protein by western blot analysis in SO/CR clones compared to the MT2 cell line (data not shown). It has been suggested that the binding of Tax to p16 could be responsible for cell transformation. Therefore p16 expression was analyzed by western blot on primary fibroblasts, SO/CR clones and spleen lymphocytes (as positive control). The expression of p16 was high in mouse lymphocytes but no expression was observed in primary fibroblasts or SO/CR clones (data not shown). Therefore, binding of Tax to p16 is not responsible for the SO/CR cell lines transformation. NF-kB activation of SO/CR cells It also has been suggested that Tax transforms cells by activating NF-kB; therefore, we examined the effect of Tax expression on the nuclear translocation of NF-kB in SO/CR clones. A plasmid containing the thymidine kinase (TK) promoter, 3 consensus NF-kB functional sites and the luciferase gene (3NF plasmid) was transfected in SO/CR clones and control cells (Fig. 5A). The same construction in which the NF-kB sites had been mutated (M6 plasmid) also was transfected as a control. After transfection with these plasmids, the luciferase activity is either NF-kB dependent (3NF plasmid) or independent (M6 plasmid). The expression due to NF-kB can be measured as the ratio of the luciferase activity obtained with the 3NF and M6 plasmids. The results presented in Figure 5A show that the NF-kB-specific luciferase activity is 6±11 times higher in SO/CR clones than in control cells. Nuclear cell extracts were prepared from SO/CR clones as well as from C3H primary fibroblasts and NIH 3T3 cells. The binding of proteins to an oligonucleotide containing a consensus NF-kB site was tested in these extracts using an electrophoretic mobility shift assay (EMSA) (Fig. 5B). The specificity of the binding assay for

FIG. 5. Analysis of NF-kB nuclear translocation in SO/CR clones. (A) Luciferase activity measured in SO/CR and control cells after transfection with the 3NF or the M6 plasmids. Luciferase activity was measured 36 h after transfection, and the ratios of the activities obtained with the 3NF and M6 plasmids were calculated. (B) Gel shift analysis for NF-kB proteins performed with 2.5 mg of nuclear extract and a 32P-labeled oligonucleotide containing a consensus NF-kB site. Binding competitions were carried out by preincubating the nuclear extracts with either 1.75 pmol of nonlabeled oligonucleotide corresponding to an NF-kB site (specific competition) or an OCT-1 site (nonspecific competition). (Lane 1) Primary tail fibroblasts. (Lane 2) 3T3 cells. (Lanes 3 and 4) SO/CR2 and SO/CR4 cells. (Lanes 5 and 6) The same SO/CR cells pretreated during 5 h with 250 ng/ml calcium ionophore and 50 ng/ml PMA. (Lane 7) Nonspecific binding competition performed with a cold oligonucleotide corresponding to an OCT-1 site. (Lane 8) Specific competition with an oligonucleotide corresponding to an NF-kB site.

the NF-kB factors was tested by competition with an OCT-1 consensus oligonucleotide. A high level of NF-kB binding activity was observed in SO/CR clones compared with control cells. These results clearly indicate that stable Tax expression induces a constitutive level of nuclear NF-kB DNA binding activity in the absence of

MECHANISM OF TAX TUMORIGENESIS IN MICE

FIG. 6. Degradation of IkB molecules in SO/CR cells. Western blot analysis of cells treated with MG-132 reveals the accumulation of IkBa and IkBb molecules in SO/CR cells. Treatment was performed with 20 mM of MG-132 for 1, 2, and 5.5 h of incubation. Lane 0 represents cell extracts without MG-132 treatment. Primary fibroblasts were used as control cells.

induction with molecules such as phorbol-12-myristate13-acetate (PMA) or calcium ionophore. Tax-induced nuclear translocation of NF-kB is due to the degradation of IkB The nuclear translocation of NF-kB factors is regulated by IkB molecules. On cell activation by phorbol esters, mitogens, or cytokines, IkB molecules are phosphorylated and subsequently degraded by proteasomes. This leads to a rapid translocation of the active NF-kB complex into the nucleus. Therefore, the activation of NF-kB observed in SO/CR clones could be due to the degradation of IkB. We examined this possibility by measuring the level of IkB molecules before and after treatment with the proteasome inhibitor N-benzyloxycarbonyl-leucylleucyl-leucinal (MG-132; synthesized by the Organic Chemistry Laboratory, Institut Pasteur, Paris, France). Cells were treated with 20 mM MG-132 and lysed after 1, 2, or 5.5 h of treatment, and IkB molecules were detected by Western blotting using specific antibodies against IkBa and IkBb. In the absence of inhibitor, no IkBa and only small amounts of IkBb were observed in SO/CR clones. The amount of both IkBa and IkBb increased with time during treatment with MG-132 (Fig. 6). In contrast, both factors could be easily detected in control primary fibroblasts, even without treatment with the proteasome inhibitor. Their levels did not increase with MG-132 treatment. This result indicates that the IkBa and IkBb proteins are constitutively degraded by proteasomes in Tax-transformed SO/CR clones.

nuclear translocation of NF-kB, the expression of this mutant IkBb molecule should cause a loss of their transformed phenotype. Because SO/CR clones could not be transfected efficiently by usual procedures (as determined with a plasmid expressing b-galactosidase), we infected the cells with a nonreplicative murine retrovirus expressing the mutated IkBb gene. The preparation of this retroviral vector, derived from vector pLXSN, is described in Materials and Methods. SO/CR clones and control NIH 3T3 and L929 cells were infected with 5-fold dilutions of the supernatant of transfected packaging cells and allowed to grow in the presence of neomycin. Two weeks after infection, G418-resistant clones were counted. The data presented in Table 2 show that the expression of the mutated IkBb did not affect the proliferation of the control cells. In contrast, the number of clones of SO/CR cells was drastically reduced by the expression of the IkBb mutant (Table 2, Fig. 7). The few colonies obtained after infection of SO/CR clones by the IkBb mutant vector were much smaller than those growing in controls, indicating inefficient growth. Their presence might be due to residual endogenous IkBb activity despite the presence of the transdominant mutant. To verify that the level of expression of the transdominant mutant was the same in control cells and the SO/CR clones, we infected these cells at a high m.o.i. with the retrovirus expressing the mutated IkBb gene. At 24 h after infection, we measured the level of the mutated IkBb by Western blot analysis, using an antibody specific for the HA tag present in the mutated IkBb construct. The mutated IkBb gene was expressed at comparable levels in SO/CR clones and in nontransformed primary fibroblasts (data not shown). As a separate control, another retroviral recombinant was constructed by incorporating the lacZ gene, instead of the mutant IkBb gene, in the pLXSN vector. After infection with this vector, the level of b-galactosidase activity was the same in control cells and in the SO/CR clones. Taken together, our results show that the SO/CR clones cannot proliferate in the absence of degradation of IkBb and indicate that the nuclear translocation of NF-kB may be critical for Tax-mediated transformation. TABLE 2 Effects of Mutated IkBb on Tax-Induced Cell Proliferation

The nuclear translocation of NF-kB is essential for transformation To determine the role of NF-kB translocation in the transformation of SO/CR clones, we stably expressed in these cells a transdominant mutant of IkBb molecule. This mutant cannot be degraded because of two point mutations in crucial phosphorylation sites (Ser19 3 Ala, and Ser23 3 Ala). The expression of this mutant irreversibly blocks the nuclear transfer of NF-kB (Weil et al., 1997). If the transformation of SO/CR clones is due to the

337

pLXSN vector pLXSN vector containing mutated IkBb Ratio

NIH 3T3

L929

SO/CR2

SO/CR4

1525

2060

370

315

1450 1:1

1975 1:0

32 11:6

44 7:2

Note. Number of G418 resistant clones obtained after infection of SO/CR and control cells by a retroviral vector (see Fig. 7). The ratio of the number of clones obtained with pLXSN control vector and pLXSNA19-A23-IkBb infection gives a quantitative estimation of the effect of NF-kB activity on cell viability and cell proliferation.

338

COSCOY ET AL.

FIG. 7. Expression of a mutated IkBb in Tax-transformed cells leads to a reversal of the transformed phenotype. Neomycin resistant clones of L929, SO/CR2 and SO/CR4 cells were obtained by infection with a recombinant retrovirus. (Column A) Infection with the control recombinant vector. (Column B) Infection by a recombinant vector expressing the mutated IkBb protein. (Row 1) L929 cells. (Row 2) SO/CR2 clone. (Row 3) SO/CR4 clone.

Interestingly, our results also show that the proliferation of L929 and NIH 3T3 cells does not require the NF-kB pathway. DISCUSSION In this work, we describe mice transgenic for the HTLV-I tax gene under the control of an HTLV-I LTR obtained from a patient with HAM/TSP. The expression of the transgene was detected mainly in the central nervous system, in the brain as well as the spinal cord, and in testis. No neurological pathology was observed in these transgenic mice. The same expression profile was observed in two different transgenic lines, indicating that the number of copies of the transgene and the integration site were not responsible for this pattern. It has been suggested that the LTR of HTLV-I might contain determinants of pathogenicity. At present, our observations speak against this idea. In a previous work, we

observed the same expression pattern as that described in the present work in mice transgenic for the LacZ gene under the control of HTLV-I promoters originating from patients with ATL or HAM/TSP (Coscoy et al., 1996). The mice had a DBA/2xC57B1/6 genetic background, whereas in the present work, the transgenic mice were inbred C3H. Thus, the genetic background of the mice, the nature of the reporter gene, and the origin of the LTR had no influence on the expression pattern driven by HTLV-I LTRs. In another work, we compared the in vitro promoter activity of LTRs isolated from patients with ATL or HAM/TSP and showed that although the activities were statistically different in some of the cell lines tested, there was no correlation between promoter activity and the disease from which the LTR was obtained (Gonzalez-Dunia et al., 1993). Furthermore, despite the fact that the promoter used in the present work was derived from a patient with HAM/TSP (Evangelista et al., 1990), the transgenic mice developed tumors, just as did mice transgenic for the tax gene with an HTLV-I promoter obtained from a patient with ATL (Nerenberg et al., 1987). Therefore, contrary to the suggestion made by Xu et al. (1996), all our observations argue against the presence of disease-specific determinants in the HTLV-I LTR. Several studies using transgenic mice have been performed to understand the role of Tax in HTLV-I pathogenesis. These studies used various promoters to control the tax gene: the human granzyme B promoter (Grossman et al., 1995), the metallothionein promoter (Saggioro et al., 1997), and the Thy-1 promoter (Nerenberg et al., 1991). The mechanism of Tax-induced transformation has also been studied in vitro in cells transfected with the tax gene under the control of various promoters (Smith and Greene, 1991; Yamaoka et al., 1996). In contrast, we studied the in vivo transformation mediated by Tax expressed under the control of its own promoter. It is now established that Tax interacts with multiple cellular proteins (Yoshida, 1996). However, the events responsible for cell transformation have not yet been clearly identified. Although results often are contradictory, several authors suggested that the expression of NF-kB is necessary for the maintenance of the malignant phenotype in Tax-transformed fibroblasts (Kitajima et al., 1992; Yamaoka et al., 1996). Our results indicate that the constitutive expression of Tax leads to a continuous activation of NF-kB in tumor-derived murine cells. We showed that this activated results from an increased degradation of both IkBa and IkBb and that blocking this degradation stops the proliferation of the cells. Interestingly, an increased degradation of IkBa has been described in T cells of patients with ATL (Kanno et al., 1994; Lacoste et al., 1995). We also investigated the involvement of p16 in Taxinduced transformation. We did not detect p16 expression in primary fibroblasts and in the SO/CR2 and SO/ CR4 cell lines, indicating that in our model, this negative regulator of the cell cycle does not play a role in cell transformation. It has also been shown that a mutated

MECHANISM OF TAX TUMORIGENESIS IN MICE

form of the tumor suppressor p53 accumulates in the nucleus of peripheral blood mononuclear cells transformed by HTLV-I (Reid et al., 1993; Yamato et al., 1993). We examined p53 levels in SO/CR2 and SO/CR4 cells. No stabilization or increased amounts of p53 protein were observed by Western blotting in these cells compared with the MT2 cell line. In conclusion, we set up a system in which the different factors involved in the transformation by Tax can be studied easily. This system allowed us to demonstrate that the Tax-induced transformation of mice fibroblasts is mediated by the degradation of IkB factor. MATERIALS AND METHODS Cell culture NIH 3T3 and L929 cells were purchased from American Type Culture Collection (Rockville, Maryland). FLYRD18 cells (Cosset et al., 1995) were kindly provided by J.-M. Heard (Institut Pasteur). Tumoral SO/CR clones were derived from the tail fibrosarcoma of a Tax II transgenic mouse. Tail tumors were removed and trypsinized with gentle shaking at 37°C. Clones were obtained by limiting dilution. These cells were propagated for up to 100 passages. As a control, nontumoral primary fibroblasts were derived from the tails of newborn nontransgenic C3H mice. These cells could not be propagated more than five to eight passages. All cells were maintained in Dulbecco's modified Eagle's medium with Glutamax II (GIBCO BRL, Grand Island, New York), supplemented with 10% fetal calf serum and antibiotics. The MT2 cell line was obtained from the National Institutes of Health AIDS Research and Reference Reagents Program.

339

transdominant mutant of IkBb is expressed under the control of the MuLV LTR. This vector also confers resistance to neomycin. Production of a retroviral vector containing IkBb mutant On transfection with pLXSN vector, packaging of the FlyRD18 cell line produces nonreplicative viral particles, which can be used for stable gene transfer and expression (Cosset et al., 1995). For transfection, FLYRD18 cells were electroporated (1000 V, 25 F) in cold PBS with 10 mg of pLXSN or pLXSN-(A19-A23)-IkBb plasmids using BioRad (Hercules, CA)GenePulse. Virus-containing supernatants were harvested 48 h post-transfection, diluted with Polybrene (8 mg/ml), and used to infect 2.5 3 105 target cells. After 8 h, the inoculum was replaced by fresh medium; 16 h later, 1 mg/ml G418 was added. Two weeks later, the cells were fixed with 10% formaldehyde for 10 min and stained with crystal violet, and G418-resistant clones were counted. Transgenic mice Manipulation of mice and eggs were carried out in the transgenesis laboratory at the Pasteur Institute using standard procedures. The LTR-tax transgenic mice were generated in an inbred C3H background. The DNA construct was linearized and purified before microinjection in fertilized C3H eggs. Founder animals were identified by PCR on DNA from tail biopsies, using LTR and taxspecific primers (Table 1). The LTR-tax founders were mated with C3H mice, and transgenic animals were screened by PCR analysis of DNA from tail biopsies.

Plasmids

RNA purification and RT-PCR

To obtain the transgene described in Figure 1, the viral LTR was amplified by PCR from peripheral blood lymphocytes of a patient with TSP/HAM and cloned in pBlueScript SK. A cDNA coding for Tax (Gonzalez-Dunia et al., 1992) was inserted downstream from the LTR. The SV40 splice site and polyadenylation signal derived from pSVK3 were cloned at the 39 end of the construct as described (Coscoy et al., 1996). Plasmids expressing the luciferase gene driven by the TK promoter (3NF and M6) were kindly provided by F. Bachelerie (Institut Pasteur). Plasmid 3NF contains three consensus NF-kB sites downstream of the promoter. These sites are mutated in the M6 plasmid, abolishing the binding of NF-kB factors. The transdominant mutant of IkBb was excised from pRc-CMV-(A19-A23)-IkBb plasmid (Weil et al., 1997) by HindIII/XbaI digestion and subcloned in a pBlueScript (SK) vector. This plasmid was then digested by XhoI, and the insert was cloned in the pLXSN retroviral vector, kindly provided by J.-M. Heard (Institut Pasteur). In the resulting plasmid, named pLXSN9-(A19-A23)-IkBb, the

Mice were anesthetized and perfused with PBS to avoid blood contamination. Organs were collected and RNA was purified according to the GAPC method (Chomczynski and Sacchi, 1987), treated with RQ1 DNase, and reverse transcribed by the AMV reverse transcriptase as described previously (Coscoy et al., 1996). Primers used for PCR are shown in Table 1. Amplified products were detected by Southern blot hybridization using [32P] 39 end-labeled specific oligonucleotide probes. Histochemical analysis Mice were anesthetized and perfused with 25 ml of PBS followed by 25 ml of 4% paraformaldehyde (PFA). The organs were fixed in 4% PFA, embedded in paraffin, and processed by standard histological procedures. In situ hybridization was performed as previously described (Brahic and Ozden, 1992; Ozden et al., 1991) using 35Slabeled riboprobes specific for Tax or Po-mRNA, a specific marker for Schwann cells.

340

COSCOY ET AL.

Immunoblotting Exponentially growing cells were lysed in 150 mM NaCl, 50 mM Tris, pH 8, 1% Nonidet P-40, 0.5 mM phenylmethylsulfonyl fluoride, and 0.2 mg/ml Pefabloc (Interbiotech/Interchim, Montluc˛ on, France). Chromatin was removed by centrifugation for 5 min at 13,000 g. Protein concentration was determined by the BCA protein assay (Pierce Chemical, Rockford, Illinois). Proteins (50 mg) were fractionated by SDS±PAGE, transferred to a cellulose membrane, incubated with an anti-Tax monoclonal antibody (168A51; National Institutes of Health AIDS Research and Reference Reagents Program) and then with an anti-mouse horseradish peroxidase-coupled IgG and detected with a chemiluminescence detection kit (Amersham). Mouse anti-IkBa, rabbit anti-IkBb, mouse antip53 antibodies, and a mouse anti-HA (a tag present in the mutant IkBb) were provided by F. Arenzana, R. Weil, F. Thierry, and C. Muchardt, respectively (Institut Pasteur, Paris). Monoclonal antibodies (sc-1661 and sc-099) from Santa Cruz (Santa Cruz, California) were used to detect mouse p16 and p53 proteins. Luciferase activity assay SO/CR and control cells (5 3 105 cells) were transfected with 1 mg of plasmid using Lipofectamine as described by the manufacturer (Life Technologies, Grand Island, New York). Cells were harvested 36 h post-transfection and lysed in 25 mM Tris, pH 7.8, 8 mM MgCl2, 1 mM dithiothreitol, 1% Triton X-100, 1% bovine serum albumin, and 15% Glycerol. Luciferase activity was determined with a Lumat LB9501 luminometer (Berthold, France) using 1 mM ATP and 0.25 mM D-luciferin (Sigma Chemical, St. Louis, Missouri) as a substrate. EMSA Exponentially growing cells were washed in PBS and treated with 50 mM NaCl, 50 mM HEPES, pH 8, 500 mM sucrose, 1 mM EDTA, 0.5 mM spermidine, 0.2% Triton X-100, 0.5 mM phenylmethylsulfonyl fluoride, and 0.2 mg/ml Pefabloc. After centrifugation, the pellets were washed in 50 mM NaCl, 10 mM HEPES, pH 8, 25% glycerol, 0.1 mM EDTA, 0.5 mM spermidine, and 0.15 mM spermine and then resuspended in 350 mM NaCl, 10 mM HEPES, pH 8, 25% glycerol, 0.1 mM EDTA, 0.5 mM spermidine, and 0.15 mM spermine. After centrifugation, supernatants were used as nuclear extract for DNA binding reactions using a gel shift assay system from Promega (Madison, Wisconsin). Gel shift analysis of NF-kB proteins were performed with 2.5 mg of nuclear extract using 32 P-labeled oligonucleotide containing a consensus NF-kB site. Competitions were carried out by preincubating extracts (for 20 min on ice) with 1.75 pmol of nonlabeled oligonucleotides corresponding either to the NF-kB site (specific competition) or the OCT-1 site (nonspecific competition). The NF-kB consensus double-

stranded oligonucleotide sequence was 59-AGTTGAGGGGACTTTCCCAGGC-39 and the OCT-1 consensus double-stranded oligonucleotide was 59-TGTCGAATGCAAATCACTAGAA-39. After incubation with the 32P-labeled probe (30 min at 37°C), nuclear extracts were loaded onto 5% nondenaturing polyacrylamide gels and run in a 0.53 Tris-borate EDTA buffer. ACKNOWLEDGMENTS We are grateful to C. Rousseau, M. Tuneskog, and V. Derreumaux for technical assistance, to R. Weil for critically reading the manuscript, and to M. Gau for her constant help. This work was supported by grants from Institut Pasteur, Agence Nationale de Recherches sur le SIDA, Centre National de la Recherche Scientifique, and EC Human Capital Mobility Program (Contract CHRX-CT94±0670).

REFERENCES Ballard, D., BoÈehnlein, E., Lowenthal, W., Wano, Y., Franza, B., and Greene, W. (1988). HTLV-I Tax induces cellular proteins that activates the KB element in the IL2 receptor a gene. Science 241, 1652±1655. Bantignies, F., Rousset, R., Desbois, C., and Jalinot, P. (1996). Genetic characterization of transactivation of the HTLV-I promoter: binding of Tax to Tax responsive element I is mediated by the cyclic AMPresponsive members of the CREB/ATF family of transcription factors. Mol. Cell. Biol. 16, 2174±2182. Baranger, A., Palmer, C., Hamm, M., Giebler, H., Brauweiler, A., Nyborg, J., and Schepartz, A. (1995). Mechanism of DNA binding enhancement by HTLV-I transactivator Tax. Nature 376, 606±608. BeÁraud, C., Sun, S., Ganchi, P., Ballard, D., and Greene, W. (1994). HTLV-I Tax associates with and is regulated by NF-kB2 p 100 gene product: implications for viral latency. Mol. Cell. Biol. 14, 1374±1382. Brahic, M., and Ozden, S. (1992). Simultaneous detection of cellular RNA and proteins. In ªIn Situ Hybridization: A Practical Approachº (R. G. Wilkinson, Eds.), pp. 85±104. IRL Press, Oxford. Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156±159. Coscoy, L., Gonzalez-Dunia, D., Chirinian-Syan, S., Brahic, M., and Ozden, S. (1996). Analysis of the expression directed by two HTLV-I promoters in transgenic mice. J. Neurovirol. 2, 336±344. Cosset, F., Takeuchi, Y., Battini, J., Weiss, R., and Collins, M. (1995). High-titer packaging cells producing recombinant retroviruses resistant to human serum. J. Virol. 69, 7430±7436. Evangelista, A., Maroushek, S., Minnigan, H., Larson, A., Retzel, E., Haase, A. T., Gonzalez-Dunia, D., McFarlin, D., Mingioli, E., Jacobson, S., Osame, M., and Sonoda, S. (1990). Nucleotide sequence analysis of a provirus derived from an individual with tropical spastic paraparesis. Microb. Pathogen. 8, 259±278. Fujii, M., Sassone-Corsi, P., and Verma, I. M. (1988). c-fos promoter trans-activation by the tax1 protein of human T cell leukemia virus type I. Proc. Natl. Acad. Sci. USA 85, 8526±8530. Fujii, M., Tsuchiya, H., Toita, M., and Yoshida, M. (1992). Interaction of HTLV-I Tax with p67SRF causes the aberrant induction of cellular immediate early genes through CArG boxes. Genes Dev. 6, 2066± 2076. Gessain, A., Barin, E., Vernant, J. C., Gout, O., Maurs, L., Calender, A., and de TheÂ, G. (1985). Antibodies to human T lymphotropic virus type-1 in patients with tropical spastic paraparesis. Lancet 2, 407± 410. Gonzalez-Dunia, D., Chirinian-Syan, S., Brahic, M., and Ozden, S. (1992). Functional analysis of two long terminal repeats from the HTLV-I retrovirus. Gene 116, 151±158. Gonzalez-Dunia, D., Komurian-Pradel, F., Chirinian-Syan, S., de TheÂ, G., Brahic, M., and Ozden, S. (1993). Comparative analysis of HTLV-I

MECHANISM OF TAX TUMORIGENESIS IN MICE promoter activities reveals no disease-linked pattern of expression. AIDS Res. Hum. Retroviruses 9, 337±341. Grassmann, R., Dengler, C., Muller-Fleckenstein, I., Fleckenstein, B., McGuire, K., Dokhelar, M., Sodorski, J., and Haseltine, A. (1989). Transformation to continuous growth of primary T lymphocyte by HTLV-I X region genes transduced by a herpesvirus saimiri vector. Proc. Natl. Acad. Sci. USA 92, 1057±1061. Grossman, W. J., Kimata, J. T., Wong, F.-H., Zutter, M., Ley, T. J., and Ratner, L. (1995). Development of leukemia in mice transgenic for the tax gene of human T cell leukemia virus type I. Proc. Natl. Acad. Sci. USA 92, 1057±1061. Hinrichs, S. H., Nerenberg, M., Reynolds, R. K., Khoury, G., and Jay, G. (1987). A transgenic mouse model for human neurofibromatosis. Science 237, 1340±1343. Hinuma, Y., Nagata, K., Hanaoka, M., Misoka, M., Nakal, T., Matsumoto, T., Kinoshita, K., Shirakawa, S., and Miyoshi, I. (1981). Adult T cell leukemia: antigen in an ATL cell line and detection of antibodies to the antigen in human sera. Proc. Natl. Acad. Sci. USA 78, 6476±6480. Inoue, J.-I., Seiki, M., Taniguchi, T., Tsuru, S., and Yoshida, M. (1986). Induction of interleukin 2 receptor gene expression by p40x encoded by human T cell leukemia virus type I. EMBO J. 5, 2883±2888. Kanno, T., Brown, K., Franzoso, G., and Siebenlist, U. (1994). Kinetic analysis of Human T cell leukemia virus type I Tax-mediated activation of NF-kB. Mol. Cell. Biol. 14, 6443±6451. Kitajima, I., Shinohara, T., Bilakovics, J., Brown, D. A., Xu, X., and Nerenberg, M. (1992). Ablation of transplanted HTLV-I Tax-transformed tumors in mice by antisense inhibition of NF-kB. Science 258, 1792±1795. Lacoste, J., Petropoulos, L., Pepin, L., and Hiscott, J. (1995). Constitutive phosphorylation and turnover of IKB in human T cell leukemia virus type I infected and Tax-expressing cells. J. Virol. 69, 564±569. Matsumoto, K., Shibata, H., Fujisawa, J., Inoue, H., Hakura, A., Tsukahara, T., and Fujii, M. (1997). HTLV-I Tax protein transforms rat fibroblastes via two distinct pathways. J Virol. 71, 4445±4451. Mochizuki, M., Watanabe, T., and Yamaguchi, K. (1992). HTLV-I uveitis: a distinct clinical entity caused by HTLV-I. Jpn. J. Cancer Res. 83, 236±239. Morgan, O., Rodgers-Johnson, P., Mora, C., and Char, G. (1989). HTLV-I and polymyositis in Jamaica. Lancet. 2, 1184±1186. Munoz, E., Courtois, S., Veschambre, P., Jalinot, P., and Israel, A. (1994). Tax induces nuclear translocation of NF-kB through dissociation of cytoplasmic complexes containing p105 or p100 but does not induce degradation of IkB alpha/MAD3. J. Virol. 68, 8035±8044. Nerenberg, M., Hinrichs, S. H., Reynolds, R. K., Khoury, G., and Jay, G. (1987). The tat gene of human T lymphotropic virus type I induces mesenchymal tumors in transgenic mice. Science 237, 1324±1329. Nerenberg, M. I., Minor, T., Price, J., Ernst, D. N., Shinohara, T, and Schwarz, H. (1991). Transgenic thymocytes are refractory to transformation by the human T cell leukemia virus type I tax gene. J. Virol. 65, 3349±3353. Nishioka, K., Maruyama, I., Sato, K., Kitajama, I., Nakajima, Y., and Osame, M. (1989). Chronic inflammatory arthropathy associated with HTLV-I. Lancet 1, 441. Osame, M., Usuku, K., Izumo, S., Ijichi, N., Amitani, H., Igata, A., Matsumoto, M., and Tara, M. (1986). HTLV-I associated myelopathy, a new clinical entity. Lancet 1, 1031±1032. Ozden, S., Aubert, C., Gonzalez-Dunia, D., and Brahic, M. (1991). In situ analysis of proteolipid protein gene transcripts during persistent Theiler's virus infection. J. Histochem. Cytochem. 39, 1305±1309. Perini, G., Wagner, S., and Green, M. (1995). Recognition of bZIP proteins by the human T cell leukaemia virus transactivator Tax. Nature 376, 602±605.

341

Poiesz, B. J., Ruscetti, F. W., Gadzar, A. F., Bunn, P. A., Minna, J. D., and Gallo, R. C. (1980). Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T cell lymphoma. Proc. Natl. Acad. Sci. USA 77, 7415±7419. Pozzati, R., Vogel, J., and Jay, G. (1990). The HTLV-I tax gene can cooperate with the Ras oncogene to induce neoplastic transformation of cell. Mol. Cell. Biol. 10, 413±417. Reid, R., Lindholm, P., Mireskandari, A., Dittmer, J., and Brady, J. (1993). Stabilization of wild-type P53 in human T lymphocytes transformed by HTLV-I. Oncogene 8, 3029±3036. Rousset, R., Desbois, C., Bantignies, F., and Jalinot, P. (1996). Effect on NK-kB1/p105 processing of the interaction between HTLV-I transactivator Tax and the proteasome. Nature 381, 328±331. Saggioro, D., Rosato, A., Esposito, G., Rosenberg, M., Harrisson, J., Felber, B., Pavlakis, G., and Chieco-Bianchi, L. (1997). Inflammatory polyarthropathy and bone remodeling in HTLV-I tax-transgenic mice. J. AIDS Hum. Retrovirol. 14, 272±280. Smith, M., and Greene, W. (1991). Type I human T cell leukemia virus Tax protein transforms rat fibroblasts through the cyclic adenosine monophosphate response element binding protein/activating transcription factor pathway. J. Clin. Invest. 88, 1038±1042. Suzuki, T., Kitao, S., Matsushime, H., and Yoshida, M. (1996). HTLV-I Tax protein interacts with cyclin-dependant kinase inhibitor p16INK4A and counteracts its inhibitory activity towards CDK4. EMBO J. 15, 1607±1614. Tanaka, A., Takahashi, C., Yamaoka, S., Nosaka, T., Maki, M., and Hatanaka, M. (1990). Oncogenic transformation by the tax gene of HTLV-I in vitro. Proc. Natl. Acad. Sci. USA 87, 1071±1075. Terada, K., Katamine, S., Eguchi, K., Moriuchi, R., Kita, M., Shimada, H., Yamashita, I., Iwata, K., Tsuji, Y., Nagataki, S., and Miyamoto, T. (1994). Prevalence of serum and salivary antibodies to HTLV-I in SjoÈgren's syndrome. Lancet 344, 1116±1119. Vernant, J. C., Buisson, G., Magdeleine, J., De Thore, J., Jouannelle, A., Neisson-Vernant, C., and Monplaisir, N. (1988). T lymphocyte alveolitis, tropical spastic paraparesis, and SjoÈgren syndrome. Lancet 1, 177. Weil, R., Laurent-Winters, C., and IsraeÈl, A. (1997). Regulation of IkBb degradation. J. Biol. Chem. 272, 9942±9949. Xu, X., Kang, S., Heidenreich, O., Da, B., and Nerenberg, M. (1996). Sequence requirements of ATF2 and CREB binding to the HTLV-I LTR R region. Virology 218, 362±371. Yamaoka, S., Inoue, H., Sakurai, M., Sugiyama, T., Hazama, M., Yamada, T., and Hatanaka, M. (1996). Constitutive activation of NF-KB is essential for transformation of rat fibroblasts by tha HTLV-I Tax protein. EMBO J. 15, 873±887. Yamato, K., Oka, T., Hiroi, M., Iwahara, Y., Sugito, S., Tsuchida, N., and Miyoshi, I. (1993). Aberrant expression of the P53 tumor suppressor gene in ATL and HTLV-I infected cells. Jpn. J. Cancer Res. 84, 4±8. Yoshida, B., Miyhoshi, I., and Hinuma, Y. (1982). Isolation and characterization of retrovirus from cell lines of human adult T cell leukaemia and its implication in the disease. Proc. Natl. Acad. Sci. USA 79, 2031±2035. Yoshida, M. (1993). Tax: regulation of gene expression and disease. Trends Microbiol. 1, 131±135. Yoshida, M. (1996). Multiple targets of HTLV-I for dysregulation of host cells. Semin. Virol. 7, 349±360. Zhao, L., and Giam, C. (1991). Interaction of HTLV-I transcriptional activator Tax with cellular factors that bind specifically to the 21 bp repeats in the HTLV-I enhancer. Proc. Natl. Acad. Sci. USA 88, 11445±11449.