Antitumor Effect of TAT-Oxygen-Dependent Degradation-Caspase-3 Fusion Protein Specifically Stabilized and Activated in Hypoxic Tumor Cells

Home , Angiogenin

[CANCER RESEARCH 62, 2013–2018, April 1, 2002] Antitumor Effect of TAT-Oxygen-dependent Degradation-Caspase-3 Fusion Protein Specifically Stabilized and Activated in Hypoxic Tumor Cells

Hiroshi Harada, Masahiro Hiraoka, and Shinae Kizaka-Kondoh1 Pharmaceuticals Division, Pharmaceutical R&D Department, POLA Chemical Industries, Inc., Totsuka-ku, Yokohama 244-0812 [H. H.]; Department of Therapeutic Radiology and Oncology, Kyoto University Graduate School of Medicine, Shogoin, Sakyo-ku, Kyoto 606-8507 [M. H.]; and Department of Molecular Oncology, Kyoto University Graduate School of Medicine, Sakyo-ku, Kyoto 606-8501 [S. K-K.], Japan

ABSTRACT The ODD domain of HIF-1␣ is located in its central region and consists of ϳ200 amino acid residues. The ODD domain controls the Human solid tumors contain hypoxic regions that have considerably degradation of HIF-1␣ by the ubiquitin-proteasome pathway, and the lower oxygen tension than normal tissues. These impart resistance to deletion of this entire region is required for DNA binding and trans- radiotherapy and anticancer chemotherapy, as well as predisposing to increased tumor metastases. To develop a potentially therapeutic protein activation in the absence of hypoxic signaling (15). Recently, the drug highly specific for solid tumors, we constructed fusion proteins product of von Hippel-Lindau tumor suppressor gene was reported to selectively stabilized in hypoxic tumor cells. A model fusion protein, mediate ubiquitination and proteasomal degradation of HIF-1␣ under oxygen-dependent degradation (ODD)-␤-galactosidase (␤-Gal), composed normoxic conditions by interaction with the core of the ODD domain of a part of the ODD domain of hypoxia-inducible factor-1␣ fused to of HIF-1␣ (16–18). ␤-Gal, showed increased stability in cultured cells under a hypoxia-mimic HIV-1 TAT has been used to deliver functional biomolecules into condition. When ODD-␤-Gal was further fused to the HIV-TAT protein cells. Although the entire protein can be used for this purpose, it is transduction domain (TAT47–57) and i.p. injected to a tumor-bearing more efficient to use the highly basic region containing residues mouse, the biologically active fusion protein was specifically stabilized in 49–57. Through covalent attachment to TAT , several proteins solid tumors but was hardly detected in the normal tissue. Furthermore, 47–57 have been delivered into cells, including an inhibitor of human pap- when wild-type (WT) caspase-3 (Casp3WT) or its catalytically inactive illomavirus type 16 (19), ovalbumin into the MHC class I pathway mutant was fused to TAT-ODD and i.p. injected to a tumor-bearing kip1 INK4a mouse, the size of tumors was reduced by the administration of TAT- (20), the cyclin-dependent kinase inhibitors p27 (21) and p16 WT WT ODD-Casp3 but not by TAT-ODD-mutant Casp3. TAT-ODD-Casp3 (22), and a Casp3 protein (23). TAT79–57 has also been successfully did not cause any obvious side effects on tumor-bearing mice, suggesting used in vivo to deliver ␤-Gal into all tissues of the mouse, including specific stabilization and activation of the fusion protein in the hypoxic the brain (24). tumor cells. These results suggest that the combination of protein therapy Casp3 is a major executioner protease in known apoptotic path- using a cytotoxic TAT-ODD fusion protein with radiotherapy and chem- ways. Casp3 remains dormant until the initiator caspases activate it by otherapy may provide a new strategy for annihilating solid tumors. direct proteolysis. Activated Casp3 cleaves the inhibitor of caspase- activated DNase, resulting in cell death (25–27). Activation of Casp3 INTRODUCTION under hypoxic conditions has been reported in several cell types, such as ventricular myocytes, endothelial cells, and squamous cell lung Ͼ Solid tumors comprise 90% of all of human malignancies. Fur- carcinomas (28–31). thermore, biologically and therapeutically significant hypoxia occurs Here we demonstrated that part of the ODD domain conferred in many solid tumor masses. The hypoxic tumor cells are more oxygen-dependent stability to a fusion protein and that the covalent resistant to radiotherapy and chemotherapy than their well-oxygen- ␤ attachment of TAT47–57 and ODD548–603 to bioactive -Gal allowed ated counterparts and are capable of stimulating angiogenesis by the quick delivery and specific stabilization in a solid tumor. Furthermore, release of several growth factors (e.g., vascular endothelial growth the replacement of ␤-Gal with Casp3 reduced tumor masses without factor, angiogenin, and platelet-derived growth factor) for endothelial any obvious side effects, demonstrating the feasibility and efficacy of cells (1–3). Although the specific delivery of anticancer gene products protein therapy using a TAT-ODD fusion protein for solid tumors. to hypoxic regions of solid tumors has been studied extensively (4–9), an effective technique has not been established. 2 ␣ The transcription factor, HIF -1 , is an important mediator of the MATERIALS AND METHODS hypoxic response of tumor cells and controls the up-regulation of a number of factors vital for solid tumor expansion, including angio- Cell Cultures and Reagents. The human lung adenocarcinoma cell line genic factors, such as vascular endothelial growth factor (10). The A549, the human embryo kidney cell line HEK293, the human pancreatic ␣ cancer cell line CF/PAC-1, and the mouse fibroblast cell line NIH/3T3 were up-regulation of HIF-1 by decreased cellular O2 concentration in- volves increased mRNA expression, protein stabilization, nuclear obtained from American Type Culture Collection. A549, HEK293, and NIH/ localization, and trans-activation (11). Perhaps the most striking reg- 3T3 were maintained in 5% FCS-DMEM, and CF/PAC-1 was maintained in 10% FCS-Iscove’s modified Dulbecco’s medium (Life Technologies, Inc., ulation occurs exclusively at the level of protein expression via ␣ Rockville, MD) supplemented with penicillin (100 units/ml) and streptomycin changes in the half-life of the HIF-1 protein in response to hypoxia (100 ␮g/ml). Cbz-LLL (Peptide Institute, Inc., Osaka, Japan) was dissolved in both in cultured cells and in vivo (12–14). DMSO to 10 mM. Pimo hydrochloride (Hypoxyprobe-1; Natural Pharmacia International, Inc., Research Triangle Park, NC) was dissolved in normal saline Received 9/21/01; accepted 1/24/02. (10 mg/ml). The costs of publication of this article were defrayed in part by the payment of page DNA Constructs. To construct pCH/ODD plasmids encoding a series of charges. This article must therefore be hereby marked advertisement in accordance with ODD-␤-Gal fusion proteins, pCH was first constructed by annealing the 18 U.S.C. Section 1734 solely to indicate this fact. Ј 1 To whom requests for reprints should be addressed, at Department of Molecular oligonucleotides containing Kozak and NLS sequences (5 -AGCTTGACAT- Oncology, Kyoto University Graduate School of Medicine, Yoshida-Konoe-cho, Sakyo- GGCGCCTAAGAAGAAGAGGAAGA-3Ј and 5Ј-GATCTCTTCCTCTTCT- ku, Kyoto 606-8501, Japan. TCTTAGGCGCCATGTCA-3Ј) and inserting the annealed DNA fragment 2 The abbreviations used are: HIF, hypoxia-inducible factor; ODD, oxygen-dependent between HindIII and BglII sites of the pCH110 vector (Amersham Pharmacia degradation; ␤-Gal, ␤-galactosidase; NLS, nuclear localization signal; Cbz-LLL, N- carbobenzoxyl-L-leucinyl-L-leucinyl-L-norvalinal; Pimo, Pimonidazole; Ab, antibody; Biotech, Milwaukee, WI). Then cDNA fragments encoding various lengths of WT, wild type; Casp3, Caspase-3; MUT, mutant. the ODD domain of human HIF-1␣ were inserted between BglII and KpnI sites 2013

of pCH. The cDNA fragments encoding ODD were amplified by PCR using pCH plasmids or treated with TAT-fusion proteins were washed three times corresponding oligonucleotides as primers from cDNA prepared from total with PBS, fixed with 1% formaldehayde-0.2% glutaraldehyde solution, RNA from A549. The ODD fragments in pCH/557-574, pCH/562–569, pCH/ washed twice with PBS, and then treated with an X-Gal staining solution (5

557-571, and pCH/560-574 were prepared by annealing the corresponding mM potassium ferrocyanide, 5 mM potassium ferricyanide, 1 mM MgCl2, and oligonucleotides. 0.04% X-gal) at 37°C. The experiments were repeated twice with similar To construct pBAD/PC, pBAD/3-0, and pBAD/557-574, which encode the results. ␤ -Gal protein fused to TAT, TAT-ODD/3-0, and TAT-ODD557–574, respec- Hypoxia. Hypoxia was achieved by incubating cells in a humidified environ- tively, pCH/TAT/PC, pCH/TAT/3-0, and pCH/TAT/557-574 were first con- ment at 37°CinCO2 incubators maintained at 94% N2,5%CO2, and 1% O2. structed. To prepare pCH/TAT/3-0, pCH/TAT/557-574, or pCH/TAT/PC, Preparation of TAT-fusion Protein and in Vitro Assay. To prepare ␤ oligonucleotides encoding TAT47–57 were inserted into the BglII site of pCH/ TAT-, TAT-ODD/3-0-, and TAT-ODD557–574- -Gal proteins, pBAD/PC, 3-0 and pCH/557-574 or substituted with the BglI-KpnI ODD/3-0 fragment of pBAD/3-0, and pBAD/557-574 were transformed into LMG194 bacteria (In- pCH/3-0, respectively. Then, pCH/TAT/3-0, pCH/TAT/557-574, and pCH/ vitrogen). TAT-fusion proteins were purified over a Ni2ϩ-NTA column (Qia- TAT/PC were digested with HindIII, blunted with klenow-fragment, and gen) following the manufacturer’s instruction manual. The ␤-Gal activities of further digested with SacI. The blunt-SacI DNA fragments were inserted into the final preparations were determined by ␤-Gal assay (Promega). To prepare blunted BamHI (413) and SacI sites of pBAD/His/LacZ (Invitrogen, Carlsbad, TAT-ODD/3-0-Casp3WT and TAT-ODD/3-0-Casp3MUT, pGEX/3-0/Casp3WT CA), resulting in pBAD/3-0, pBAD/557-574, and pBAD/PC. To construct and pGEX/3-0/Casp3MUT were transformed into BL21(DE3)pLysS (Novagen, pGEX/3-0/Casp3WT encoding TAT-ODD/3-0-Casp3, Casp3 cDNA was am- Madison, WI), and the fusion proteins were purified using Glutathione Sepha- plified by PCR using corresponding oligonucleotides as primers from cDNA rose 4B and PreScission Protease (Amersham Pharmacia Biotech) following prepared from total RNA from A549 and substituted for ␤-Gal cDNA in the manufacturer’s instruction manual. pCH/TAT/3-0, and a DNA fragment encoding TAT-ODD/3-0-Casp3WT was Xenograft Tumor Growth, TAT-fusion Protein Injection, and Immu- inserted to SmaI and SalI sites of pGEX-6P-3 (Amersham Pharmacia Biotech, nohistochemical Analysis. CF/PAC-1, human pancreatic cancer cells Uppsala, Sweden). Casp3MUT encoding a mutant Casp3(C163M) was con- (2 ϫ 106) were s.c. injected into the flanks of 6–8-week-old nude mice (CLEA structed as described previously (23) and then substituted for Casp3 in pGEX/ Japan, Inc., Tokyo, Japan). The mice carrying tumors were i.p. injected with 3-0/Casp3WT. 100 ␮l of TAT-␤-Gal, TAT-ODD-␤-Gal protein (containing 100 units of DNA Transfection, ␤-Gal Assay, and X-gal Staining. For ␤-Gal assays, ␤-Gal activity), or TAT-ODD-Casp3 protein 2–3 weeks after the xenograft. To cells (1 ϫ 104/well) were seeded onto 24-well plates and cotransfected with 2 detect the hypoxic regions, the tumor-bearing mice were i.p. injected with ␮g of pCH plasmids (Fig. 1, A and B) and 0.2 ␮g of pRL-CMV (Promega, Hypoxyprobe-1 (60 mg/kg) 2 h after TAT-fusion protein administration. Later Madison, WI) as described (32). The following day, the cells were trypsinized, (2 h), the tumors (ϳ300 mm3) and the livers were surgically excised, fixed in and aliquots were plated in duplicate onto 24-well plates (one well without 4% formaldehyde, and embedded in paraffin for hematoxylin and anti-␤-Gal or ␮ Cbz-LLL and the other with 5 M Cbz-LLL). Later (24 h), the cells were lysed antipimonidazol staining or were dropped into liquid N2 immediately after with 200 ␮lof1ϫ Reporter Lysis Buffer (Promega). The lysates (50 ␮l) were excision and kept at Ϫ80°C until staining for X-gal. Immunostaining of the subjected to the calorimetric enzyme assay in triplicate using a ␤-Gal Enzyme tissue sections for the hypoxia marker, Pimo, and ␤-Gal was carried out with

Assay System (Promega), and obtained A405 nm was normalized by lumines- a Hypoxyprobe-1 kit (Natural Pharmacia International, Inc., Belmont, MA) cence counts (10 s) of 10-␮l lysates assayed with a Luciferase Assay System and rabbit anti-␤-Gal Ab (ICN Pharmaceuticals, Inc., Aurora, OH) by the (Promega). The relative ␤-Gal activities were calculated by dividing the indirect immunoperoxidase methods as described previously (33). X-gal stain- normalized counts of the lysate from the cells cultured with Cbz-LLL with the ing of the tissue sections was done as described previously (34). ones of the corresponding lysate from the untreated cells. The experiments Assay for Apoptosis. NIH/3T3 cells (1 ϫ 104/well) were seeded onto were repeated at least three times. For X-gal staining, the cells transfected with 24-well plates and preincubated for 24 h in the presence of the solvent (DMSO) or 5 ␮M Cbz-LLL. Then, TAT-ODD-Casp3 proteins (7.5 ␮g/30 ␮l) were added to the fresh culture medium, and the cells were further incubated

for 24 h under normoxic (20% O2) or hypoxic (1% O2) conditions. The photos were taken under an inverted microscope. Assay for Casp3 Activity. TAT-ODD-Casp3 proteins (7.5 ␮g/30 ␮l) were added to the culture of HEK293 cells (5 ϫ 105 Cells/0.5 ml) and incubated for 4 h in the presence of the solvent (DMSO) or 10 ␮M Cbz-LLL. Then the cells were harvested, and Casp3 activity in a 50-␮g cell lysate was detected as the amount of chromophore p-nitroanilide after cleavage from the labeled sub- strate DEVD-p-nitroanilide (CPP32/Casp3 colorimetric protease assay kit; MBL, Nagoya, Japan) and quantified using a microtiter plate reader at 405 nm.

RESULTS Determination of the Core of the ODD Domain Responsible for ODD of an ODD-fusion Protein. Because the stability of HIF-1␣ is regulated through the ODD domain, we assumed that the stability of a protein fused to an appropriate region of ODD domain would be under the same control as HIF-1␣. To identify the region responsible for ODD of an ODD-fusion protein, we constructed a series of fusion proteins composed of parts of the ODD domain of HIF-1␣ and ␤-Gal as a bioactive reporter. The plasmids encoding the NLS and various ␤ Fig. 1. Identification of HIF-1␣ ODD regions responsible for ODD of ␤-Gal fusion lengths of the ODD domain fused to the NH2 terminal of -Gal (Fig. proteins. A, schematic representation of ODD-␤-Gal reporter plasmid. B, summary of the 1, A and B) were transfected to HEK293 cells. The ␤-Gal activity of ␤ effects of different ODD regions on ODD regulation of ODD- -Gal fusion proteins. The these cells was examined by X-gal staining after culturing in the amino acid numbers indicated for each ODD fragment correspond to the amino acid number of HIF-1␣. The ODD regulation refers to the increased ␤-Gal activity and presence or absence of Cbz-LLL, a peptide aldehyde that inhibits the increased number and thickness of blue color in X-gal-stained cells cultured with 5 ␮M major peptidase activities of proteasomes and reduces the degradation the ODD regulations were observed, but ,ءCbz-LLL as compared with no treatment. ϩ relatively higher ␤-Gal activity was observed in the transfectants cultured without 5 ␮M of proteins (35). Therefore, culture with Cbz-LLL renders the cells in Cbz-LLL. a hypoxia-mimic condition (36). The results of X-gal staining were 2014

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2002 American Association for Cancer Research. ANTITUMOR EFFECT OF TAT-ODD FUSION PROTEIN summarized as the ODD regulation, which refers to the increased ␤-Gal activity and increased number and thickness of blue color in X-gal-stained cells in Fig. 1B. In the cells transfected with the plasmids encoding fusion proteins consisting of ␤-Gal and the HIF-1␣ residues 401–447 (pCH/0-1), 401–497 (pCH/0-2), 401–547 (pCH/0- 3), 448–497 (pCH/1-2), 448–547 (pCH/1-3), or 579–603 (pCH/4-0), the X-Gal staining pattern was same between the Cbz-LLL-treated and untreated cells (Figs. 1B and 2A). On the other hand, in the cells transfected with the plasmids encoding the fusion proteins consisting of ␤-Gal and the HIF-1␣ residues 401–603 (pCH/0-0), 448–603 (pCH/1-0), 498–603 (pCH/2-0), 548–603 (pCH/3-0), or 548–583 (pCH/3-4), more cells were stained deeper with X-gal in the Cbz- LLL-treated cells than the untreated ones (Figs. 1B and 2A). To determine the core of ODD region responsible for ODD of the ODD-fusion proteins, we further narrowed down the ODD region. The residues 548–583 have 18 amino acid sequences (557–574) conserved among the related HIFs and between species. We constructed a plasmid encoding the fusion protein consisting of ␤-Gal and ␣ the HIF-1 residues 557–574 (ODD557–574) and examined its ODD regulation as above. Although a considerable number of cells was stained with X-gal in the absence of Cbz-LLL, more cells were stained deeper in the Cbz-LLL-treated cells than the untreated ones (Fig. 2A), suggesting that ODD557–574 still retained the ODD function. How- ever, the stability of fusion proteins containing the HIF-1␣ residues of 562–569 (pCH/562–569), 557–571 (pCH/557–571), or 560–574 (pCH/560–574) were no longer under the control of oxygen tension

(Fig. 1B), indicating that ODD557–574 contained the minimum ODD residues necessary for the ODD regulation of an ODD-fusion protein. Furthermore, to examine the importance of the NLS for ODD regulation, we constructed pCH/557–574 ⌬ NLS and pCH/3-0 ⌬ NLS, which had the NLS sequence deleted from their vector constructs (Fig. 1A), and examined the stability of fusion proteins with ␤-Gal assay after culturing their transfectants in the presence or absence of Cbz-LLL. The ODD regulation was significantly decreased in the NLS-less fusion proteins (Fig. 2B), confirming the importance of the nuclear localization of proteins for the ubiquitin/ proteasome-mediated degradation through the ODD domain of HIF- 1␣. For additional study, we used the fusion proteins containing NLS ␣ and HIF-1 residues 548–603 (ODD/3-0) or ODD557–574. Construction of TAT-ODD Fusion Protein and Its Oxygen- dependent Instability in Vitro. To apply the ODD-fusion proteins to ␤ in vivo study, we fused TAT47–57 to the NH2 terminal of ODD- -Gal and created a TAT-ODD-␤-Gal triple fusion protein; proteins without TAT transmembrane sequences are not delivered to any tissues (24). We first examined its ODD regulation in vitro. When the control TAT-␤-Gal fusion protein was added to the culture medium of A549 Fig. 2. In vitro study of ODD regulation of ␤-Gal fusion proteins. HEK293 cells were transfected with the plasmids indicated in the figure and treated with or without 5 ␮M cells, 100% of cells were stained by X-Gal, irrespective of the oxygen Cbz-LLL for 24 h. pCH is the control plasmid encoding ␤-Gal and no ODD sequences. condition (Fig. 2C, a and c), indicating that TAT-␤-Gal fusion protein The cells were stained with X-gal (A) or applied for ␤-Gal assay (B). In pCH/3-0⌬NLS ⌬ was efficiently delivered and stably localized in the cells. The addition and pCH/557–574 NLS plasmids, NLS sequences (see Fig. 1a) were deleted from pCH/3-0 and pCH/557-574, respectively. Relative ␤-Gal activity ϭ (normalized ␤-Gal of 10 times more ␤-Gal protein alone to the culture medium resulted activity in the lysates from the cells treated with Cbz-LLL)/(normalized ␤-Gal activity in in no staining of the cells (data not shown). When the optimal amount the lysates from the untreated cells). In C, TAT-fusion protein (1 unit of ␤-Gal activity/ml) of TAT-ODD/3-0-␤-Gal was added to the culture medium of A549 was added to the culture medium of A549 cells. The cells were cultured for 18 h under normoxic (20% O2) or hypoxic (1% O2) conditions and then stained with X-gal for 2 days. cells, the cells were stained significantly milder with X-Gal under normoxic condition than the cells under hypoxic condition (Fig. 2C, b and d). Similar results were obtained by the experiment using TAT- stabilized in the hypoxic regions of the tumors. Mice bearing tumors ␤ ODD557–574- -Gal (data not shown). However, the degradation of the resulting from xenografts of a human pancreatic cancer cell line ␤ fusion protein under normoxic condition (20% O2) or in the absence CF/PAC-1 were i.p. injected with a control protein, TAT- -Gal, and of Cbz-LLL was less efficient in the experiment using the pCH/557– the tumors were subsequently analyzed for the delivery and stability 574 plasmid (Fig. 2, A and B). Therefore, we used TAT-ODD/3-0-␤- of the control protein with X-Gal staining and anti-␤-Gal Ab. Entire Gal for in vivo study. tumors were stained with both X-Gal and anti-␤-Gal Ab, indicating Selective Stabilization of TAT-ODD-␤-Gal in the Hypoxic Re- that bioactive ␤-Gal was delivered efficiently (Fig. 3A, a and c). gions of the Solid Tumor. Next, we applied TAT-ODD/3-0-␤-Gal to However, when TAT-ODD/3-0-␤-Gal was used for the same exper- in vivo study to examine if it was delivered to tumors and specifically iment, only parts of tumors were stained with X-Gal and anti-␤-Gal 2015

ptosis when NIH/3T3 and A549 cells were treated with TAT-ODD/ 3-0-Casp3WT under hypoxic condition (Fig. 5A and data not shown). Moreover, significant increase of Casp3 activity was observed in the cells treated with TAT-ODD/3-0-Casp3WT only in the presence of Cbz-LLL (Fig. 5B). We then injected TAT-ODD/3-0-Casp3WT and TAT-ODD/3-0-Casp3MUT into the tumor-bearing mice. TAT-ODD/ 3-0-Casp3WT significantly and dose dependently reduced the tumor masses for a couple of days, whereas TAT-ODD/3-0-Casp3MUT did not influence tumor growth (Fig. 5C). These results suggest that Casp3 was stabilized and activated in the cells under hypoxic conditions, resulting in the hypoxic tumor cell death. The application of fusion protein Յ20 mg/kg did not cause any obvious side effects, such as weight loss, nor was the appearance or behavior of the tumor- bearing mice altered during the observation period.

DISCUSSION We have presented a novel protein therapy targeting solid tumors. To design a protein drug, we used the TAT protein transduction domain, which quickly and efficiently delivered a bioactive protein to all tissues and took advantage of the selective stabilization of the ODD-fusion protein under hypoxic conditions. The TAT-␤-Gal- fusion protein was delivered successfully in vivo into all tissues of the mouse (Ref. 24; Figs. 3A, a and c; and 4, a, c, and e), whereas TAT-ODD-␤-Gal was detected in only a part of the solid tumors (Fig. 3A, b and d). To examine if the regions where TAT-ODD-␤-Gal was detected were hypoxic, we compared them with the regions detected with Ab against a hypoxia marker (anti-Pimo Ab). The regions were similar, although not identical, and anti-␤-Gal Ab was detected in an apparently larger area than the regions detected with anti-Pimo Ab (Fig. 3B). We speculated that the regions surrounding the hypoxic Fig. 3. Selective stabilization of TAT-ODD-␤-Gal in solid tumors. In A, the enzymatic activity and quantification of ␤-Gal protein in the tumor sections were assessed by X-gal regions in the tumor would have a relatively low oxygen tension and staining (a and b) and anti-␤-Gal Ab (c and d) after i.p. injection of TAT-␤-Gal (a and c) would therefore have slower degradation of TAT-ODD/3-0-␤-Gal. or TAT-ODD/3-0-␤-Gal (b and d) fusion proteins. In B, Pimo was i.p. injected 2 h after i.p. injection of fusion proteins. The samples were analyzed with anti-Pimo Ab (right)or anti-␤-Gal Ab (left). The Ab-stained samples (Aa and Ac; B) were also stained with hematoxylin. Bar, 100 ␮m.

Ab (Fig. 3A, b and d). To evaluate if these areas in the tumor corresponded to the hypoxic regions, the tumor-bearing mice were injected i.p. with Pimo, a hypoxia marker (37), after the injection of the fusion proteins, and their tumors were stained with anti-Pimo and anti-␤-Gal Ab. The regions detected with monoclonal anti-Pimo and rabbit anti-␤-Gal Ab were indeed similar (Fig. 3B). These data suggest that although TAT-ODD/3-0-␤-Gal was delivered into the tumors, it was degraded in normoxic regions and lingered in the hypoxic tumor cells. To examine the stability of TAT-ODD/3-0-␤-Gal in normal tissue, we examined the liver of the tumor-bearing mice with X-gal staining and anti-␤-Gal Ab. The entire liver from the mice injected with TAT-␤-Gal stained with both X-Gal and anti-␤-Gal Ab, confirming that bioactive ␤-Gal was delivered efficiently to the tissues (Fig. 4A, a, c, and e). On the other hand, ␤-Gal activity, as well as ␤-Gal protein, was hardly detected in the liver from the mice injected with TAT-ODD/3-0-␤-Gal (Fig. 4A, b, d, and f). These results suggest that TAT-ODD/3-0-␤-Gal was degraded efficiently in the normal tissue. Antitumor Effect of TAT-ODD-Casp3 Fusion Protein. Because TAT-ODD/3-0-␤-Gal was stabilized specifically in the hypoxic solid ␤ tumors, a cytotoxic TAT-ODD/3-0 fusion protein should have an Fig. 4. Rapid degradation of TAT-ODD- -Gal fusion protein in normal tissue. Anal- ysis of ␤-Gal protein and enzymatic activity was assessed by anti-␤-Gal Ab (a–d) and WT MUT antitumor effect. Casp3 and Casp3 (C163M), encoding a cat- X-gal staining (e and f), respectively, in the liver. The livers were removed 4 h after i.p. alytically inactive Casp3 (23), were fused to the COOH terminal of injection of TAT-␤-Gal (a, c, and e) or TAT-ODD/3-0-␤-Gal (b, d, and f) fusion proteins to tumor-bearing mice and then stained with X-gal for 3 days. c and d, the higher TAT-ODD/3-0, and their Casp3 activities and ODD regulation were magnification of a and b, respectively. The sections for X-gal staining (e and f) were not first examined in vitro. Significant numbers of cells underwent apo- stained with hematoxylin. Bar, 100 ␮m. 2016

Because the hypoxic regions with the lowest oxygen tension would be just a few percentage of the total tumor mass, removal of the hypoxic regions alone would not influence the tumor mass significantly. The administration of TAT-ODD/3-0-Casp3WT, however, not only suppressed the tumor growth but also reduced the tumor masses (Fig. 5C), suggesting that TAT-ODD/3-0-Casp3WT must have stabilized in the hypoxic regions, as well as their outskirts, in a distribution similar to TAT-ODD/3-0-␤-Gal. When applying a cytotoxic TAT-ODD fusion protein as a protein drug, we must minimize the damage to normal tissues. To achieve this, at least three things have to be considered for additional study: (a) dosage; (b) administration; and (c) characteristics of the cytotoxic protein fused to TAT-ODD. Addressing the first point, determination of the optimal dose of the fusion protein is the crucial part of this protein therapy. Because the degradation of ODD-fusion proteins depends on the cellular ubiquitin/proteasome system, their degradation rate is restricted by the capacity of the cellular ubiquitin/proteasome system. In vitro experiments using TAT-ODD-␤-Gal fusion protein indicated that the ODD regulation was evident at the optimal concentration of the ODD-fusion proteins, but at higher concentra- tions, the ODD regulation was hardly observed (data not shown). The fusion of an additional ODD domain to the COOH-terminal end of the protein might increase the efficiency of the ODD degradation. Ad- dressing the second point, although we gave only one injection of the TAT-ODD fusion protein to the tumor-bearing mice, repeated admin- istrations might be effective in reducing the tumor mass. The timing of the administration and the combination with radiotherapy and chemotherapy constitute the next phase of this study. Lastly, to minimize the side effects, the cytotoxic activity should preferably be indirect, and a certain level of accumulation of proteins should be needed to achieve cytotoxicity. We used a dormant form of Casp3, a nuclear executioner caspase, as a cytotoxic protein fused to TAT- ODD. We expected that Casp3 would be active in solid tumor cells under hypoxic conditions, because Casp3 activation has been observed under hypoxic conditions in vivo as well as in vitro (Refs. 28–31 and Fig. 5AB). It should be emphasized that although TAT- ODD-Casp3 was able to induce apoptosis in mouse cells under hypoxic condition (Fig. 5A), 20 mg/kg TAT-ODD/3-0-Casp3WT did not cause any obvious side effects in the tumor-bearing mice, con- sistent with the data that ␤-Gal activity and protein were hardly detected in the normal tissue (Fig. 4). Although we successfully demonstrated the antitumor effect of TAT-ODD-Casp3, there might be better cytotoxic protein candidates. Therefore, it is imperative to conduct an additional study. We determined the minimum ODD residues (557–574) that can regulate the ODD of a fusion protein. This region is also located

within the minimal NH2-terminal trans-activation domain (532– 585; Refs. 38 and 39), which has been identified recently as the region critical for interaction with von Hippel-Lindau to mediate the ubiqutin-proteasomal degradation of HIF-1␣ under normoxic conditions (16–18). The ODD residues required for the ODD regulation of HIF-1␣ stability (36) were different from the ones Fig. 5. The antitumor effect of TAT-ODD-Casp3. In A, NIH/3T3 cells were treated with WT or Mut TAT-ODD-Casp3 proteins or the buffer and cultured for 24 h under required for the regulation of ODD-fusion proteins (Fig. 1B). normoxic (20% O2) or hypoxic (1% O2) conditions. In B, WT or Mut TAT-ODD- Srinivas et al. (40) obtained similar results using a Gal4/hHIF-1␣ Casp3 proteins or the buffer were added to the culture of HEK293 cells and incubated for4hinthepresence of the solvent only (dark bars) or Cbz-LLL (gray bars). The (530– 652) fusion protein. At present, we are unable to explain cells were harvested, and Casp3 activity in a 50-␮g cell lysate was measured and these differences. The ODD regulation of HIF-1␣ per se might shown in the figure. C, tumor-bearing mice were i.p. injected with the buffer only (0 need an additional regulatory factor. Because the ODD regulation mg/kg) or 2 or 20 mg/kg TAT-ODD/3-0-Casp3 fusion proteins (Mut or WT). Average 3 total volume of the tumors was ϳ300 mm . The size of tumors was measured every of ODD/3-0 was tighter than those of ODD557–574 and ODD548– day after injection, and the relative tumor mass compared with the mass on the 583, there might be some regulatory element within the region injected day was indicated in the figure. Tumors were measured with calipers, and tumor volume was calculated as 0.5 ϫ L ϫ W2. The results are the mean of six 584– 603. The molecular mechanism of the ODD regulation in independent tumors Ϯ SD. ODD-fusion proteins is under investigation. 2017

ACKNOWLEDGMENTS 21. Nagahara, H., Vocero-Akbani, A. M., Snyder, E. L., Ho, A., Latham, D. G., Lissy, N. A., Becher-Hapak, M., Ezhevsky, S. A., and Dowdy, S. F. Transduction of Kip1 We thank Drs. Thoru Shibata and Gen Kondoh for discussion; Dr. Nobtake full-length TAT fusion proteins into mammalian cells: TAT-p27 induces cell migration. Nat. Med., 4: 1449–1452, 1998. Akiyama, Ataru Kobayashi, and Shigeki Masui for technical advice and 22. Gius, D. R., Ezhevsky, S. A., Becker-Hapak, M., Nagahara, H., Wei, M. C., and valuable reagents; and Naoko Murakami for technical assistance. Dowdy, S. F. Transduced p16INK4a peptides inhibit hypophosphorylation of the retinoblastoma protein and cell cycle progression prior to activation of Cdk2 com- plexes in late G1. Cancer Res., 59: 2577–2580, 1999. REFERENCES 23. Vocero-Akbani, A. M., Heyden, N. V., Lissy, N. A., Ratner, L., and Dowdy, S. F. 1. Ho¨ckel, M., and Vaupel, P. Tumor hypoxia: definitions and current clinical, biologic Killing HIV-infected cells by transduction with an HIV protease-activated caspase-3 and molecular aspects. J. Natl. Cancer Inst. (Bethesda), 93: 266–276. 2001. protein. Nat. Med., 5: 29–33, 1999. 2. Mueller-Klieser, W. Tumor biology and experimental therapeutics Crit. Rev. Oncol. 24. Schwarze, S. R., Ho, A., Vocero-Akbani, A., and Dowdy, S. F. In vivo protein Hematol., 36: 123–139, 2000. transduction: delivery of a biologically active protein into the mouse. Science (Wash. 3. Teicher, B. A. Hypoxia and drug resistance. Cancer Metastasis Rev., 13: 139–168. DC), 285: 1569–1572, 1999. 1994. 25. Woo, M., Hakem, R., Soengas, M. S., Duncan, G. S., Shahinian, A., Kagi, D., Hakem, 4. Kung, A. L., Wang, S., Klco, J. M., Kaelin, W. G., Jr., and Livingston, D. M. A., McCurrach, M., Khoo, W., Kaufman, S. A, et al. Essential contribution of caspase Suppression of tumor growth through disruption of hypoxia-inducible transcription. 3/CPP32 to apoptosis and its associated nuclear changes. Genes Dev., 12: 806–819, Nat. Med., 6: 1335–1340, 2000. 1998. 5. Brown, J. M. Exploiting the hypoxic cancer cell: mechanisms and therapeutic strat- 26. Enari, M., Sakahira, H., Yokoyama, H., Okawa, K., Iwamatsu, A., and Nagata, S. A egies. Mol. Med. Today, 6: 157–162, 2000. caspase-activated DNase that degrades DNA during apoptosis and its inhibitor ICAD. 6. Lemmon, M. J., van Zijl, P., Fox, M. E., Mauchline, M. L., Giaccia, A. J., Minton, Nature (Lond.), 391: 43–50, 1998. N. P., and Brown, J. M. Anaerobic bacteria as a gene delivery system that is 27. Liu, X., Li. P., Widlak, P., Zou, H., Luo, X., Garrard, W. T., and Wang, X. The controlled by the tumor microenvironment. Gene Ther., 4: 791–796, 1997. 40-kDa subunit of DNA fragmentation factor induces DNA fragmentation and chro- 7. Dachs, G. U., Coralli, C., Hart, S. L., and Tozer, G. M. Gene delivery to hypoxic cells matin condensation during apoptosis. Proc. Natl. Acad. Sci. USA, 95: 8461–8466, in vitro. Br. J. Cancer, 83: 662–667, 2000. 1998. 8. Koshikawa, N., Takenaga, K., Tagawa, M., and Sakiyama, S. Therapeutic efficacy of 28. de Moissac, D., Gurevich, R. M., Zheng, H., Singal, P. K., and Kirshenbaum, L. A. the suicide gene driven by the promoter of vascular endothelial growth factor gene Caspase activation and mitochondrial cytochrome C release during hypoxia-mediated against hypoxic tumor cells. Cancer Res., 60: 2936–2941, 2000. apoptosis of adult ventricular myocytes. J. Mol. Cell. Cardiol., 32: 53–63, 2000. 9. Binley, K., Iqball, S., Kingsman, A., Kingsman, S., and Naylor, S. An adenoviral 29. Derrick, M., He, J., Brady, E., and Tan, S. The in vitro fate of rabbit fetal brain cells vector regulated by hypoxia for the treatment of ischaemic desease and cancer. Gene after acute in vivo hypoxia. J. Neurosci., 21: RC138, 1–5, 2001. Ther., 6: 1721–1727, 1999. 30. Yamamoto, K., Morishita, R., Hayashi, S., Matsushita, H., Nakagami, H., Moriguchi, 10. Ryan, H. E., Lo, J., and Johnson, R. S. HIF-1␣ is required for solid tumor formation A., Matsumoto, K., Nakamura, T., Kaneda, Y., and Ogihara, T. Contribution of Bcl-2, and embryonic vascularization. EMBO J., 17: 3005–3015, 1998. but not Bcl-xL and Bax, to antiapoptotic actions of hepatocyte growth factor in 11. Semenza, G. L. Regulation of mammalian O2 homeostasis by hypoxia-inducible hypoxia-conditioned human endothelial cells. Hypertension, 37: 1341–1348, 2001. factor 1. Annu. Rev. Cell Dev. Biol., 15: 551–578, 1999. 31. Volm, M., Mattern, J., and Koomagi, R. Inverse correlation between apoptotic (Fas 12. Jiang, B. H., Semenza, G. L., Bauer, C., and Marti, H. H. Hypoxia-inducible factor ligand, caspase-3) and angiogenic factors (VEGF, microvessel density) in squamous 1 levels vary exponentially over a physiologically relevant range of O2 tension. Am. J. cell lung carcinomas. Anticancer Res., 19: 1669–1671, 1999. Physiol., 271: C1172–C1180, 1996. 32. Chen, C., and Okayama, H. High-efficiency transformation of mammalian cells by 13. Huang, E., Arany, Z., Livingston, D. M., and Bunn, H. F. Activation of hypoxia- plasmid DNA. Mol. Cell. Biol., 7: 2745–2752, 1987. inducible transcription factor depends primarily upon redox-sensitive stabilization of ␣ 33. Raleith, J. A., Chou, S-C., Tables, L., Suchindran, S., and Varia, M. A. Relationship its subunit. J. Biol. Chem., 271: 32253–32259, 1996. of hypoxia to metallothionein expression in murine tumors. Int. J. Radiat. Oncol. Biol. 14. Yu, A. Y., Frid, M. G., Shimoda, L. A., Wiener, C. M., Stenmark, K., and Semenza. Phys., 42: 727–730, 1998. G. L. Temporal, spatial, and oxygen-regulated expression of hypoxia-inducible 34. Akiyama, N., Matsuo, Y., Sai, H., Noda, M., and Kizaka-Kondoh, S. Identification of factor-1 in the lung. Am. J. Physiol. Lung Cell. Mol. Physiol., 275: L818–L826, a series of transforming growth factor ␤-responsive genes by retrovirus-mediated 1998. gene trap screening. Mol. Cell. Biol., 20: 3266–3273, 2000. 15. Huang, L. E., Gu, J. Schau, M., and Bunn, H. F. Regulation of hypoxia-inducible 35. Rock, K. L., Gramm, C., Rothstein, L., Clark, K., Stein, R., Dick, L., Hwang, D., and factor 1␣ is mediated by an O -dependent degradation domain via the ubiquitin- 2 Goldberg, A. L. Inhibitors of the proteasome block the degradation of most cell proteasome pathway. Proc. Natl. Acad. Sci. USA, 95: 7987–7992, 1998. 16. Tanimoto, K., Makino, Y., Pereira, T., and Poellinger, L. Mechanism of regulation of proteins and the generation of peptides presented on MHC class I molecules. Cell, 78: the hypoxia-inducible factor-1␣ by the von Hippel-Lindau tumor suppressor protein. 761–771, 1994. ␣ EMBO J., 19: 4298–4309, 2000. 36. Sutter, C. H., Laughner, E., and Semenza, G. L. Hypoxia-inducible factor 1 protein 17. Ohh, M., Park, C. W., Ivan, M., Hoffman, M. A., Kim, T. Y., Huang, L. E., Pavletich, expression is controlled by oxygen-regulated ubiquitination that is disrupted by N., Chau, V., and Kaelin, W. G. Ubiquitination of hypoxia-inducible factor requires deletions and missense mutations. Proc. Natl. Acad. Sci. USA, 97: 4748–4753, 2000. direct binding to the ␤-domain of the von Hippel-Lindau protein. Nat. Cell Biol., 2: 37. Varia, M. A., Calkins-Adams, D. P., Rinker, L. H., Kennedy, A. S., Novotny, D. B., 423–427, 2000. Fowler, W. C., Jr., and Raleigh, J. A. Pimonidazole: a novel hypoxia marker for 18. Maxwell, P. H., Wiesener, M. S., Chang, G. W., Clifford, S. C., Vaux, E. C., complementary study of tumor hypoxia and cell proliferation in cervical carcinoma. Cockman, M. E., Wykoff, C. C., Pugh, C. W., Maher, E. R., and Ratcliffe, P. J. The Gynecol. Oncol., 71: 270–277, 1998. tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-depen- 38. Jiang, B. H., Zheng, J. Z., Leung, S. W., Roe, R., and Semenza, G. L. Transactivation dent proteolysis. Nature (Lond.), 399: 271–275, 1999. and inhibitory domains of hypoxia-inducible factor 1␣. Modulation of transcriptional 19. Pepinsky, R. B., Androphy, E. J., Corina, K., Brown, R., and Barsoum, J. Specific activity by oxygen tension. J. Biol. Chem., 272: 19253–19260, 1997. inhibition of a human papillomavirus E2 trans-activator by intracellular delivery of its 39. Pugh, C. W., O’Rourke, J. F., Nagao, M., Gleadle, J. M., and Ratcliffe, P. J. repressor. DNA Cell Biol., 13: 1011–1019, 1994. Activation of hypoxia-inducible factor-1; definition of regulatory domains within the 20. Kim, D. T., Mitchell, D. J., Brockstedt, D. G., Fong, L., Nolan, G. P., Fathman, C. G., ␣ subunit. J. Biol. Chem., 272: 11205–11214, 1997. Engleman, E. G., and Rothbard, J. B. Introduction of soluble proteins into the MHC 40. Srinivas, V., Zhang, L., Zhu, X., and Caro, J. Characterization of an oxygen/redox- class I pathway by conjugation to an HIV tat peptide. J. Immunol., 159: 1666–1668, dependent degradation domain of Hypoxia-Inducible Factor ␣ (HIF-␣) proteins. 1997. Biochem. Biophys. Res. Commun., 260: 557–561, 1999.

2018

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2002 American Association for Cancer Research. Antitumor Effect of TAT-Oxygen-dependent Degradation-Caspase-3 Fusion Protein Specifically Stabilized and Activated in Hypoxic Tumor Cells

Hiroshi Harada, Masahiro Hiraoka and Shinae Kizaka-Kondoh

Cancer Res 2002;62:2013-2018.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/62/7/2013

Cited articles This article cites 40 articles, 17 of which you can access for free at: http://cancerres.aacrjournals.org/content/62/7/2013.full#ref-list-1

Citing articles This article has been cited by 16 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/62/7/2013.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/62/7/2013. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.