Tumor Necrosis Factor-a, a New Tumor Promoter, Engendered by Biochemical Studies of Okadaic Acid 1

Transcription

1 JB Review J. Biochem. 115, 1-5 (1994) Tumor Necrosis Factor-a, a New Tumor Promoter, Engendered by Biochemical Studies of Okadaic Acid 1 irota Fujiki* and Masami Suganuma** 'Saitama Cancer Center Research Institute, Ina, Kitaadachi-gun, Saitama 362, and "National Cancer Center Research Institute, TsiOujh Chuo-ku, Tokyo 104 Received for publication, August 31, 1993 Okadaic acid is a potent tumor promoter on mouse skin and in rat glandular stomach, and an inhibitor of PP-1 and PP-2A. ow okadaic acid biochemically induces tumor promotion in these tissues was reviewed. Okadaic acid bound to a catalytic subunit of PP-1 and PP-2A and induced hyperphosphorylation of proteins, such as vimentin, cytokeratins, SP 27, and tumor suppressor gene products. Since one of the okadaic acid class compounds, microcystin-lr, induced tumor promotion in rat liver, the okadaic acid pathway mediated through inhibition of PP-1 and PP-2A is seen to be a general biochemical process of tumor promotion in various organs. The biochemical mimicry of okadaic acid by TNF-a led us to find that TNF-a: is an endogenous tumor promoter. The study of tumor promotion in two-stage carcinogenesis experiments with the okadaic acid class of compounds engendered a new tumor promoter applicable to human cancer development. Key words: cell transformation, okadaic acid, protein phosphatase 1 and 2A, TNF-a, tumor promotion. ow cancer cells grow from initiated cells is now called tumor promotion and tumor progression. Since the discovery of TPA or PMA as a tumor promoter by E. ecker and B.L. Van Duuren in the late 1970s {1,2), tumor promotion has long been studied with various chemical tumor promoters (3, 4). TPA, teleocidin isolated from Streptomyces, and aplysiatoxin isolated from blue-green algae, which are structurally different from each other, induced tumor promotion on mouse skin initiated with 7,12-dimethylbenz [a]anthracene mediated through activation of PKC (Fig. 1) (5, 6). owever, these tumor promoters have classical limitations in tumor promotion, such as organ and tissue specificities (7). We thought that it is important for the study of tumor promotion in human cancer development to find a general biochemical process of tumor promotion beyond these limitations. Thus, we further screened for non-tpa type tumor promoters which do not activate PKC and act differently from TPA-types. Okadaic acid, isolated from the black sponge alichondria okadai, and dinophysistoxin-1, which 1 This work was supported in part by Grants-in-Aid for Cancer Research from the Ministry of Education, Science and Culture, a grant from the program for the Comprehensive 10-Year Strategy for Cancer Control, the Ministry of ealth and Welfare of Japan, and by grants from the Foundation for the Promotion of Cancer Research, the Uehara Memorial Life Science Foundation, the Princess Takamatsu Cancer Research Fund, the Smoking Research Foundation and M.O. A ealth Science Center. The results presented in the text were obtained at Cancer Prevention Division, National Cancer Center Research Institute, where the author (.F.) had worked. Abbreviations: CK, cytokeratin peptide; SP 27, heat shock protein 27; PKC, protein kinase C; PP-1 and PP-2A, protein phosphatase 1 and protein phosphatase 2A; FTP, protein tyrosine phosphatase; TNF-<*, tumor necrosis factor-^ ; TPA or PMA, 12-0-tetradecanoylphorbol-13-acetate or phorbol-myristate-acetate. is 35-methylokadaic acid isolated from hepatopancreas of mussels, induced tumor promotion on mouse skin as potently as TPA, teleocidin and aplysiatoxin (8, 9). Interestingly, okadaic acid and dinophysistoxin-1 are potent inhibitors of PP-1 and PP-2A (10). In addition, okadaic acid induced tumor promotion in rat glandular stomach initiated with TST-methyl-iV'-nitro-Ar-nitrosoguanidine (11). Microcystin-LR, which was isolated from cyanobacteria and was the strongest inhibitors of PP-1 and PP-2A (12), induced tumor promotion in rat liver initiated with diethylnitrosamine (13). Thus, inhibitors of PP-1 and PP-2 A induced tumor promotion in three different organs initiated with three different initiators (14). These results led us to conclude that the okadaic acid pathway mediated through inhibition of PP-1 and PP-2 A is a unique mechanism of tumor promotion and a general biochemical process of tumor promotion applicable to various organs (Fig. 1). If the okadaic acid pathway can be extended to cover tumor promotion in human cancer development, it becomes important how the effects brought about by the okadaic acid pathway are biochemically reflected in human tissues. This idea engendered an interesting result that tumor necrosis factor-a acts as a tumor promoter in BALB/3T3 cell transformation (15). This paper reviews a biochemical process of tumor promotion by the okadaic acid class compounds in relation to inhibition of PP-1 and PP-2 A in vitro, in cells, and in vivo, and briefly describes a possible link between okadaic acid and TNF- a within a conceptual framework of tumor promotion in humans. Okadaic Acid Receptors, PP-1 and PP-2 A Before the biochemical nature of the okadaic acid receptors had been identified we found that a radioactive okadaic Vol. 115, No. 1, 1994

2 . FujiM and M. Suganuma TPA-type tumor promoters I Protein kinase C,N ' OOC Dephosphorylation Signal i 1 if IOO,N COO Phosphorylation,N COO Signal -X- Dephosphorylation t Okadaic acid class tumor promoters Fig. 1. Schematic illustration of mechanism of action of the TPA-type tumor promoters and the okadaic acid class tumor promoters. acid, okadaic acid (14 Ci/mmol), specifically bound to the particulate and cytosolic fractions of various mouse tissues: skin, brain, lung, and colon; the Ka values were 27.1 nm for the particulate and 1.0 nm for cytosolic fractions of mouse skin (16). The specific 3 -okadaic acid binding to the particulate fraction, for example, was inhibited dose-dependently by okadaic acid, but not by okadaic acid tetramethyl ether, an inactive compound (Fig. 2). In 1987, Takai and his colleagues first reported that okadaic acid is a potent inhibitor of PP-2A, times stronger than PP-1 (20). Based on the evidence, we studied the interaction of okadaic acid with PP-2A, which was isolated from bovine brain and consists of two regulatory subunits of 65 and 42 kda and a catalytic subunit of 37 kda. Covalent binding of a photoaflinity probe, methyl 7- O-(4-azidobenzoyl)okadaate, to the catalytic subunit was found after irradiation with UV light of the reaction mixture, but not to the other regulatory subunits (17). The results first indicated that the okadaic acid receptors are catalytic subunits of PP-2A, and probably of PP-1 as well. Since okadaic acid is reported to act as a non-competitive inhibitor of PP-1 and PP-2A (10), okadaic acid is assumed not to bind to the active site, yet probably binds not far from that site. By using the receptor binding test with 3 -okadaic acid, three other types of compounds, calyculin, microcystin, and tautomycin were found to be potent inhibitors of PP-1 and PP-2A (Fig. 3) (18). Their ICBM for a catalytic subunit of PP-1 purified from rabbit skeletal muscle ranged from 0.1 to 3.4 nm, and the order of their potencies was microcystin-lr> calyculin A>tautomycin>okadaic acid (18). The ICM, for PP-2A purified from human erythrocytes ranged from 0.07 to 0.65 nm, and their potency order was okadaic acid>microcystin-lr>calyculin A>tautomycin (Table I) (18). Thus, the okadaic acidbinding site of the catalytic subunit is also available for binding by these inhibitors. Calyculin A, isolated from a marine sponge, induced tumor promotion on mouse skin as strongly as okadaic acid (29), but its tumor promoting activity in rat glandular stomach has not yet been tested, due to a limited availability of the compound. Tumor promotion of tautomycin, isolated from Streptomyces, is under investigation. Four types of the okadaic acid class compounds at doses 10* 10' 10* 10* 10* 10* Concentration ( nm ) Fig 2. Inhibition of the specific '-okadalc acid binding to the particulate fraction by okadaic acid ( ), bat not by okadaic acid tetramethyl ether (O). up to 10 //M did not inhibit protein tyrosine phosphatase prepared from recombinant rat brain protein tyrosine phosphatase 1 (Table I) (18). Therefore, the okadaic acid class compounds represent a structurally diverse group associated with strong inhibitory activity toward PP-1 and PP-2A. yperphosphorylation of Proteins Inhibition of PP-1 and PP-2A by okadaic acid results in inhibition of dephosphorylation of phosphoserine and phosphothreonine of proteins which were previously phosphorylated by various serine/threonine protein kinases in the cells. Incubation of the cytosolic fraction containing various serine/threonine protein kinases and PP-2A with y- 32 P-ATP and okadaic acid stimulated incorporation of 32 P into histone EII-S more strongly than it did without okadaic acid (20). This increased phosphorylation by okadaic acid was called "apparent activation" of protein kinases because protein kinases were not directly activated (20). To study the target proteins of "apparent activation" of protein kinases, cells were treated with okadaic acid. Treatment of primary human fibroblasts with 100 nm okadaic acid for 2 h induced a strongly phosphorylated protein with a molecular weight of 58 kda, that is, hyperphosphorylation of vimentin (21), which is reported to be phosphorylated by p34 cdc2 (22). In a similar way, hyper - phosphorylated cytokeratins such as CK 5, CK 6, CK 7, CK 16, and CK 19 were found in human keratinocytes treated with okadaic acid (23), while hyperphosphorylated CK 8 and CK 18 were in rat primary cultured hepatocytes treated with microcystin-lr (24). The hyperphosphorylation of intermediate filaments, such as vimentin and cytokeratins, was closely associated with morphological changes in the cells. This is one of the pleiotropic effects induced by a tumor promoter. In addition to intermediate filaments, we found that treatment of primary human fibroblasts with okadaic acid induced hyperphosphorylation of the Rb protein and p53 (25), which are reported to regulate cell growth in a negative manner. This we think is an alternative explanation for the loss of function of both tumor suppressor gene products, which might take place in tumor promotion of okadaic acid in tissues (26). The results indicated that J. Biochem.

3 and Okadaic Acid OR, 'OR Okadaic acid Dinophysistoxin-1 Ri R 2 C 3 3 CO? 0 Calyculin A O O C3 OC, jc Cj O 0 Y? O 0C3 C 3 3 C COO Microcystin-LR Tautomycin Fig. 3. Four structurally different types of the okadaic acid class compounds. TABLE I Effects of the okadaic acid class compounds on PP-1, PP-2A, and PTP. Compounds PP-1 PP-2A PTP Okadaic acid Calyculin A Microcystin-LR Tautomycin Na vanadate , >l, C 3 M II I COO N okadaic acid inhibited PP-1 and PP-2A in the presence of serine/threonine protein kinases in nuclei of cells. Although we were not yet able to directly show penetration of okadaic acid into nuclei, okadaic acid induced hyperphosphorylation of the Rb protein and p53 by treatment of nuclei isolated from rat regenerating liver, but not by treatment of nuclei isolated from normal rat liver (25). This tells us that okadaic acid reacted with PP-1 and PP-2A Fig. 4. Superimposition of the three-dimensional structures of okadaic acid, calyculin A, and microcystin-lr. Vol 115, No. 1, 1994

4 present in nuclei of the cells subsequent to penetration. As described previously, okadaic acid and calyculin A have a similarly strong inhibitory activity toward PP-2A, but okadaic acid has not toward PP-1, while calyculin A has (Table I). Since they have similarly potent tumor promoting activity on mouse skin, tumor promotion seems to be well correlated with inhibition of PP-2A (19). Guy and his colleagues reported that okadaic acid treatment stimulated phosphorylation of over 140 proteins in human fibroblasts on two-dimensional gel electrophoresis (27). Therefore, inhibition of PP-1 and PP-2A stimulates phosphorylation of various proteins, and the phosphorylation of intermediate filaments was particularly apparent due to large amounts of the proteins or to the turn-over rate of phosphorylation/dephosphorylation of the proteins. Gene Expression and Transcriptional Regulation Okadaic acid, like TPA, induces expression of various genes, including a family of primary response genes, the TPA induced sequence (TIS) genes, in 3T3 cells; however, the time-course of the induction of TIS mrna was different from that by TPA (28). Namely, the maximum mrna accumulation was found at 3 h for okadaic acid and at 60 min for TPA. Thus, the delayed induction by okadaic acid seems to be associated with its biochemical mechanism of action. Okadaic acid stimulated expression of the human collagenase gene partly through the AP-1 complex binding site, and it induced transcription of the c-fos gene through a region containing multiple regulatory elements such as the serum-responsible element or a camp-like responsive element (29). Okadaic acid stimulated expression of the human immunodeficiency virus LTR-CAT gene through the NF-kB binding site (30). All these results support the suggestion that the phosphorylation of transcription factors induced by okadaic acid is involved in the transcriptional regulation of a variety of genes. Recently a TPA inactive, mutated TPA-responsive element ( C TGACTCC A ) was identified and named the okadaic acid response element (ORE), which is transactivated by both c-jun A and jun B (31). Transcription factors which bind to ORE are now under investigation. Induction of early response genes such as c-jun, jun B, c-fos, and fos B by okadaic acid showed sustained mrna accumulation in human keratinocytes. The similar induction was also observed in primary cultured rat hepatocytes by microcystin-lr and nodularin (Sueoka et al., manuscript in preparation). The results led us to postulate that the cells in various organs commonly undergo the same biochemical and molecular processes by treatment with the okadaic acid class of tumor promoters. ow stable these mrna accumulations induced by okadaic acid are compared with those induced by TPA should be investigated in relation to AU-rich sequence of mrna and phosphorylation of its binding proteins. Structural Features of the Compounds Over 30 compounds of the okadaic acid class were roughly classified into four different types, okadaic acid, calyculin, microcystin, and tautomycin (14). Since their representative compounds have nearly the same strong specific. Fujiki and M. Suganuma activity in inhibition of PP-1 and PP-2A, they are assumed to have a common functional structure in their molecules responsible for this activity. Computer-assisted molecular modeling of the three compounds, okadaic acid, calyculin A, and microcystin-lr, by Quinn and his associates allowed two main common parts to be recognized, that is, the functional part for inhibition of PP-1 and PP-2A and the binding part to catalytic subunit of PP-1 and PP-2A (Fig. 4) (32). For example, the receptor binding sites appear to be a long side chain of the three compounds (32). It was of interest that these three compounds have a similarly circularized shape, formed by an intramolecular hydrogen bond between C-l carbonyl and C-24 hydroxyl groups of okadaic acid (33), and hydrogen bonds between phosphoric acid and various parts of calyculin A, and cyclic peptides of the microcystins and nodularin (34, 35). The presence of the common structure confirmed our important evidence that okadaic acid, calyculin A, and microcystin-lr belong functionally to the same okadaic acid class. TNF-a as a New Tumor Promoter A tumor promoter is always applied to the carcinogeninitiated tissues with a certain continuity. ow do human tissues produce similar patterns in response to the sustained effects of a tumor promoter? To that question we thought the answer is as follows: One of the induced proteins, X, acts as an endogenous tumor promoter to the cell through autocrine and paracrine, and X induces similar effects to those of okadaic acid in the cells. This was supported by recent investigations that okadaic acid mimics TNF-a/EL-1 in inducing a protein phosphorylation pattern and expression of early response genes in human fibroblasts and NF-kB in Jurkat cells (27, 36). TNF-a, with a molecular weight of 17 kda is produced mainly by activated macrophages and is a cytokine inducing hemorrhagic necrosis of transplanted solid tumors in mice, from the study of L.J. Old at the Memorial Sloan-Kettering Cancer Center in New York in 1975 (37). To test the okadaic acid mimicry of TNF, we utilized hyperphosphorylation of vimentin and SP 27 in primary human fibroblasts (15). uman TNF-or at a concentration of 25 ng/ml (1.47 nm) increased the maximum level of vimentin phosphorylation in the cells 15 min after treatment to twice that of its basal level, whereas okadaic acid increased it 10 times over that of its basal level 2 h after treatment (15, 21). Thus, the results indicated that human TNF-a induces phosphorylation of the same proteins as does okadaic acid, although TNF-a directly activates various protein kinases and okadaic acid inhibits PP-1 and PP-2A. Interestingly, we found that TNF- a release into the medium was induced by treatment of mouse BALB/3T3 cells with okadaic acid (15), as was mouse TNF-a mrna (Suganuma et al., manuscript in preparation). uman TNF-a significantly stimulated transformation of BALB/ 3T3 cells initiated with 3-methylcholanthrene (25). Based on molar concentration, human TNF-a was about 1,000 times stronger than TPA in cell transformation assays. In addition, it is of great importance to note that human TNF-a at concentrations from pm to 0.6 nm specifically induced growth of Bhas 42 cells, which were v-a- ras transfected BALB/3T3 cells, whereas human TNF-a was not significantly effective in BALB/3T3 cells. These results J. Biochem.

5 TNF-a and Okadaic Acid clearly indicated that human TNF-a is responsible for clonal expansion of the initiated cells (15). Thus, we think that chemical tumor promoters such as TPA, teleocidin, and okadaic acid are TNF- a inducers, and that TNF- a acts as an endogenous tumor promoter and a central mediator of tumor promotion. If TNF-a- is now accepted as an endogenous tumor promoter, many historical puzzles, such as its relationships to inflammation, wound healing and clonal expansion of the initiated cells in various organs can be resolved. In this regard, the new tumor promotion pathway mediated through TNF-a has put us one step closer towards defining the tumor promotion process in human cancer development. REFERENCES 1. ecker, E. (1971) Methods Cancer Res. 6, Van Duuren, B.L (1969) Prog Exp. Tumor Res 11, Boutwell, R.K. (1977) in Origins of uman Cancer (iatt,.., Watson, J.D., & Winston, D., eds.) pp , Cold Spring arbor Laboratory, Cold Spring arbor, NY 4. Fujiki,. & Sugimura, T (1987) in Advances in Cancer Research (Klein, G. & Weinhouse, S, eds) Vol. 49, pp , Academic Press, San Diego 5 Nishizuka, Y. (1984) Nature 308, Fujiki,, Tanaka, Y., Miyake, R., Kikkawa, U, Nishizuka, Y., & Sugiraura, T. (1984) Biochem. Biophys. Res. Commun. 120, Berenblum, I (1978) in Carcinogenesis A Comprehensive Survey: Mechanisms of Tumor Promotion and Carcinogenesis (Slaga, T.J, Sivak, A., & Boutwell, R.K., eds.) Vol. 2, pp 1-10, Raven Press, New York 8. Suganuma, M., Fujiki,., Suguri,., Yoshizawa, S., irota, M., Nakayasu, M., Ojika, M., Wakamatsu, K., Yamada, K., & Sugimura, T. (1988) Proc Nad Acad. Sd. USA 86, Fujiki,., Suganuma, M., Suguri,, Yoshizawa, S., Takagi, K., Uda, N., Wakamatsu, K, Yamada, K., Murata, M, Yasumoto, T., & Sugimura, T. (1988) Jpn J Cancer Res. (Gann) 79, Takai, A., Biolajan, C, Troschka, M, &Rflegg, J.C. (1987) FEBS Lett. 217, Suganuma, M, Tatematsu, M, Yatsunami, J., Yoshizawa, S., Okabe, S., Uemura, D., & Fujiki,. (1992) Carcinogenesis 13, Yoshizawa, S., Matsushima, R., Watanabe, M.F., arada, K.-L, Ichihara, A., Carmichael, W W, & Fujiki,. (1990) J. Cancer Res Chn. OncoL 116, Nishiwaki-Matsushima, R., Ohta, T., Nishiwaki, S, Suganuma, M., Kohyama, K., Ishikawa, T., Carmichael, W.W., & Fujiki,. (1992) J Cancer Res. Clin. Oncol. 118, Fujiki,. & Suganuma, M. (1993) in Advances in Cancer Research (Vande Woude, G F. & Klein, G, eds.) Vol. 61, pp , Academic Press, San Diego 15. Komon, A, Yatsunami, J, Suganuma, M., Okabe, S., Abe, S., Sakai, A., Sasaki, K, & Fujiki,. (1993) Cancer Res. 53, Suganuma, M., Suttajit, M., Sugun,., Ojika, M., Yamada, K., & Fujiki, (1989) FEBS Lett. 260, Nishiwaki, S, Fujiki,., Suganuma, M., Ojika, M, Yamada, K, & Sugimura, T. (1990) Biochem. Biophys. Res. Commun. 170, Suganuma, M., Fujiki,, Okabe, S., Nishiwaki, S, Brautigan, D., Ingebritsen, T.S., & Rosner, M.R. (1992) Toxicon 30, Suganuma, M., Fujiki,., Furuya-Sugun,., Yoahizawa, S., Yasumoto, S., Kato, Y, Fusetam, N., & Sugimura, T. (1990) Cancer Res. 50, Sassa, T, Richter, W.W, Uda, N., Suganuma, M., Suguri,., Yoshizawa, S., irota, M., & Fujiki, (1989) Biochem. Biophys. Res. Commun. 159, Yatsunami, J., Fujiki,, Suganuma, M., Yoshizawa, S., Eriksson, J.E., Olson, M.O.J., & Goldman, R.D. (1991) Biochem. Biophys. Res Commun. 177, Chou, Y.-., Bischoff, J R., Beach, D., & Goldman, R.D (1990) Cell 62, Yatsunami, J., Komon, A., Ohta, T., Suganuma, M., Yuspa, 5., & Fujiki,. (1993) Cancer Res 53, Ohta, T., Nishiwaki, R., Yataunami, J., Komon, A., Suganuma, M., & Fujiki,. (1992) Carcinogenesis 13, Yatsunami, J., Komori, A., Ohta, T., Suganuma, M., ftfujiki, (1993) Cancer Res 53, Fujiki,. (1992) MoL Carcinogen. 5, Guy, G.R., Cao, X., Chua, S.P., & Tan, Y.. (1992) J Bwl. Chem. 267, erschman,.r., Lim, R.W., Brankow, D.W., & Fujiki, (1989) Carcinogenesis 10, Kim, S.-J., Lafyatis, R., Kim, K.Y., Angel, P, Fujiki,., Karin, M., Sporn, M B., & Roberts, A.B. (1990) CelL ReguL 1, Thevenin, C, Kim, S -J., Rieckmann, P., Fujiki,., Norcross, M.A., Sporn, M B, Fauci, A.S., & Kehrl, J.. (1990) New BwL 2, Levy, J.P., Fujiki,., Angel, P, & Bowden, G.T. (1991) Proc. Am. Assoc. Cancer Res. 32, Quinn, R.J, Taylor, C, Suganuma, M., & Fujiki,. (1993) BioMed. Chem. Lett. 3, Uemura, D. & irata, Y (1989) in Studies m Natural Products Chemistry (Rahman, A.U., ed.) pp , Elsevier, Amsterdam 34. Taylor, C, Quinn, R.J., McCulloch, R, Nishiwaki-Matsuahima, R., & Fujiki,. (1992) BioMed. Chem. Lett. 2, Nishiwaki-Matsushima, R, Fujiki,., arada, K.-L, Taylor, C, 6 Quinn, R.J. (1992) BioMed. Chem. Lett. 2, Thevenin, C, Kim, S.-J., & Kehl, J.. (1991) J Biol. Chem. 266, Old, L.J. (1985) Science 230, Vol. 115, No. 1, 1994