- Email: [email protected]
Thioredoxin redox control of cell growth and death and the effects of inhibitors
Chemico-Biological Interactions 111 – 112 (1998) 23 – 34
Thioredoxin redox control of cell growth and death and the effects of inhibitors Garth Powis a,*, D. Lynn Kirkpatrick b, Miguel Angulo a, Amanda Baker a a
Arizona Cancer Center, Uni6ersity of Arizona, 1515 North Campbell A6enue, Tucson, AZ 85724 -5024, USA b Department of Chemistry, Uni6ersity of Regina, Regina, Sask. S4S OA2, Canada
Abstract Thioredoxin is a redox protein found over-expressed in some human tumors. Thioredoxin is secreted by tumor cells and stimulates cancer cell growth. Redox activity is essential for growth stimulation by thioredoxin. Cells transfected with thioredoxin cDNA show increased tumor growth and decreased apoptosis in vivo and decreased sensitivity to apoptosis induced by a variety of agents both in vitro and in vivo. Cells transfected with a redox-inactive mutant thioredoxin show inhibited tumor growth in vivo. Thus, thioredoxin offers an attractive target for anticancer drug development. A class of disulfide inhibitors of thioredoxin has been identified. These disulfides inhibit cancer cell growth in culture and have antitumor activity against some human tumor xenografts in animals. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Thioredoxin reductase; Thioredoxin; Apoptosis; Disulfides
1. The thioredoxin system The thioredoxin redox system comprised of two proteins, thioredoxin reductase and thioredoxin, is widely distributed in prokaryotes and eukaryotes. The * Corresponding author. [email protected]
Tel.:
+1
520
6266408;
fax:
+1
0009-2797/98/$19.00 © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S0009-2797(97)00148-8
520
6264848;
e-mail:
24
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
flavoprotein thioredoxin reductase catalyzes the transfer of reducing equivalents from NADPH to the second protein thioredoxin which then reduces a number of protein thiol acceptors. The structures of bacterial and human thioredoxin and thioredoxin reductase are shown in Fig. 1. Both proteins are found widely expressed in human tissues with highest levels of thioredoxin reductase in testis and the uterus and highest levels of thioredoxin in the intestine and kidney [1].
2. Thioredoxin reductase We purified thioredoxin reductase, a selenium containing flavoprotein that catalyzes the NADPH-dependent reduction of thioredoxin [2] from human placenta [3]. Using unique terminal and internal peptide sequences we have also cloned and sequenced the full length cDNA [4]. The predicted amino acid sequence of human thioredoxin reductase shows 44% identity to human glutathione reductase but only 24% identity to bacterial thioredoxin reductase, from which it differs by an additional 150 amino acid dimer interface domain. The human thioredoxin reductase gene (TRXRD1) is found at chromosome 12q 23q24.1 [5]. The C-terminal sequence that we predicted for human placental thioredoxin reductase was based on the presence of a UGA stop codon after the bases coding for the C-terminal Gly494-Cys495 [4]. Subsequently, a selenoprotein was isolated from a human lung adenocarcinoma cell line having a C-terminal amino acid sequence -Gly494-Cys495SeCys496-Gly497 [6] was found to have thioredoxin reductase activity [7]. Thus, the SeCys in human thioredoxin reductase appears to be encoded by the mRNA stop
Fig. 1. Comparison of E. coli and human thioredoxin and thioredoxin reductase.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
25
codon UGA [8]. For UGA to escape recognition by the protein translation machinery and to code for selenocysteine incorporation in eukaryotes requires SeCys incorporation sequence (SECIS) stem-loop element in the mRNA 3%-untranslated region. From work with type 1 deiodinase it appears that the UGA of selenoprotein-encoding mRNAs can alternately function to encode selenocysteine or as a stop codon, depending on conditions. These conditions include limiting selenium [9,10] or excess selenoprotein DNA [10]. A purine immediately following the UGA codon increases the frequency of termination relative to selenocysteine incorporation. Human thioredoxin reductase has a G immediately following UGA which is also found in glutathione peroxidase [11]. This ‘terminator’ purine has been suggested to create a site where termination is favored over selenocysteine incorporation when selenocysteinyl-tRNA is limiting [12] and provides an explanation for the decrease in glutathione peroxidase levels upon nutritional selenium deprivation [13]. Alternatively, cysteine (Cys) may be incorporated in place of SeCys which usually gives an enzyme with less activity than the wild-type protein [14]. Selenium deficiency also directly regulates glutathione peroxidase mRNA levels [15]. The SeCys of thioredoxin reductase lies at the C-terminal end of the protein and is not part of the predicted catalytic site [4]. This contrasts other selenoproteins that catalyze oxidation reduction reactions where SeCys forms part of the active site [16]. It appears that SeCys is essential for the activity of the enzyme because recombinant mutant human thioredoxin reductases without the C-terminal SeCysGly or with SeCys replaced by Cys will not reduce thioredoxin but will reduce the non-specific substrate dithionitrobenzoic acid (unpublished observations). The role SeCys plays in the biological activity thioredoxin reductase remains a matter of conjecture. Our other studies have shown that Cys73 of thioredoxin outside the conserved catalytic site is critical for the interaction of thioredoxin with human thioredoxin reductase [5] so that interaction between the SeCys of thioredoxin reductase and thioredoxin may occur at the Cys73 residue. The amount of available selenium in cells controls thioredoxin reductase activity. Both a concentration-dependent increase in cellular thioredoxin reductase activity and concentration-dependent increase in thioredoxin protein measured by Western blotting [14] were observed with increasing concentrations of sodium selenite added to serum-free growth medium. However, the maximum increase in protein at any selenium concentration is considerably less than the increase in thioredoxin reductase activity. Thus, in the absence of selenium, there appears to be premature termination or incorporation of Cys into thioredoxin reductase in place of SeCys. This results in a decrease in the specific activity of the enzyme. There is also a decrease in the amount of thioredoxin reductase mRNA and protein.
3. Thioredoxin as a growth factor The Cys residues at the conserved -Cys32-Gly-Pro-Cys35-Lys active site of the low molecular weight redox protein thioredoxin [17], undergo reversible oxidation-re-
26
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
duction catalyzed thioredoxin reductase [18]. Human thioredoxin is a 11.5 kDa protein and contains three additional Cys residues not found in bacterial E. coli thioredoxin with which it has 27% amino acid identity. These additional Cys residues give human thioredoxin unique biological properties [19]. Thioredoxin was first studied for its ability to act as a reducing co-factor for ribonucleotide reductase, the first unique step in DNA synthesis [20]. More recently, thioredoxin has been shown to exert redox control over some transcription factors to modulate their binding to DNA and thus, to regulate gene transcription. Transcription factors that are regulated by thioredoxin include NF-kB [21], the glucocorticoid receptor [22] and indirectly through a nuclear redox factor Ref-1/HAP1, thioredoxin can regulate AP-1 (Fos/Jun heterodimer) [23]. Thioredoxin is also a growth factor with a unique mechanism of action. Human recombinant thioredoxin stimulates the proliferation of both normal fibroblasts and a wide variety of human solid and leukemic cancer cell lines, even in minimal medium in the absence of serum [24,25]. Thioredoxin undergoes spontaneous oxidation in the absence of reducing agents [26], initially through oxidation to monomeric forms that retain the ability to be reduced by thioredoxin reductase but that cannot stimulate cell proliferation. This is followed by a slower spontaneous conversion of thioredoxin to a homodimer form that is not a substrate for reduction by thioredoxin reductase and that also cannot stimulate cell proliferation. X-ray crystallography has shown that human thioredoxin forms a dimer through a highly conserved interface domain that is stabilized by an intermolecular Cys73-Cys73 disulfide bond [27]. We have shown that a Cys73 Ser mutant thioredoxin monomer was just as effective as the wild-type protein at stimulating cell growth and did not show spontaneous or oxidant induced loss of activity. Furthermore, this mutant did not form a dimer [26]. Thus, Cys73 of thioredoxin, although not necessary for the growth stimulating activity of thioredoxin, is involved in inactivation of the wild-type protein. Whether this inactivation plays a role in the physiological actions of thioredoxin remains to be determined. The studies with 125I-thioredoxin have revealed no high affinity binding sites that might suggest receptors for thioredoxin on the surface of cancer cells [28]. Instead the studies demonstrate that thioredoxin stimulates cell proliferation by increasing the sensitivity of the cells to growth factors secreted by the cells themselves [28]. This potentiation can be over 100-fold for example with basic fibroblast growth factor and is seen at concentrations of thioredoxin found in human serum which are in the range of 10–30 nM [29]. Cancer cells secrete thioredoxin [1,30]. The pleiotropic effects of thioredoxin on cell growth, taken together with the observations that thioredoxin is over-expressed by a number of human primary tumors, suggests that thioredoxin could be an important factor in human cancer cell growth [28].
4. Thioredoxin prevents apoptosis Transfection of cells with thioredoxin is able to prevent apoptosis. Apoptosis is a process of programmed cell death that mediates normal cell turnover, hormone-
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
27
Table 1 Effect of trx and bcl-2 transfection on apoptosis induced by different agents Relative apoptosis
Dexamethasone Staurosporine Etoposide Thapsigargin
Neo
Trx5
Trx6
W.Hb12
2.99 0.4 58.39 9.8 162.5 9 12.1 4.39 0.9
0.9 9 0.0* 5.3 9 0.0* 6.0 9 0.6* 2.3 9 1.8*
0.3 90.0* 9.9 9 1.1* 20.9 94.1* 1.8 91.2*
1.0 90.1* 4.8 90.7* 4.5 90.7* 0.9 90.6*
Vector-alone transfected WEHI7.2 cells (Neo); Trx5 and Trx6 trx transfected WEHI7.2 cells and bcl-2 transfected WEHI7.2 cells (W.Hb12) were treated with 1 mM dexamethasone for 24 h, 1 mM staurosporine for 21 h, 60 mM N-acetyl-sphingosine for 42 h, 1 mM etoposide for 15 h or 50 nM thapsigargin for 24 h. The times were determined to be optimum for detecting apoptosis with each agent. Apoptosis was measured by flow cytometry and expressed as a ratio of the sum of regions representing early and late apoptotic cells divided by region representing live non-apoptotic cells, and normalized to the ratio for vehicle treated vector-alone transfected cells. Values are the mean of determinations9S.E. Statistical analysis was by linear regression with indicator values for drugs and cells. *PB0.05 compared to vector-alone transfected control cells.
induced tissue atrophy, cell mediated killing in immunity and tumor regression. WEHI7.2 mouse thymoma-derived cells were stably transfected with human trx cDNA and had levels of thioredoxin mRNA up to 1.8-fold compared to endogenous levels of mouse thioredoxin mRNA [31]. Immunofluorescent staining determined that the thioredoxin transfected cells also have increased levels of thioredoxin protein, with approximately 60% of the staining found in the nucleus and 60% in the cytoplasm. These thioredoxin transfected WEHI7.2 cells were resistant to apoptosis induced by dexamethasone compared to wild-type or vector-alone transfected cells (Table 1). Thioredoxin has been reported to be necessary for assembly of the glucocorticoid receptor [22], however, glucocorticoid receptor activity measured using a glucocorticoid receptor/chloramphenicol acetyltransferase reporter plasmid was not altered in the thioredoxin transfected cells. The effect of thioredoxin transfection on other agents known to induce apoptosis was studied (Table 1). Thioredoxin transfected cells were resistant to apoptosis, compared to vector-alone transfected control cells, induced by staurosporine, a general kinase inhibitor [32], thapsigargin, which blocks the uptake of intracellular Ca2 + , resulting in an increase in intracellular free Ca2 + concentration [33] and by etoposide, an inhibitor of topoisomerase II [34]. WEHI7.2 cells transfected with the bcl-2 anti-apoptotic proto-oncogene (W.HB12 cells) showed a similar pattern of protection against apoptosis induced by the various agents as did the thioredoxin transfected cells.
28
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
5. Thioredoxin stimulates tumor growth and inhibits apoptosis in vivo When inoculated into severe combined immunodeficient (SCID) mice, the tumors that formed from the thioredoxin transfected WEHI7.2 cells grew more rapidly than the wild-type cell tumors (Fig. 2). Histological examination of the wild-type cell tumors showed apoptosing cells with condensed and marginated chromatin and a dense cytoplasm accompanied by vacuolization. In contrast, tumors that formed from the thioredoxin transfected cells showed few apoptosing cells. Dexamethasone treatment inhibited the growth of the wild-type cell tumors (Fig. 2), however, not that of the hioredoxin transfected cell tumors. It was also of interest for us to compare the thioredoxin transfected WEHI7.2 cells to Bcl-2 transfected WEHI7.2 cells in vivo. The Bcl-2 transfected cells formed tumors that grew faster than tumors formed by wild-type WEHI7.2 cells but still considerably slower than the thioredoxin transfected cell tumors. Surprisingly, the growth of the bcl-2 transfected cell tumors was still inhibited by dexamethasone treatment. Evidence from transgenic mice over-expressing bcl-2 under transcriptional regulation of the Ig heavy chain enhancer has shown that the mice develop benign lymphoma that eventually
Fig. 2. Tumor formation in SCID mice by (, ) wild-type WEHI7.2 cells; (, ) bcl-2 transfected WEHI7.2 (W.Hb12) cells; and ( , ) Trx6 trx transfected WEHI7.2 cells. A total of 20 mice were injected s.c. with 2× 107 cells and 0.1 ml matrigel into the flank. Tumor size was measured every 3 – 4 days with calipers and tumor volumes calculated. Half the mice were injected s.c. into the opposite flank with 1 mg/kg/day dexamethasone (, , ) or with vehicle alone (, , ) 9 days after tumor cell injection. Mice were euthanized the first measurement that tumor volume exceeded 2 cm3. Values are the mean of ten mice in each group and bars are S.E.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
29
progresses to high grade malignant disease [35]. This has been taken to indicate that bcl-2 provides a survival advantage to cells in vivo but that an additional change, most frequently rearrangement of myc [35], is necessary for tumor growth [36]. In contrast to Bcl-2, thioredoxin itself offers a clear growth advantage to tumors in vivo.
6. Thioredoxin in human cancer Expression of thioredoxin mRNA was elevated in almost half (5/12) of the human primary lung cancers examined compared to normal lung tissue from the same subject [19] and in half (6/12) of the human primary colorectal tumors compared to normal colonic mucosa [1]. Others have reported increased levels of thioredoxin protein in human squamous cervical cancer and hepatoma [37,38] and we have found increased levels in human acute lymphocytic leukemia (3/6 subjects). An important caveat to these studies is that the thioredoxin mRNA was extracted from pieces of tumor giving no indication of thioredoxin’s distribution within the tumor, whether from malignant cells, stromal or hypoxic core cells. This is significant because hypoxia is known to lead to an increase in thioredoxin expression [1]. The preliminary immunohistochemical studies have shown that in normal human colon thioredoxin is found in the dividing basal crypt cells while in primary colon cancer thioredoxin is over-expressed exclusively by the cancer cells. It remains to be determined whether thioredoxin over-expression is related proliferation or to apoptosis in human primary tumors and if it has a role as a prognostic indicator. While human tumors show many fold over-expression of thioredoxin mRNA, up to 100-fold in one patient with colon cancer, the cell transfection studies have given, at most, a 2-fold over-expression of mRNA. Similarly, we have been able to obtain only very low levels of thioredoxin over-expression in the transgenic mouse studies (not shown). These results suggests that high levels of thioredoxin over-expression may be toxic to cells. If this is so, human tumors appear to have become resistant to high levels of thioredoxin expression or may have mutant thioredoxin.
7. Thioredoxin as a target for drug development Since the molecular studies provide proof-of-principle that the thioredoxin system is a rational target for anticancer drug development, the initial approach was to develop agents which might selectively inhibit the thioredoxin system and thioredoxin-dependent cell proliferation. A series of alkyl 2-imidazolyl disulfides, typical of which is 1-methylpropyl 2-imidazolyl disulfide (IV-2) were studied (Fig. 3). Short-chain alkyl analogues of this series were found to be substrates for the enzymes, thioredoxin reductase while analogues with a branched alkyl chain or benzyl group are competitive inhibitors. Prolonged incubation of thioredoxin with the alkyl 2-imidazolyl disulfide results in irreversible inhibition of the thioredoxin as a substrate for reduction by thioredoxin reductase [39]. The inhibition appears to be
30
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
Fig. 3. Structures of 2-imidazolyl disulfides.
specific for Cys73 of thioredoxin and is not seen with a mutant Cys73 Ser thioredoxin, which remains a substrate for thioredoxin reductase even after prolonged incubation with the disulfides. In culture, the disulfides show greater inhibition of thioredoxin- than serum-dependent cell proliferation, suggesting that they may be producing their effects by inhibiting the thioredoxin system [39].
8. Inhibition of cancer cell growth Besides inhibiting the growth of a number of cancer cell lines in culture [24,40,41], studies with the National Cancer Institute (NCI) 60 cancer cell line panel [42] show a number of the disulfides exhibit selectivity for inhibition of leukemia, colon and renal cell growth. A significant negative correlation between disulfide activity and thioredoxin reductase mRNA in the cell lines was revealed using the COMPARE algorithm [42]. Furthermore, one of the disulfides, IV-2, was found to inhibit the growth of a number of human primary tumors in soft agarose, showing selectivity for myeloma, cervical and breast cancer [43].
9. Disulfide drugs cause apoptosis that is inhibited by thioredoxin Regulation of cancer cell growth by the disulfides via a redox role is also suggested by their induction of apoptosis. Two of the disulfides, III-2 and IV-2, induced HL60 cells to undergo apoptosis within 8 h of exposure, whereas the disulfide, while another analogue, IX-2 did not cause apoptosis at the same concentration in the same time frame [41,44]. These results correlated with the
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
31
thioalkylating ability of these disulfides, IX-2 being three logs slower than III-2 in undergoing thiol/disulfide exchange reaction with thioredoxin [45]. The apoptosis produced by III-2 could be prevented with a 13 h pretreatment of the HL60 cells with the antioxidant, N-acetylcysteine. The 2-imidazolyl disulfides are among the most potent inducers of apoptosis found from among a series of known apoptosisinducing agents we tested.
10. In vivo antitumor activity Several of the disulfides have been shown to exhibit antitumor activity against human MCF-7 breast cancer xenografts growing in SCID mice (Fig. 4).
11. Conclusion The studies have shown that the low molecular weight redox protein, thioredoxin, is a rational target for drug development. It stimulates cell proliferation, it
Fig. 4. Antitumor activity of 2-imidazolyl disulfides against human MCF-7 breast cancer xenografts growing in estrogen supplemented female SCID mice. Mice were inoculated s.c. with 107 MCF-7 cells on day 0 and daily i.p. administration of the disulfides started 24 h later for 14 days. () Vehicle alone; () VI-2 12 mg/kg; () DLK-36 15 mg/kg; and () IV-2 15 mg/kg. Tumor volumes were measured with calipers on the days shown. There were eight animals in each group. Bars are S.E.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
32
prevents apoptosis and it is found over-expressed in many human cancers. Antitumor activity against human xenografts has been observed for a number of alkyl 2-imidazolyl disulfides which have been shown to interact with thioredoxin in vitro. The agent, 1-methylpropyl 2-imidazolyl disulfide, has provided the lead for drug discovery and preclinical studies with this agent are continuing.
Acknowledgements Supported by NIH grants CA48725 and CA17094.
References [1] M. Berggren, A. Gallegos, J.R. Gasdaska, P.Y. Gasdaska, J. Warneke, G. Powis, Thioredoxin and thioredoxin reductase gene expression in human tumors and cell lines, and the effects of serum stimulation and hypoxia, Anticancer Res. 16 (1996) 3459 – 3466. [2] P. Gasdaska, J.E. Oblong, G. Powis, Cloning and expression of thioredoxin (T) cDNA from human colon cancer cells. Identity with ADF and levels of T mRNA in human tumors, Proc. Am. Assoc. Cancer Res. 34 (1993) 62. [3] J.E. Oblong, P.Y. Gasdaska, K. Sherrill, G. Powis, Purification of human thioredoxin reductase: properties and characterization by absorption and circular dichroism spectroscopy, Biochemistry 32 (1993) 7271–7277. [4] P.Y. Gasdaska, J.R. Gasdaska, S. Cochran, G. Powis, Cloning and sequencing of human thioredoxin reductase, FEBS Lett. 373 (1995) 5 – 9. [5] J.R. Gasdaska, P.Y. Gasdaska, A. Gallegos, G. Powis, Human thioredoxin reductase gene localization to chromosomal position 12q23 q24.1 and mRNA distribution in human tissue, Genomics 37 (1996) 257–259. [6] V.N. Gladyshev, K. Jeang, T.C. Stadtman, Selenocysteine, identified as the penultimate C-terminal residue in human T-cell thioredoxin reductase, corresponds to TGA in the human placental gene, Proc. Natl. Acad. Sci. USA 93 (1996) 6146 – 6174. [7] T. Tamura, T.C. Stadtman, A new selenoprotein from human lung adenocarcinoma cells: purification, properties, and thioredoxin reductase activity, Proc. Natl. Acad. Sci. USA 93 (1996) 1006 – 1011. [8] G.D. Paine, C.W. Taylor, R.A. Curtis, M.H.A. Lopez, R.T. Dorr, D. Roe, C.Y. Funk, F. Thompson, E.M. Hersh, Human tumor models in the severe combined immunue deficient scid mouse, Cancer Chemother. Pharmacol. 40 (1997) 209 – 214. [9] D. DePalo, W.B. Kinlaw, C. Zhao, H. Engelberg-Kulka, D.L. St. Germain, Effect of selenium deficiency on type I 5%-deodinase, J. Biol. Chem. 269 (1994) 16223 – 16228. [10] M.J. Berry, J.W. Harney, T. Ohama, D.L. Hatfield, Selenocysteine insertion or termination: factors affecting UGA codon fate and complementary anticodon:codon mutations, Nucl. Acids Res. 22 (1994) 3753–3759. [11] S.C. Low, M.J. Berry, Knowing when not to stop: selenocysteine incorporation in eukaryotes, Trends Biochem. Sci. 21 (1996) 203– 207. [12] K.K. McCaughan, C.M. Brown, M.E. Dalphin, M.J. Berry, W.P. Tate, Translational termination efficiency in mammals is influenced by the base following the stop codon, Proc. Natl. Acad. Sci. USA 92 (1995) 5431–5435. [13] J.A. Milner, Effect of selenium on virally induced and transplantable tumor models, Fed. Proc. 44 (1995) 2568–2572. [14] M. Berggren, A. Gallegos, J. Gasdaska, G. Powis, Cellular thioredoxin reductase activity is regulated by selenium, Anticancer Res. 17 (1997) 3377 – 3380.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
33
[15] R.D. Baker, S.S. Baker, K. LaRosa, C. Whitney, P.E. Newburger, Selenium regulation of glutathione peroxidase in human hepatoma cell line Hep3B, Arch. Biochem. Biophys. 304 (1993) 53–57. [16] T.C. Stadtman, Selenocysteine, Ann. Rev. Biochem. 65 (1996) 83 – 100. [17] A. Holmgren, Thioredoxin and glutaredoxin systems, J. Biol. Chem. 264 (1989) 13963 – 13966. [18] M. Luthman, A. Holmgren, Rat liver thioredoxin and thioredoxin reductase: purification and characterization, Biochemistry 21 (1982) 6628 – 6633. [19] P.Y. Gasdaska, J.E. Oblong, I.A. Cotgreave, G. Powis, The predicted amino acid sequence of human thioredoxin is identical to that of the autocrine growth factor human adult T-cell derived factor (ADF): thioredoxin mRNA is elevated in some human tumors, Biochim. Biophys. Acta 1218 (1994) 292–296. [20] T.C. Laurent, E.C. Moore, P. Reichard, Enzymatic synthesis of deoxyribonucleotides VI. Isolation and characterization of thioredoxin, the hydrogen donor from Escherichia coli B, J. Biol. Chem. 239 (1964) 3436–3444. [21] J.R. Matthews, N. Wakasugi, J.-L. Virelizier, J. Yodoi, R.T. Hay, Thioredoxin regulates the DNA binding activity of NF-kB by reduction of a disulphide bond involving cysteine 62, Nucl. Acids Res. 20 (1992) 3821–3830. [22] J.F. Grippo, W. Tienrungroj, M.K. Dahmer, P.R. Housley, W.B. Pratt, Evidence that the endogenous heat-stable glucocorticoid receptor-activating factor is thioredoxin, J. Biol. Chem. 258 (1983) 13658–13664. [23] C. Abate, L. Patel, F.J. Rauscher III, T. Curran, Redox regulation of fos and jun DNA-binding activity in vitro, Science 249 (1990) 1157 – 1161. [24] G. Powis, J.E. Oblong, P.Y. Gasdaska, M. Berggren, S. Hill, D.L. Kirkpatrick, The thioredoxin/ thioredoxin reductase redox system and control of cell growth, Oncol. Res. 6 (1994) 539 – 544. [25] J.E. Oblong, M. Berggren, P.Y. Gasdaska, G. Powis, Site-directed mutagenesis of active site cysteines in human thioredoxin produces competitive inhibitors of human thioredoxin reductase and elimination of mitogenic properties of thioredoxin, J. Biol. Chem. 269 (1994) 11714 – 11720. [26] J.R. Gasdaska, S.R. Hill, L. Kirkpatrick, W. Montfort, A. Weichsel, M. Kuperis, M. Berggren, G. Powis, Oxidative inactivation of thioredoxin as a cellular growth factor and protection by a Cys73 Ser mutation, Biochem. Pharmacol. 52 (1996) 1741 – 1747. [27] A. Weichsel, J.R. Gasdaska, G. Powis, W.R. Montfort, Crystal structures of reduced, oxidized and mutated human thioredoxins: evidence for a regulatory homodimer, Structure 4 (1996) 735 – 751. [28] J.R. Gasdaska, M. Berggren, G. Powis, Cell growth stimulation by the redox protein thioredoxin occurs by a novel helper mechanism, Cell Growth Differ. 6 (1995) 1643 – 1650. [29] Y. Kitaoka, K. Sorachi, H. Nakamura, H. Masutani, A. Mitsui, F. Kobayashi, T. Mori, J. Yodoi, Detection of adult T-cell leukemia-derived factor/human thioredoxin in human serum, Immunol. Lett. 41 (1994) 155–161. [30] A. Gallegos, J.R. Gasdaska, C.W. Taylor, G.D. Paine-Murrieta, D. Goodman, P.Y. Gasdaska, M. Berggren, M.M. Briehl, G. Powis, Transfection with human thioredoxin increases cell proliferation and a dominant negative mutant thioredoxin reverses the transformed phenotype of breast cancer cells, Cancer Res. 56 (1996) 5765–5770. [31] A. Baker, M. Briehl, C. Payne, G. Powis, Thioredoxin, a gene found over-expressed in human cancer, inhibits apoptosis in vitro and in vivo, Cancer Res. 57 (1997) 5162 – 5167. [32] Y. Kondo, E. Woo, A.E. Michalska, A.C. Choo, J.S. Lazo, Metallothionein null cells have increased sensitivity to anticancer drugs, Cancer Res. 55 (1995) 2021 – 2023. [33] M. Lam, G. Dubyak, L. Chen, G. Nun˜ez, R.L. Miesfeld, C.W. Distelhorst, Evidence that BCL-2 represses apoptosis by regulating endoplasmic reticulum-associated Ca2 + fluxes, Proc. Natl. Acad. Sci. USA 91 (1994) 6569–6573. [34] Y. Onishi, Y. Azuma, Y. Sato, Y. Mizuno, T. Tadakuma, H. Kizaki, Topoisomerase inhibitors induce apoptosis in thymocytes, Biochim. Biophys. Acta 1175 (1993) 147 – 154. [35] T.J. McDonnell, S.J. Korsmeyer, Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14:18), Nature 349 (1991) 254 – 256. [36] G. Merlino, Regulatory imbalances in cell proliferation and cell death during oncogenesis in transgenic mice, Semin. Cancer Biol. 5 (1994) 13 – 20.
34
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
[37] S. Fujii, Y. Nanbu, H. Nonogaki, I. Konishi, T. Mori, H. Masutani, J. Yodoi, Co-expression of adult T-cell leukemia-derived factor, a human thioredoxin homologue and human papillomavirus DNA in neoplastic cervical squamous epithelium, Cancer 68 (1991) 1583 – 1591. [38] N. Kawahara, T. Tanaka, A. Yokomizo, H. Nanri, M. Ono, M. Wada, K. Kohno, K. Takenaka, K. Sugimachi, M. Kuwano, Enhanced co-expression of thioredixin and high mobility group protein 1 genes in human hepatocellular carcinoma and the possible association with decreased sensitivity to cisplatin, Cancer Res. 56 (1996) 5330 – 5333. [39] D.L. Kirkpatrick, M. Kuperus, M. Dowdeswell, M. Kunkel, M. Berggren, M. Angulo, Mechanisms of inhibition of the thioredoxin growth factor system by antitumor 2-imidazolyl disulfides, Biochem. Pharmacol. (1998) in press. [40] D.L. Kirkpatrick, K.M. King, I. Abu Sa’da, R.A. Kelln, M.E. Richardson, D.W. Sieman, Effect of a hypoxic tumor cell cytotoxic disulfide on the membrane and DNA of tumor cells in culture, Anticancer Drugs 3 (1992) 651–658. [41] D.L. Kirkpatrick, I. Abu Sa’da, W. Chernoff, M. Kuperus, Disulfide cytotoxicity under hypoxia, Oncol. Res. 6 (1994) 545–552. [42] M. Kunkel, D.L. Kirkpatrick, G. Johnson, J.I. Johnson, K.T. Paul, T. Myers, G. Powis, Cell line-directed screening assay for inhibitors of thioredoxin reductase signalling as potential anticancer drugs, Anti-Cancer Drug Design 12 (1997) 659 – 670. [43] G. Powis, D.L. Kirkpatrick, M. Angulo, J.R. Gasdaska, M.B. Powell, S.E. Salmon, W.R. Montfort, Thioredoxin redox signalling: a novel target for anticancer drug development, Anticancer Drugs 7 (3) (1996) 121–126. [44] D.L. Kirkpatrick, J. McCabe, M. Kuperus, Stimulation of apoptosis by a redox active disulfide, Proc. Am. Assoc. Cancer Res. 36 (1995) 41. [45] M. Kuperus, D.L. Kirkpatrick, Interaction of redox active disulfides with the autocrine growth factor, human thioredoxin, Proc. Am. Assoc. Cancer Res. 36 (1995) 42.
.
Thioredoxin redox control of cell growth and death and the effects of inhibitors Garth Powis a,*, D. Lynn Kirkpatrick b, Miguel Angulo a, Amanda Baker a a
Arizona Cancer Center, Uni6ersity of Arizona, 1515 North Campbell A6enue, Tucson, AZ 85724 -5024, USA b Department of Chemistry, Uni6ersity of Regina, Regina, Sask. S4S OA2, Canada
Abstract Thioredoxin is a redox protein found over-expressed in some human tumors. Thioredoxin is secreted by tumor cells and stimulates cancer cell growth. Redox activity is essential for growth stimulation by thioredoxin. Cells transfected with thioredoxin cDNA show increased tumor growth and decreased apoptosis in vivo and decreased sensitivity to apoptosis induced by a variety of agents both in vitro and in vivo. Cells transfected with a redox-inactive mutant thioredoxin show inhibited tumor growth in vivo. Thus, thioredoxin offers an attractive target for anticancer drug development. A class of disulfide inhibitors of thioredoxin has been identified. These disulfides inhibit cancer cell growth in culture and have antitumor activity against some human tumor xenografts in animals. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Thioredoxin reductase; Thioredoxin; Apoptosis; Disulfides
1. The thioredoxin system The thioredoxin redox system comprised of two proteins, thioredoxin reductase and thioredoxin, is widely distributed in prokaryotes and eukaryotes. The * Corresponding author. [email protected]
Tel.:
+1
520
6266408;
fax:
+1
0009-2797/98/$19.00 © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S0009-2797(97)00148-8
520
6264848;
e-mail:
24
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
flavoprotein thioredoxin reductase catalyzes the transfer of reducing equivalents from NADPH to the second protein thioredoxin which then reduces a number of protein thiol acceptors. The structures of bacterial and human thioredoxin and thioredoxin reductase are shown in Fig. 1. Both proteins are found widely expressed in human tissues with highest levels of thioredoxin reductase in testis and the uterus and highest levels of thioredoxin in the intestine and kidney [1].
2. Thioredoxin reductase We purified thioredoxin reductase, a selenium containing flavoprotein that catalyzes the NADPH-dependent reduction of thioredoxin [2] from human placenta [3]. Using unique terminal and internal peptide sequences we have also cloned and sequenced the full length cDNA [4]. The predicted amino acid sequence of human thioredoxin reductase shows 44% identity to human glutathione reductase but only 24% identity to bacterial thioredoxin reductase, from which it differs by an additional 150 amino acid dimer interface domain. The human thioredoxin reductase gene (TRXRD1) is found at chromosome 12q 23q24.1 [5]. The C-terminal sequence that we predicted for human placental thioredoxin reductase was based on the presence of a UGA stop codon after the bases coding for the C-terminal Gly494-Cys495 [4]. Subsequently, a selenoprotein was isolated from a human lung adenocarcinoma cell line having a C-terminal amino acid sequence -Gly494-Cys495SeCys496-Gly497 [6] was found to have thioredoxin reductase activity [7]. Thus, the SeCys in human thioredoxin reductase appears to be encoded by the mRNA stop
Fig. 1. Comparison of E. coli and human thioredoxin and thioredoxin reductase.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
25
codon UGA [8]. For UGA to escape recognition by the protein translation machinery and to code for selenocysteine incorporation in eukaryotes requires SeCys incorporation sequence (SECIS) stem-loop element in the mRNA 3%-untranslated region. From work with type 1 deiodinase it appears that the UGA of selenoprotein-encoding mRNAs can alternately function to encode selenocysteine or as a stop codon, depending on conditions. These conditions include limiting selenium [9,10] or excess selenoprotein DNA [10]. A purine immediately following the UGA codon increases the frequency of termination relative to selenocysteine incorporation. Human thioredoxin reductase has a G immediately following UGA which is also found in glutathione peroxidase [11]. This ‘terminator’ purine has been suggested to create a site where termination is favored over selenocysteine incorporation when selenocysteinyl-tRNA is limiting [12] and provides an explanation for the decrease in glutathione peroxidase levels upon nutritional selenium deprivation [13]. Alternatively, cysteine (Cys) may be incorporated in place of SeCys which usually gives an enzyme with less activity than the wild-type protein [14]. Selenium deficiency also directly regulates glutathione peroxidase mRNA levels [15]. The SeCys of thioredoxin reductase lies at the C-terminal end of the protein and is not part of the predicted catalytic site [4]. This contrasts other selenoproteins that catalyze oxidation reduction reactions where SeCys forms part of the active site [16]. It appears that SeCys is essential for the activity of the enzyme because recombinant mutant human thioredoxin reductases without the C-terminal SeCysGly or with SeCys replaced by Cys will not reduce thioredoxin but will reduce the non-specific substrate dithionitrobenzoic acid (unpublished observations). The role SeCys plays in the biological activity thioredoxin reductase remains a matter of conjecture. Our other studies have shown that Cys73 of thioredoxin outside the conserved catalytic site is critical for the interaction of thioredoxin with human thioredoxin reductase [5] so that interaction between the SeCys of thioredoxin reductase and thioredoxin may occur at the Cys73 residue. The amount of available selenium in cells controls thioredoxin reductase activity. Both a concentration-dependent increase in cellular thioredoxin reductase activity and concentration-dependent increase in thioredoxin protein measured by Western blotting [14] were observed with increasing concentrations of sodium selenite added to serum-free growth medium. However, the maximum increase in protein at any selenium concentration is considerably less than the increase in thioredoxin reductase activity. Thus, in the absence of selenium, there appears to be premature termination or incorporation of Cys into thioredoxin reductase in place of SeCys. This results in a decrease in the specific activity of the enzyme. There is also a decrease in the amount of thioredoxin reductase mRNA and protein.
3. Thioredoxin as a growth factor The Cys residues at the conserved -Cys32-Gly-Pro-Cys35-Lys active site of the low molecular weight redox protein thioredoxin [17], undergo reversible oxidation-re-
26
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
duction catalyzed thioredoxin reductase [18]. Human thioredoxin is a 11.5 kDa protein and contains three additional Cys residues not found in bacterial E. coli thioredoxin with which it has 27% amino acid identity. These additional Cys residues give human thioredoxin unique biological properties [19]. Thioredoxin was first studied for its ability to act as a reducing co-factor for ribonucleotide reductase, the first unique step in DNA synthesis [20]. More recently, thioredoxin has been shown to exert redox control over some transcription factors to modulate their binding to DNA and thus, to regulate gene transcription. Transcription factors that are regulated by thioredoxin include NF-kB [21], the glucocorticoid receptor [22] and indirectly through a nuclear redox factor Ref-1/HAP1, thioredoxin can regulate AP-1 (Fos/Jun heterodimer) [23]. Thioredoxin is also a growth factor with a unique mechanism of action. Human recombinant thioredoxin stimulates the proliferation of both normal fibroblasts and a wide variety of human solid and leukemic cancer cell lines, even in minimal medium in the absence of serum [24,25]. Thioredoxin undergoes spontaneous oxidation in the absence of reducing agents [26], initially through oxidation to monomeric forms that retain the ability to be reduced by thioredoxin reductase but that cannot stimulate cell proliferation. This is followed by a slower spontaneous conversion of thioredoxin to a homodimer form that is not a substrate for reduction by thioredoxin reductase and that also cannot stimulate cell proliferation. X-ray crystallography has shown that human thioredoxin forms a dimer through a highly conserved interface domain that is stabilized by an intermolecular Cys73-Cys73 disulfide bond [27]. We have shown that a Cys73 Ser mutant thioredoxin monomer was just as effective as the wild-type protein at stimulating cell growth and did not show spontaneous or oxidant induced loss of activity. Furthermore, this mutant did not form a dimer [26]. Thus, Cys73 of thioredoxin, although not necessary for the growth stimulating activity of thioredoxin, is involved in inactivation of the wild-type protein. Whether this inactivation plays a role in the physiological actions of thioredoxin remains to be determined. The studies with 125I-thioredoxin have revealed no high affinity binding sites that might suggest receptors for thioredoxin on the surface of cancer cells [28]. Instead the studies demonstrate that thioredoxin stimulates cell proliferation by increasing the sensitivity of the cells to growth factors secreted by the cells themselves [28]. This potentiation can be over 100-fold for example with basic fibroblast growth factor and is seen at concentrations of thioredoxin found in human serum which are in the range of 10–30 nM [29]. Cancer cells secrete thioredoxin [1,30]. The pleiotropic effects of thioredoxin on cell growth, taken together with the observations that thioredoxin is over-expressed by a number of human primary tumors, suggests that thioredoxin could be an important factor in human cancer cell growth [28].
4. Thioredoxin prevents apoptosis Transfection of cells with thioredoxin is able to prevent apoptosis. Apoptosis is a process of programmed cell death that mediates normal cell turnover, hormone-
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
27
Table 1 Effect of trx and bcl-2 transfection on apoptosis induced by different agents Relative apoptosis
Dexamethasone Staurosporine Etoposide Thapsigargin
Neo
Trx5
Trx6
W.Hb12
2.99 0.4 58.39 9.8 162.5 9 12.1 4.39 0.9
0.9 9 0.0* 5.3 9 0.0* 6.0 9 0.6* 2.3 9 1.8*
0.3 90.0* 9.9 9 1.1* 20.9 94.1* 1.8 91.2*
1.0 90.1* 4.8 90.7* 4.5 90.7* 0.9 90.6*
Vector-alone transfected WEHI7.2 cells (Neo); Trx5 and Trx6 trx transfected WEHI7.2 cells and bcl-2 transfected WEHI7.2 cells (W.Hb12) were treated with 1 mM dexamethasone for 24 h, 1 mM staurosporine for 21 h, 60 mM N-acetyl-sphingosine for 42 h, 1 mM etoposide for 15 h or 50 nM thapsigargin for 24 h. The times were determined to be optimum for detecting apoptosis with each agent. Apoptosis was measured by flow cytometry and expressed as a ratio of the sum of regions representing early and late apoptotic cells divided by region representing live non-apoptotic cells, and normalized to the ratio for vehicle treated vector-alone transfected cells. Values are the mean of determinations9S.E. Statistical analysis was by linear regression with indicator values for drugs and cells. *PB0.05 compared to vector-alone transfected control cells.
induced tissue atrophy, cell mediated killing in immunity and tumor regression. WEHI7.2 mouse thymoma-derived cells were stably transfected with human trx cDNA and had levels of thioredoxin mRNA up to 1.8-fold compared to endogenous levels of mouse thioredoxin mRNA [31]. Immunofluorescent staining determined that the thioredoxin transfected cells also have increased levels of thioredoxin protein, with approximately 60% of the staining found in the nucleus and 60% in the cytoplasm. These thioredoxin transfected WEHI7.2 cells were resistant to apoptosis induced by dexamethasone compared to wild-type or vector-alone transfected cells (Table 1). Thioredoxin has been reported to be necessary for assembly of the glucocorticoid receptor [22], however, glucocorticoid receptor activity measured using a glucocorticoid receptor/chloramphenicol acetyltransferase reporter plasmid was not altered in the thioredoxin transfected cells. The effect of thioredoxin transfection on other agents known to induce apoptosis was studied (Table 1). Thioredoxin transfected cells were resistant to apoptosis, compared to vector-alone transfected control cells, induced by staurosporine, a general kinase inhibitor [32], thapsigargin, which blocks the uptake of intracellular Ca2 + , resulting in an increase in intracellular free Ca2 + concentration [33] and by etoposide, an inhibitor of topoisomerase II [34]. WEHI7.2 cells transfected with the bcl-2 anti-apoptotic proto-oncogene (W.HB12 cells) showed a similar pattern of protection against apoptosis induced by the various agents as did the thioredoxin transfected cells.
28
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
5. Thioredoxin stimulates tumor growth and inhibits apoptosis in vivo When inoculated into severe combined immunodeficient (SCID) mice, the tumors that formed from the thioredoxin transfected WEHI7.2 cells grew more rapidly than the wild-type cell tumors (Fig. 2). Histological examination of the wild-type cell tumors showed apoptosing cells with condensed and marginated chromatin and a dense cytoplasm accompanied by vacuolization. In contrast, tumors that formed from the thioredoxin transfected cells showed few apoptosing cells. Dexamethasone treatment inhibited the growth of the wild-type cell tumors (Fig. 2), however, not that of the hioredoxin transfected cell tumors. It was also of interest for us to compare the thioredoxin transfected WEHI7.2 cells to Bcl-2 transfected WEHI7.2 cells in vivo. The Bcl-2 transfected cells formed tumors that grew faster than tumors formed by wild-type WEHI7.2 cells but still considerably slower than the thioredoxin transfected cell tumors. Surprisingly, the growth of the bcl-2 transfected cell tumors was still inhibited by dexamethasone treatment. Evidence from transgenic mice over-expressing bcl-2 under transcriptional regulation of the Ig heavy chain enhancer has shown that the mice develop benign lymphoma that eventually
Fig. 2. Tumor formation in SCID mice by (, ) wild-type WEHI7.2 cells; (, ) bcl-2 transfected WEHI7.2 (W.Hb12) cells; and ( , ) Trx6 trx transfected WEHI7.2 cells. A total of 20 mice were injected s.c. with 2× 107 cells and 0.1 ml matrigel into the flank. Tumor size was measured every 3 – 4 days with calipers and tumor volumes calculated. Half the mice were injected s.c. into the opposite flank with 1 mg/kg/day dexamethasone (, , ) or with vehicle alone (, , ) 9 days after tumor cell injection. Mice were euthanized the first measurement that tumor volume exceeded 2 cm3. Values are the mean of ten mice in each group and bars are S.E.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
29
progresses to high grade malignant disease [35]. This has been taken to indicate that bcl-2 provides a survival advantage to cells in vivo but that an additional change, most frequently rearrangement of myc [35], is necessary for tumor growth [36]. In contrast to Bcl-2, thioredoxin itself offers a clear growth advantage to tumors in vivo.
6. Thioredoxin in human cancer Expression of thioredoxin mRNA was elevated in almost half (5/12) of the human primary lung cancers examined compared to normal lung tissue from the same subject [19] and in half (6/12) of the human primary colorectal tumors compared to normal colonic mucosa [1]. Others have reported increased levels of thioredoxin protein in human squamous cervical cancer and hepatoma [37,38] and we have found increased levels in human acute lymphocytic leukemia (3/6 subjects). An important caveat to these studies is that the thioredoxin mRNA was extracted from pieces of tumor giving no indication of thioredoxin’s distribution within the tumor, whether from malignant cells, stromal or hypoxic core cells. This is significant because hypoxia is known to lead to an increase in thioredoxin expression [1]. The preliminary immunohistochemical studies have shown that in normal human colon thioredoxin is found in the dividing basal crypt cells while in primary colon cancer thioredoxin is over-expressed exclusively by the cancer cells. It remains to be determined whether thioredoxin over-expression is related proliferation or to apoptosis in human primary tumors and if it has a role as a prognostic indicator. While human tumors show many fold over-expression of thioredoxin mRNA, up to 100-fold in one patient with colon cancer, the cell transfection studies have given, at most, a 2-fold over-expression of mRNA. Similarly, we have been able to obtain only very low levels of thioredoxin over-expression in the transgenic mouse studies (not shown). These results suggests that high levels of thioredoxin over-expression may be toxic to cells. If this is so, human tumors appear to have become resistant to high levels of thioredoxin expression or may have mutant thioredoxin.
7. Thioredoxin as a target for drug development Since the molecular studies provide proof-of-principle that the thioredoxin system is a rational target for anticancer drug development, the initial approach was to develop agents which might selectively inhibit the thioredoxin system and thioredoxin-dependent cell proliferation. A series of alkyl 2-imidazolyl disulfides, typical of which is 1-methylpropyl 2-imidazolyl disulfide (IV-2) were studied (Fig. 3). Short-chain alkyl analogues of this series were found to be substrates for the enzymes, thioredoxin reductase while analogues with a branched alkyl chain or benzyl group are competitive inhibitors. Prolonged incubation of thioredoxin with the alkyl 2-imidazolyl disulfide results in irreversible inhibition of the thioredoxin as a substrate for reduction by thioredoxin reductase [39]. The inhibition appears to be
30
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
Fig. 3. Structures of 2-imidazolyl disulfides.
specific for Cys73 of thioredoxin and is not seen with a mutant Cys73 Ser thioredoxin, which remains a substrate for thioredoxin reductase even after prolonged incubation with the disulfides. In culture, the disulfides show greater inhibition of thioredoxin- than serum-dependent cell proliferation, suggesting that they may be producing their effects by inhibiting the thioredoxin system [39].
8. Inhibition of cancer cell growth Besides inhibiting the growth of a number of cancer cell lines in culture [24,40,41], studies with the National Cancer Institute (NCI) 60 cancer cell line panel [42] show a number of the disulfides exhibit selectivity for inhibition of leukemia, colon and renal cell growth. A significant negative correlation between disulfide activity and thioredoxin reductase mRNA in the cell lines was revealed using the COMPARE algorithm [42]. Furthermore, one of the disulfides, IV-2, was found to inhibit the growth of a number of human primary tumors in soft agarose, showing selectivity for myeloma, cervical and breast cancer [43].
9. Disulfide drugs cause apoptosis that is inhibited by thioredoxin Regulation of cancer cell growth by the disulfides via a redox role is also suggested by their induction of apoptosis. Two of the disulfides, III-2 and IV-2, induced HL60 cells to undergo apoptosis within 8 h of exposure, whereas the disulfide, while another analogue, IX-2 did not cause apoptosis at the same concentration in the same time frame [41,44]. These results correlated with the
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
31
thioalkylating ability of these disulfides, IX-2 being three logs slower than III-2 in undergoing thiol/disulfide exchange reaction with thioredoxin [45]. The apoptosis produced by III-2 could be prevented with a 13 h pretreatment of the HL60 cells with the antioxidant, N-acetylcysteine. The 2-imidazolyl disulfides are among the most potent inducers of apoptosis found from among a series of known apoptosisinducing agents we tested.
10. In vivo antitumor activity Several of the disulfides have been shown to exhibit antitumor activity against human MCF-7 breast cancer xenografts growing in SCID mice (Fig. 4).
11. Conclusion The studies have shown that the low molecular weight redox protein, thioredoxin, is a rational target for drug development. It stimulates cell proliferation, it
Fig. 4. Antitumor activity of 2-imidazolyl disulfides against human MCF-7 breast cancer xenografts growing in estrogen supplemented female SCID mice. Mice were inoculated s.c. with 107 MCF-7 cells on day 0 and daily i.p. administration of the disulfides started 24 h later for 14 days. () Vehicle alone; () VI-2 12 mg/kg; () DLK-36 15 mg/kg; and () IV-2 15 mg/kg. Tumor volumes were measured with calipers on the days shown. There were eight animals in each group. Bars are S.E.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
32
prevents apoptosis and it is found over-expressed in many human cancers. Antitumor activity against human xenografts has been observed for a number of alkyl 2-imidazolyl disulfides which have been shown to interact with thioredoxin in vitro. The agent, 1-methylpropyl 2-imidazolyl disulfide, has provided the lead for drug discovery and preclinical studies with this agent are continuing.
Acknowledgements Supported by NIH grants CA48725 and CA17094.
References [1] M. Berggren, A. Gallegos, J.R. Gasdaska, P.Y. Gasdaska, J. Warneke, G. Powis, Thioredoxin and thioredoxin reductase gene expression in human tumors and cell lines, and the effects of serum stimulation and hypoxia, Anticancer Res. 16 (1996) 3459 – 3466. [2] P. Gasdaska, J.E. Oblong, G. Powis, Cloning and expression of thioredoxin (T) cDNA from human colon cancer cells. Identity with ADF and levels of T mRNA in human tumors, Proc. Am. Assoc. Cancer Res. 34 (1993) 62. [3] J.E. Oblong, P.Y. Gasdaska, K. Sherrill, G. Powis, Purification of human thioredoxin reductase: properties and characterization by absorption and circular dichroism spectroscopy, Biochemistry 32 (1993) 7271–7277. [4] P.Y. Gasdaska, J.R. Gasdaska, S. Cochran, G. Powis, Cloning and sequencing of human thioredoxin reductase, FEBS Lett. 373 (1995) 5 – 9. [5] J.R. Gasdaska, P.Y. Gasdaska, A. Gallegos, G. Powis, Human thioredoxin reductase gene localization to chromosomal position 12q23 q24.1 and mRNA distribution in human tissue, Genomics 37 (1996) 257–259. [6] V.N. Gladyshev, K. Jeang, T.C. Stadtman, Selenocysteine, identified as the penultimate C-terminal residue in human T-cell thioredoxin reductase, corresponds to TGA in the human placental gene, Proc. Natl. Acad. Sci. USA 93 (1996) 6146 – 6174. [7] T. Tamura, T.C. Stadtman, A new selenoprotein from human lung adenocarcinoma cells: purification, properties, and thioredoxin reductase activity, Proc. Natl. Acad. Sci. USA 93 (1996) 1006 – 1011. [8] G.D. Paine, C.W. Taylor, R.A. Curtis, M.H.A. Lopez, R.T. Dorr, D. Roe, C.Y. Funk, F. Thompson, E.M. Hersh, Human tumor models in the severe combined immunue deficient scid mouse, Cancer Chemother. Pharmacol. 40 (1997) 209 – 214. [9] D. DePalo, W.B. Kinlaw, C. Zhao, H. Engelberg-Kulka, D.L. St. Germain, Effect of selenium deficiency on type I 5%-deodinase, J. Biol. Chem. 269 (1994) 16223 – 16228. [10] M.J. Berry, J.W. Harney, T. Ohama, D.L. Hatfield, Selenocysteine insertion or termination: factors affecting UGA codon fate and complementary anticodon:codon mutations, Nucl. Acids Res. 22 (1994) 3753–3759. [11] S.C. Low, M.J. Berry, Knowing when not to stop: selenocysteine incorporation in eukaryotes, Trends Biochem. Sci. 21 (1996) 203– 207. [12] K.K. McCaughan, C.M. Brown, M.E. Dalphin, M.J. Berry, W.P. Tate, Translational termination efficiency in mammals is influenced by the base following the stop codon, Proc. Natl. Acad. Sci. USA 92 (1995) 5431–5435. [13] J.A. Milner, Effect of selenium on virally induced and transplantable tumor models, Fed. Proc. 44 (1995) 2568–2572. [14] M. Berggren, A. Gallegos, J. Gasdaska, G. Powis, Cellular thioredoxin reductase activity is regulated by selenium, Anticancer Res. 17 (1997) 3377 – 3380.
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
33
[15] R.D. Baker, S.S. Baker, K. LaRosa, C. Whitney, P.E. Newburger, Selenium regulation of glutathione peroxidase in human hepatoma cell line Hep3B, Arch. Biochem. Biophys. 304 (1993) 53–57. [16] T.C. Stadtman, Selenocysteine, Ann. Rev. Biochem. 65 (1996) 83 – 100. [17] A. Holmgren, Thioredoxin and glutaredoxin systems, J. Biol. Chem. 264 (1989) 13963 – 13966. [18] M. Luthman, A. Holmgren, Rat liver thioredoxin and thioredoxin reductase: purification and characterization, Biochemistry 21 (1982) 6628 – 6633. [19] P.Y. Gasdaska, J.E. Oblong, I.A. Cotgreave, G. Powis, The predicted amino acid sequence of human thioredoxin is identical to that of the autocrine growth factor human adult T-cell derived factor (ADF): thioredoxin mRNA is elevated in some human tumors, Biochim. Biophys. Acta 1218 (1994) 292–296. [20] T.C. Laurent, E.C. Moore, P. Reichard, Enzymatic synthesis of deoxyribonucleotides VI. Isolation and characterization of thioredoxin, the hydrogen donor from Escherichia coli B, J. Biol. Chem. 239 (1964) 3436–3444. [21] J.R. Matthews, N. Wakasugi, J.-L. Virelizier, J. Yodoi, R.T. Hay, Thioredoxin regulates the DNA binding activity of NF-kB by reduction of a disulphide bond involving cysteine 62, Nucl. Acids Res. 20 (1992) 3821–3830. [22] J.F. Grippo, W. Tienrungroj, M.K. Dahmer, P.R. Housley, W.B. Pratt, Evidence that the endogenous heat-stable glucocorticoid receptor-activating factor is thioredoxin, J. Biol. Chem. 258 (1983) 13658–13664. [23] C. Abate, L. Patel, F.J. Rauscher III, T. Curran, Redox regulation of fos and jun DNA-binding activity in vitro, Science 249 (1990) 1157 – 1161. [24] G. Powis, J.E. Oblong, P.Y. Gasdaska, M. Berggren, S. Hill, D.L. Kirkpatrick, The thioredoxin/ thioredoxin reductase redox system and control of cell growth, Oncol. Res. 6 (1994) 539 – 544. [25] J.E. Oblong, M. Berggren, P.Y. Gasdaska, G. Powis, Site-directed mutagenesis of active site cysteines in human thioredoxin produces competitive inhibitors of human thioredoxin reductase and elimination of mitogenic properties of thioredoxin, J. Biol. Chem. 269 (1994) 11714 – 11720. [26] J.R. Gasdaska, S.R. Hill, L. Kirkpatrick, W. Montfort, A. Weichsel, M. Kuperis, M. Berggren, G. Powis, Oxidative inactivation of thioredoxin as a cellular growth factor and protection by a Cys73 Ser mutation, Biochem. Pharmacol. 52 (1996) 1741 – 1747. [27] A. Weichsel, J.R. Gasdaska, G. Powis, W.R. Montfort, Crystal structures of reduced, oxidized and mutated human thioredoxins: evidence for a regulatory homodimer, Structure 4 (1996) 735 – 751. [28] J.R. Gasdaska, M. Berggren, G. Powis, Cell growth stimulation by the redox protein thioredoxin occurs by a novel helper mechanism, Cell Growth Differ. 6 (1995) 1643 – 1650. [29] Y. Kitaoka, K. Sorachi, H. Nakamura, H. Masutani, A. Mitsui, F. Kobayashi, T. Mori, J. Yodoi, Detection of adult T-cell leukemia-derived factor/human thioredoxin in human serum, Immunol. Lett. 41 (1994) 155–161. [30] A. Gallegos, J.R. Gasdaska, C.W. Taylor, G.D. Paine-Murrieta, D. Goodman, P.Y. Gasdaska, M. Berggren, M.M. Briehl, G. Powis, Transfection with human thioredoxin increases cell proliferation and a dominant negative mutant thioredoxin reverses the transformed phenotype of breast cancer cells, Cancer Res. 56 (1996) 5765–5770. [31] A. Baker, M. Briehl, C. Payne, G. Powis, Thioredoxin, a gene found over-expressed in human cancer, inhibits apoptosis in vitro and in vivo, Cancer Res. 57 (1997) 5162 – 5167. [32] Y. Kondo, E. Woo, A.E. Michalska, A.C. Choo, J.S. Lazo, Metallothionein null cells have increased sensitivity to anticancer drugs, Cancer Res. 55 (1995) 2021 – 2023. [33] M. Lam, G. Dubyak, L. Chen, G. Nun˜ez, R.L. Miesfeld, C.W. Distelhorst, Evidence that BCL-2 represses apoptosis by regulating endoplasmic reticulum-associated Ca2 + fluxes, Proc. Natl. Acad. Sci. USA 91 (1994) 6569–6573. [34] Y. Onishi, Y. Azuma, Y. Sato, Y. Mizuno, T. Tadakuma, H. Kizaki, Topoisomerase inhibitors induce apoptosis in thymocytes, Biochim. Biophys. Acta 1175 (1993) 147 – 154. [35] T.J. McDonnell, S.J. Korsmeyer, Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14:18), Nature 349 (1991) 254 – 256. [36] G. Merlino, Regulatory imbalances in cell proliferation and cell death during oncogenesis in transgenic mice, Semin. Cancer Biol. 5 (1994) 13 – 20.
34
G. Powis et al. / Chemico-Biological Interactions 111–112 (1998) 23–34
[37] S. Fujii, Y. Nanbu, H. Nonogaki, I. Konishi, T. Mori, H. Masutani, J. Yodoi, Co-expression of adult T-cell leukemia-derived factor, a human thioredoxin homologue and human papillomavirus DNA in neoplastic cervical squamous epithelium, Cancer 68 (1991) 1583 – 1591. [38] N. Kawahara, T. Tanaka, A. Yokomizo, H. Nanri, M. Ono, M. Wada, K. Kohno, K. Takenaka, K. Sugimachi, M. Kuwano, Enhanced co-expression of thioredixin and high mobility group protein 1 genes in human hepatocellular carcinoma and the possible association with decreased sensitivity to cisplatin, Cancer Res. 56 (1996) 5330 – 5333. [39] D.L. Kirkpatrick, M. Kuperus, M. Dowdeswell, M. Kunkel, M. Berggren, M. Angulo, Mechanisms of inhibition of the thioredoxin growth factor system by antitumor 2-imidazolyl disulfides, Biochem. Pharmacol. (1998) in press. [40] D.L. Kirkpatrick, K.M. King, I. Abu Sa’da, R.A. Kelln, M.E. Richardson, D.W. Sieman, Effect of a hypoxic tumor cell cytotoxic disulfide on the membrane and DNA of tumor cells in culture, Anticancer Drugs 3 (1992) 651–658. [41] D.L. Kirkpatrick, I. Abu Sa’da, W. Chernoff, M. Kuperus, Disulfide cytotoxicity under hypoxia, Oncol. Res. 6 (1994) 545–552. [42] M. Kunkel, D.L. Kirkpatrick, G. Johnson, J.I. Johnson, K.T. Paul, T. Myers, G. Powis, Cell line-directed screening assay for inhibitors of thioredoxin reductase signalling as potential anticancer drugs, Anti-Cancer Drug Design 12 (1997) 659 – 670. [43] G. Powis, D.L. Kirkpatrick, M. Angulo, J.R. Gasdaska, M.B. Powell, S.E. Salmon, W.R. Montfort, Thioredoxin redox signalling: a novel target for anticancer drug development, Anticancer Drugs 7 (3) (1996) 121–126. [44] D.L. Kirkpatrick, J. McCabe, M. Kuperus, Stimulation of apoptosis by a redox active disulfide, Proc. Am. Assoc. Cancer Res. 36 (1995) 41. [45] M. Kuperus, D.L. Kirkpatrick, Interaction of redox active disulfides with the autocrine growth factor, human thioredoxin, Proc. Am. Assoc. Cancer Res. 36 (1995) 42.
.