Rhodanese–thioredoxin system and allyl sulfur compounds Implications in apoptosis induction Renato Sabelli 1 , Egidio Iorio 2 , Angelo De Martino 3 , Franca Podo 2 , Alessandro Ricci 2 , Giuditta Viticchie ` 1 , Giuseppe Rotilio 3 , Maurizio Paci 1 and Sonia Melino 1 1 Department of Sciences and Chemical Technologies, University of Rome ‘‘Tor Vergata’’, Italy 2 Department of Cellular Biology and Neurosciences, Istituto Superiore di Sanita ` , Rome, Italy 3 Department of Biological Sciences, University of Rome ‘‘Tor Vergata’’, Italy The induction of programmed cell death by sulfane sulfur compounds [alk(en)yl thiosulfate, selenodigluta- thione, allyl disulfide, etc.] poses significant questions concerning their metabolism and on the role of the enzymes involved in the cancerogenesis. Toohey [1] suggested that uncontrolled proliferation of neoplastic cells is a result of sulfane sulfur deficiency or overacti- vity of the enzymes involved in their metabolism. Several organosulfur compounds (OSCs) such as diallyl disulfide, diallyl trisulfide, allicin and (more Keywords garlic; mobile lipids; sodium 2-propenyl thiosulfate (2-PTS); sulfane sulfur; sulfurtransferase Correspondence S. Melino, Dipartimento di Scienze e Tecnologie Chimiche, University of Rome ‘‘Tor Vergata’’, via della Ricerca Scientifica 00133-Rome, Italy Fax: +39 0672594328 Tel: +39 0672594449 E-mail: melinos@uniroma2.it E. lorio, Department of Cellular Biology and Neurosciences, Istituto Superiore di Sanita ´ , Viale Regina Elena, 299 00161 Rome, Italy Fax: +39 06 49387143 Tel: +39 06 49902548 E-mail: egidio.iorio@iss.it (Received 7 March 2008, revised 30 May 2008, accepted 3 June 2008) doi:10.1111/j.1742-4658.2008.06535.x Sodium 2-propenyl thiosulfate, a water-soluble organo-sulfane sulfur com- pound isolated from garlic, induces apoptosis in a number of cancer cells. The molecular mechanism of action of sodium 2-propenyl thiosulfate has not been completely clarified. In this work we investigated, by in vivo and in vitro experiments, the effects of this compound on the expression and activity of rhodanese. Rhodanese is a protein belonging to a family of enzymes widely present in all phyla and reputed to play a number of dis- tinct biological roles, such as cyanide detoxification, regeneration of iron– sulfur clusters and metabolism of sulfur sulfane compounds. The cytotoxic effects of sodium 2-propenyl thiosulfate on HuT 78 cells were evaluated by flow cytometry and DNA fragmentation and by monitoring the progressive formation of mobile lipids by NMR spectroscopy. Sodium 2-propenyl thio- sulfate was also found to induce inhibition of the sulfurtransferase activity in tumor cells. Interestingly, in vitro experiments using fluorescence spec- troscopy, kinetic studies and MS analysis showed that sodium 2-propenyl thiosulfate was able to bind the sulfur-free form of the rhodanese, inhibit- ing its thiosulfate:cyanide-sulfurtransferase activity by thiolation of the catalytic cysteine. The activity of the enzyme was restored by thioredoxin in a concentration-dependent and time-dependent manner. Our results sug- gest an important involvement of the essential thioredoxin–thioredoxin reductase system in cancer cell cytotoxicity by organo-sulfane sulfur com- pounds and highlight the correlation between apoptosis induced by these compounds and the damage to the mitochondrial enzymes involved in the repair of the Fe–S cluster and in the detoxification system. Abbreviations DCF-DA, 2’, 7’-dichlorodihydrofluorescein diacetate dye; 2-PTS, sodium 2-propenyl thiosulfate; ESI, electrospray ionization; GSH, reduced glutathione; GSSG, oxidized glutathione; OSC, organosulfur compound; PCho, phosphocholine; RhdA(E), sulfur-free form of rhodanese; RhdA(ES), sulfur-loaded rhodanese; RhdA, recombinant rhodanese from Azotobacter vinelandii; RhdA-PS, propexylsulfide-form of rhodanese; ROS, reactive oxygen species; Trd, thioredoxin reductase; Trx, thioredoxin; TST, thiosulfate:cyanide sulfurtransferase. 3884 FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS recently), n-propyl thiosulfate and sodium 2-propenyl thiosulfate (2-PTS) have been shown to suppress the proliferation of tumor cells in vitro through the induc- tion of apoptosis [2–7]. The biochemical mechanisms underlying the antitumorigenic and antiproliferative effects of garlic-derived OSCs are not yet fully under- stood, although it seems likely that the rate of clear- ance of allyl sulfur groups from cells is a determinant of the overall response. It has also been hypothesized that a reduced or incorrect functionality of enzymes involved in the metabolism of OSCs may cause an excess of seleno or sulfane sulfur compounds in the cell, inducing the onset of degenerative states and apoptosis. The in vitro study of their interaction with enzymes probably involved in the metabolism of sulfur may explain their in vivo effects. Sulfurtransferases, enzymes that act at the borderline between sulfur and selenium metabolism, may have an important role in the processes induced by these sulfur compounds. In particular, rhodanese displays a sulfurtransferase activ- ity in vitro, transferring the sulfane sulfur atom from thiosulfate to cyanide to produce thiocyanate by a double-displacement mechanism [thiosulfate:cyanide sulfurtransferase EC 2.8.1.1, (TST)] [8]. In this activity, rhodanese cycles between two stable intermediates, namely a sulfur-loaded form (ES) and a sulfur-free form (E). The observed abundance of potentially func- tional rhodanese-like proteins suggests that members of this homology superfamily (accession number: PF00581; http://sanger.ac.uk/cgi-bin/Pfam) may play distinct biological roles [9,10]. Rhodanese modules, either alone or in combination with other proteins, can perform a variety of roles, ranging from transport mechanisms of sulfur ⁄ selenium in a biologically avail- able form [11–13], to the modulation of general detoxification processes [8] and to the restoration of iron–sulfur centres in Fe–S proteins, such as ferredoxin [14]. In the last few years, many different studies have supported the hypothesis that the rhodanese-like pro- teins have roles in ‘managing’ stress tolerance and in maintaining redox homeostasis [15–18]. It is also note- worthy that a study, using microarray to examine the gene expression in colonic mucosa from cancerous and normal tissues, hypothesized that a possible cause of colorectal cancer carcinogenesis might be located in the mitochondria and identified the rhodanese gene as one out of three mitochondrial genes that had a statis- tically significant decrease in expression from normal tissue to tumor at every Dukes’ stage A–D [19,20]. More recently an increase in TST expression was observed in colonocyte differentiation. In a human colon cancer cell line, TST activity and expression were significantly increased by butyrate and by histone- deacetylase inhibitors, which promote the differentia- tion of these cells [21]. Thus, butyrate could protect the colonocytes from sulfide-induced cytotoxicity, thus promoting TST expression. [21]. Down-regulation of rhodanese gene expression has also been observed in diseases such as Friedreich’s ataxia [22]. The associa- tion between the expression of rhodanese and degener- ative states might make rhodanese a potential tumor ⁄ disease biomarker and treatment target. Not all cells are equally susceptible to the deleterious effects of the garlic sulfur compounds and, in particular, non- neoplastic cells are less susceptible [3,23]. Therefore, the greater sensitivity of tumor cells to these com- pounds may be related to a down-regulated expression of TST, which leads to a reduced rate of clearance of these compounds; notably, epidemiological investiga- tion has revealed that increased consumption of garlic diminishes the risk of stomach and colorectal cancers [24–27]. To explore the importance of rhodanese in the metabolism of the allyl-sulfur compounds, we investi- gated, in the present work, the effect of some natural sulfur constituents of garlic on TST activity. In particular, we studied the interaction of 2-PTS with rhodanese and investigated the effect of this OSC on the expression and activity of rhodanese. Sodium 2-propenyl thiosulfate is present in aqueous garlic extract [28] and has an anti-aggregator effect in vitro on both canine and human platelets as a result of the inhibition of cyclooxygenase activity [29]. The anti- tumor effect of 2-PTS, resulting from induction of the apoptosis process, has been recently investigated [7]. The data reported here show that 2-PTS is able to bind to the active site of rhodanese, resulting in TST inhibition. We also investigated the role of reduced thioredoxin (Trx) as a possible sulfur acceptor in this reaction to restore TST activity. Moreover, the cyto- toxic effect of 2-PTS on HuT 78 cells was also evalu- ated using flow cytometry analysis after treatment and by monitoring the progressive formation of mobile lipids by NMR spectroscopy. All results obtained provide evidence that highlights the possible role of rhodanese in the management of the cytotoxicity of reactive OSCs in tumor cells and may contribute to the design of a scheme of the mechanism of action of 2-PTS as an apoptosis inducer. Results Inhibition of cell cycle progression and induction of apoptosis by 2-PTS The effects of 2-PTS on the human T-lymphoblastoid cell line, HuT 78, were analysed and a typical time- R. Sabelli et al. Sulfurtransferases and apoptotic natural sulfane sulfur compounds FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS 3885 dependent and dose-dependent inhibition of cell growth of these cells as a result of the presence of 2-PTS was observed. In fact, a statistically significant decrease in the number of viable cells, at 0.25, 0.5 and 1 mm concen- trations of 2-PTS and at different time-points, was found when comparing control and thiosulfate-treated cells (data not shown). About 75 and 10% of the HuT 78 cells were viable following exposure to 0.5 mm 2-PTS for 24 and 48 h respectively (Fig. 1A). The growth- inhibitory effects of 2-PTS were determined by using the Trypan Blue dye-exclusion assay. Flow cytometric anal- ysis of HuT 78 cells after 24 and 48 h of treatment with 0.5 mm 2-PTS, following staining with propidium iodide, resulted in a statistically significant increase in the fraction of subG1 compared with the control (Fig. 1B), showing a characteristic feature of apoptosis. Sodium 2-propenyl thiosulfate induced an increase in the fraction of HuT 78 cells in the G 2 ⁄ M phase after 24 h of treatment, reaching a value of 67.43% (Fig. 1B). After 48 h, the percentage of the cell population that showed apoptotic features (subG 1 region) was $ 39%, suggesting that an antiproliferative activity of 2-PTS against HuT 78 cells (blockage in G 2 ⁄ M phase) results in the triggering of apoptotsis. The effect of treatment with 2-PTS on DNA frag- mentation was assessed by analysis of the agarose-gel electrophoretic pattern of HuT 78 cells (see supplemen- tary Fig. S1). A typical DNA-fragmentation pattern was observed as early as at 24 h after the addition of 2-PTS, thus confirming the apoptosis-inducing effects of 2-PTS. The role of reactive oxygen species (ROS) as potential mediators of 2-PTS-induced cytotoxicity in HuT 78 cells was also investigated. Figure 1C shows the production of ROS during the first 3 h of treatment with 2-PTS. HuT 78 cells underwent an increase of ROS flux as early as 1 h after addition of 2-PTS. These results implied a strict association between 2-PTS effects and oxidative imbalance. Reduced glutathione (GSH) is the most important physiological antioxidant, both by directly reacting with ROS and indirectly by preserving cysteine residues of proteins from irreversible oxidations, so giv- ing rise to GS-R. The intracellular reduced (GSH) and oxidized (GSSG) glutathione levels were measured using HPLC chromatography in order to show the potential involvement of this redox system in the resistance of HuT 78 cells to treatment with 2-PTS. Untreated cells showed an average intracellular GSH concentration of $ 55 nmolÆ mg )1 of protein, whereas 2-PTS-treated cells showed a rapid and sustained increase of GSH levels (118.5 nmolÆmg )1 protein) up to 12 h, probably as a result of to the detoxification process. The intracellular GSSG content was not significantly affected by treat- ment with 2-PTS and remained at very low levels. Formation of 1 H-NMR-visible mobile lipids during 2-PTS-induced apoptosis The apoptotic parameters were quantitatively moni- tored by NMR spectroscopy carried out on intact HuT 78 cells and on their aqueous cell extracts after cell expo- sure to 2-PTS. Detection of the progressive formation of mobile lipids in intact cells indicated the induction of apoptosis after treatment with 2-PTS (from 6 to 48 h). Resonances centered at d = 1.3 p.p.m., as a result of the saturated fatty acyl chain methylene segments -(CH 2 ) n -, and at d = 0.9 p.p.m., arising from methyl A B C Fig. 1. In vitro effect of 2-PTS on the growth of HuT 78 cells. (A) Viability of HuT 78 cells in culture after treatment with 0.5 m M 2-PTS. Trypan Blue staining was used to differentiate viable from nonviable cells. Data are expressed as mean ± SD. *P < 0.005, **P < 0.001 (n = 8). (B) Percentage of the cell cycle distribution of HuT 78 cells after 24 and 48 h of treatment with 0.5 m M 2-PTS. (C) Intracellular production of ROS in HuT 78 cells after 1 and 3 h of treatment with 0.5 m M 2-PTS, detected by measurement of DCF fluorescence using a FACSCalibur flow cytometer. Data are expressed as mean ± SD. *P < 0.001, **P < 0.05 (n = 3). Sulfurtransferases and apoptotic natural sulfane sulfur compounds R. Sabelli et al. 3886 FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS groups, were followed. Quantitative analysis of mobile lipid spectral profiles (see the supplementary material) was obtained by measuring the peak area (a) ratio R chains = a[(CH 2 ) n ] lip ⁄ a[CH 3 ] tot , according to our previ- ous work [30], where a[(CH 2 ) n ] lip is the integral of the mobile lipids, (CH 2 ) n is the resonance and a(CH 3 ) tot is the integral of the total CH 3 resonance at 0.9 p.p.m. caused by amino acids and lipids. The value R chains in untreated control cells was 0.20 ± 0.1, not significantly different from that measured in HuT 78 cells exposed to 0.25 mm 2-PTS at any time-point. The average R chains value at 24 h of exposure to 2-PTS increased with the apoptotic fraction from 1.25 to 1.56 in cells treated with 0.5 and 1.0 mm 2-PTS, respectively. The spectral profiles of cells treated with either 0.5 or 1.0 mm 2-PTS for 48 h were mainly characterized by signals from mobile lipids (R chains = 4.45 ± 1.1), while all resonances caused by small aqueous metabolites decreased to very low levels as a result of the massive apoptosis of treated HuT 78 cells, probably associated with loss of cell integrity. Quantitative analysis on 1 H-NMR spectra of cell extracts after exposure to 0.5 mm 2-PTS for an interme- diate time interval (24 h, not yet associated with late apoptosis) showed that the intracellular level of taurine (measured from the triplet centred at d = 3.45 p.p.m.) decreased about two-fold from a basal control level of 13.0 ± 2.4 to 6.5 ± 1.0 nmol per 10 6 cells). Under the same experimental conditions, individual choline- containing metabolites (tCho, detected under a reso- nance band centered at a chemical shift of $ 3.22 p.p.m.) underwent differential changes; in fact phosphocholine (PCho) decreased by 45% (from 15.1 ± 1.4 to 8.1 ± 0.8 nmol per 10 6 cells), while free choline and glycerophosphocholine both increased by about three-fold. The changes in the levels of water- soluble choline-containing metabolites observed in cells treated with 2-PTS for 24 h probably reflect the progres- sive activation of phospholipases and phosphodiester- ases. On the other hand, a complex network of pathways may, in principle, contribute to the measured decrease in PCho, a metabolite particularly sensitive to conditions determining a block in cell proliferation and ⁄ or to the activation of enzymes involved in choline– phospholipid degradation. In fact, both increases and decreases in PCho have been reported to occur in differ- ent systems of apoptotic induction, according to partic- ular experimental conditions [30]. Effect of 2-PTS on TST expression and activity in HuT 78 cells The effects of 2-PTS on the expression and activity of TST in HuT 78 cells were analyzed. HuT 78 cells were treated without and with 2-PTS at various concentra- tions (0.25, 0.5 and 1 mm). Figure 2A shows the western blot of the HuT 78 lysates after 8 and 24 h of treatment with 0.5 mm 2-PTS. Densitometry measure- ments of western blots, corrected for actin or for glyceraldehyde-3-phosphate dehydrogenase (see the supplementary material) expression, show that no sig- nificant variation of the expression of TST with respect to the control was induced by treatment with 2-PTS (see Fig. 2A). By contrast, a reduction of TST activity was observed after 24 h, as shown in Fig. 2B. Interaction of 2-PTS and allyl compounds with rhodanese In vitro interaction and kinetic studies were performed using the recombinant Azotobacter vinelandii rhoda- nese (RhdA) [31–33]. RhdA has similar properties and kinetic behaviour and a high structural homology with bovine rhodanese [34], which is the rhodanese consid- ered as an appropriate model for using to study human rhodanese [32,33,35]. The A. vinelandii rhoda- A B Fig. 2. Effect of 2-PTS on TST expression and activity in HuT 78 cells. (A) Western immunoblotting showing the expression of TST in HuT 78 cell lysates after treatment with 2-PTS. Thirty micro- grams of protein lysates from untreated cells (CTRL) incubated for 8 and 24 h, and from cells treated with 0.5 m M 2-PTS for 8 and 24 h, were analysed. A monoclonal anti-actin Ig was used as a con- trol of the protein concentrations; and densitometry measurements of western immunoblotting were calculated by comparison with the intensity of actin expression. (B) Percentage the TST activity of HuT 78 cellular extracts [untreated cells (CTRL) and cells treated with 0.5 m M 2-PTS (2-PTS) after 8, 24 and 48 h of incubation]. The control after 24 h was used as the reference value. *P < 0.05. R. Sabelli et al. Sulfurtransferases and apoptotic natural sulfane sulfur compounds FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS 3887 nese has been shown to contain only one cysteine resi- due, which is essential for the catalytic reaction, and therefore it represents a good model for using to study the possible interaction of TST proteins with other proteins or molecules involved in sulfur and selenium metabolism [35,36]. In order to investigate the action of 2-PTS on sulfur- transferase activity, we analyzed the effect of 2-PTS by monitoring the changes of intrinsic fluorescence that occur when the rhodanese cycles between the sulfur- free [RhdA(E)] form and the sulfur-loaded [RhdA(ES)] form and that are caused by long-range energy trans- fer and local conformational changes in the protein [36–40]. No fluorescence changes were observed when 2-PTS was added to RhdA(ES) (Fig. 3A). By contrast, the addition of 2-PTS induced an evident quenching of intrinsic fluorescence of RhdA(E) (Fig. 3B), in a con- centration-dependent manner, indicating a specific interaction of 2-PTS with the active site. Moreover, the quantum yield of intrinsic fluorescence of RhdA(E) after the addition of 2-PTS was lower than that obtained following the addition of thiosulfate (DF% = 27.2 compared with thiosulfate DF% = 15.7) (Fig. 3B). This major fluorescence quenching was probably a result of the closeness of the allyl group to the Trp residues present in the protein active site, and is probably responsible for the differential quenching of the intrinsic fluorescence in the two states of the enzyme [37,39,41]. Interestingly, cyanide did not restore the unloaded form of the enzyme (Fig. 3C) and a significant loss of the TST activity of the enzyme was observed. Inactivation of sulfur-free rhodanese by 2-PTS and characterization of the derivative protein In order to study chemical modifications of sulfur-free rhodanese, it is necessary first to remove the persulfide from the enzyme by adding an excess of cyanide to the protein, thus forming the thiocyanate product and sulfur-free rhodanese. As sulfur-free rhodanese is somewhat unstable, it is customary to add the modify- ing reagent immediately after cyanide treatment. We performed the experiment in this way to analyze the effect of the 2-PTS on the sulfur-free form of the enzyme. Figure 4 shows the time course of inactivation of RhdA(E) caused by an interaction with 2-PTS. Pre- incubation of RhdA in the presence of a three-fold molar excess of cyanide and a 200-fold molar excess of 2-PTS induced a decrease of sulfurtransferase activity over time, and a complete loss of sulfurtransferase activity was observed after 90 min at 37 °C (Fig. 4A). The activity of treated RhdA was not restored after dialysis, indicating that stable binding occurs between 2-PTS and RhdA(E). By contrast, no inhibition of the TST activity was observed after pre-incubation of RhdA(ES) in the presence of the same concentration of 2-PTS without CN - . TST activity of the propenyl- sulfide-form of rhodanese, RhdA-PS, was restored by treatment with dithiothreitol. In fact, 53.6% of the TST activity, with respect to the TST activity of A B C Fig. 3. Intrinsic fluorescence changes of RhdA following the addi- tion of substrates. (A) RhdA(ES) (2.3 l M)( ____ ), after the addition of 460 l M 2-PTS (- __ -) and after the addition of 460 lM CN - (- - -); (B) RhdA(E) (4.6 l M)( _____ ), with a molar ratio E: thiosulfate 1 : 400 (- - - -), with a molar ratio E:2-PTS 1 : 200 ( ___ ) and E:2-PTS 1 : 400 ( __ - __ ); and (C) RhdA(E) (5 lM)( _____ ), in the presence of 2-PTS (E: 2-PTS 1 : 250 c ⁄ c) ( ___ ), E: 2-PTS 1 : 500 c ⁄ c(- __ -) and after the addition of CN - (E: 2-PTS: CN 1 : 500 : 1000 c ⁄ c ⁄ c) (- - - -). a.u., arbitrary units. Sulfurtransferases and apoptotic natural sulfane sulfur compounds R. Sabelli et al. 3888 FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS RhdA(ES) before treatment, was recovered after 30 min of incubation at room temperature (23°C) (see Fig. 4B). These results were in agreement with an oxi- dation state of the catalytic Cys. Thiosulfate sulfur- transferase activity of RhdA-PS was also restored when Trx protein was added to the solution and this activity was dependent on the Trx concentration (Fig. 5A). Thioredoxin was able to re-establish 66% of the TST activity of RhdA-PS when was incubated with Trx at an RhdA-PS ⁄ Trx molar ratio of 1 : 2. The yield of the recovery was faster and higher when thioredoxin reductase (Trd) and NADPH were also present in the solution. Figure 5B shows the recovery of the TST activity of RhdA-PS in the presence of Trd, NADPH, and different molar concentrations of reduced Trx. A total recovery of the TST activity was obtained after 70 min of incubation with 0.5 lm Trx. No increase of thiocyanate production was observed in the presence of Trx, Trd and NADPH without rhodanese. These results, together with the fluorescence spectra, indicate an interaction of 2-PTS with the catalytic Cys. Thus, either the reduced Trx or dithiothreitol can reduce a disulfide bond between the OSC and the catalytic cys- teine of RhdA, and thereby restore the enzyme to its active state. This hypothesis was confirmed by MS analysis of the new form of the enzyme. RP-HPLC chromatography of RhdA-PS and RhdA(E) was per- formed, followed by electrospray ionization (ESI) MS analysis. Although the two forms of the protein showed the same retention times in RP-HPLC (see the supplementary material), they showed different molec- ular mass peaks. In fact, RhdA-PS and RhdA(E) had molecular mass values that corresponded to m ⁄ z 31138.9 ± 3.19 and m ⁄ z 31063.8 ± 3, respec- tively. These results are consistent with thiolation of A B Fig. 4. Inhibition of the sulfur-free form of RhdA by 2-PTS. (A) Time-dependent decrease of the TST activity of 48.3 l M RhdA(E) in 50 m M Tris–HCl buffer, pH 8.0, 0.3 M NaCl. RhdA(E) was treated with a three-fold molar excess of cyanide and a 200-fold molar excess of 2-PTS at 37 °C. (B) Effects of dithiothreitol on the TST activity of RhdA(ES) and RhdA-PS. The proteins were incubated with 4 m M dithiothreitol at room temperature and the TST activity was assayed after 0, 15 and 30 min. RhdA(ES) and RhdA-PS were treated identically, except that cyanide and 2-PTS were absent dur- ing the treatment of RhdA(ES). The TST activity of the enzyme before treatment was taken to represent 100% activity. Each value represents the average of three independent determinations. DTT, dithiothreitol. A B Fig. 5. Recovery of the TST activity of 17 lM RhdA-PS detected using the So ¨ rbo assay. (A) Recovery of TST activity after 2 h of incubation at 25 °C in the absence and in the presence of thiore- doxin at molar ratios 1 : 0, 1 : 0.5, 1 : 1 and 1 : 2 c ⁄ c RhdA-PS ⁄ Trx. All values are expressed as a percentage of the TST activity value of RhdA(ES). (B) Recovery of the TST activity of 8.1 l M RhdA-PS after incubation in the presence of 0.1 U Trd, 50 l M NADPH and different concentrations of Trx (0, 0.05, 0.15, 0.25 and 0.5 l M). R. Sabelli et al. Sulfurtransferases and apoptotic natural sulfane sulfur compounds FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS 3889 the catalytic cysteine of the enzyme with a propenylsul- fide (m ⁄ z 73), confirming modification of the protein at the catalytic site. Sensitivity of RhdA-PS to proteolysis To characterize the RhdA-PS form in greater detail, limited proteolysis of RhdA-PS was performed. Lim- ited proteolysis of globular proteins generally occurs at sites that contain the most flexible regions of the poly- peptide chain within a domain or at the flexible hinges between domains. Therefore, a limited trypsin diges- tion of RhdA-PS was performed to investigate the flex- ibility of this modified enzyme. As previously observed, the sulfur-loaded form of RhdA [RhdA(ES)] [42] appeared to be quite resistant to limited proteoly- sis. In fact, Rhd(ES), which was treated in the same manner as for RhdA-PS, but in the absence of cyanide and 2-PTS, was not proteolyzed by trypsin and remained intact, even after incubation overnight (Fig. 6A). By contrast, RhdA-PS showed a higher sen- sitivity to proteolysis than RhdA(ES), as shown in Fig. 6B,and a band rapid digestion was observed. A stable daughter band (b band), of about 17.3 kDa, appeared after a few minutes of proteolysis and was present also after many hours of digestion. These data suggest that RhdA-PS is more flexible than RhdA(ES), probably as a result of local conformational differ- ences. RhdA-PS showed behaviour very similar to that previously observed for the alkylated and oxidated forms of the bovine rhodanese [43,44]. In fact, limited proteolysis of the alkylated and oxidized forms yielded fragments that were about half of the apparent molec- ular mass of the protein, as a result of cleavage at the interdomain tether that connects the two domains into which the single polypeptide chain protein is folded [42,43,45]. Thus, the inactivation of the rhodanese by 2-PTS induces local conformational changes that make the enzyme much more sensitive to proteolytic degra- dation. RhdA-PS form catalysed the Trx oxidation Thioredoxin oxidation analyses were performed with the aim of clarifying the mechanism involved in restor- ing the TST activity of RhdA-PS. The oxidation of Trx by RhdA was observed by NADPH oxidation. The change in NADPH concentration was caused by oxidation of the reduced Trx to its disulfide form, which was reduced by NADPH in the reaction cata- lyzed by Trd. The presence of RhdA(ES) at an equi- molar concentration of Trx caused oxidation of the reduced Trx at equilibrium with NADPH, as previ- ously observed for the bovine rhodanese [46,47] even in the absence of a sulfur donor (Fig. 7A). Thus, also in this case, reduced Trx behaves as sulfur-acceptor substrate. Control experiments showed that no oxida- tion of NADPH occurred in the presence of rhodanese when reduced Trx was absent. The addition of 2-PTS (as a substrate of RhdA) to the solution resulted in further oxidation of NADPH (Fig. 7A). 2-PTS was also able to oxidize Trx, but the presence of RhdA caused an increase in the NADPH oxidation rate (Fig. 7A). In addition, RhdA-PS was able to catalyze Trx oxidation. A rapid decrease in the concentration of NADPH was observed when an equimolar A B Fig. 6. Time course of trypsin digestion of RhdA-PS and RhdA(ES). Three-hundred micrograms of enzyme was subjected to limited digestion with 1% (w ⁄ w) trypsin in 1 mL of 50 m M Tris–HCl buffer, pH 8, at room temperature. After the reaction the samples were subjected to SDS-PAGE. (A) Lanes 2–7, proteolysis products of RhdA(ES) (lane 1) at 0, 5, 10, 15, 20 and 30 min and overnight incu- bation. (B) Lanes 2–8, proteolysis products of RhdA-PS (lane 1) at 0, 5, 10, 15, 20, 30, 60 min and overnight incubation, respectively. ‘a’ and ‘b’ bands are the parent and daughter bands, at about 29.7 and 17.3 kDa respectively. Molecular markers are on the left. STDs, molecular mass standards. Sulfurtransferases and apoptotic natural sulfane sulfur compounds R. Sabelli et al. 3890 FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS concentration of RhdA-PS was added to the stabilized solution containing Trx, 0.1 U Trd and 50 lm NADPH (Fig. 7B). Moreover, the subsequent addition of 2-PTS to the solution led to a further increase in NADPH oxidation. These results indicate that the rhodanese– Trx–Trd system is not inhibited by 2-PTS and that this system has an antioxidant action against OSCs as well a sulfane sulfur detoxification system. Discussion Natural compounds, which improve detoxification enzymes and ⁄ or reduce the expression and activity of the carcinogen activating enzymes, are good candidates for cancer chemoprevention. The controlled prolifera- tion of the cell may be a result of the presence of a sul- fane sulfur compound that has the ability to reduce or promote the activity of important proteins implicated in the cellular process. Many OSCs present in allium vegetables have been shown to be able to inhibit the proliferation of cancer cells [4,48–51]. The induction of programmed cell death by sulfane sulfur compounds raises relevant questions about the role of enzymes involved in their metabolism. A feature of a neoplastic cell is the residual activity of cysteine aminotransfer- ase, 3-mercaptopyruvate sulfurtransferase and rhoda- nese, as well as the total lack of cystathionase activity [52]; in fact, the biosynthesis and transport of com- pounds from the sulfane sulfur pool does not occur in these cells [53]. Our studies presented here, on HuT 78 cells, showed that the viability of these cells is reduced significantly upon 24 h of exposure to 2-PTS and that this reduced growth rate was related to a blockage in the G 2 ⁄ M phase of the cell cycle. The detection, by NMR, of increasing amounts of mobile lipids in 2-PTS-treated HuT 78 cells, further supports mitochondrial dysfunction in these cells. In fact, several studies have reported the appearance of mobile lipids in a variety of cells induced to apoptosis [33,54–56], in which the characteristic apoptotic pheno- type generally results in loss of the mitochondrial membrane potential, release of cytochrome c and mito- chondrial-dependent activation of effector caspases. Recently, alterations of mitochondrial functions by different uncouplers of the respiratory chain have been found to be responsible for the accumulation of intra- cellular lipid bodies and therefore for the detection of NMR-visible mobile lipids in intact HuT 78 cells (E. lorio, C. Testa, A. Stringaro, M. Condello, G. Ara- ncia, E. Lococo, R. Carnevale, R. Strom, L. Lenti & F. Podo, unpublished data). Furthermore, mobile lipid signals have been reported in tumor cells exposed to lipophilic cationic compounds that cause mitochon- drial damage [57]. Also the observed decrease of intra- cellular taurine in cells exposed to 2-PTS for 24 h (i.e. before the occurrence of massive cell death at 48 h) seems to be in general agreement with the view of pro- gressive mitochondrial impairment in these cells. In fact, Hansen et al. [58] recently suggested a new role of taurine in mitochondrial function by acting as a modu- lator of pH in the mitochondrial matrix and therefore altering the overall oxidative capacity of this subcellu- lar organelle, including a reduction in fatty acyl b-oxi- dation. According to this hypothesis, the simultaneous decrease in taurine and an increase in NMR-detected mobile lipids suggest a substantial 2-PTS-induced loss of mitochondrial function with subsequent accumula- tion of long fatty acyl chains in triglycerides in cyto- plasmatic lipid bodies. In recent years, it has become A B Fig. 7. Trx oxidation by 2-PTS, in the presence and in the absence of rhodanese, by measurement of NADPH (50 l M) consumption (absorbance at 340 nm), in 50 m M Tris–HCl buffer, pH 8.0, 1 mM EDTA, with 4 lM Trx, 0.1 U Trd and in the presence of 0.5 mM 2-PTS. (A) Tris–HCl buffer (a) or 4 lM RhdA(ES) (b) was added to the solution after 30 min at 37 °C, and, after stabilization, 2-PTS was added. The data were normalized against an A 340 of 0.093 (which represents 100%). (B) Tris–HCl buffer (a) or 4 l M RhdA-PS (b) was added to the solution after 15 min of incubation at 37 °C and, after stabilization (about 10 min), 2-PTS was added. The data were normalized against an A 340 of 0.085 (which represents 100%). R. Sabelli et al. Sulfurtransferases and apoptotic natural sulfane sulfur compounds FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS 3891 apparent that mitochondria are integrally involved in the mechanism of cell death, and anticancer drugs, which inhibit the functions of mitochondria can sensi- tize the cells to undergo apoptosis [59,60]. It is impor- tant to consider that rhodanese is an important mitochondrial enzyme in mammalian cells [61] and, certainly, the rhodanese–Trx system is involved in the integrity of this main energy-generator organelle. It is noteworthy that recently it has also been dem- onstrated that rhodanese deficiency affects the activity of Fe–S enzymes of the tricarboxylic acid cycle, such as the aconitase in A. vinelandii [15]. These studies sug- gest that in tumor cells in which there is a decrease in expression of the TST gene [19,20], the involvement of rhodanese in the detoxification of OSCs could lead to an inhibition of the normal function of the cyanide detoxification and Fe–S repair activities of this enzyme, thus inhibiting the normal functions of the mitochondria. The elucidation of interactions between OSCs and proteins, such as rhodanese or Trx that are involved in maintaining the redox homeostasis of the cell, could help to explain the mechanism whereby they can induce programmed cell death in tumor cells. In the present work we investigated the effect of 2-PTS on the activity and expression of TST. We performed our experiments in vitro using the A. vinelandii rhoda- nese. The data reported here demonstrate that 2-PTS is able to interact with the Cys of RhdA(E), inducing a covalent modified form (RhdA-PS). 2-PTS interacts with the active site of the rhodanese enzyme by thiola- tion of the catalytic cysteine, therefore inhibiting its TST activity. A peculiar characteristic of the A. vine- landii enzyme is the presence of only one cysteine resi- due, which is the catalytic site. The structural similarity of RhdA with the bovine rhodanese led us to speculate that 2-PTS can induce a thiolation at the level of the catalytic Cys residue of the human rhoda- nese. The ability of this garlic compound to thiolate an internal Cys, such as that of the active site of rhoda- nese, is an important observation that should be borne in mind when considering the mechanism of action of OSCs. We showed that the thiolation of a mitochon- drial enzyme, which is involved in ‘managing’ the redox state of the cell, could be a relevant event in the mechanism of action of this compound. The major sensitivity of RhdA-PS to degradation by proteolysis is an important factor that should be con- sidered. Limited proteolysis of RhdA-PS shows that, as well the alkylation of bovine rhodanese [43], local conformational changes occur when the enzyme is modified by 2-PTS interactions, and high flexibility of the enzyme with respect to RhdA(ES) is induced. These data were also in agreement with a significant decrease in the expression of TST after 48 h of treat- ment with 2-PTS (data not shown). Therefore, the major flexibility of RhdA-PS may also explain its ability to interact with and oxidize the Trx compared to RhdA(ES). Our results highlight a direct interaction between OSCs from garlic and rhodanese. The bond of the propenyl sulfide with the catalytic cysteine shows the characteristic of a disulfide bond (i.e. it is not cleavable by nucleophilic attack of the cyanide). The propenyl- sulfur protein has a low redox potential, so Trx and dithiothreitol reduce it to restore the TST activity. These results provide evidence that, in vitro, reduced Trx also regulates TST activity via redox regulation. 2-PTS alone was also able to oxidize the reduced Trx, but in the presence of RhdA(ES) or RhdA-PS Trx was oxidized more rapidly. The results presented here sup- port a possible mechanism (Fig. 8) where the reduced Trx could be the sulfide acceptor of the rhodanese in the intracellular system. Trd was able to reduce the oxidized Trx, and thus this process may be a potential sulfane sulfur detoxification system present in the cell. This result highlights the antioxidant action of a rhodanese– Trx–Trd system against OSCs. As described in previous work, reduced Trx is a sulfur-acceptor substrate for rho- danese. It has been also hypothesized that a primary function of the rhodanese could be Trx-linked oxygen radical detoxification [46,47]. The data presented here shed new light on the role of this enzyme in the sulfane sulfur detoxification system and on the involvement of the Trx–Trd system in this process. Recently, evidence for the down-regulation of TST expression in some cancer cells has been reported [20,21]. Our data show that no change in the expression of rhodanese occurs after 8 and 24h of treatment with 2-PTS. By contrast, O | | H C 2 HC- H C = 2 O -S -S - - a N + dhR - S _ _ || O STP-2 S _ _ S _ _ | HC 2 dhR H S53 xrT S 23 - S2 3 __ S _ _ H C 2 __ HC=HC 2 xrT S5 3 __ H dhR - S __ 2 | HC | | HC 2 HP DAN dT xrT S53 S23 | SH __ HC 2 __ HC =HC 2 T r d Fig. 8. Scheme of the proposed reactions of the interaction between rhodanese (Rhd) and 2-PTS, and of the restoration of rho- danese activity by the thioredoxin (Trx)–thioredoxin reductase (Trd) system in the cell. Sulfurtransferases and apoptotic natural sulfane sulfur compounds R. Sabelli et al. 3892 FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS a significant reduction of TST activity was observed dur- ing treatment, indicating that cyanide detoxification of rhodanese was reduced by the presence of 2-PTS. Thus, the data suggest that rhodanese could be a target enzyme of the garlic OSCs and that the reduced TST activity could be caused by an increase of the sulfur- detoxification activity of the enzyme, which also involves the Trx system. The ability of 2-PTS to inhibit, either in vitro or in the cell, the TST activity of the rhodanese and to oxi- dize the Trx in vitro, both in the absence and in the presence of rhodanese, can be related to its ability to induce apoptosis in the cell. Our studies showed that the viability of HuT 78 cells is reduced significantly after 24 h of exposure to 2-PTS and this reduced growth rate is related to a blockage in the G 2 ⁄ M phase of the cell cycle. HuT 78 cells underwent an early increase of ROS flux after the addition of 2-PTS, and 2-PTS-treated cells showed a rapid and sustained increase of GSH levels up to 12 h, most probably because of a detoxification process. These results implied a strict correlation between the apoptotic effects of 2-PTS and oxidative imbalance, and this can also be linked to a reduction in the ability of the rhoda- nese–Trx system to detoxify by oxygen radicals in vivo. Interestingly, the blockage in the G 2 ⁄ M phase of the cell cycle was linked to an early increase of ROS flux. Cells treated with 2-PTS showed a rapid and sustained increase in GSH levels up to 12 h; this was attributable to a detoxification process. These results imply a strict correlation between 2-PTS apoptotic effects and oxida- tive unbalance, and this can be also linked to a reduc- tion in the activity of the oxygen radicals detoxification of the rhodanese–thioredoxin system. These results are in agreement with previous studies reported by Chang et al. [7], where it was observed that 2-PTS suppresses, in a dose-dependent manner, the growth of HL-60 cells through the induction of apoptosis initiated by oxidative stress. Although there might be several mechanisms involved in the apoptosis of cancer cells, we believe that the effects of this sulfur compound, and probably also of other OSCs, may be linked to mitochondrial expression levels and activity both of rhodanese and of Trx in the cancer cells and that the RNA interference technique could be used to demonstrate this hypothesis. Interestingly, high levels of Trx have also been associ- ated with cancer that is resistant to therapy, and low Trx levels have been associated with apoptosis in gastric carcinomas [62–64]. The Trx ⁄ Trd system is considered to act as an endogenous antioxidant system in all living cells, in addition to the glutathione system, so the mal- function of this antioxidant system in mitochondria can lead to an increase of intracellular ROS [65]. Mitochon- drial rhodanese–Trx ⁄ Trd-system oxidation by OSCs could lower the normal reducing activity of Trx and thus inhibit the activity of important enzymes, such as perox- iredoxin 3, an enzyme involved in H 2 O 2 metabolism in mitochondria, whose oxidation plays an important role in the promotion of apoptosis [66]. In conclusion, these studies contribute to extend the knowledge on the physiological role of rhodanese and may represent a relevant starting point to elucidate the implication of the rhodanese–Trx ⁄ Trd system in che- moprevention therapy approaches using sulfane sulfur compounds, whose biochemical metabolism and certain biological effects warrant further investigation. Materials and methods 2-PTS synthesis Sodium 2-propenyl thiosulfate was synthesized according to a method described by Chapelet et al. [67]. The product was dried in vacuo , extracted with methanol and the extract was purified by silica gel chromatography (methanol ⁄ chloroform; 45 : 55, v ⁄ v). The purity and structure of the compound were evaluated by RP-HPLC, LC-MS and 1 H-NMR. Cell proliferation assay HuT 78 human T-lymphoblastoid cells were purchased from the ISS (Istituto Superiore di Sanita ` , Rome, Italy). Approximately 0.2 · 10 6 HuT 78 cells were pre-incubated for 24 h in RPMI 1640 (GIBCO, Milan, Italy) in the pres- ence of 1% glutamine, 10% heat-inactivated fetal bovine serum and antibiotics (1% penicillin and streptomycin sulfate) at 37 °C in air supplemented with 5% CO 2 and were then exposed to 2-PTS for 24 and 48 h. The cells were collected and counted, after staining with Trypan Blue (0.4% Tripan blu solution; Sigma-Aldrich, Milan, Italy), by optical microscopy using a Thoma chamber. The rate of growth inhibition by 2-PTS was calculated with respect to the control culture taken as 100% growth. Cell cycle analysis The cell cycle distribution of HuT 78 cells was measured by flow cytometry. Approximately 0.5 · 10 6 harvested cells were stained with 50 lgÆmL )1 of propidium iodide (Sigma- Aldrich) in NaCl ⁄ P i with 0.1% Triton X-100 and 1mgÆmL )1 of sodium citrate. Then, the cells were imme- diately analysed using a flow cytometer (FACSCalibur; Becton Dickinson, San Jose ` , CA, USA) and the percentage of cells in each phase of the cell cycle was evaluated accord- ing to Nicoletti et al. [68]. R. Sabelli et al. Sulfurtransferases and apoptotic natural sulfane sulfur compounds FEBS Journal 275 (2008) 3884–3899 ª 2008 The Authors Journal compilation ª 2008 FEBS 3893 [...]... performed using a Bruker win-nmr software package Protein extraction and western blot analysis Proteins were extracted from HuT 78 cells in 200 lL of 50 mm Tris–HCl, pH 7.09, containing a protease inhibitors cocktail (Sigma-Aldrich) and sonicated over four steps of 5 s and 1 min of pause in ice The samples were centrifuged (10 min, 5000 g, 4 °C) The protein contents were determined using the bicinchoninic... class of enzyme and of other proteins containing rhodanese domains for which biochemical characterization is still lacking A dysregulation of TST expression and activity induced by hydrogen sulfide (H2S) was observed, and this could result in the inability to detoxify effectively and be a factor in the cell loss and in ammation observed in ulcerative colitis and in colorectal cancer [16] Acknowledgements... during the organic synthesis of the 2-PTS compound, and Dr C Capo and Dr V Brunetti for technical support in some experimental studies We are grateful to Professor Silvia Pagani for kindly giving us the plasmid encoding the RhdA protein This work was supported, in part, by grants PRIN and FIRB of Italian MIUR References 1 Toohey JI (1986) Persulfide sulfur is a growth factor for cells defective in sulfur. .. The instrument was calibrated using myoglobin Proteolysis of RhdA-PS Limited proteolysis of RhdA was performed using 1% (w ⁄ w) trypsin (Sigma-Aldrich) in 50 mm Tris–HCl, pH 8, 0.3 m NaCl The reaction was stopped at different timepoints by the addition of SDS-PAGE sample buffer and boiling the samples for 2 min Subsequently, the samples were analysed using SDS-PAGE with a 15% resolving gel Thioredoxin... enzymes in Azotobacter vinelandii FEBS Lett 278, 151–154 Sulfurtransferases and apoptotic natural sulfane sulfur compounds 33 Pagani S, Sessa G, Sessa F & Colnaghi R (1993) Properties of Azotobacter vinelandii rhodanese Biochem Mol Biol Int 29, 595–604 34 Bordo D, Deriu D, Colnaghi R, Carpen A, Pagani S & Bolognesi M (2000) The crystal structure of a sulfurtransferase from Azotobacter vinelandii highlights... equilibria between intermediates in rhodanese catalysis J Biol Chem 258, 7894–7896 42 Melino S, Cicero DO, Forlani F, Pagani S & Paci M (2004) The N-terminal rhodanese domain from Azotobacter vinelandii has a stable and folded structure independently of the C-terminal domain FEBS Lett 577, 403–408 43 Horowitz PM & Criscimagna NL (1988) Sulfhydryldirected triggering of conformational changes in the enzyme... enzyme in 50 mm Tris–HCl buffer, pH 7.2 Kinetic analysis Rhodanese activity was measured by the discontinuous colorimetric assay described by Sorbo [71], where the pro¨ duction of thiocyanate from thiosulfate and cyanide was followed at 460 nm using a Perkin-Elmer spectrometer Briefly, 0.5 lg of enzyme (RhdA) was incubated in a reaction mixture containing 58 mm KCN and 58 mm sodium thiosulfate in 50... (2005) Growth inhibitory effect of alk(en)yl thiosulfates derived from onion and garlic in human immortalized and tumor cell lines Cancer Lett 223, 47–55 8 Westley B & Rochefort H (1980) A secreted glycoprotein induced by estrogen in human breast cancer cell lines Cell 20, 353–362 3896 R Sabelli et al 9 Hofmann K, Bucher P & Kajava AV (1998) A model of Cdc25 phosphatase catalytic domain and Cdk-interaction... Herman-Antosiewicz A, Marynowski SW & Singh SV (2006) c-Jun NH(2)-terminal kinase signaling axis regulates diallyl trisulfide-induced generation of reactive oxygen species and cell cycle arrest in human prostate cancer cells Cancer Res 66, 5379–5386 51 Hosono T, Fukao T, Ogihara J, Ito Y, Shiba H, Seki T & Ariga T (2005) Diallyl trisulfide suppresses the proliferation and induces apoptosis of human colon cancer... beta-tubulin J Biol Chem 280, 41487–41493 52 Wlodek L, Wrobel M & Czubak J (1993) Transamination and transsulphuration of L-cysteine in Ehrlich ascites tumor cells and mouse liver The nonenzymatic reaction of L-cysteine with pyruvate Int J Biochem 25, 107–112 53 Iciek MB, Rokita HB & Wlodek LB (2001) Effects of diallyl disulfide and other donors of sulfane sulfur on the proliferation of human hepatoma cell line . Rhodanese–thioredoxin system and allyl sulfur compounds Implications in apoptosis induction Renato Sabelli 1 , Egidio Iorio 2 , Angelo De Martino 3 ,. 2B. Interaction of 2-PTS and allyl compounds with rhodanese In vitro interaction and kinetic studies were performed using the recombinant Azotobacter vinelandii rhoda- nese