Chemosphere 63 (2006) 942–949 www.elsevier.com/locate/chemosphere
Cyanide detoxification by recombinant bacterial rhodanese Rita Cipollone a, Paolo Ascenzi
a,b,c
, Emanuela Frangipani a, Paolo Visca
a,b,*
a Dipartimento di Biologia, Universita` ‘Roma Tre’, Viale G. Marconi 446, 00146 Rome, Italy Istituto Nazionale per le Malattie Infettive I.R.C.C.S. ‘Lazzaro Spallanzani’, Via Portuense 292, 00149 Rome, Italy Laboratorio Interdipartimentale di Microscopia Elettronica, Universita` ‘Roma Tre’, Via della Vasca Navale 79, 00146 Rome, Italy b
c
Received 25 July 2005; received in revised form 1 September 2005; accepted 7 September 2005 Available online 22 November 2005
Abstract Cyanide is a major environmental pollutant of the chemical and metallurgical industries. Although extremely toxic, cyanide can enzymatically be converted to the less toxic thiocyanate by rhodaneses (thiosulfate:cyanide sulfurtransferases, EC 2.8.1.1). We engineered a genetic system to express high levels of recombinant Pseudomonas aeruginosa rhodanese (r-RhdA) in Escherichia coli, and used this organism to test the role of r-RhdA in cyanide detoxification. Inducible expression of the rhdA gene under the control of the hybrid T7-lacO promoter yielded active r-RhdA over a 4-h period, though r-RhdA-expressing E. coli showed decreased viability starting from 1 h post-induction. At this time, Western blot analysis and enzymatic assay showed r-RhdA partition between the cytoplasm (95%) and the periplasm (5%). The accessibility of thiosulfate to r-RhdA was a limiting step for the sulfur transfer reaction in the cellular system, but cyanide conversion to thiocyanate could be increased upon permeabilization of the bacterial membrane. Specific r-RhdA activity was higher in the whole-cell assay than in the in vitro assay with pure enzyme (2154 vs. 816 lmol min1 mg1 r-RhdA, respectively), likely reflecting enzyme stability. The r-RhdA-dependent cyanide detoxification resulted in increased resistance of r-RhdA overexpressing E. coli to 5 mM cyanide. Bacterial survival was paralleled by release of thiocyanate into the medium. Our results indicate that cyanide detoxification by engineered E. coli cells is feasible under laboratory conditions, and suggest that microbial rhodaneses may contribute to cyanide transformation in natural environments. 2005 Elsevier Ltd. All rights reserved. Keywords: Bioremediation; Protein expression; Pseudomonas aeruginosa; Sulfurtransferase
1. Introduction Hydrogen cyanide (HCN) is produced on large scale worldwide (e.g., 1 million tons in United States yearly) to fulfill the demand of major industrial countries. Main * Corresponding author. Address: Dipartimento di Biologia, Universita` ÔRoma TreÕ, Viale G. Marconi 446, 00146 Rome, Italy. Tel.: +39 06 5517 6347; fax: +39 06 5517 6321. E-mail address:
[email protected] (P. Visca).
users of cyanides are steel, electroplating, mining, and chemical industries. Potassium and sodium cyanide are massively employed in industrial operations, particularly for the recovery of precious metals (e.g., gold and silver) from mineral ores (ATSDR, 1997). Cyanide is extremely toxic to aerobic forms of life as it tightly binds cytochrome oxidase thereby inhibiting respiration (Solomonson, 1981). In mammals the lethal dose of cyanide is in the range 0.5–3.5 mg kg1 body weight (ATSDR, 1997). Therefore, cyanide release from
0045-6535/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2005.09.048
R. Cipollone et al. / Chemosphere 63 (2006) 942–949
industries represents a serious threat for environmental health, and cyanide-containing wastes must appropriately be treated before being discharged. An accidental catastrophe resulting from uncontrolled release of cyanide in the environment occurred in January 2000 when huge amounts of cyanide spilled from an uncovered industrial reservoir swept through the Hungarian plains coming from the Szamos and Tisza rivers, continuing the way to the Danube and the Black Sea. In that circumstance, no cases of occupational poisoning were reported but an unprecedented near-total death of all life forms took place. Detoxification efforts remained futile, but the accident prompted an intense debate on the efficacy of remediation strategies (Gacsi et al., 2005, and references therein). Current chemical methods for treatment of cyanide-polluted wastewater include alkaline chlorination, ozonation, wet-air oxidation, and sulfur-based technologies. Each one of these technologies has a relatively high cost and challenges the environment for the release of chemical agent(s) potentially causing secondary pollution (Akcil and Mudder, 2003). Despite being a metabolic inhibitor, cyanide and chemically related compounds are also synthesized, excreted and degraded in nature by hundreds of species of bacteria, algae, fungi, plants and insects (Knowles, 1976). Fungi and bacteria are the principal cyanide producers. In certain ecosystems (e.g., the soil and the rhizosphere) cyanogenic microorganisms represent a large part (up to 50%) of the microbial community (Kremer and Souissi, 2001). Consequently, microorganisms have evolved a multiplicity of resistance and/or assimilation strategies to face with the toxicity of either endogenous or exogenous cyanide. Rhodaneses (thiosulfate:cyanide sulfurtransferases or TSTs; EC 2.8.1.1) are highly conserved and widespread enzymes, currently regarded as one of the mechanism evolved for cyanide detoxification (Raybuck, 1992). In vitro rhodaneses catalyze the irreversible transfer of a sulfane sulfur atom from a suitable donor (i.e., thiosulfate) to cyanide, leading to formation of less toxic sulfite and thiocyanate. Kinetic studies support a scheme where the overall catalysis occurs in two steps in which the enzyme cycles between two intermediates, the sulfur-free and the persulfurated forms, the Cys catalytic residue undergoing the (de)persulfuration cycle (Alexander and Volini, 1987; Bordo et al., 2001; Cipollone et al., 2004). The enzyme activity is modulated by phosphate ions and divalent anions that have been shown to interact with the active site (Alexander and Volini, 1987; Bordo et al., 2001). The role of rhodanese in cyanide detoxification is supported by the high concentration of this enzyme in some mammalian tissues and organs (e.g., the liver) exposed to cyanide (Sylvester and Sander, 1990). Rhodaneses have been also identified in a variety of bacterial species including Escherichia coli (Alexander and Volini,
943
1987), Azotobacter vinelandii (Colnaghi et al., 1996) and several species of Thiobacillus (Westley, 1973), but in prokaryotes the physiological function of the enzyme has not been univocally addressed. Constitutive rhodanese activity has been also documented in the cyanideproducing bacterium Pseudomonas aeruginosa (Ryan and Tilton, 1977), and a rhodanese (RhdA, PA4956 at www.pseudomonas.com) has recently been cloned and biochemically characterized (Cipollone et al., 2004). Here, we investigated the possibility of using recombinant (r-) RhdA produced by E. coli to detoxify cyanide from water solutions. Factors affecting the enzyme expression and the catalytic properties of r-RhdA, as well as the effect of r-RhdA on viability of cyanideexposed E. coli cells were also considered.
2. Materials and methods 2.1. Gene cloning, protein expression and cell fractionation To clone the P. aeruginosa rhdA gene under the control of the T5-lacO promoter, ORF PA4956 (rhdA at http://www.pseudomonas.com) was amplified by PCR with primers 5 0 -GGGGGGATCCGCATGTCCGTTTTCTCCGA-3 0 and 5 0 -GGGCAAGCTTCCTCAAACCTCTACAGGGG-3 0 , using P. aeruginosa PAO1 genomic DNA as the template. Primers were designed to introduce BamHI and HindIII restriction sites (underlined) at the 5 0 - and 3 0 -ends, respectively. The digested PCR product was directionally cloned in pQE-32 (Qiagen), downstream of the His6-coding sequence and checked by DNA sequencing. The resulting plasmid, pQErhdA, was introduced in E. coli SG13009 carrying the pREP4 repressor plasmid (lacIq) by transformation (Gottesman et al., 1981). Cloning of P. aeruginosa rhdA gene under the control of the T7-lacO promoter, downstream of the His6-coding sequence into the expression vector pET-28a (Novagen) has previously been described (Cipollone et al., 2004). The resulting plasmid, pETrhdA, was introduced into the host strain E. coli BL21(DE3) by transformation (Sambrook et al., 1989). For protein expression, both SG13009(pREP4)(pQErhdA) and BL21(DE3)(pETrhdA) were grown in Luria-Bertani (LB) medium containing appropriate antibiotics [100 lg ml1 ampicillin for SG13009(pREP4)(pQErhdA); 25 lg ml1 kanamycin for BL21(DE3)(pETrhdA)] at 37 C until OD600 = 0.60. Protein expression was usually obtained by 4-h induction with 0.1 mM isopropyl thio-b-D-galactopyranoside (IPTG) unless otherwise stated. The recombinant His6-tagged RhdA (r-RhdA) was purified from E. coli BL21(DE3)(pETrhdA) cells as previously described (Cipollone et al., 2004).
944
R. Cipollone et al. / Chemosphere 63 (2006) 942–949
Rhodanese produced in inclusion bodies, named ri-RhdA, was extracted from E. coli SG13009(pREP4)(pQErhdA) cells in 0.1 M sodium phosphate buffer containing 6.0 M guanidine–HCl (pH 8.0), and purified by nickel(II)-chelate affinity chromatography under denaturing conditions using a stepwise (8.0–4.0) pH gradient in 8.0 M urea, 10 mM Tris–HCl according to the manufacturerÕs protocol (Qiagen). Renaturation of ureasolubilized ri-RhdA was performed at 4 C by either extensive dialysis against 10 mM Tris–HCl, pH 7.3, or slow discontinuous dilution in 50 mM Tris–HCl, pH 7.3. The periplasmic fraction of E. coli was obtained by osmotic shock (Ausubel et al., 1989). Briefly, the cell pellet was pre-treated for 10 min at 25 C with 20% sucrose in 30 mM Tris–HCl, 1 mM EDTA, pH 8.0, then centrifuged and suspended in ice-cold 5 mM MgSO4 for 10 min. After centrifugation at 10 000g for 10 min, the supernatant containing the shock-sensitive periplasmic fraction was recovered. Protein expression and localization were determined by running cell fractions in 12% (w/v) SDS-PAGE with Coomassie blue staining (Sambrook et al., 1989). Immunoblot analysis was performed with either murine anti-RGS(H)4 antibody (Qiagen) or the polyclonal anti-RhdA antiserum (1:1000), using alkaline phosphatase-conjugate anti-mouse IgG secondary antibodies (1:7500) (Calbiochem) as described elsewhere (Sambrook et al., 1989). For quantitative estimation, the immunoblots were calibrated using known quantities of the purified enzyme. 2.2. Polyclonal antiserum production Purified ri-RhdA (430 lg) was emulsioned with complete FreundÕs adjuvant (Sigma) and used to immunize a female Balb/c mouse by intramuscular injection. After 5 weeks, a second boost of ri-RhdA (120 lg) in incomplete FreundÕs adjuvant (Sigma) was given, followed by a third boost (40 lg in saline) at the seventh week. The animal was bled 2 weeks later and the serum (ca. 2 ml) was stored at 4 C. Animal experiments were performed according to the legislative decree 116/92 by the Italian Ministry of Health. 2.3. Determination of protein and cyanide concentration Total soluble protein concentrations of the extracts were determined by a colorimetric assay (Bradford, 1976) using bovine serum albumin as the standard. The concentration of pure r-RhdA was estimated using the extinction coefficient (e280 nm = 58 320 M1 cm1), calculated according to the deduced amino acid composition (Cipollone et al., 2004), and verified colorimetrically (Bradford, 1976). Cyanide concentration was determined by stoichiometric titration with human hemoglobin (Cipollone et al., 2004).
2.4. Enzyme activity assay The enzyme activity was determined in 50 mM Tris–HCl, pH 7.3, at 25 C, using either whole cells or cell-free enzyme preparations. The estimated enzyme amount ranged between 12 and 210 nM, and cyanide concentration ranged between 0.3 and 68 mM. The reaction product (thiocyanate) was determined by the So¨rbo method (So¨rbo, 1953). Product formation was linear with the time assay (3 min). The linearity of product appearance with time and its direct proportionality to the enzyme concentration was tested between 0 C and 60 C, and in the presence of 1–150 mM KH2PO4. The spontaneous reaction between sulfur donor and cyanide was taken into account in the determination of catalytic parameters. Steady-state kinetics has been analyzed according to the classical Michaelis-Menten equation and data were fitted using the GraFit 5.0 software (Erithacus Software). 2.5. Determination of E. coli viability upon r-RhdA expression and cyanide treatment A stationary-phase culture of E. coli BL21(DE3)(pETrhdA) was diluted in LB medium to OD600 = 0.10, then grown at 37 C with vigorous shaking. At OD600 = 0.60, the culture was induced with 0.1 mM IPTG for 4 h. At hourly intervals, the bacterial viability was assessed by counting the colony forming units (cfu) according to the plate dilution method. Cyanide susceptibility of E. coli BL21(DE3)(pET28a) was preliminarily determined after 1-h IPTG induction. Induced cultures were centrifuged at 6000g for 10 min and suspended at OD600 = 0.50 (corresponding to 1.2 · 107 cfu ml1) in IPTG-free fresh LB medium containing potassium cyanide between 0.05 and 10 mM. Cell suspensions were incubated statically at 37 C, and the OD600 was measured during 1 h. Viable counts were performed 40 min after exposure to cyanide. Cyanide-free cultures were used as the control. The effect of r-RhdA expression on cyanide susceptibility of E. coli BL21(DE3)(pETrhdA) was determined in LB supplemented with 5 mM potassium cyanide at 37 C, using both induced (1 h, 0.1 mM IPTG) and uninduced cultures. For relative growth calculation, viable counts were performed 20, 40 and 60 min after cyanide treatment. To quantify the extent of r-RhdA protection from cyanide toxicity, the ratio (l) of viable count from a culture subjected to cyanide (lcyanide) to the respective control culture (lcyanide-free) was calculated for both induced and uninduced E. coli BL21(DE3)(pETrhdA) cells. The same procedure was followed in experiments where 10 mM thiosulfate (2:1 thiosulfate:cyanide ratio) was added to cyanide-containing cell suspensions. Experiments were repeated at least three times, with assays performed in duplicate.
R. Cipollone et al. / Chemosphere 63 (2006) 942–949
One-way analysis of variance (ANOVA) and DunnettÕs multiple comparison test were applied for statistical analysis.
3. Results 3.1. E. coli expression systems for production of recombinant rhodanese In a preliminary attempt to obtain high-level expression of active rhodanese from E. coli SG13009(pREP4)(pQErhdA), the majority of ri-RhdA was recovered in the insoluble fraction (Fig. 1, compare lanes 3 and 4) and the corresponding rhodanese activity in the crude lysate was only 1.5-fold higher than the uninduced control (0.064 vs. 0.042 lmol min1 mg1 total proteins). Attempts to increase the intracellular solubility of the protein by testing different media and culture conditions (i.e., temperature, aeration, inoculum size, IPTG concentration and incubation time) negligibly affected riRhdA recovery from the soluble cell fraction (data not
945
shown). However, inclusion bodies provided relatively pure protein even prior to purification, attaining ca. 15% of the total cellular proteins (Fig. 1, lane 2). Affinity-chromatography purification of ri-RhdA from solubilized inclusion bodies yielded greater than 95% pure protein which eluted in 8 M urea in a pH range between 5.9 and 4.0. Protein identity and purity were assessed by SDS-PAGE and Western blot analysis with a commercial anti-RGS(H)4 antibody (Fig. 1a and b, lane 5). Repeated attempts at ri-RhdA refolding failed since very low specific activity (0.2–0.3 lmol min1 mg1) was obtained in all instances. However, ri-RhdA purified from inclusion bodies served as the source of highly purified antigen for production of polyclonal antibodies. Expression of the rhdA gene in E. coli BL21(DE3)(pETrhdA) resulted in high-level production of a soluble 32-kDa protein upon 0.1 mM IPTG-induction for 4 h at 37 C (Fig. 1a, lanes 7 and 8). The overexpressed protein was localized in the soluble cell fraction (Fig. 1a, compare lanes 8 and 9) and could be purified to homogeneity by single-step affinity chromatography under native conditions (Fig. 1a, lane 10). Immunoblot analysis with antibodies against ri-RhdA recognized a 32-kDa band, thereby confirming the identity of r-RhdA (Fig. 1b, lanes 7–10). Coherently, the rhodanese activity was 450-fold increased in lysates of IPTG-induced E. coli BL21(DE3)(pETrhdA) cells relative to the uninduced control (91 vs. 0.2 lmol min1 mg1 total proteins), and the purified protein showed high specific activity (816 lmol min1 mg1 r-RhdA). 3.2. Catalytic properties of r-RhdA-producing E. coli cells
Fig. 1. Expression and purification of recombinant rhodanese in E. coli. (a) SDS-PAGE and (b) Western blot analysis of enzyme preparations. The gel was stained with Coomassie brilliant blue, then blotted onto a nitrocellulose filter and hybridized with either the RGS(H)4 monoclonal antibodies (lanes 1–5) or the polyclonal antiserum raised against ri-RhdA (lanes 6–10). M, molecular mass marker with sizes shown on the left; lanes 1–2 and 6–7, total cellular proteins (30 lg each) from uninduced (lanes 1 and 6) and IPTG-induced (lanes 2 and 7) E. coli SG13009(pREP4)(pQErhdA) or E. coli BL21(DE3)(pETrhdA), respectively; lanes 3–4 and 8–9 soluble fractions (lanes 3 and 8; 20 lg proteins each) and insoluble fractions (lanes 4 and 9; 10 lg proteins each) from IPTG-induced E. coli SG13009(pREP4)(pQErhdA) or E. coli BL21(DE3)(pETrhdA), respectively; lane 5, ri-RhdA purified under denaturing conditions (2 lg); lane 10, r-RhdA purified under native conditions (2 lg). The arrow indicates overexpressed RhdA. For details, see text.
The specific r-RhdA activity in whole cells of E. coli BL21(DE3)(pETrhdA) after 4-h induction was 1960 lmol min1 mg1 r-RhdA, and the final enzyme yield was ca. 32 r-RhdA lg mg1 bacterial proteins, as determined by Western blot analysis (Fig. 2). However, induction of r-RhdA expression caused progressive bacterial death, attaining 2.7 Log reduction of viable cells (i.e., cfu counts) in 4 h. Nevertheless, reduction of cell viability upon r-RhdA overexpression was negligible at 1 h postinduction (Fig. 2a). At this time, uninduced and induced cultures showed identical OD600 values to which corresponded similar cfu counts (1.25 · 108 and 1.1 · 108, respectively). Assuming that the viable count of the uninduced culture is a reliable estimation of total (both viable and dead) induced cells, the production of r-RhdA at 1 h post-induction was ca. 6 · 107 r-RhdA molecules cell1 (Fig. 2b, lane 7). r-RhdA activity at 1 h post-induction was mainly recovered from the cell-associated fractions. Western blot analysis and enzymatic assay on subcellular fractions showed partition of r-RhdA between the periplasm
946
R. Cipollone et al. / Chemosphere 63 (2006) 942–949
Fig. 2. (a) Time-course of r-RhdA expression in E. coli BL21(DE3)(pETrhdA): effect on cell viability (—s—; ni cfu, viable counts of uninduced cells; i cfu, viable counts of IPTGinduced cells), and yields of cell-associated (—n—) and extracellular (—m—) rhodanese activity. (b) quantitative immunoblot analysis of r-RhdA levels in E. coli BL21(DE3)(pETrhdA) at 1 h post-induction. Using fixed amounts of proteins from total cell lysates and periplasmic fractions, the relative densitometric intensities of the immunoblot signals were compared with the signal generated by known enzyme quantities from a serial dilutions of purified r-RhdA. Lanes 1–5, purified r-RhdA (1600, 800, 400, 200 and 100 ng, respectively); lanes 6 and 7, total cellular proteins (3 lg each) from uninduced and IPTG-induced E. coli BL21(DE3)(pETrhdA), respectively; lane 8, periplasmic fraction (30 lg) from IPTG-induced cells. For details, see text.
(ca. 5%) and the cytoplasm (ca. 95%) (Fig. 2b, compare lanes 7 and 8, then normalize by the different quantities of loaded proteins). After prolonged induction (4 h), rhodanese activity could also be detected in the culture supernatant (6 lmol min1 mg1 total proteins), likely as the result of cell death and lysis (Fig. 2a). Irrespective of its localization, r-RhdA was similarly active in catalyzing the sulfur transfer reaction, being the specific activity 819 and 731 lmol min1 mg1 r-RhdA for cytoplasmic and periplasmic enzyme, respectively. Interestingly, r-RhdA activity assay on whole cells at 1 h post-induction showed K KCN ¼ 3:5 mM and Vmax = m 88 lmol min1 mg1, at 25 C (Fig. 3, inset). A significant increase in Vmax (2154 lmol min1 mg1) was observed after permeabilization of cells with toluene (Fig. 3), attaining a value similar to that observed for non-permeabilized cells 4 h post-induction (1960 lmol min1 mg1 r-RhdA). In all cases, no activity was detected in the absence of thiosulfate in the reaction mixture.
Fig. 3. Effect of cyanide concentration on the initial velocity (vi) for the r-RhdA catalyzed sulfur-transfer reaction in intact cells (—s—; see inset) and toluene-permeabilized cells (—d—), at 25 C and pH 7.3. The continuous lines were calculated according to the Michaelis-Menten equation with values of catalytic parameters given in the text.
The effect of temperature and phosphate ions on r-RhdA catalytic properties was also investigated. The enzyme showed a temperature-dependent activity with a maximum between 20 C and 32 C (Fig. 4). At 25 C and pH 7.3, r-RhdA activity in permeabilized bacteria was inhibited by high phosphate concentrations, being 1130 lmol min1 mg1 and 2154 lmol min1 mg1 in 50 mM KH2PO4 and 50 mM Tris–HCl, respectively. Phosphate inhibition was not observed at 630 mM KH2PO4.
Fig. 4. Temperature dependence of r-RhdA catalyzed (—s—) and abiotic (—d—) sulfur-transfer reaction. The 100% activity value corresponds to 816 lmol min1 mg1 r-RhdA. For details, see text.
R. Cipollone et al. / Chemosphere 63 (2006) 942–949
3.3. r-RhdA-dependent cyanide detoxification In order to estimate to which extent r-RhdA would contribute to cyanide detoxification in vivo, growth assays were performed with induced and uninduced E. coli BL21(DE3)(pETrhdA) cells exposed to inhibitory cyanide concentrations. As previously reported, P1 mM cyanide causes significant inhibition of aerobically growing E. coli (Weigel and Englund, 1975). In a preliminary test, E. coli BL21(DE3)(pET-28a) showed dose-dependent growth inhibition during 40-min exposure to cyanide concentrations P1 mM. Hence, a potassium cyanide concentration of 5 mM was chosen which caused ca. 70% reduction of viable cells after 40-min exposure (data not shown). To assess any protective effect of r-RhdA against cyanide toxicity, E. coli BL21(DE3)(pETrhdA) cells were challenged with 5 mM cyanide 1 h after IPTG induction. Bacteria showed r-RhdA-dependent cyanide detoxification, as deduced by 100% viability of the IPTG-induced culture relative to uninduced control at 20 min postexposure. At this time, the maximum protection exerted by r-RhdA was recorded in the presence of 10 mM thiosulfate (Fig. 5), i.e. when the sulfur donor was exogenously supplemented. Bacterial survival was paralleled by release of thiocyanate (the reaction product) as inferred by colorimetric detection in the cell suspension (data not shown). However, reduced r-RhdA-dependent
Fig. 5. Effect of potassium cyanide on E. coli BL21(DE3)(pETrhdA) viability in the absence or in the presence of 10 mM thiosulfate (+ Na2S2O3). The l values indicate the ratios of growth of induced (+ IPTG) or uninduced ( IPTG) E. coli exposed to 5 mM cyanide to the corresponding cyanide-free culture. l of the same cultures treated with 10 mM thiosulfate are also reported (for details, see text). Results are the mean of at least three independent experiments performed in duplicate. Standard deviation bars are shown. For statistical significance, data were analysed by ANOVA (F = 10.85) and compared to data for the control culture by DunnettÕs multiple comparison test (confidence interval, 99%). At each time point, letters above the bars denote significant differences (P 6 0.05) between induced and uninduced cultures.
947
detoxification was observed 40 and 60 min post-exposure, with bacterial survival rates that resembled those obtained in the absence of thiosulfate. Increasing thiosulfate concentrations did not enhance bacterial survival (data not shown).
4. Discussion Enzymes involved in protection against cyanide include b-cyanoalanine synthase, c-cyano-a-aminobutyric acid synthase, cyanide hydratase, and sulfurtransferases, namely rhodanese and 3-mercaptopyruvate sulfurtransferase (Raybuck, 1992). While sulfurtransferases appear to be a primary strategy for cyanide detoxification in mammals (Sylvester and Sander, 1990), a similar role in prokaryotes has not yet been documented. The cyanogenic properties of P. aeruginosa led us to investigate the role of RhdA in cyanide detoxification. For this purpose, we developed a suitable r-RhdA expression system and used E. coli as a prokaryotic host to test the role of r-RhdA in protection from cyanide toxicity. Two different expression systems made it possible to overproduce recombinant P. aeruginosa rhodanese in E. coli, but only one system yielded a soluble and catalytically active enzyme. In general, recombinant protein recovery in the soluble cell fraction depends on a variety of factors, including host metabolism, differential biosynthesis rates, and chaperonin-assisted protein folding, which ultimately affect the equilibrium between in vivo protein aggregation and solubilization (reviewed in Sørensen and Mortensen, 2005). Given the critical role of chaperonins in building up mature rhodanese (Weber and Hayer-Hartl, 2000), we speculate that an unbalance between the rate of gene expression (transcription 7and translation) and assisted folding could have led to formation of ri-RhdA aggregates in E. coli SG13009(pREP4)(pQErhdA). Noticeably, the specific activity of r-RhdA was higher in permeabilized E. coli cells than in Tris-buffered solution at cytosolic pH (2154 vs. 816 lmol min1 mg1 r-RhdA, respectively). This difference is likely to reflect the micro-environment in which the sulfur transfer reaction takes place, since compartmentalization in the reducing cytoplasmic environment could prevent oxidative inactivation of the enzyme (Westley, 1973). However, r-RhdA was also present in the periplasm, albeit to a much lesser extent than in the cytoplasm. This is consistent with the observation that rhodanese activity was associated with the periplasmic fraction in P. aeruginosa (Ryan et al., 1979). The accessibility of thiosulfate to r-RhdA is a limiting step for the sulfur transfer reaction in the cellular system. While cyanide is highly diffusible (ATSDR, 1997), thiosulfate cannot permeate bacterial membrane, and requires an ABC-type system for active transport in
948
R. Cipollone et al. / Chemosphere 63 (2006) 942–949
E. coli (Sirko et al., 1995). Because active transport is induced by sulfur starvation, the r-RhdA activity observed in intact E. coli cells exposed to an excess of thiosulfate could be ascribed to the periplasmic fraction of the enzyme. In fact, membrane permeabilization enabled the sulfur donor request of the cytoplasmic r-RhdA fraction to be completely fulfilled, and this resulted in maximum enzyme activity. It should also be mentioned that r-RhdA overexpressed in E. coli is completely persulfurated (Cipollone et al., 2004) indicating that the sulfur pool is physiologically funneled towards r-RhdA during enzyme biogenesis. However, this process does not ensure adequate sulfur supply to sustain the catalytic cycle in intact cells. In our experimental conditions, the uppermost thiocyanate concentration achievable from a single cycle of sulfur transfer to cyanide would be 23 nM, a concentration 40 000-fold below the lower detection limit of rhodanese colorimetric assay. This emphasizes the need of exogenous thiosulfate supply for the r-RhdA reaction in the cellular system. The r-RhdA detoxification capability is reflected by the increased viability of r-RhdA producing E. coli cell in the presence of cyanide. Our data suggest that r-RhdA overexpressing E. coli has the potential of being used as an active cyanide-degrading cell factory, and that the loss of cyanide toxicity to E. coli is the result of an r-RhdA-driven process rather than the effect of cyanide natural attenuation (e.g., volatilization). However, the detoxification activity exerted by r-RhdA was restricted to a short time (i.e., 20 min), possibly as the result of inefficient enzyme recycling to the persulfurated state. Indeed, the in vitro thiosulfate-dependent persulfuration of r-RhdA is kinetically slow by comparison with the sulfur transfer reaction (Cipollone et al., 2004). In view of a possible application of the system presented in this study, critical parameters affecting microbial processes, such as nutrient availability and temperature have also been considered. Phosphate is a limiting nutrient for bacteria in natural environment. In gold mill effluents, where microbiological detoxification is successfully applied, phosphate is conveniently added in concentrations ranging from about 1 to 5 mg l1 using concentrated phosphoric acid, to promote bacterial growth (Akcil and Mudder, 2003). On the other hand, high phosphate concentrations have been reported to inhibit A. vinelandii rhodanese (Bordo et al., 2001) which shares a 79% identity with r-RhdA (Cipollone et al., 2004). While phosphate could reduce the rate of cyanide degradation below the optimum, the phosphate level required to sustain bacterial nutrition (ca. 10 mM) is lower than the observed r-RhdA inhibitory concentration (P50 mM). Although some rhodaneses, namely beef liver and tapioca leaf rhodaneses, show optimal activity at P50 C, r-RhdA shows optimum activity at mesophylic
temperature, as reported for other prokaryotic sources of the enzyme (www.brenda.uni-koeln.de) and, above all, under conditions in which the non-enzymatic reaction between thiosulfate and cyanide would be too slow to concur to cyanide detoxification. The r-RhdA-dependent thiocyanate production implies ca. 150-fold toxicity reduction (in rats, the oral LD50 is 3–8 mg kg1 for cyanide vs. 854 mg kg1 for thiocyanate) (ATSDR, 1997) but the possibility of causing secondary pollution has to be considered. To date, no information on the ecological effect of thiocyanate is available (ATSDR, 1997). However, the r-RhdA-mediated cyanide detoxification process might conceivably be integrated by the association of r-RhdA overexpressing bacteria with thiocyanate-degrading microrganisms which are known to convert thiocyanate to ammonia (Yamasaki et al., 2002). The development of syntrophic bacterial consortia whose members cooperate for a complete biodegradative process is a strategy that has recently found successful application (Yu et al., 2005). Taken together, our results open the way to a possible exploitation of r-RhdA for an environmentally friendly and kinetically advantageous bioremediation process of cyanide-polluted fluids. To our best knowledge, this is the first study considering a possible application of rhodanese to environmental biotechnology, with the exception of those aimed to the development of cyanide-specific biosensors (Mattiasson and Mosbach, 1977; Fonong, 1987; Ikebukuro et al., 1996). The construction of E. coli strains genetically engineered for bioremediation purposes has been already reported for environmental cleanup of Hg(II) and Ni(II) (Chen and Wilson, 1997; Krishnaswamy and Wilson, 2000). Our results indicate that cyanide detoxification through r-RhdA overexpressing E. coli is a feasible process, at least under laboratory conditions, and suggest that microbial rhodaneses can contribute to cyanide transformation in natural environments. Engineering endemic bacterial species (e.g., environmental pseudomonads) for r-RhdA production would lower the ecological impact of the process and provide a promising alternative to chemical or physical cyanide treatment.
Acknowledgement This work was supported by grants from ISPESL, MIUR-COFIN and Ministero della Salute-Ricerca Corrente to P.V.
References Agency for Toxic Substances and Disease Registry (ATSDR), 1997. Toxicological Profile for Cyanide. US Department of Health and Human Services, Atlanta, GA.
R. Cipollone et al. / Chemosphere 63 (2006) 942–949 Akcil, A., Mudder, T., 2003. Microbial destruction of cyanide wastes in gold mining: process review. Biotechnol. Lett. 25, 445–450. Alexander, K., Volini, M., 1987. Properties of an Escherichia coli rhodanese. J. Biol. Chem. 262, 6595–6604. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K., 1989. Current Protocols in Molecular Biology vol. 2. John Wiley & Sons, New York. Bordo, D., Forlani, F., Spallarossa, A., Colnaghi, R., Carpen, A., Bolognesi, M., Pagani, S., 2001. A persulfurated cysteine promotes active site reactivity in Azotobacter vinelandii rhodanese. Biol. Chem. 382, 1245–1252. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248– 254. Chen, S., Wilson, D.B., 1997. Genetic engineering of bacteria and their potential for Hg(II) bioremediation. Biodegradation 8, 97–103. Cipollone, R., Bigotti, M.G., Frangipani, E., Ascenzi, P., Visca, P., 2004. Characterization of a rhodanese from the cyanogenic bacterium Pseudomonas aeruginosa. Biochem. Biophys. Res. Comm. 325, 85–90. Colnaghi, R., Pagani, S., Kennedy, C., Drummond, M., 1996. Cloning, sequence analysis and overexpression of the rhodanese gene of Azotobacter vinelandii. Eur. J. Biochem. 236, 240–248. Fonong, T., 1987. Enzyme method for the spectrophotometric determination of micro-amounts of cyanide. Analyst 112, 1033–1035. Gacsi, M., Czegeny, I., Nagy, G., Banfalvi, G., 2005. Survival of fish upon removal of cyanide from water. Environ. Res. 97, 293–299. Gottesman, S., Halpern, E., Trisler, P., 1981. Role of sulA and sulB in filamentation by lon mutants of Escherichia coli K12. J. Bacteriol. 148, 265–273. Ikebukuro, K., Shimomura, M., Onuma, N., Watanabe, A., Nomura, Y., Nakanishi, K., Arikawa, Y., Karube, I., 1996. A novel biosensor system for cyanide based on a chemiluminescence reaction. Anal. Chim. Acta 329, 111–116. Knowles, C.J., 1976. Microorganisms and cyanide. Bacteriol. Rev. 40, 652–680. Kremer, R.J., Souissi, T., 2001. Cyanide production by rhizobacteria and potential for suppression of weed seedling growth. Curr. Microbiol. 43, 182–186. Krishnaswamy, R., Wilson, D.B., 2000. Construction and characterization of an Escherichia coli strain genetically
949
engineered for Ni(II) bioaccumulation. Appl. Environ. Microbiol. 66, 5383–5386. Mattiasson, B., Mosbach, K., 1977. Application of cyanidemetabolizing enzymes to environmental control; enzyme thermistor assay of cyanide using immobilized rhodanese and injectase. Biotechnol. Bioeng. 19, 1643–1651. Raybuck, S.A., 1992. Microbes and microbial enzymes for cyanide degradation. Biodegradation 3, 3–18. Ryan, R.W., Tilton, R.C., 1977. The isolation of rhodanese from Pseudomonas aeruginosa by affinity chromatography. J. Gen. Microbiol. 103, 197–199. Ryan, R.W., Gourlie, M.P., Tilton, R.C., 1979. Release of rhodanese from Pseudomonas aeruginosa by cold shock and its localization within the cell. Can. J. Microbiol. 25, 340– 351. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning. Cold Spring Harbor Laboratory Press, New York. Sirko, A., Zatyka, M., Sadowy, E., Hulanicka, D., 1995. Sulfate and thiosulfate transport in Escherichia coli K-12: evidence for a functional overlapping of sulfate- and thiosulafatebinding proteins. J. Bacteriol. 177, 4134–4136. Solomonson, L.P., 1981. Cyanide as a metabolic inhibitor. In: Vannesland, B., Conn, E.E., Knowles, C.J., Westley, J., Wissing, F. (Eds.), Cyanide in Biology. Academic Press, London and New York, pp. 11–28. So¨rbo, B.H., 1953. Crystalline rhodanese. I. Purification and physicochemical examination. Acta Chem. Scand. 7, 1129– 1136. Sørensen, H.P., Mortensen, K.K., 2005. Advances genetic strategies for recombinant protein expression in Escherichia coli. J. Biotechnol. 115, 113–128. Sylvester, M., Sander, C., 1990. Immunohistochemical localization of rhodanese. Histochem. J. 22, 197–200. Weber, F., Hayer-Hartl, M., 2000. Prevention of rhodanese aggregation by the chaperonin GroEL. Methods Mol. Biol. 140, 111–115. Weigel, P.H., Englund, P.T., 1975. Inhibition of DNA replication in Escherichia coli by cyanide and carbon monoxide. J. Biol. Chem. 250, 8536–8542. Westley, J., 1973. Rhodanese. Adv. Enzymol. Relat. Areas Mol. Biol. 39, 327–368. Yamasaki, M., Matsushita, Y., Namura, M., Nyunoya, H., Katayama, Y., 2002. Genetic and immunochemical characterization of thiocyanate-degrading bacteria in lake water. Appl. Environ. Microbiol. 68, 942–946. Yu, S.H., Ke, L., Wong, Y.S., Tam, N.F., 2005. Degradation of polycyclic aromatic hydrocarbons by a bacterial consortium enriched from mangrove sediments. Environ. Int. 31, 149–154.
