Tellurite effects on Rhodobacter capsulatus cell viability and superoxide dismutase activity under oxidative stress conditions

Research in Microbiology 156 (2005) 807–813 www.elsevier.com/locate/resmic

Tellurite effects on Rhodobacter capsulatus cell viability and superoxide dismutase activity under oxidative stress conditions Francesca Borsetti a , Valentina Tremaroli a , Francesca Michelacci a , Roberto Borghese b , Christine Winterstein b , Fevzi Daldal b , Davide Zannoni a,∗ a Department of Biology, Microbiology Unit, University of Bologna, 40126 Bologna, Italy b Department of Biology, University of Pennsylvania, Philadelphia, PA 19104-6018, USA

Received 10 January 2005; accepted 18 March 2005 Available online 23 May 2005

Abstract Cells of the facultative photosynthetic bacterium Rhodobacter capsulatus (MT1131 strain) incubated with 10 µg ml−1 of the toxic oxyanion tellurite (TeO2− 3 ) exhibited an increase in superoxide dismutase activity. The latter effect was also seen upon incubation with sublethal amounts of paraquat, a cytosolic generator of superoxide anions (O2·− ), in parallel with a strong increase in tellurite resistance (TeR ). A mutant strain (CW10) deficient in SenC, a protein with similarities to peroxiredoxin/thiol:disulfide oxidoreductases and a homologue of mitochondrial Sco proteins, was constructed by interposon mutagenesis via the gene transfer agent system. Notably, the absence of SenC affected R. capsulatus resistance to periplasmic O·2− generated by xanthine/xanthine oxidase but not to cytosolic O2·− produced by paraquat. Further, the absence of SenC did not affect R. capsulatus tellurite resistance. We conclude that: (1) cytosolic-generated O2·− enhances TeR of this bacterial species; (2) small amounts of tellurite increase SOD activity so as to mimic the early cell response to oxidative stress; (3) SenC protein is required in protection of R. capsulatus against periplasmic oxidative stress; and finally, (4) SenC protein is not involved in TeR , possibly because tellurite does not generate O·− 2 at the periplasmic space level.  2005 Elsevier SAS. All rights reserved. Keywords: Oxidative stress; Rhodobacter capsulatus; SenC protein; Superoxide dismutase; Tellurite

1. Introduction Tellurium oxyanions such as tellurite (TeO2− 3 ) are highly toxic toward both Gram-negative and -positive bacteria [33]. The majority of tellurite-resistant bacteria convert tellurite into less toxic elemental tellurium (Te0 ) which is accumulated intracellularly as black inclusions [33], although tellurite resistance (TeR ) without metal accumulation has also been observed in several phototrophic bacteria [39]. It has been proposed that tellurite toxicity results from the ability of this metalloid to act as a strong oxidizing agent toward any specific components of the cell [24,33]. Further, the possibility that free-radical production takes place as a byproduct * Corresponding author.

E-mail address: [email protected] (D. Zannoni). 0923-2508/$ – see front matter  2005 Elsevier SAS. All rights reserved. doi:10.1016/j.resmic.2005.03.011

of tellurite reduction has also been suggested [38]. Reactive oxygen species (ROS), such as hydrogen peroxide (H2 O2 ), superoxide radical anions O2·− and hydroxyl radicals (OH·), are typical unavoidable byproducts of aerobic metabolism [7,19] and also result from exposure of the cells to freeradical-generating compounds such as metals and metalloids [6,14]. Among the numerous enzymes required to protect cells against the deleterious effects of ROS, superoxide dismutases (SODs) are a ubiquitous class of metalloenzymes catalyzing the detoxification of the superoxide radical anion (O2·− ) to hydrogen peroxide (H2 O2 ) and molecular oxygen [14,17,31,32]. Recently, a class of proteins named Sco (Sco stands for synthesis of cytochrome c oxidase) has been shown to be involved in protection against oxidative stress in pathogenic Neissera spp. [28] as sco− mutants are highly sensitive to

808

F. Borsetti et al. / Research in Microbiology 156 (2005) 807–813

oxidative killing by paraquat, a cytosolic generator of superoxide anions [26]. Sco from yeast (Sco1 protein) [1] and its homologue from Rhodobacter sphaeroides (PrrC protein) [13] are able to bind Cu ions, in agreement with a role for these proteins in the assembling of the cytochrome c oxidase (COX) [11,23,25,27]. It has also been shown that the PrrC protein of R. sphaeroides (homologous of SenC in R. capsulatus) acts as an oxygen sensor [5,26]; however, secondary structure predictions for Sco, PrrC- and SenC-proteins reveal that they have similarities to peroxiredoxin/thiol:disulfide oxidoreductases [8] which might suggest an additional role in protecting cells against oxidative stress. The actual evidence for tellurite-induced imbalance of cellular homeostasis in favor of pro-oxidants is at present lacking. In this respect, facultative photosynthetic bacteria such as those belonging to the genera Rhodobacter represent an ideal system for testing the potential oxidative effect of tellurite, since they are sensitive to the oxyanion under aerobic conditions of growth, while strongly resistant under anaerobic/light growth conditions [2]. In this work we have examined the effect of ROS generators (paraquat and/or xanthine/xanthine oxidase) and/or tellurite on the cell viability of R. capsulatus MT1131 (wildtype strain) and its SenC− mutant derivative (CW10 strain). We show here that the senC product has no significant role in protecting R. capsulatus against cytosolic O2·− but it is necessary for protection against periplasmic O2·− . Notably, MT1131 cells treated with sublethal amounts of paraquat exhibited a strong increase in tellurite resistance, a phenomenon that is associated with an increase in SOD activity. As the latter phenomenon was observed upon exposure of cells to low amounts of tellurite (10 µg ml−1 ), it is therefore proposed that the effects of tellurite on R. capsulatus cells mimic the early cell responses to oxidative stress, as seen by increased SOD activity.

C. Bauer (Indiana University, Bloomington, IL, USA), constructed by deleting the coding region of SenC from amino acid position 6–178, and replacing it in the appropriate orientation with a kanamycin resistance cartridge that is not polar to the downstream genes [5], was used. This senC::kan allele thus obtained was introduced into MT1131 selecting for kanamycin-resistant colonies on MPYE-enriched medium (MgCl2 , 1 mM; CaCl2 , 1 mM; peptone, 0.3%; yeast extract, 0.3%) containing plates. Similarly, one KanR colony that can grow on MPYE and MedA [29] plates under both photosynthetic and respiratory conditions, and that shows an NADI slow (see below Section 2.4) staining phenotype [18] was retained and shown to contain the senC::kan allele on its chromosome by PCR analysis.

2. Materials and methods

2.4. Phenotype characterization

2.1. Bacterial strains and culture conditions R. capsulatus MT1131, CW10 (SenC− mutant, see Section 2.2) and GK32 (cbb3 oxidase minus mutant [18]) strains were grown aerobically in the dark at 30 ◦ C in YPS medium (yeast extract, 0.3%; peptone, 0.3%, CaCl2 , 2 mM; MgSO4 , 2 mM, in deionized water at pH 6.8) as described previously [40]. Cells were grown to early log phase (A660 = 0.3) and used within 2–4 h after harvesting. 2.2. Construction of the senC-deficient strain Strain SenC− (named CW10) is a derivative of R. capsulatus wild-type strain MT1131, and was obtained using interposon mutagenesis via the gene transfer agent GTA [30]. For this purpose a senC::kan allele kindly provided by

2.3. Cell viability test Oxidative killing assays using paraquat or xanthine/xanthine oxidase were performed essentially as described by Tseng et al. [36]. Paraquat (PQ2+ ; 1,1 -4,4 -bipyridinium dichloride) is a redox compound that is reduced to its free radical form (PQ·+ ) by low potential electron donors within the cell. The paraquat free radical is then oxidized by dioxygen, leading to generation of superoxide anions (O2·− ) in the cell cytoplasm. The xanthine/xanthine oxidase (X/Xox) assay, used to generate periplasmic O2·− , was performed using 2.15 mM xanthine (Sigma) and 0.15 U ml−1 of xanthine oxidase (Sigma). Cells were harvested at their early log phase of aerobic growth (OD660 = 0.3) and washed once. 104 cells were suspended in YPS medium containing variable concentrations of PQ2+ or X/Xox or tellurite. After incubation at 30 ◦ C with shaking at 150 rpm, 10 µl of the assay were plated at variable experimental time periods (0, 30 and 60 min). CFU counting on plates was done after 48 h of incubation at 30 ◦ C. Each experiment was performed in duplicate and repeated at least three times.

The NADI-test was performed by addition of α-naphthol and DMPD (N ,N -dimethyl-1,4-phenylene diamine monohydrochloride) to colonies, enabling visual detection of cytochrome c oxidase activity by production of a blue compound [21]. 2.5. Detection of SOD activity by non-denaturating PAGE R. capsulatus cells were grown aerobically in the dark to OD660 of 0.3, harvested, washed, and incubated with paraquat or xanthine/xanthine oxidase or tellurite for variable time periods. Then, cells were disrupted by ultrasonication (MSE Soniprep 150) at 4 ◦ C (three bursts of 40 s each, 20 s off, 4 power, 10 µm of amplitude, titanium exponential probe of 3 mm end diameter). Supernatants were cleared by centrifugation (100 000 g, 1 h) and total soluble protein was estimated by the method of Lowry [20] using

F. Borsetti et al. / Research in Microbiology 156 (2005) 807–813

809

bovine serum albumin (BSA) as a standard. Gel assays measuring SOD activity were performed by an in situ staining procedure [12,36], after electrophoresis of the corresponding cleared lysates in non-denaturating 8% polyacrylamide gels. Each lane contained approximately 35 µg of proteins and bands corresponding to SOD activity were analyzed by an Image Analyzer FLA-3000 (Fujifilm, Japan). 2.6. Detection of SOD activity by the xanthine/xanthine oxidase-cytochrome c assay The assay was performed by essentially following the method described by McCord and Fridovich [22], except that 15 µM horse-heart cytochrome c, 500 µM xanthine and 20 mU ml−1 of microbial xanthine/oxidase were used in the presence of 35 µg ml−1 of lysate-proteins. Reduction of cytochrome c was determined spectrophotometrically at 550 nm in the presence of 0.1 mM KCN to inhibit the cytochrome c oxidase activity. 2.7. Heme staining Membrane fragments (chromatophores) were isolated by ultracentrifugation after French-pressure cell disruption as described previously [40] and analyzed using a 16.5% sodium dodecyl sulfate–polyacrylamide gel electrophoresis, according to Koch et al. [18]. C-type hemes were revealed according to Thomas et al. [34]. 2.8. Other biochemical assays Respiratory activities of membrane fragments were determined with a Clark type oxygen electrode YSI 53 (Yellow Springs Inst. Inc., Yellow Springs, OH, USA) as detailed elsewhere [40]. Reduced minus oxidized difference spectra were recorded at room temperature using a Jasco 7800 spectrophotometer. Absorption coefficients, ε561–575 of 22 mM−1 cm−1 and ε551–540 of 19 mM−1 cm−1 were used for b- and c-type cytochromes, respectively.

3. Results 3.1. Effect of tellurite on survival of R. capsulatus in the presence or absence of a superoxide anion generator It has been proposed that tellurite toxicity results from the ability of this oxyanion to generate free radicals as a byproduct of tellurite reduction [38]. This suggests that an increase in tellurite resistance might be linked to activation of cells defense against ROS. We tested this hypothesis by determining the survival of R. capsulatus cells in the presence of increasing tellurite amounts upon cell incubation with subinhibitory concentrations of PQ2+ under assay conditions in which sodB gene induction was reported [9]. Recently, we have shown that R. capsulatus cells grown anaerobically

Fig. 1. R. capsulatus MT1131 (wild-type) cell viability as a function of time and of different tellurite concentrations in the presence ( dashed lines) or absence (continuous lines) of 1 mM paraquat. Symbols: ((") 1000, (Q) 2500 µg ml−1 ). See text for details.

in the light are resistant to high TeO2− 3 amounts [minimal inhibitory concentration of 400 µg ml−1 (1.6 mM)] [2]. Fig. 1 ((") continuous line) shows that R. capsulatus cells can survive to tellurite concentrations up to 1000 µg ml−1 (4 mM) within a 60 min experimental time; conversely, the cell density drops from 2.5 × 104 to 2.5 × 101 CFU in the presence of 2500 µg ml−1 tellurite ((Q) continuous line). However, the survival capacity of R. capsulatus is strongly enhanced after 2 h pre-incubation with 1 mM PQ2+ , as cell survival, which decreases from 2.7 × 104 to 1.9 × 104 CFU in the presence of 1000 µg ml−1 ((") continuous line), is not affected by tellurite in cells pre-treated with PQ2+ ((") dashed line). Notably, protection by PQ2+ is even more significant at 2500 µg ml−1 tellurite, as the cell viability rises from 2.5 × 101 to 1.6 × 103 CFU ((Q) dashed line). 3.2. Respiratory features of the SenC mutant It has been recently reported that Neisseria spp. sco− mutants catalyze cytochrome c oxidase activity (COX) similarly to wild-type cells [28]. R. capsulatus colonies with a functional cytochrome c complex (cbb3 oxidase) are NADI+ , which is indicative of the COX-catalyzed synthesis of indophenol (a blue-dark compound) from α-naphtol and DMPD (see Section 2). Wild-type MT1131 colonies catalyze the synthesis of indophenol in less than 1 min (<1 min, NADI+ ), whereas the SenC mutant CW10 (senC is homologous to sco) colonies retain their original color for some time, and then begin to show a noticeable blue color after ca. 4 min (>4 min, NADI slow). The NADI slow phenotype suggests that in R. capsulatus, unlike the case in Neisseria spp., senC mutation affects COX activity, which is reduced but not absent in CW10 cells. To better quantitatively define COX activity, we measured the ability of isolated membrane vesicles (chromatophores) from MT1131 and CW10 cells to oxidize various exogenous electron donors. Table 1 reports

810

F. Borsetti et al. / Research in Microbiology 156 (2005) 807–813

Table 1 Cytochrome c oxidase activities, NADI test and cytochrome content of R. capsulatus MT1131 and CW10 mutant strain Respiratory activities

MT1131 wild type

CW10 SenC− mutant

Na-ascorbate–DCIP oxidase Na-ascorbate–TMPD oxidase Na-ascorbate–DAD oxidase Na-ascorbate–HHCyt c oxidase NADI test Cytochromes c-Type hemes b-Type hemes b-Heme/c-heme ratio

7.7 7.7 11.1 21.6 <1 min

1.1 1.1 1.1 3.3 >4 min

0.6 nmol (mg protein)−1 0.4 nmol (mg protein)−1 0.68

0.39 nmol (mg protein)−1 0.44 nmol (mg protein)−1 1.13

Respiratory activities and NADI test were performed in isolated membrane fragments and intact cells, respectively. Symbols, abbreviations and concentrations: <1 min (NADI positive); >4 min (NADI slow); Na-ascorbate (5 mM); TMPD (250 µM); DCIP (250 µM); DAD, diaminodurene (100 µM); HHCyt c, horseheart cytochrome c (100 µM); cyt c, c-type heme; cyt b, b-type heme. Oxidase activities are expressed as µmoles of oxygen consumed per hour per mg of protein.

ble cytochrome c (c2 ) and membrane-bound cytochromes c1 and cy [16]. 3.3. Survival of wild-type (MT1131) and the two mutant strains CW10 and GK32 in oxidative killing assays

Fig. 2. Cytochrome c oxidase (cbb3 -type) subunits in R. capsulatus MT1131 (w.t.) and CW10 (SenC− mutant) membrane fragments (chromatophores) as detected by TMBZ/SDS–PAGE analysis. Total membrane proteins (100 µg) were loaded per lane. Cytochromes cp and c0 correspond to subunits II and III of cyt cbb3 oxidase, and c1 and cy correspond to the cyt c1 subunit of the cyt bc1 complex and the membrane attached electron carrier cyt cy , respectively. Soluble cytochrome c2 is the electron donor to the cyt cbb3 oxidase. Lane A, MT1131; lane B, CW10. See text for further details.

that the absence of SenC protein (CW10) strongly affects both COX activity (7–10 times less than MT1131) and total c-type heme content (35% less in the mutant). Furthermore, five distinct membrane-bound c-type hemes with Mr of 32 (cp ), 31 (c1 ), 29 (cy ), 28 (c0 ) and 12 (c2 ) kDa are detected by SDS–TMBZ gel staining in membrane vesicles of the wildtype strain MT1131 [5] (see also Fig. 2, lane A). However, as shown in lane B of Fig. 2, membranes from CW10 (SenC− ) contained very low amounts of the c-type cytochromes cp and c0 corresponding to c-type subunits of cytochrome cbb3 oxidase, respectively, along with normal amounts of solu-

The effect of paraquat (PQ2+ ) and/or xanthine/xanthine oxidase (X/Xox) as generators of superoxide anions on survival of wild-type and the CW10 mutant strain is reported in Figs. 3A and 3B. In the presence of 10 mM PQ2+ no effect was seen after 60 min of incubation of cells from either the wild-type MT1131 or the mutant strain CW10 ((") and (2) traces in panel A), while a significant decrease in survival of CW10 (trace (2), panel B) was observed after incubation with X/Xox (2.15 mM and 0.15 U ml−1 , respectively). Indeed, the cell viability of CW10 (SenC− ) decreased from 2.5 × 104 to 103 CFU after 60 min incubation with X/Xox (trace (2), panel B), indicating that the absence of SenC protein affected only periplasmic O2·− sensitivity. Notably, the cell viability of R. capsulatus GK32, a cbb3 (COX) oxidasedeficient mutant [18] (trace (Q), panel B), was not affected by X/Xox, supporting the concept that decrease in survival of CW10 is actually due to the absence of SenC and does not depend on the absence of a fully active COX. 3.4. Effect of paraquat and tellurite on SOD activity Lysates from R. capsulatus MT1131 cells treated with PQ2+ (1 mM) and tellurite (10 µg ml−1 ) were assayed for SOD activity by non-denaturating PAGE. As shown in Fig. 4 (lanes A–C), both PQ2+ (lane B) and tellurite (lane C) increased superoxide dismutase activity, as detected by the in situ staining procedure [12,36]. In parallel, the cell lysates were tested for in vitro cytochrome c reduction by superoxide anions generated by X/Xox [22], with this assay being an indirect measurement of the endogenous SOD activity which competes with cytochrome c for oxygen-reacting species. Table 2 indicates that in the soluble fractions of R. capsulatus cells treated with either tellurite (10 and 200 µg ml−1 ) or PQ2+ (1 mM), cytochrome c reduction activities were

F. Borsetti et al. / Research in Microbiology 156 (2005) 807–813

811

Table 2 Ferricytochrome c reduction assay in lysates from R. capsulatus cells incubated for 2 h in the presence of either paraquat (PQ2+ ) or tellurite (TeO2− 3 )

Control + PQ2+ (1 mM) −1 + TeO2− 3 (10 µg ml ) −1 + TeO2− 3 (200 µg ml )

Activitya (nmol cyt c reduced mg protein per min)

% of inhibition

5.22 ± 0.22 4.69 ±0.21 3.91 ± 0.18

0 10 25

4.02 ± 0.21

23

a Values are the means of three independent measurements in cell ly-

sates obtained from different cell cultures. Assays were performed using 35 µg ml−1 of proteins (see text and Section 2 for further experimental details).

4. Discussion

Fig. 3. Cell viability of R. capsulatus MT1131 (w.t.), CW10 (SenC− mutant) and GK32 (COX− mutant) as a function of time in the presence of superoxide anion generators such as paraquat (panel A) and xanthine/xanthine oxidase (panel B). Symbols: panels A and B ((") MT1131; (2) CW10; (Q) GK32). See text for details.

Fig. 4. Non-denaturating polyacrylamide gel of superoxide dismutase activity in R. capsulatus MT1131 (wild type) cells incubated for 2 h in the presence of either paraquat or tellurite. Lanes: A, control; B, paraquat 1 mM; C, tellurite 10 µg ml−1 . See text for details.

lower (23–25% and 10%, respectively) than corresponding activities of lysates from non-treated cells. This latter result confirms the in situ staining measurements reported in Fig. 4, indicating that SOD activity is enhanced by both PQ2+ (1 mM) and low amounts of tellurite (10 µg ml−1 ).

Recently, we have demonstrated that tellurite oxyanions are rapidly taken up by R. capsulatus cells and reduced to Te0 in the cytoplasm [4]. The uptake mechanism mainly depends on
pH [4] while the cytosolic reduction process is likely to involve the generation of diglutathione tellurite leading to cytosolic free-radical production [15,38]. The apparent Km for tellurite uptake is quite low (0.08 mM or 20 µg ml−1 ) with a transport rate of 1.6 ± 0.1 µg min−1 per mg of protein under conditions of continuous illumination [4]. Furthermore, we have recently reported that tellurite/cell ratios on the order of 1 ng/5000 are sufficient to induce severe damage to the cytoplasmic membrane [2]. These data, taken together, demonstrate that tellurite toxicity can rapidly be counteracted by R. capsulatus cells, by removing the oxyanion from the periplasmic space on the condition that tellurite is
50 µg ml−1 ; indeed, at higher tellurite amounts, the uptake is totally inhibited, since serious membrane damage takes place, irreversibly affecting both membrane potential and cell viability [2,3]. R. capsulatus contains a cambialistic Mn/Fe superoxide dismutase coded by the sodB gene [9] which is strongly induced upon exposure of cells to air or superoxide generators [9]. In line with the latter observation, we have shown here that R. capsulatus cells treated with sublethal amounts of paraquat induce an increase in SOD activity (Fig. 4 and Table 2) which is paralleled by a significant increase in tellurite resistance (Fig. 1). Interestingly, we also report that SOD activity is increased by the addition of 10 µg ml−1 of tellurite following a 2 h incubation time, while no further increase in SOD activity was seen with higher tellurite amounts. This result can be explained on the basis of our previous observations on the kinetics of tellurite uptake (t1/2 of 5 min) and membrane damage induced by this toxic oxyanion [2,4]. The overall findings indicate that tellurite shows pro-oxidant properties and that SOD mediates an early cell response to exogenous tellurite concentrations in the range of the apparent Km value for the oxyanion uptake. While the toxic effects of tellurite in the cytosol of R. capsulatus cells can be rationalized in terms of oxidative dam-

812

F. Borsetti et al. / Research in Microbiology 156 (2005) 807–813

Fig. 5. Cell viability of R. capsulatus MT1131 (wild type) and CW10 (SenC mutant ) as a function of time and of different tellurite concentrations. Empty symbols (CW10), full symbols (MT1131) ((2), (1), 200; ("), (!), 400; (Q), (P), 1000 µg ml−1 ). See text for details.

age induced by superoxide anions linked to tellurite uptake and reduction [15,38], tellurite toxicity at the level of the periplasmic space is far from being elucidated [37]. In particular, actual tellurite activity toward soluble and membrane-bound proteins, along with tellurite effects on physical/chemical features of the membrane lipids, remain to be explained [10]. In recent years, several reports have indicated the possible interaction between membrane-bound redox complexes and tellurite [3,34,35]. In particular, it has been shown that the electron transport chain of cells of R. capsulatus grown in the presence of low concentrations of tellurite, i.e.,
50 µg ml−1 , synthesize very low amounts of both membrane- and soluble c-type hemes [3] with a drastic repression of cytochrome c oxidase activity; conversely, the electron transport rate through the quinol oxidase pathway is not affected by tellurite growth [3]. These data demonstrate that c-type cytochromes are targets for a direct or indirect effect of tellurite toxicity, since it is quite unlikely that tellurite growth adaptation by R. capsulatus cells would be favored by repression of the membrane potential [2] and of the cytochrome c oxidase activity [3]. In this respect, a class of proteins named Sco (Sco stands for synthesis of cytochrome c oxidase) were shown to bind Cu ions, in agreement with a role for these proteins in assembling the cytochrome c oxidase [11,23,25,27]. It has also been shown that homologues of Sco proteins in photosynthetic bacteria (PrrC protein of R. sphaeroides and SenC protein in R. capsulatus) act as oxygen sensors [5,26]; additionally, secondary structure predictions for Sco, PrrC and SenC proteins reveal that they have similarities to peroxiredoxin/thiol:disulfide oxidoreductases [8] which might suggest an additional role in protecting the cells against oxidative stress. Indeed, the SenC homologue (Sco protein) in pathogenic Neisseria spp. was reported to protect the cells against oxidative killing [28]. Here we show that strain CW10 (SenC− mutant) of R. capsulatus is more sensitive

to periplasmic O2·− -generated by X/Xox than cytosolic O2·− produced by PQ2+ (Fig. 3B). This result is therefore in line with the predicted topology of Sco proteins which are membrane-attached proteins protruding into the periplasmic space [13]. On the other hand, the absence of SenC protein from R. capsulatus CW10 is also linked to a drastic decrease in cbb3 oxidase activity and c-type hemes (cp - and c0 -type subunits) (Table 1 and Fig. 2). This might suggest that the higher sensitivity of CW10 to oxidative stress is not due to SenC deficiency but to the absence of a fully active cbb3 complex. However, cells from R. capsulatus GK32, a cbb3 mutant strain [18], show a response to oxidative stress generated by X/Xox similar to MT1131 wildtype cells (Fig. 3B) and this supports the concept that SenC is actually involved in protection against periplasmic O2·− . Our results with R. capsulatus CW10 (SenC− ) therefore differ from those reported for a sco mutant of Neisseria spp. [28] which is, however, endowed with a fully active cbb3 complex and is also sensitive to cytosolic O2·− produced by PQ2+ [28]. This particular aspect thus requires further investigation. If SenC is involved in protection of cells of R. capsulatus against periplasmic oxidative stress, one would expect to observe a decrease in tellurite resistance in SenC mutant cells on the condition that tellurite toxicity in the periplasmic space is due to O2·− . However, the data in Fig. 5 show that in the absence of SenC, the tellurite resistance of R. capsulatus does not vary significantly between 200 and 1000 µg ml−1 tellurite; this result is therefore interpreted as evidence that tellurite does not generate superoxide oxyanions at the periplasmic space. Experiments are in progress aiming to define the capacity of tellurite to directly interact with membrane redox components and/or to alter the mechanism involved in c-type heme assembly.

Acknowledgements This work was financed by MIUR (PRIN2003) to D.Z. and by DOE grant 91ER2005 to F.D. Authors like to thank R. Turner (University of Calgary) for helpful discussion during the preparation of the manuscript.

References [1] J. Beers, D.M. Glerum, A. Tzagaloff, Purification and characterization of yeast Sco1p, a mitochondrial copper protein, J. Biol. Chem. 277 (2002) 22185–22190. [2] R. Borghese, F. Borsetti, P. Foladori, G. Ziglio, D. Zannoni, Effects of the metalloid oxyanion tellurite on growth characteristics of the phototrophic bacterium Rhodobacter capsulatus, Appl. Env. Microbiol. 70 (2004) 6595–6602. [3] F. Borsetti, R. Borghese, F. Francia, M.R. Randi, S. Fedi, D. Zannoni, Reduction of potassium tellurite to elemental tellurium and its effect on the plasma membrane redox components of the facultative phototroph Rhodobacter capsulatus, Protoplasma 221 (2003) 153–161.

F. Borsetti et al. / Research in Microbiology 156 (2005) 807–813

[4] F. Borsetti, A. Toninello, D. Zannoni, Tellurite uptake by cells of the facultative phototroph Rhodobacter capsulatus is a
pH dependent process, FEBS Lett. 554 (2003) 315–318. [5] J. Buggy, C.E. Bauer, Cloning and characterization of SenC, a gene involved in both aerobic respiration and photosynthesis gene expression in Rhodobacter capsulatus, J. Bacteriol. 177 (1995) 6958–6965. [6] E. Cadenas, Biochemistry of oxygen toxicity, Annu. Rev. Biochem. 58 (1989) 79–110. [7] R. Cannio, G. Fiorentino, A. Morana, M. Rossi, S. Barolucci, Oxygen: A friend or foe? Archaeal superoxide dismutases in the protection of intra- and extracellular oxidative stress, Front. Biosci. 5 (2000) 768– 776. [8] Y.V. Chinenov, Cytochrome c oxidase assembly factors with a thioredoxin fold are conserved among prokaryotes and eukaryotes, J. Mol. Med. 78 (2000) 239–242. [9] N. Cortez, N. Carrillo, C. Pasternak, A. Balzer, G. Klug, Molecular cloning and expression analysis of the Rhodobacter capsulatus sodB gene, encoding an iron superoxide dismutase, J. Bacteriol. 180 (1998) 5413–5420. [10] B. Deuticke, P. Lutkemeier, B. Poser, Tellurite-induced damage of erythrocyte membrane: Manifestations and mechanisms, Biochim. Biophys. Acta 1109 (1992) 97–107. [11] E.K. Dickinson, D.L. Adams, E.A. Schon, D.M. Glerum, A human SCO2 mutation helps define the role of Sco1p in the cytochrome oxidase assembly pathway, J. Biol. Chem. 275 (2000) 26780–26785. [12] J.L. Donahue, C. Moses-Okpodu, C.L. Cramer, E.A. Grabau, R.G. Alsscher, Responses of antioxidants to paraquat in pea leaves, Plant Physiol. 113 (1997) 249–257. [13] J.M. Eraso, S. Kaplan, From redox flow to gene regulation: Role of PrrC protein of Rhodobacter sphaeroides 2.4.1, Biochemistry 39 (2000) 2052–2062. [14] I. Fridovich, Superoxide radical and superoxide dismutases, Annu. Rev. Biochem. 64 (1995) 97–112. [15] P. Garber, L. Engman, V. Tolachev, H. Lundquist, R.G. Gerdes, I.A. Cotgreave, Binding of tellurium to hepatocellular selenoproteins during incubation with inorganic tellurite: Consequences for the activity of selenium-dependent glutathione peroxidase, Int. J. Biochem. Cell Biol. 31 (1999) 291–301. [16] K.A. Gray, M. Grooms, H. Myllikallio, C. Moomaw, C. Slaughter, F. Daldal, Rhodobacter capsulatus contains a novel cb-type cytochrome c oxidase without CuA center, Biochemistry 33 (1994) 3120–3127. [17] D.H. Kho, S.-B. Yoo, J.-S. Kim, E.-J. Kim, J.K. Lee, Characterization of Cu- and Zn-containing superoxide dismutase of Rhodobacter sphaeroides, FEMS Lett. 234 (2004) 261–267. [18] H.-G. Koch, C. Winterstein, S.A. Saribas, J.O. Albel, F. Daldal, Roles of the ccoGHIS gene products in the biogenesis of the cbb3-type cytochrome c oxidase, J. Mol. Biol. 297 (2000) 49–65. [19] J.A. Imlay, S. Linn, DNA damage and oxygen radical toxicity, Science 240 (1988) 1302–1309. [20] O.H. Lowry, N.J. Rosebrough, A.L. Farr, R.J. Randall, Protein measurement with the Folin phenol reagent, J. Biol. Chem. 193 (1951) 265–275. [21] B.L. Marrs, H. Gest, Regulation of bacteriochlorophyll synthesis by oxygen in respiratory mutants of Rhodopseudomonas capsulata, J. Bacteriol. 114 (1973) 1045–1051. [22] J. McCord, I. Fridovich, Superoxide dismutase, an enzymic function for erythrocuprein (hemoprotein), J. Biol. Chem. 244 (1969) 6049– 6055.

813

[23] A.G. McEwan, A. Lewin, S.L. Davy, R. Boetzel, A. Leech, D. Walker, T. Wood, G.R. Moore, PrrC from Rhodobacter sphaeroides, a homologue of eukaryotic Sco proteins, is a copper-binding protein and may have a thiol-sulfide oxidoreductase activity, FEBS Lett. 518 (2002) 10–16. [24] M.D. Moore, S. Kaplan, Identification of intrinsic high-level resistance to rare-earth oxides and oxyanions in members of the class Proteobacteria: Characterization of tellurite, selenite, and rhodium sesquioxide reduction in Rhodobacter sphaeroides, J. Bacteriol. 174 (1992) 1505– 1514. [25] T. Nittis, G.N. George, D.R. Winge, Yeast Sco1, a protein essential for cytochrome c oxidase function is a Cu(I)-binding protein, J. Biol. Chem. 276 (2001) 42520–42526. [26] J. Oh, S. Kaplan, in: D. Zannoni (Ed.), Respiration in Archaea and Bacteria: Diversity of Prokaryotic Electron Transport Carriers, vol. 15, Kluwer Academic, Dordrecht, 2004, pp. 287–309. [27] A. Rentzsch, G. Krummeck-Weiss, A. Hofer, A. Bartuschka, K. Ostermann, G. Rodel, Mitochondrial copper metabolism in yeast: Mutational analysis of Sco1p involved in the biogenesis of cytochrome c oxidase, Curr. Genet. 35 (1999) 103–108. [28] K.J. Seib, M.P. Jennings, A.G. McEwan, A Sco homologue plays a role in defense against oxidative stress in pathogenic Neisseria, FEBS Lett. 546 (2003) 411–415. [29] W. Siström, A requirement for sodium in the growth of Rhodopseudomonas sphaeroides, J. Gen. Microbiol. 22 (1960) 778–785. [30] M. Solioz, H.-C. Yen, B.L. Marrs, Release and uptake of gene transfer agent by Rhodopseudomonas capsulata, J. Bacteriol. 123 (1975) 651– 672. [31] G. Storz, L.A. Tartaglia, S.B. Farr, B.N. Ames, Bacterial defenses against oxidative stress, Trends Genet. 6 (1990) 363–368. [32] L.C. Tabares, C. Bittel, N. Carrillo, A. Bortolotti, N. Cortez, The single superoxide dismutase of Rhodobacter capsulatus is a cambialistic, manganese containing enzyme, J. Bacteriol. 185 (2003) 3223– 3227. [33] D.E. Taylor, Bacterial tellurite resistance, Trends Microbiol. 7 (1999) 111–115. [34] P.E. Thomas, D. Ryan, W. Lewin, An improved staining procedure for the detection of the peroxidase activity of cytochrome P-450 on sodium dodecyl sulfate polyacrylamide gels, Anal. Biochem. 75 (1976) 168–176. [35] A.M. Trutko, V.K. Akimenko, N.E. Suzina, L.A. Anisimova, M.G. Shlyapnikov, B.P. Baskunov, V.I. Duda, A.M. Boronin, Involvement of the respiratory chain of Gram-negative bacteria in the reduction of tellurite, Arch. Microbiol. 173 (2000) 178–186. [36] A.L. Tseng, Y. Srikhanta, A.G. McEwan, M.P. Jennings, Accumulation of manganese in Neisseria gonorrhoeae correlates with resistance to oxidative killing by superoxide anion and is independent of superoxide dismutase activity, Mol. Microbiol. 40 (2001) 1175– 1186. [37] R.J. Turner, J.H. Weiner, D. Taylor, Tellurite-mediated thiol oxidation in Escherichia coli, Microbiology 145 (1999) 2549–2557. [38] R.J. Turner, Tellurite toxicity and resistance in Gram-negative bacteria, Rec. Res. Dev. Microbiol. 5 (2001) 69–77. [39] V.V. Yurkov, J.T. Beatty, Aerobic anoxygenic phototrophic bacteria, Microbiol. Mol. Biol. Rev. 62 (1998) 695–724. [40] D. Zannoni, P. Jasper, B.L. Marrs, Light induced oxygen uptake as a probe of electron transport between respiratory and photosynthetic components in membranes of Rhodopseudomonas capsulata, Arch. Biochem. Biophys. 191 (1978) 625–631.