Respiratory Enzymes of Thiobacillus ferrooxidans A KINETIC STUDY OF ELECTRON TRANSFER BETWEEN IRON AND RUSTICYANIN IN SULFATE MEDIA*

Thiobacillus ferrooxidans is a chemolithotrophic bacterium capable of fulfilling all of its energy require- ments from the oxidation of soluble ferrous sulfate. Rusticyanin is a soluble blue copper protein found in abundance in the periplasmic space of this bacterium. The one-electron transfer reaction between soluble iron and purified rusticyanin has been studied by stopped flow spectrophotometry in acidic solutions containing sulfate. Second order rate constants for the reduction of rusticyanin by Fez+, FeHSO:, and FeS09 were 0.022, 0.73, and 2.30 M-’ s-’, respectively. The pseudo-first order rate constant for the reduction of rusticyanin exhibited substrate saturation when the concentration of the total ferrous ion was varied in solutions of limiting sulfate. This saturation behavior was quantitatively described using the values of the second order rate constants listed above and the distribution of the total ferrous ion into its water-, bisul- fate-, and sulfate-coordinated forms. Second order rate constants for the oxidation of rusticyanin by Fe3+ and FeSO: were 0.73 and 0.26 M-’ s-’, respectively. The electron transfer reactions between iron and rusticyanin monitored in vitro were far too slow to support the hypothesis that rusticyanin is the primary oxidant of ferrous ions in the iron-dependent respiratory electron transport chain of T. ferrooxidans. pH. When grown in the absence of additional copper ions, all 13 strains released approximately 1.5 mg of cell-free rusticyanin/g (wet weight) of bacterial cells. When grown in the presence of additional copper ions, the strain identified as ATCC 23270 (the T. ferrooxidans-type strain) consistently released roughly 3.0 mg of cell-free rusticy- anin/g (wet weight) of bacterial cells. Large scale growth of ATCC 23270 for the purification of rusticyanin was subsequently achieved by batch culture at ambient temperatures in the acidified ferrous sulfate medium (22) supplemented with 1.6 mM cupric sulfate. Purification Rusticyanin-Rusticyanin was prepared by modifi-cation a published procedure (20). Briefly, crude extracts were prepared by sonic oscillation of whole cell suspensions in sulfuric 2.0. The crude was then to ammonium sulfate precipitation. The bulk of the rusticyanin precipitated between 45 and 95% saturated ammonium sulfate. The precipitated material containing rusticyanin was dialyzed extensively against 0.01 M sodium acetate, pH 5.5, adsorbed to a column of Sephadex C- 25 equilibrated with the dialysis buffer, and eluted from the column in a stepwise fashion using the acetate buffer containing additional NaCl. The identity of the solution used to dissolve the ammonium pellet had a curious influence on the subsequent behavior of the rusticyanin on the ion exchange resin. Dissolution of the precipitated rusticyanin in 0.01 M acetate, pH 5.5, resulted in a single peak of rusticyanin when the resin was washed with acetate buffer containing 100 D M NaC1; dissolution in 0.01 N sulfuric acid produced two peaks of rusticyanin when the resin was washed with acetate buffer plus 500 mM NaCl, and dissolution in 0.001 N sulfuric acid produced a single, symmetrical peak of rusticyanin when the resin was washed with acetate buffer plus 200 mM NaCl. The last treatment produced the highest yield of blue protein and was used routinely. of the solution from each driving syringe. Spectral changes were linear to an absorbance of 1.8. A typical M of 0.06 provided acceptable signal to noise characteristics. Absorbance Spectra-Absorbance spectra were obtained on a Cary 14 dual-beam spectrophotometer rebuilt and modified by On-Line Instrument Systems, Inc. Instrument control and data analysis were accomplished via a Zenith ZF 118 computer interfaced to the rebuilt spectrophotometer. Extinction coefficients were calculated by com-paring the absorbance of a given sample with the amount of protein in the sample. Protein was quantified by the method of Bradford (23) using purified azurin as the standard. Extinction coefficients deter- mined from four separate preparations of purified protein were char-acterized by a standard deviation of +6%. Materials-Ferrous perchlorate, ferric perchlorate, and sodium perchlorate were obtained from Morton Thiokol, Inc. (Alfa Products). Ferrous sulfate, ferric sulfate, sodium sulfate, sodium bisulfate, and sulfuric acid were obtained from Fisher. Purified azurin was obtained from Sigma. All other chemicals were reagent grade.

Thiobacillus ferrooxidans is the most extensively characterized member of a group of chemolithotrophic bacteria that inhabit ore-bearing geological formations exposed to the atmosphere and obtain all of their energy for growth from the dissolution and oxidation of minerals within the ore. These Gram-negative, obligately acidophilic bacteria display optimal growth from pH 1.5 to 3.5 (2)(3)(4). T. ferrooxidans takes its species name from its ability, unique among the thiobacilli, to grow in the laboratory using soluble ferrous sulfate as its sole source of energy (5)(6)(7). Energy is derived from oxidative phosphorylation coupled to respiratory electron transfer; molecular oxygen is the ultimate electron acceptor. Cytochromes c and a have been identified in T. ferrooxidans, and numerous reports have demonstrated that both types of cytochromes are reduced in intact cells exposed to Fe(I1) (8)(9)(10)(11).
Contrary to earlier reports of a cytoplasmic iron-cytochrome c reductase (12)(13)(14), it is now generally accepted that *This research was supported by Grant DE-FG05-85ER13339 from the United States Department of Energy. Preliminary reports of part of this investigation have been presented (1). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. the initial electron transfer reaction from Fe(I1) to a cellular component must occur at the outer surface of the plasma membrane in the periplasmic space (15)(16)(17)(18). The agent thought to be the primary initial electron acceptor from Fe(I1) is a soluble blue copper protein called rusticyanin. Rusticyanin is a type I copper protein with an apparent molecular mass of 16,300 daltons (19)(20)(21). The assignment of rusticyanin as the initial electron acceptor from Fe(I1) is based on the following observations: 1) rusticyanin is reduced when intact cells are exposed to Fe(I1); 2) the purified protein is reduced directly by Fe(I1); 3) the protein is very stable at pH 1.5 (the pH of the growth medium) and very unstable at the cytosolic pH of neutrality; and 4) the protein exhibits a redox potential of 680 mV (9, 15), which compares favorably with the value of 770 mV for the uncomplexed ferrous/ferric couple (21). Electrons derived from the oxidation of Fe(I1) are then thought to cross the plasma membrane via an as yet undefined cytochrome chain to the inner surface (9, 15, 21), where the reduction of molecular oxygen is catalyzed by a cytochrome oxidase. Because the reaction with molecular oxygen consumes protons that must be replenished from the bulk phase utilizing a pH difference of some 4-5 units, a proton gradient sufficient to drive the phosphorylation of ADP is created as a direct, inevitable consequence of the suggested sequence and topography of the electron transfer chain components (see Ref. 15 for a detailed review).
Most of the evidence for proposing the roles and sequence of respiratory chain components responsible for the oxidation of Fe(I1) has been obtained from spectrophotometric and inhibition studies with intact cells and crude cell-free preparations. More direct evidence, the separation of the respiratory chain into purified components and attempted reconstitution of portions of the system with the purified components, has only recently begun to appear in the literature (11,(19)(20)(21). The present paper describes a detailed kinetic study of the transfer of electrons between iron and purified rusticyanin. These kinetic experiments were conducted in the presence of sulfate, the principal anion in the bacteria's natural environment and the anion of choice in laboratory culture. The results presented indicate that the rate of electron transfer from Fe(I1) to rusticyanin is far too slow to account for the facile Fe(I1)-dependent reduction of molecular oxygen in the intact organism.

EXPERIMENTAL PROCEDURES
Growth of T. ferrozidans-Thirteen purified strains of T. ferrooxidam (seven available from the ATCC and six (TFI 1, 4, 10, 29, 35, and 35SA) from the private collection of Dr. Olli Tuovinen, Department of Microbiology, Ohio State University) were obtained and cultured in the acidified ferrous sulfate growth medium described by Tuovinen and Kelly (22). Each of the 13 strains was grown in sufficient yield to quantify the amount of rusticyanin that could be solubilized by limited sonic oscillation of the cell suspensions at acidic Electron Transfer between Iron and Rusticyanin pH. When grown in the absence of additional copper ions, all 13 strains released approximately 1.5 mg of cell-free rusticyanin/g (wet weight) of bacterial cells. When grown in the presence of additional copper ions, the strain identified as ATCC 23270 (the T. ferrooxidanstype strain) consistently released roughly 3.0 mg of cell-free rusticyanin/g (wet weight) of bacterial cells. Large scale growth of ATCC 23270 for the purification of rusticyanin was subsequently achieved by batch culture at ambient temperatures in the acidified ferrous sulfate medium (22) supplemented with 1.6 mM cupric sulfate.
Purification of Rusticyanin-Rusticyanin was prepared by modification of a previously published procedure (20). Briefly, crude extracts were prepared by sonic oscillation of whole cell suspensions in sulfuric acid, pH 2.0. The crude extract was then subjected to ammonium sulfate precipitation. The bulk of the rusticyanin precipitated between 45 and 95% saturated ammonium sulfate. The precipitated material containing rusticyanin was dissolved, dialyzed extensively against 0.01 M sodium acetate, pH 5.5, adsorbed to a column of Sephadex C-25 equilibrated with the dialysis buffer, and eluted from the column in a stepwise fashion using the acetate buffer containing additional NaCl. The identity of the solution used to dissolve the ammonium sulfate pellet had a curious influence on the subsequent behavior of the rusticyanin on the ion exchange resin. Dissolution of the precipitated rusticyanin in 0.01 M acetate, pH 5.5, resulted in a single peak of rusticyanin when the resin was washed with acetate buffer containing 100 D M NaC1; dissolution in 0.01 N sulfuric acid produced two peaks of rusticyanin when the resin was washed with acetate buffer plus 500 mM NaCl, and dissolution in 0.001 N sulfuric acid produced a single, symmetrical peak of rusticyanin when the resin was washed with acetate buffer plus 200 mM NaCl. The last treatment produced the highest yield of blue protein and was used routinely.
The rusticyanin in the effluent fractions was concentrated to approximately 20 mg of protein/ml by ultrafiltration through an Amicon PM-10 membrane at 4 "C. The concentrated rusticyanin was then subjected to gel filtration on a column of Bio-Gel P-100 equilibrated and developed with 0.001 N sulfuric acid. Rusticyanin prepared in this fashion appeared electrophoretically homogeneous when subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions. This purification procedure reproducibly yielded roughly 1.0 mg of pure rusticyanin/g (wet weight) of T.
ferrooxidans cell paste. The purified rusticyanin could be stored in 0.001 N sulfuric acid for at least 4 months at 4 "C without appreciable deleterious effects.
Stopped Flow Measurements-Kinetic measurements were performed on a stopped flow spectrophotometer based on a Kinetic Instruments (Ann Arbor, MI) absorbance mixing apparatus and a data acquisition and analysis system designed and constructed by On-Line Instrument Systems, Inc. (Jefferson, GA). The optical system of this apparatus is comprised of an Instruments S. A., Inc. H-10 single holographic grating monochromator and a tungsten source powered by an OLIS XL150 adjustable power supply. A computercontrolled stepping motor attached to the monochromator permits acquisition of the absorbance spectrum of the contents of the observation cell (2-cm path length). The signal from the phototube (Hammamatsu Corp. 955) is fed in turn to an OLIS HVA-SF amplifier, OLIS Model 3920 interface, and finally to a Zenith ZF 118 computer equipped with a dual floppy disc drive. Data collection software permits signal averaging of input decays and control of data input timing.
The rusticyanin and the ferrous or ferric ion were prepared in identical solutions of acidified sulfate media and added to separate syringes of the stopped flow spectrophotometer. The temperature of the driving syringes was maintained at 25 rt 1 'C by circulating water. Room temperature solutions were allowed to equilibrate for 10 min in the driving syringes. Reactions were initiated by rapidly mixing 0.1 ml of the solution from each driving syringe. Spectral changes were linear to an absorbance of 1.8. A typical M of 0.06 provided acceptable signal to noise characteristics.
Absorbance Spectra-Absorbance spectra were obtained on a Cary 14 dual-beam spectrophotometer rebuilt and modified by On-Line Instrument Systems, Inc. Instrument control and data analysis were accomplished via a Zenith ZF 118 computer interfaced to the rebuilt spectrophotometer. Extinction coefficients were calculated by comparing the absorbance of a given sample with the amount of protein in the sample. Protein was quantified by the method of Bradford (23) using purified azurin as the standard. Extinction coefficients determined from four separate preparations of purified protein were characterized by a standard deviation of +6%.
Materials-Ferrous perchlorate, ferric perchlorate, and sodium per-chlorate were obtained from Morton Thiokol, Inc. (Alfa Products). Ferrous sulfate, ferric sulfate, sodium sulfate, sodium bisulfate, and sulfuric acid were obtained from Fisher. Purified azurin was obtained from Sigma. All other chemicals were reagent grade.

RESULTS
An immediate and striking property of the rusticyanin molecule is its intense blue color. The visible absorbance spectrum of oxidized rusticyanin is represented by the solid line in Fig. 1. The absorbance peak at about 600 nm is accompanied by two less intense peaks on either side, one a t about 445 nm and one at around 740 nm. The spectrum shown in Fig. 1 is representative of those exhibited by a class of copper-containing proteins which includes azurin (24), plastocyanin (25), stellacyanin (26), amicyanin (27), plantacyanin (28), and umecyanin (29). The current interpretation of spectra such as the one in Fig. 1 is that the intense blue absorbance at 600 nm is due to charge transfer between the x-s orbitals of a cystyl residue and the (dx2-yZ) orbitals of the proteinbound Cu(II), while the band a t 740 nm is due to u-s and Cu(d.~-~z) orbitals (30,31). There is some disagreement regarding the interpretation of the lower wavelength adsorption (31). All three absorbance peaks disappear upon the oneelectron reduction of the protein. The visible absorbance spectrum of reduced rusticyanin is represented by the dashed line in Fig. 1. The blue absorbance of the copper center in oxidized rusticyanin thus constitutes an intrinsic spectrophotometric probe whereby transient changes in the redox state of the molecule may be monitored with great sensitivity. The goal of the experiments discussed below was to exploit this intrinsic spectrophotometric probe to study the kinetic mechanism of the one-electron transfer between iron and rusticyanin in acidified sulfate media. The purpose of achieving this goal was to relate kinetic studies performed in. vitro to the proposed physiological role of the rusticyanin in the iron respiratory electron transport chain of the bacterium.
Reduction of Rusticyanin by Fe(II)-The object of these experiments was to obtain detailed kinetic data on the oneelectron reduction of rusticyanin by Fe(I1). Fig. 2 shows the representative pseudo-first order behavior observed when oxidized rusticyanin was mixed with different concentrations of Fe(I1) in the stopped flow spectrophotometer. Under experimental conditions where the concentration of Fe(I1) was always in a 10-fold or greater excess to that of the protein concentration (pseudo-first order conditions), each kinetic trace obtained with Fe(1I) could be described mathematically ."""  Although stability constants for the above equilibria are not reported for the exact ionic strengths and solution conditions employed in the present study, reasonable estimates for the values of these constants based on data acquired from solutions of similar composition are available and are listed in Table I. The values of the equilibrium constants listed in Table I were used, along with Equation 3, to calculate the distribution of the total ferrous ion into its three forms under any set of solution conditions. It was of interest to determine whether the kinetic behavior observed in Figs. 3 and 4 could be attributed to different kinetic properties of the different forms of the ferrous ion present in acidified sulfate media. Accordingly, solution conditions were defined which would permit us to determine the contribution of each stable form of the ferrous ion to the overall rate of reduction of the rusticyanin. The results of these kinetic experiments are shown in Fig. 5. Regardless of the identity of the principal reductant, the reduction of the rusticyanin appeared first order with an apparent rate constant, kobs, as follows. hoba = kl(reductant) + kz Fig. 5A shows the dependence of the rate of reduction of the rusticyanin upon the concentration of the aqueous, hexoaquo ferrous ion. Purified rusticyanin was extensively dialyzed against 1.0 M perchlorate, pH 2.0, to remove traces of sulfate ion from the protein preparation. Ferrous perchlorate was then dissolved in 1.0 M perchlorate, pH 2.0, to permit the subsequent electron transfer to rusticyanin to be monitored in the complete absence of sulfate. Perchlorate was chosen as the anion in these experiments because it complexes very  poorly, if at all, with the ferrous cation (38). The values of kl and k, extracted from the slope and the ordinate intercept, respectively, of the plot in Fig. 5A are shown in Table 11.  and the data in Table I indicated that experimental conditions of 1.0 M sulfate and pH 3.5 would ensure that 98% of the total ferrous ion was in the FeSO%q form, while the remaining 2% was in the FeHSO: form; the concentration of the Fe2+uq was negligibly small under these conditions. Fig. 5C shows the dependence of the rate of reduction of the rusticyanin upon the concentration of the positively charged ferrous-bisulfate complex. The experimental conditions of 1.0 M sulfate and pH 0.5 dictated that 95% of the total ferrous ion was in the FeHSO: form, while the remaining 5% was in the FeSO%q form. The plots of k b s uersus the concentration of either FeSO2aq or FeHSO: were linear over a 100-fold concentration range. Values of kl and k2 extracted from the plots in Fig. 5, B and C, are listed in Table 11.  Table I. It would seem that the substrate saturation behavior observed at lower sulfate concentrations in Fig. 3 may be accommodated by a mechanism where the sulfate concentration limits the amount of FeSOtaq (a much faster reducing agent for rusticyanin than Fez+uq) formed at pH 2.0. It is unnecessary to postulate a rate-determining protein activation mechanism to account quantitatively for the experimental observations shown above.
Oxidation of Rusticyunin by Fe(III)-The object of these experiments was to obtain detailed kinetic data on the oneelectron oxidation of reduced rusticyanin by Fe(II1). Oxidation of the reduced rusticyanin was monitored in the stopped flow spectrophotometer as an increase in the absorbance at 597 nm. Under experimental conditions where the concentration of Fe(II1) was always in a 10-fold or greater excess to that of the protein concentration, each kinetic trace obtained with Fe(II1) could be described mathematically as a single exponential function of time. Accurate values for the pseudofirst order rate constant describing each kinetic trace were obtained as described above.
Rate constants for the oxidation of rusticyanin were a function of the Fe(II1) concentration. It is clear from the results discussed above that the effect of sulfate complexation must be considered in any kinetic study involving Fe(I1) or Fe(II1). The known ferric-sulfate complexes are FeSO:, FeHSO:+, Fe(S04);, and FeHS04S04 (38). At a pH value of 2.0 the order of predominance with increasing sulfate is Fe3+uq, FeSO:, and Fe(S04); (39); insignificant quantities of FeHSO:' and FeHS04S04 are found above a pH of 1.0. The effects of Fe3+uq and FeSO: on the oxidation of rusticyanin are presented here.
Variation of the pH from 1.0 to 3.0 had no appreciable effect on the rate constant for the oxidation of rusticyanin by either Fe3+aq or FeSO: (data not shown). The contribution of Fe3+uq and FeSO: to the rate of oxidation of rusticyanin are shown in Fig. 6. Regardless of the identity of the principal oxidant, the oxidation of rusticyanin appeared first order with an apparent rate constant as follows.
There was no evidence for limiting kinetic behavior when the concentration of either Fe3+uq or FeSO: was varied. Fig. 6A shows the dependence of the rate of oxidation of the rusticyanin upon the concentration of the aqueous ferric ion. Purified rusticyanin was reduced by reacting a 2-fold excess of ferrous

Electron Transfer between Iron and Rusticyanin
pyrophosphate with the protein and extensively dialyzing against 1.0 M perchlorate, p H 2.0, to remove the excess reducing agent and traces of sulfate ion. Reduced rusticyanin is remarkably stable to air oxidation. Samples of the reduced protein have been stored in 0.001 N sulfuric acid for up to 6 months at 4 "C before air-oxidized rusticyanin could be detected. Ferric perchlorate was then dissolved in 1.0 M perchlorate, pH 2.0, to permit the subsequent electron transfer from rusticyanin to be monitored in the complete absence of sulfate. Fig. 6B shows the dependence of the rate of oxidation of rusticyanin upon the concentration of the positively charged ferric-sulfate complex. The concentrations of FeSO+ were calculated from the data in Table  I. Values of kl and k2 obtained from the slopes and intercepts, respectively, of the plots in Fig. 6 are listed in Table 11. The data on oxidation of the rusticyanin are consistent with the following simple mechanism: where the oxidant may be either Fe3+ or FeSO:.
It is of interest to note that the second order rate constants for reduction of the rusticyanin increase as the amount of positive charge on the reductant decreases, while the opposite trend is observed in the second order rate constants for oxidation of the rusticyanin. The former observation could be taken as evidence that electrostatic repulsion between the positively charged reductant and the positively charged protein (rusticyanin will have no negative charges on its surface at pH 2.0 or below) has an influence on the rate constant for reduction. However, the latter observation implies that charge-dependent forces between the protein and the inorganic reagent have no significant bearing on the rate constant for electron transfer. A more important influence on the value of the rate constant may be the thermodynamic driving force for electron transfer. Sulfate anions bind with a greater affinity to Fe(II1) than to Fe(I1) ( Table I). This discriminatory binding serves to lower the effective redox potential of the ferrous/ferric couple from 770 mV (uncomplexed) to 650 mV (complexed with sulfate). The data shown in Table I1 suggest that chelation of the metal ion plays an important role in the mechanism of electron transfer with rusticyanin. The second order rate constants for reduction of the rusticyanin increase as the reduction potential of the metal chelates decrease, while the second order rate constants for oxidation of the rusticyanin increase as the oxidation potentials of the metal chelates decrease. This statement is consistent with the qualitative application of relative Marcus theory (40) to the electron transfer reactions studied here. The quantitative application of relative Marcus theory to rusticyanin-dependent electron transfer reactions would require a more extensive set of kinetic data.

DISCUSSION
A high concentration of electron transport chain components is typical of chemolithotrophs that utilize substrates of high redox potential (41). Because such substrates are poor sources of redox potential energy, the organisms require a high throughput of the growth substrate to meet their energy requirements. Ferrous ion-dependent oxygen consumption is very rapid in intact cells of T. ferrooxiduns. Kinetic studies on intact cells in acidified ferrous sulfate media yielded maximal rates of oxygen consumption of up to 2.0 mL of molecular oxygen/h/mg of total cellular protein, with an apparent K , for Fe(I1) between 2.0 and 5.0 mM (42)(43)(44)(45). If 5% of the total cellular protein is assumed to be rusticyanin (15), then the apparent turnover number for the transfer of electrons from Fe(I1) to molecular oxygen by whole cells may be expressed as approximately 10 s"/molecule of rusticyanin in the cell. In order for rusticyanin to serve as the primary initial electron acceptor in the iron-dependent respiratory chain, the apparent rate constant for the Fe(I1)-dependent reduction of NSticyanin must be greater than or equal to the overall turnover number for the entire process, 10 s-'. The data presented here indicate that such is certainly not the case. Under solution conditions of optimal sulfate concentrations and Fe(I1) concentration 10-fold higher than the apparent K , for Fe(I1) in intact cells, the effective rate constant for the reduction of purified rusticyanin approaches 0.1 s" (Fig. 5 ) , some 2 orders of magnitude too low.
The apparent inability of rusticyanin to serve as the initial oxidant of Fe(I1) in intact T. feroxiduns has been noted by other investigators working under conditions of sulfate limitation (36, 37). Two hypotheses have been advanced which identify other possible primary oxidants of Fe(I1). The first hypothesis is that an Fe(II1)-sulfate chelate forms a polynuclear layer around the outer cell wall of this Gram-negative organism and serves as the initial electron acceptor from Fe(I1) in the bulk solution (46). Cells of T. ferroxiduns contain a polynuclear Fe(II1) complex (a system of Fe(II1) ions interacting with each other such that the behavior of the individual ions cannot be observed) detectable by EPR with paramagnetic properties similar to those of phosvitin containing bound Fe(II1) (9, 47). Reducing equivalents derived from solution Fe(I1) could move rapidly through the polynuclear Fe(II1) coat until they reached rusticyanin or some alternative oxidant at the periplasmic surface. Rusticyanin could then serve to shuttle electrons from the outer cell layer across the periplasmic space to the cytoplasmic membrane. The polynuclear Fe(II1) coat would function in a manner analogous to a lightening rod by conducting electrons encountered in the cell's environment to the periplasmic surface. This is an attractive proposal for an organism which must extract electrons from insoluble, semiconducting mineral sulfides in its native habitat. Nonetheless, it should be noted that no direct functional evidence in support of this hypothesis has appeared in the literature to date.
The second hypothesis is that a putative iron-sulfur protein recently detected in whole cells by EPR spectroscopy is responsible for the direct oxidation of Fe(I1) to Fe(II1) (48). Rusticyanin in this hypothesis could again serve to shuttle electrons from the iron-sulfur protein to the cytoplasmic membrane. Although preliminary evidence placing the ironsulfur protein on the reducing side of rusticyanin has been presented, it is disturbing that extensive EPR experiments reported by other investigators (9, 15) have failed to reveal the existence of this iron-sulfur protein.
A third hypothesis may be constructed around the kinetic behavior of an acid-stable cytochrome c present in crude extracts of T. ferrooxiduns. We have obtained a partially purified cytochrome c preparation that catalyzes the Fe(I1)dependent reduction of purified rusticyanin to a remarkable degree (49). Preliminary stopped flow spectrophotometric experiments suggest that the cytochrome c may indeed be the agent which mediates the transfer of electrons from Fe(I1) to rusticyanin.
A final hypothesis recognizes that the role of rusticyanin as the primary oxidant of Fe(I1) has not been entirely eliminated. Conditions experienced by the rusticyanin in the microenvironment of the periplasmic space may not have been duplicated in the in vitro kinetic experiments presented above and by others. It is possible that the kinetic behavior of the rusticyanin is influenced by complexation of the rusticyanin with other species in the periplasm, such as its physiological oxidant. Alternatively, the reductant for rusticyanin in the periplasm could be Fe(I1) in a different form from the solvated sulfate complex. If the Fe(I1) were to exchange outer-sphere coordination ligands upon entry to the periplasm, the presence of the different ligand could greatly alter the kinetic behavior of the ferrous complex. The effect of different anions on the Fe(I1)-dependent reduction of rusticyanin is currently under investigation in this laboratory.