Growth of Thiobacillus ferrooxidans on Elemental Sulfur

APPLIED AND ENVIRONMFNTAL MICROBIOLOGY. Aug. 1987. p. 1907-1912 Vol. 53. No. 8 0099-224)0/87/081907-06S52.00/0 Growth of Thiobacillus ferrooxidans ROMILIO T. ESPEJOt* AND on Elemental Sulfur PEDRO ROMERO Il.stitlito de Ivwestigaciones BiotnMdicas, Unihersidad Nacional Aitt6noinal de vxi(wo, Mexi-ho City, D.F. 04510, Me.vico Received 17 February 1987/Accepted 25 May 1987 Enhanced leaching of sulfide minerals by Thiobachills ferrcooxidanis may proceed in at least two different ways, direct and indirect. Indirect leaching entails oxidation of the minerals by the ferric iron produced by bacterial oxidation of soluble ferrous iron; direct leaching proceeds through a frontal attack by the bacteria on the sulfide, or on the sulfur deposited on the particles after oxidation of the sulfide by ferric iron, and requires intimate contact of the bacteria with the mineral (see. for example, the review by Hutchins et al. [5]). It is well documented that the physical attachment of thiobacilli to sulfur particles is necessary and plays an important role in the microbial oxidation rate of sulfur. Vogler and Umbreit (15) showed the necessity for direct contact in sulfur oxidation of Thiobacillus thiooxidatns. Schaeffer and Umbreit (11) identified phosphatidylinositol as a "wetting agent" of sulfur in T. thiooxidanis. Takakuwa et al. (13) showed that thiol groups may be essential for the adhesion process because adhesion was inhibited by sulfhydryl-binding reagents and this inhibition was released by sulfhydryl donors. They also showed that adhesion ability seemed energy dependent. On the other hand, Bryant et al. (3) have shown that for a different species of thiobacilli (T. albertis), adhesion to elemental sulfur is due to the organism's threadlike glycocalyx material interacting with the sulfur surface, and it is not dependent on physiological conditions such as pH, cellular energy, or peripheral cell envelope sulfhydryl groups. T. albertis was shown to colonize S" surfaces, forming microcolonies (3), and Laishley et al., using spherical S( prills, showed that the rate of S5 oxidation by T. alben'tis is a function of surface area per unit weight of sulfur (6). The mechanism, bioenergetics, and parameters of the growth of T. Jerrooxicdanis on ferrous iron are much better known than those for the growth of this bacterium on sulfides or elemental sulfur; this is mainly due to the diffi- culties for quantification of both the adsorbed bacteria and the amount of substrate actually available to the microorganism when utilizing the insoluble sulfides or sulfur as energy source. Since the importance of direct leaching of minerals is well recognized, understanding of the growth of the adsorbed bacteria may be of value for better management of commercial leaching operations. In this manuscript we describe the kinetics of growth of T. e'rrooxidans on St prills of known surface area (6). MATERIALS AND METHODS T. Jferroo.xidans ATCC 19859, purified after colony plating and passage for three or more times in medium containing elementary sulfur as the sole energy source, was used in these experiments. Growth was measured in basal medium MS9b containing the following salts (grams per liter): (NH4)2SO4, 0.1; K2HPO4 2H20, 0.04; MgSO4 7H20, 0.25; and FeSO4 7H20, 0.003. The pH was adjusted to 1.6 with H2S04, and the medium was then autoclaved. This basal medium was supplemented with elementary sulfur or FeSO4. Spherical prills, 0.15 to 0.17 cm in diameter, of elementary sulfur (sublimed; J. T. Baker) were prepared as described (6) and added to MS9b medium at a concentration of 0.04 g/ml (approximately 12 prills per ml). The prills were then sterilized by heating the medium at 105°C for 0.5 h on 2 successive days. FeSO4 was supplied as a 1.25 M FeSO4 7H.O solution previously sterilized by passing through a GS membrane filter (Millipore). Media were inoculated with the supernatant obtained, after allowing the rapid sedimentation at 1 x g of the sulfur prills, from a culture containing 1 x 108 to 4 x 108 bacteria per ml of supernatant; 0.01 ml of supernatant was added per ml of medium. Incubation was performed at 28°C in a Gyrotory water bath shaker at about 200 rpm. The total number of bacteria in suspension was measured with a Petroff-Hausser bacteria counting chamber in a phasecontrast microscope. Colony-forming bacteria were counted by plating in a two-layer agarose gel at a pH of 2.5, as described by Harrison (4); colonies were counted under a light microscope at x40 after 7 days of incubation at 28°C. Six fields, containing in total more than 100 colonies, were Corresponding author. Present address: Department of Chemical Engineering. Faculty of Physical and Mathematical Sciences. University of Chile. Casilla 2777. Santiago. Chile. * t 1907 Downloaded from https://aem.asm.org/ on April 20, 2021 by guest Growth kinetics of Thiobacillus ferrooxidans in batch cultures, containing prills of elementary sulfur as the sole energy source, were studied by measuring the incorporation of radioactive phosphorus in free and adsorbed bacteria. The data obtained indicate an initial exponential growth of the attached bacteria until saturation of the susceptible surface was reached, followed by a linear release of free bacteria due to successive replication of a constant number of adsorbed bacteria. These adsorbed bacteria could continue replication provided the colonized prills were transferred to fresh medium each time the stationary phase was reached. The bacteria released from the prills were unable to multiply, and in the medium employed they lost viability with a half-life of 3.5 days. The spreading of the progeny on the surface was followed by staining the bacteria on the prills with crystal violet; this spreading was not uniform but seemed to proceed through distortions present in the surface. The specific growth rate of T. ferrooxidans ATCC 19859 was about 0.5 day-1, both before and after saturation of the sulfur surface. The growth of adsorbed and free bacteria in medium containing both ferrous iron and elementary sulfur indicated that T. ferrooxidans can simultaneously utilize both energy sources. 1908 ESPEJO AND ROMERO APPL. ENVIRON. MICROBIOL. IKx -i 2CL. I x sz ._L C. xlo Cp usually counted, and the average value was used to calculate the total number of colonies contained in the 6-cm-diameter plates. Bacteria were labeled by adding about 0.1 mCi of radioactive phosphorus to the medium. 32P radioactivity was measured by counting Cerenkov radiation. To measure 32p incorporation in bacteria in suspension, a portion of the liquid (usually 0.2 ml) was diluted into 2 ml of cold MS9b medium and subsequently filtered through a 2.5-cm-diameter GS type Millipore filter, which was previously washed twice with 2 ml of MS9b medium; the bacteria retained in the wetted filter were then washed five times with 2.5 ml of MS9b medium. To measure 32p incorporation in adsorbed bacteria, the sulfur prills were washed in a tube five times with 2.5 ml of MS9b medium. Filters or prills were then placed in vials and counted in a Packard scintillation counter. The amount of radioactivity found at different times was corrected for the normal radioactive decay of 32p. For staining, prills washed as described above were placed in 2% crystal violet, prepared as for Gram staining (9), and, after 1 min, exhaustively washed with water. Micrographs of the stained prills were taken in a Zeiss DRC stereo microscope provided with an MC63 camnera, using a Kodak Plus X Pan, 125 ASA film. Ferrous and total iron were determined by the phenanthroline method (8). The 32p used was carrier-free Pi in dilute HCl solution, from Amersham International. RESULTS Kinetics of T. ferrooxidans growth was studied in batch cultures containing S° prills (6) as the sole energy source, in medium supplemented with radioactive phosphorus (32p). The increase of free bacteria, measured both by direct counting under a microscope and by plating, together with the incorporation of 32p in both liquid and S° prills is shown in Fig. 1A. The radioactivity in the material suspended in the liquid phase but retained by the 0.22-,um-pore-size filters parallels the increase of bacteria measured by either microscopy or plating, indicating that incorporation of 32p, determined in this manner, is a good estimate of bacterial growth. The following two control experiments indicated that the 32p detected in the prills after washing was due to bacterial growth. (i) Only a small amount of 32p (less than 5% of that found in colonized prills) was detected in the prills when the flask was incubated with heat-killed bacteria (heated at 90°C for 15 min), and this amount remained unchanged throughout the experiment. (ii) The 32p in the prills was of the same nature as the 32P found in bacteria present in the liquid phase. In this control experiment, bacteria pelleted from the medium and those adsorbed to washed prills were lysed with EDTA and sodium dodecyl sulfate and subsequently treated with pronase. This treatment released 90 to 100% of both 32p attached to the prills atnd 32p incorporated into sedimentable bacteria. Upon treatment with phenol-chloroform, 84 and 91% of the 32p in free and adsorbed bacteria, respectively, remained in the aqueous phase. Forty-eight and 36% of this radioactivity was precipitated with ethanol in free and adsorbed bacteria, respectively. Finally, the 32p precipitated with ethanol showed similar patterns after both samples were banded in CsCl; 15 and 11% of the precipitated isotope in free and adsorbed bacteria, respectively, banded at the buoyant density of T. ferrooxidans DNA. Considering, then, that all the phosphorus found after washing the prills is due to bacterial growth, the growth curve observed for T. ferrooxidans on elementary sulfur (Fig. 1) can be simply explained by an initial exponential growth of the adsorbed bacteria until a limit number of bacteria adsorbed per milliliter of culture (Ns) is reached. Upon saturation of the prills, a lineal increase of free bacteria follows due to the release of the progeny resulting from the successive replication of a constant number of Downloaded from https://aem.asm.org/ on April 20, 2021 by guest T I M E ( da ys ) FIG. 1. Growth curve of T. ferrooxidans on elementary sulfur prills suspended in MS9b medium containing radioactive phosphorus. (A) Symbols: 0, counts per minute retained in 0.22-p.m-pore-size filters after filtration and washing of 0.2 ml of medium; *, counts per minute retained per prill after washing (average of two prills); +, bacteria per milliliter as determined by microscopy; *, CFU per milliliter. (B) Symbols: 0, bacteria per milliliter of liquid, calculated from the amount of radioactive 32P retained in the filters; 0, bacteria in prills contained in 1 ml of medium, calculated from the amount of radioactive 32P found per prill. The radioactivity per bacterium was calculated from the amount of radioactivity retained in the filter and the number of bacteria determined in the liquid after 10 days at incubation. VOL. 53. 1987 7V. FERROOXIDANS GROWTH ON SULFUR 1909 A D FIG. 2. Stereo microscopy of sulfur prills incubated for the indicated times with T. fierroo.vidans and subsequently stained with crystal violet. (A) 0 h; (B) 8 days, (C) 30 days: and (D) 3 months with six successive transfers to fresh medium. adsorbed bacteria: this lineal increase is better observed in Fig. lB. The limit number Ns would be proportional to the number of adsorption sites on the surface of the sulfur prills; if the whole calculated surface, of about 0.08 cm-, of the prills (0.15 to 0.17 cm in diameter) were colonized by the bacteria (0.5 by 1 p.m), around 1.6 x 107 bacteria could be adsorbed to a single prill. Since the limit number refers to adsorbed bacteria per milliliter of culture, and since in the experiment shown in Fig. 1 there were 12 prills per ml, the expected Ns value would be 1.9 x 108 bacteria per ml if the whole surface were colonized, a value higher than that experimentally observed (Fig. 1B). The exponential phase of the growth curve would be described by equation 1 for limited growth: dNa - pNa Ns -N (1) Ns dt where Na is the number of adsorbed bacteria, Ns is the limit value of adsorbed bacteria, and ,. is the specific growth rate. The calculated p. from the initial slope of the curve showing the 3-P remaining in the prills (Fig. 1) is 0.5 day-'. The linear increase of bacteria in the liquid phase would be described by equation 2: dN1dXt p= Ns (2) where N- is the number of free bacteria in the liquid. The specific growth rate at this stage should then be calculated according to equation 3: AN1 (3) At x Ns p. calculated by this equation with the data in Fig. 1B is 0.58 day-t, suggesting that the specific growth rate does not change significantly once the available surface is saturated. The adsorption of the bacteria and their subsequent growth on the prills can be visualized by staining with crystal violet. Figure 2 shows the pattern of staining observed with prills incubated with T. jerrooxidans for different times: these patterns suggest that adsorption does not occur at random at any site and that colonization does not proceed uniformly on the surface of the prills; the bacteria instead seem to adsorb and proceed with colonization at limited sites, possibly distortions on the surface which are more favorable for bacterial adhesion and growth. As expected from the suggested model for growth of T. fe,rroo.xidans on sulfur, the bacteria in the liquid phase of the cultures did not increase after removal of the prills; instead, they lost viability with a half-life of 3.5 days (results not shown). On the other hand, the colonized sulfur prills continued releasing bacteria at a linear rate into the liquid phase when incubated with fresh medium. Figure 3 shows Downloaded from https://aem.asm.org/ on April 20, 2021 by guest C 1910 APPL. ENVIRON. MICROBIOL. ESPEJO AND ROMERO 30 X 20 X W0 e CL U .2 observed in the culture containing only trace amounts of ferrous iron (Fig. 5) (6), suggesting that T. ferrooxidans, or at least this strain, can simultaneously utilize both ferrous iron and elemental sulfur as energy sources. However, further experiments are necessary to obtain an unambiguous answer on the capacity of T. ferrooxidans to utilize both substrates simultaneously. 30% DISCUSSION il gi 0 0 20 X IOt 0 m cr. 2 4 6 0 2 4 6 the results obtained when prills from a culture at the lineal stage of growth were transferred to fresh medium with the same specific radioactivity present in the original culture. The calculated value for the specific growth rate from the data shown in Fig. 3 was 0.61 day-'. When the same experiment was repeated using different concentrations of colonized prills, the increase of free bacteria in the medium was directly proportional to the number of colonized prills per milliliter of medium. Similar values of 1L, calculated by equation 3, were obtained when using 8 (p. = 0.45 day-'), 4 (,u = 0.6 day-'), and 2 (,. = 0.67 day-) prills per ml. When the radioactive prills were incubated in nonradioactive medium, the 32p was released with a half-life of 1.5 days (Fig. 4). The half-life expected for the 32P remaining in the prills, if the bacteria continued dividing in this nonradioactive medium at a of 0.5 day-' and if the 32p were equally distributed between parent and daughter cells, would be equal to the duplication time, 0.69/0.5 day-', or 1.4 days. The decrease of 32P remaining in the prills, shown in the semilog plot in Fig. 4, did not follow the simple exponential decrease expected for a first-order process. Therefore the observed curve was considered the result of two components, first, that which represents the actual decrease of the amount of isotope due to duplication of the labeled cells in nonradioactive medium, and second, that due to the presence of isotope in nonreplicating bacteria, which would amount to the 10% still remaining after 12 days of incubation. Since this 32p would not be released from the prills, the amount of radioactivity present at the end of the experiment was subtracted from the amount found at earlier times; the curve obtained in this manner approached that expected for a first-order process with a half-life of 1.5 days. Figure 4 also shows the amount of radioactivity released from the prills after different times of incubation. To study the effect of ferrous iron on growth on elemental sulfur, the increase of free and adsorbed bacteria was measured in a batch culture containing, besides sulfur prills, 25 mM FeSO4. Even though the bacteria oxidized the ferrous iron (the arrow shows the time at which the ferrous iron had been completely oxidized), the adsorption and the increase of the bacteria adsorbed to the sulfur prills, measured by 32p incorporation, proceeded at the same rate 5X13' .' 0 0 o 0 + \+ + C 16 ~~~~~~~~~~+ a~~~~~~~~~~~~ I0 1 0 2 4 6 8 10 12 TIM E (days ) FIG. 4. Release of 32P from colonized SO prills transferred from radioactive to nonradioactive medium. SO prills from a 20-day culture containing radioactive phosphorus were transferred into nonradioactive medium. Upon transfer, the release of 32p from the prills and its accumulation in the medium was measured as described in Materials and Methods. 0, 32P remaining in prills; the values obtained for the two prills measured at each time are indicated in the figure to show differences usually obtained. +. 3'P remaining in the prills after subtracting the amount remaining after 12 days of incubation. 0, 32P from 0.15 ml of medium retained in the filters. Downloaded from https://aem.asm.org/ on April 20, 2021 by guest TI E (days) FIG. 3. Release of bacteria from colonized S° prills. SI prills from a 20-day culture containing radioactive phosphorus were transferred into fresh medium with the same specific radioactivity. Symbols are as explained in the legend of Fig. 1. Although it has been accepted that the physical attachment of thiobacilli to sulfur particles is necessary for growth when utilizing this substrate as the sole energy source, the particular growth kinetics expected in this condition has not been considered, even in cases when duplication times have been estimated (12, 14). The data on the kinetics of growth presented in this manuscript indicate that only the bacteria attached to the sulfur are able to grow and that the bacteria released are unable to carry on any further duplication. A further duplication of nonadsorbed bacteria could have been possible if the progeny were released with enough stored energy (or energy source), or if soluble sulfur compounds susceptible to further oxidation were released into the medium. The equations describing the growth kinetics observed may be oversimplified: the equation for the exponential phase implies that all the progeny become immediately adsorbed, and the equation for the lineal growth contains the assumption that the surface of the prills becomes saturated and that further increase of actively replicating bacteria is not significant after about 10 days of incubation. In fact, as suggested by the results, specifically the increasing staining of the prills and the amount of radioactivity found in the prills in Fig. 5, Ns seems to continue increasing, even though slowly. However, in spite of the oversimplifications, the equations seem appropriate for understanding growth on VOL. 53, 1987 T. FERROOXIDANS GROWTH ON SULFUR 1911 4D a. I o 0 / C.0 - 5X 04 E~~~~67 5 C. 0/4 60 3 /2 a r=~~~ ~ ~ I .0 a or a 2 I~~~~~~~TIM rgs FIG. 5. Growth curves of T. ferrooxidans in MS9b medium containing sulfur prills alone and sulfur prills plus 25 mM FeSO4. The arrow indicates the time at which the ferrous iron had been completely oxidized. Symbols: O and O- indicate counts per minute retained in 0.22-,um-pore-size filters per milliliter of liquid from medium containing, respectively, sulfur prills only and both sulfur prills and FeSO4; and *, counts per minute in prills contained in 1 ml of culture with sulfur and sulfur plus FeSO4, respectively. nonsoluble substrates and for workable estimation and comparison of specific growth rates. The estimation of the number of adsorbed bacteria, necessary to calculate specific growth rate, contains the assumption that the 32P retained after the prills are washed represents the number of actively replicating bacteria; this assumption implies that washing does not release adsorbed bacteria or that adsorption can be considered irreversible, and that the remaining 32p iS mainly that present in actively duplicating bacteria. Both conditions seem to hold: the amount of radioactivity remained practically constant after the second washing, and only less than 5% of the colonyforming bacteria, estimated according to the amount of radioactivity present at this time in the prills, were released in the subsequent washes (results not shown). Also, the isotope remaining in the prills after washing behaved like the isotope incorporated into bacteria present in the liquid phase. Finally, when the radioactive prills were incubated in nonradioactive medium, about 90% of the isotope was released at the rate expected for 32p incorporated in actively dividing bacteria. These types of experiments were not performed with prills incubated in radioactive medium for periods much longer than 20 days, and it is not possible to extrapolate this result for similar prills incubated for months, on which a considerably larger colonization was apparent after staining. The apparent pattern of colonization observed in the prills incubated for different periods after staining with crystal violet suggests that adsorption and growth of bacteria do not occur homogeneously on the surface but in regions which might offer advantageous conditions. Even on prills which had been incubated for longer than 4 months, it was possible to observe apparently intact regions remaining on the surface while in other places growth seemed to have progressed deeply. The erosion of sulfur crystal surfaces by T. ferrooxidans has been previously studied by electron microscopy (1, 10). The formation of pits in characteristic patterns, reported by Bennett and Tributsch (2), in pyrite crystal incubated for 2 years with T. ferrooxidans suggested that bacterial distribution is critically dependent on crystal structure and on deviations in the crystal order of the substrate. The observations reported by Lashley et al. (6) also suggested that bacterial action on elementary sulfur is more sensitive to the crystal microstructure of the sulfur than to its specific molecular composition. Since the sulfur employed could contain an appreciable amount of impurities, the number of deviations in the crystal order of the substrate is probably very large. The intensity and amount of staining observed in the prills suggest that the number of adsorbed bacteria continues increasing after the initial rapid colonization observed in the first days when measuring radioactivity in adsorbed bacteria. One of the most important observations from the described experiments seems to be that growth of adsorbed bacteria can continue after reaching the stationary phase if the colonized prills are transferred to fresh medium. It has been previously reported that T. ferrooxidans can simultaneously oxidize elementary sulfur and ferrous iron (7). The results presented in this manuscript indicate that T. ferrooxidans not only can oxidize but can efficiently grow on ferrous iron after successive growth in sulfur even in the presence of the latter. They also show that these bacteria can adsorb to sulfur in the presence of ferrous iron and that after depletion of the ferrous iron the number of bacteria in the liquid phase can continue increasing at a rate similar to that observed in the absence of ferrous iron, suggesting that T. ferrooxidans can simultaneously utilize both energy sources. A rapid early increment of bacteria in the liquid phase of the culture containing sulfur prills only was observed in this and in other experiments. In fact, this sudden increment is expected when the prills become saturated with bacteria with some degree of synchronization. The findings reported in this manuscript and the interpretation of the growth kinetics of T. ferrooxidans on elementary sulfur are important considerations for the management of mineral bioleaching. Downloaded from https://aem.asm.org/ on April 20, 2021 by guest a 11 E /~~~~~~~~~~~~~~~~~~~ /00 10 w 25 0 1912 ESPEJO AND ROMERO ACKNOWLEDGMENTS Staining and light microscopy of colonized sulfur prills were performed by M. Louzada. We are grateful to S. Lopez and E. Lamoyi for critical reading of this manuscript. 8. Muir, M. K., and T. Anderson. 1977. Determination of ferrous iron in copper-process metallurgical solutions by the Dphenanthroline colorimetric method. Metallurg. Trans. 8B: 517-518. 9. Paik, G., and M. T. Suggs. 1974. Reagents. stains, and miscellaneous test procedures, p. 930-950. In E. H. Lennette, E. H. Spaulding, and J. P. Truant (ed.), Manual of clinical microbiology, 2nd ed. American Society for Microbiology, Washington, D.C. 10. Schaeffer, W. I., P. E. Holbert, and W. W. Umbreit. 1963. Attachment of Thiohacill//s thiooxidans to sulfur crystals. J. Bacteriol. 85:137-140. 11. Schaeffer, W. I., and W. W. Umbreit. 1962. Phosphotidylinositol as a wetting agent in sulfur oxidation by T/1iobacill//s thioo.ridans. J. Bacteriol. 85:492-493. 12. Silver, M. 1970. Oxidation of elemental sulfur compounds and CO, fixation by Fel,robicilli/s ferrooxidans (Thiohbacill/s fierrooxidan.s). Can. J. Microbiol. 16:845-849. 13. Takakuwa, S., T. Fujimori, and H. Iwasaki. 1979. Some properties of cell-sulfur adhesion in Thiobacillus thiooxidans. J. Gen. AppI. Microbiol. 25:21-29. 14. Unz, R. F., and D. G. Lundgren. 1961. A comparative nutritional study of three chemoautotrophic bacteria: Ferrobaicill//s ferr ooxidains, T/hiobacillus Jerrooxidans and Thtiobaci/us thiooxidain1s. Soil Sci. 92:302-313. 15. Vogler, K. G., and W. W. Umbreit. 1941. The necessity for direct contact in sulphur oxidation by Tliiobacill//s thiooxidains. Soil Sci. 51:331-337. Downloaded from https://aem.asm.org/ on April 20, 2021 by guest LITERATURE CITED 1. Baldensperger, J., L. J. Guarraia, and W. J. Humpheys. 1974. Scanning electron microscopy of thiobacilli grown on colloidal sulfur. Arch. Microbiol. 99:323-329. 2. Bennett, J. C., and H. Tributsch. 1978. Bacterial leaching patterns on pyrite crystal surfaces. J. Bacteriol. 134:310-317. 3. Bryant, R. D., J. W. Casterton, and E. J. Laishley. 1984. The role of Thiobacillus albertis glycocalyx in the adhesion of cells to elemental sulfur. Can. J. Microbiol. 30:81-90. 4. Harrison, A. P., Jr. 1984. The acidophilic thiobacilli and other acidophilic bacteria that share their habitat. Annu. Rev. Microbiol. 38:265-292. 5. Hutchins, S. R., M. S. Davidson, J. A. Brierly, and C. L. Brierley. 1986. Microorganisms in reclamation of metals. Annu. Rev. Microbiol. 40:311-336. 6. Laishley, E. J., R. Bryant, B. W. Kobryn, and J. B. Hyne. 1986. Microcrystalline structure and surface area of elemental sulphur as factors influencing its oxidation by Thiobacill/s albe rtis. Can. J. Microbiol. 32:237-242. 7. Landesman, J., D. W. Duncan, and C. C. Wolden. 1966. Oxidation of inorganic sulfur compounds by washed cell suspension of Thiobacl/i/s ferrooxidans. Can. J. Microbiol. 12:957-964. APPL. ENVIRON. MICROBIOL.