Isolation and Physiological Characterization of

Journal of General Microbiology (1988), 134, 1407-1417.

Printed in Great Britain

1407

Isolation and Physiological Characterization of Autotrophic Sulphur Bacteria Oxidizing Dimethyl Disulphide as Sole Source of Energy By N E I L A. S M I T H A N D D O N P. KELLY* Department of Biological Sciences, University of Warwick, Coventry CV4 7AL, UK (Received 2 November 1987; revised 4 February 1988) The isolation of a number of strains of bacteria able to grow on dimethyl disulphide and dimethyl sulphide as sole source of energy is described. The isolates came from diverse habitats, including soil, peat, marine mud and a freshwater pond. The isolates were morphologically and physiologically best described as thiobacilli, capable of growth as Calvin cycle autotrophs on inorganic sulphur compounds, methylated sulphides or thiocyanate. They could not grow heterotrophically or methylotrophically. One isolate (E6) was examined in detail. Substrate oxidation kinetics indicated that methanethiol, sulphide, formaldehyde and formate, but not dimethyl sulphide, could be implicated as intermediates in dimethyl disulphide metabolism. Apparent K , values for the oxidation of dimethyl disulphide and methanethiol were 2.5 and 3.2 p . respectively. ~ Growth yields in chemostat culture on dimethyl disulphide with and without thiosulphate indicated that energy conservation was probably coupled to the oxidation of formaldehyde and sulphide (derived from dimethyl disulphide via methanethiol) to C 0 2 and sulphate. Maximum growth yield (Y,,,) on dimethyl disulphide was 17 g cell-carbon per mol of dimethyl disulphide. At one dilution rate (0.078 h-l), the biomass of a culture limited by dimethyl disulphide increased when thiosulphate was also supplied, indicating a thiosulphatedependent yield of 2-45 g cell-carbon mol-l. This is the first demonstration of the isolation of organisms into pure culture that are capable of growth on dimethyl disulphide as sole energy substrate, and of degrading it completely to C 0 2 and sulphate.

INTRODUCTION

Volatile organic sulphur compounds are now recognized as having an important role in the biogeochemical cycling of sulphur through the atmosphere (Bremner 8z Steele, 1978; Kelly, 1988). The most important of these are carbon disulphide, carbonyl sulphide (COS), methanethiol (MT), dimethyl sulphide (DMS) and dimethyl disulphide (DMDS). DMS in particular has been shown to be produced in large amounts in aquatic environments (Zinder et al., 1977; Bechard 8z Rayburn, 1979; Ferek et al., 1986; Dacey 8z Wakeham, 1986; Franzmann et al., 1987), and large scale biological production of MT, DMDS (and probably tri- and tetrasulphide) have been reported (Bremner & Steele, 1978; Konig et al., 1980; Deprez et al., 1986; Drotar et al., 1987). DMDS was shown to be produced by numerous bacterial strains isolated from activated sludge (Tomita et al., 1987) and to be reduced to methanethiol in salt marsh sediments (Kiene & Capone, 1988). Methylated sulphides have been shown to be absorbed by soils and to be oxidized by Thiobacillus-likebacteria in biofilters treating wood pulping effluents (Bremner 8z Steele, 1978; Sivela, 1980). Anaerobic oxidation of DMS to dimethyl sulphoxide (DMSO) by purple phototrophs has also been reported (Zeyer et al., 1987). Some Hyphomicrobium and Thwbacillus strains have been shown to grow aerobically on DMS, oxidizing it to C 0 2and sulphate (De Bont _ _ _ _ _ _ _ _ _ _ _ _ ~

Abbreviations: DMS, dimethyl sulphide; DMDS, dimethyl disulphide; DMSO,dimethyl sulphoxide; MT, methanethiol; COS, carbonyl sulphide; TOC, total organic carbon. 0001-4508

0 1988 SGM Downloaded from www.microbiologyresearch.org by IP: 86.151.64.96 On: Fri, 20 Nov 2015 11:36:41

1408

N . A . SMITH AND D . P . K E L L Y

et al., 1981 ; Suylen & Kuenen, 1986; Suylen et al., 1986; Kanagawa & Kelly, 1986). None of these was originally isolated on DMS as the enrichment substrate, because of its reported toxicity (Suylen & Kuenen, 1986): the hyphomicrobia were obtained originally as DMSO-users, and the Thiobacillus as a component of a mixed culture able to degrade a pesticide breakdown product (Kanagawa et al., 1982). The hyphomicrobia oxidized DMS through MT, formaldehyde and formate to C 0 2 , assimilated carbon by means of the serine pathway and also obtained energy from the oxidation of the sulphide moiety to sulphate (De Bont et al., 1981;Suylen et al., 1986). The Thiobacillus was an obligate autotroph, obtaining carbon exclusively by the fixation of C 0 2 ,while oxidizing DMS completely to C 0 2 and sulphate to provide energy (Kanagawa & Kelly, 1986). There have been no reports of any organism able to oxidize DMDS as its sole source of energy : Thiobacillus thioparus TK-m (Kanagawa & Kelly, 1986; N. A. Smith, unpublished data) and Hyphornicrobium S (De Bont et al., 1981) cannot use it, although they do grow on DMS; and a mixed culture enriched on DMSO could only oxidize DMDS at 8 % of the rate of DMS oxidation (Suylen & Kuenen, 1986). This suggested that the ability to grow on DMDS was not a property necessarily common to the DMS-using bacteria and indicated that specific enrichment using DMDS as a substrate was necessary to detect DMDS-oxidizing strains. We now report the isolation into pure culture of bacteria able to use DMDS, and DMS, as their sole source of energy, and discuss the physiological relationships of bacteria oxidizing organic and inorganic sulphur compounds to provide energy for growth. METHODS

Enrichment culture sample sources. The following materials were used to initiate enrichment cultures for the isolation of organic-sulphide-oxidizingbacteria : E 1, commercially available gardeners' peat ; E2, red clay-soil, Burton-upon-Trent area; E3, garden compost; E4, cattle manure; E5, marine mud, Plymouth Sound; E6, pond water, Gibbet Hill, University of Warwick campus; E7, moss from a deodorization unit; E8, marine sediment, Salcombe Bay. Culture media. The mineral salts medium used for enrichment cultures, and for all batch and continuouscultures of pure isolates, contained (g 1-l): KH2P04(2); K2HP04(2); NH4C1(0.4); Na2C03(0.4); MgC12. 6 H 2 0 (0.2); 3 ml vitamin mixture (Kanagawa et al., 1982) and 1 ml trace metal solution (Tuovinen & Kelly, 1973); initially at pH 7.1. Substrates were added as described below. Solid medium for plates and slopes was prepared by adding substrate and agar (1-5%, w/v) to the mineral salts medium. Pure cultures were maintained on agar medium with 20 mM-thiosulphate. Enrichment cultures. The following were added to 200 ml unsterilized mineral salts medium, with or without yeast extract (0.1 g l-l), in 500 ml bottles: soil, compost and peat, 20 g air dried wt; manure, 5 g wet wt; marine mud and sediment, 10 g wet wt; water samples were diluted 50 :50. The bottles were then sealed with rubber Subaseal stoppers. DMS or DMDS were injected through the Suba-seal using a Hamilton syringe to give an initial concentration of 0.5 mM. The bottles were incubated at 30 "C for 3 d during which time any fall in pH was noted and cultures readjusted to pH 6-8-7.0 with NaOH. After 3 d incubation the flask contents were allowed to settle for 30 min after which 40 ml of the upper phase was replaced with 40 ml from mineral salts medium. The bottles were then re-stoppered and a further 0-5 mM-DMS or DMDS injected. This medium replacement procedure was repeated every 3 d until the upper phase became visibly turbid at which point the substrate concentration was progressively raised by 0.5 m steps to a final concentration of 2 mM, over a period of 2-4 months. When the enrichment cultures were oxidizing 2 mM substrate in 2-3 d, 25 ml of the upper phase was removed and transferred to sterile 250 ml Quickfit flasks containing 25 ml sterile mineral salts medium. After sealing with Suba-seal stoppers DMS or DMDS (2 mM) were injected. Subculturing was subsequently carried out every 5-7 d (lo%, v/v, inocula). Isolation and purification of DMS- and DMDS-oxidizingorganisms was attempted by streaking liquid culture samples on to solid media containing one of the following: thiosulphate (20 m);methylamine or dimethylamine (20 mM); formate (20 m); glucose (10 m);or on to nutrient agar. Plates were incubated at 30 "C. Single colonies were then transferred to DMS or DMDS liquid medium (10 ml). This procedure was repeated several times on isolates which grew successfully until purity was assured. Isolation of pure cultures by this enrichment procedure took up to 10 months. Oxidation of one-carbon and sulphur compounds. Oxygen uptake rates were determined using a Teflon-coated Clark-type oxygen electrode cell (Rank) linked to a chart recorder. Organism suspensions (2 ml; 38 pg protein D = 0.074 h-l). Substrate-dependentoxygen ml-l) were taken directly from a DMDS-limited chemostat (2 m ; uptake was determined at 30 "C, the reaction being initiated by the injection of substrate (5-20 pl). Results are expressed as nmol O2min-l (mg protein)-* and corrected for endogenous oxygen uptake.

Downloaded from www.microbiologyresearch.org by IP: 86.151.64.96 On: Fri, 20 Nov 2015 11:36:41

Oxidation of methylated sulphides

1409

Analytical methods. Growth of cultures was monitored by measuring optical density values at 440 nm in a Unicam SP 1700 spectrophotometer.Samplesfrom the chemostat were used to prepare calibration curves relative to OD440 of biomass in terms of (mg 1-l): total organic carbon (TOC), protein or dry wt. Dry wts were determined by drying a suspension of cells to constant weight at 105 "C. Pellets of cells centrifuged from 4 ml culture were assayed for TOC and protein. TOC was determined in aqueous suspensions of pellets using a Beckman Total Organic Carbon Analyzer, model 905B. Protein was determined according to Lowry after dissolving the cells in 0.5 M-NaOH for 30 mins. MT solutions (approximately 300 mM) were prepared as described by Suylen et al. (1986), the actual concentration being determined by TOC measurement and diluted to 2.7 mM for use. Sulphate was determined by atomic absorption spectrophotometry, by measuring residual barium following sulphate precipitation by barium chloride (Varian Manual). DMDS in aqueous solution (1 ml) was extracted with hexane (2 ml). The DMDS peak height of the upper (hexane) phase was measured at 260 nm against a hexane blank, and the DMDS concentration determined by reference to a calibrationcurve. Thiosulphate was determined according to Kelly et al. (1969). Ribulose-1,s-bisphosphatecarboxylaseactivity was assayed as described by Kelly & Wood (1982), except that the cell permeabilizing agent used was cetyltrimethylammoniumbromide (Leadbeater et al., 1982) (0-1%,v/v) rather than Triton X-100. Growth of organisms in batch and continuous culture. Batch cultures (50 ml) were grown in 250 ml Quickfit flasks sealed with rubber Suba-seal stoppers. DMS, DMDS, or CS2 (2 mM unless otherwise stated) were injected into cultures using a Hamilton syringe. MT and COS were added either as unsterilized gas or in solution. Elemental sulphur was added to the mineral salts medium as 'flowers of sulphur' (10 g 1-l). Sulphidewas filter sterilized prior to addition to sterile mineral salts medium. All other substrates were sterilized in the mineral salts medium. To minimize any auto-oxidation of sulphide, several additions of freshly prepared sulphide (1-2 m)were made during growth. Possible auto-oxidationof MT to DMDS could not be discounted but appeared to be unimportant. Radiolabellingexperiments to measure the incorporation of 14C02into isolate E6 growing on 2 m-DMDS were done in sealed flasks as described above. Cultures (50 ml), from which Na2C03 had been omitted, were supplemented with either 2-5or 10 m-NaH14C03.Samples (2 ml) were filtered through Whatman membrane filters (0.2 pm), washed with water and counted in Optiphase Safe scintillant (LKB; 10 ml) using a Beckman LS 7000 spectrometer. A chemostat culture (750 ml culture volume) of isolate E6 was established on 2 m-DMDS in a 1 litre waterjacketed glass vessel (LH Engineering).The culture was maintained at a temperature of 30 & 1 "C with stirring at 750 r.p.m. and aeration with air at 600 ml min-l. pH control was not necessary since the oxidation of 2 m-DMDS produced a steady state culture pH of 6-7. Unsterilized DMDS was added to the medium reservoir to give a final concentration of 2 mM. Constant stirring of the medium in the reservoir ensured homogeneous solution of the DMDS and prevented the Occurrence of any concentrationgradients. An air filled balloon was attached to the air inlet to prevent loss of DMDS and allow for decrease in volume of the residual medium in the reservoir. Black butyl or Viton tubing (Watson-Marlow,Falmouth, Cornwall, UK) was used for all connections from the reservoir to the vessel. A Suba-seal port was positionedclose to the medium inlet to facilitate sampling of the medium and to enable accurate determinationof the input DMDS concentration.The DMDS-medium was forced into the culture through the aeration tube below the impeller, thus ensuring instant mixing and the prevention of any loss of DMDS from the culture. No d o u r of DMDS was detectable in the effluent air flow. Additional substrates were metered separately into the vessel. Chemicals. MT, CS2 and thiocyanate were obtained from Fisons; DMS and COS from Aldrich; DMDS from Sigma; and Na, 14C03from Amersham.

RESULTS

Enrichment and isolation of organisms capable of growth on DMS and DMDS Enrichment cultures on each of DMS and DMDS were run with inocula from all eight materials given in Methods. The soil enrichments (E2) were least successful and were discarded. Enrichments for the peat (El) and Salcombe sediment (ES) samples were successful on DMS, but pure cultures have not yet been obtained. Considerable difficulty was experienced in attempting to obtain pure cultures from enrichments : using various heterotrophic media or mineral salts medium with methylamine resulted in the production of numerous colonies on plates, but these failed to grow when returned to liquid medium with DMS or DMDS as sole substrates. Single colony production from the enrichments was attempted by inoculating mineral salts agar and incubating at 22 "C in sealed jars containing air saturated with DMS or DMDS. Tiny colonies were observed at 7-10 d, but it was not possible to transfer these into liquid culture. The only successful isolates of DMS- and DMDS- using bacteria came from colonies grown on thiosulphate-mineral-salts agar. It was concluded that the growth seen on Downloaded from www.microbiologyresearch.org by IP: 86.151.64.96 On: Fri, 20 Nov 2015 11:36:41

1410

N . A . SMITH A N D D. P. K E L L Y

Table 1. Growth of isolates E3 to E7 on inorganic and organic sulphur compounds E3, E4 and E5 were isolated from DMS-enrichments and E6 and E7 from DMDS-enrichments. Growth after 8 d incubation Substrate (mM) Sulphide (2) Sulphur (10 g l-l)* Thiosulphate (10) Tetrathionate (10) Thiocyanate (3) Methanethiol (2) Dimethyl sulphide (2) Dimethyl disulphide (2) Carbon disulphide (2) Carbonyl sulphide (2) Thiourea (2)

+

r

E3

+ + + + + ++ -

E4

+ + + + + + +

++ +

E5

++ + + + + ++-

E6

+ + + + + + + ++-

1

E7

+ + + + + + + ++-

* Growth on sulphur was scored or - by visual inspection of turbidity after allowing sulphur to settle. No growth of isolates E3 and E5 occurred during 3 weeks incubation.

organic nutrient media was of bacteria incapable of growth on methylated sulphides. These bacteria could have been sustained in the enrichments as commensals using excreted products from the DMS- and DMDS-using organisms, which could not themselves grow on the organic media. Consequently, five organisms were isolated in pure culture by transfer from thiosulphate agar. These organisms were coded as for the sample sources: E3, E4 and E5 enriched originally on DMS, and E6 and E7 on DMDS. All five were routinely maintained as autotrophic cultures on thiosulphate as the sole source of energy. Cultures could also be maintained on agar supplemented with DMDS (2-3 mM) or DMS (2-4 mM). Characteristics of strains E3-E7 All five pure cultures were small, Gram-negative rods (0.5 x 1-2 pm) capable of growth on inorganic and organic sulphur compounds (Table 1). None grew on sucrose, fructose, glucose, acetate, citrate or formate (each at 10 mM), formaldehyde (0.5 mM), methylamine, dimethylamine or trimethylamine (5 mM), methanol, dimethyl sulphate, DMSO, ethanethiol, diethyl sulphide, diethyl sulphate (each at 2 mM) or methane. Physiologically, these strains resembled chemolithotrophic thiobacilli in that all grew on thiosulphate and sulphide and some also used elemental sulphur or thiocyanate (Table 1). In contrast to most thiobacilli, all were able to grow on methylated sulphides and COS, but not on CS2. Strain E4 alone grew on thiourea and could also use thioacetamide. These organisms are the first to be reported to be able to grow on DMDS as the sole substrate in pure culture. Using organisms previously grown on DMDS, simultaneous measurement of increase in cell carbon and fixation of I4CO2 during the first 8 h growth of strains E3 and E4 on DMS and E5, E6 and E7 on DMDS showed that up to 93% of the cell carbon of these strains was derived directly from C 0 2 . This indicated that growth of all five strains was autotrophic rather than methylotrophic in terms of the pathway of carbon assimilation. A more detailed study was made of strain E6 originally isolated from a DMDS enrichment, which grew on all of the methylated sulphides and inorganic sulphur substrates tested. This strain exhibited lags of 5-6d when transferred to DMS, DMDS or thiosulphate after prior growth on one of the other substrates. Batch culture of strain E6 on DMDS :physiological characteristics Strain E6 had a pH optimum of 6.7-6.9, although there was very little difference in growth rates ( p = 0.08409 h-l) on either DMDS (2 m ~ or)thiosulphate (5 m ~ between ) initial pH values of 6.2-7.2. No growth occurred below pH 5.5 or above pH 8.2. A final pH of 6.3-6.4 was


1411

Oxidation of methylated sulphides 0.5

4

2.0

0.4

1.5

z E

n

W

n Q

3

3 0.3

W

2

8

5

8

3zE

1.0

3

0.2

g

i3

0.5

s 0

2 2 a c (

m

1

20

40 60 Time (h) Time (h) Fig. 1 . Growth of strain E6 in batch culture on DMDS. The inoculum culture was grown to the end of the log phase on DMDS (2 m).(a)Growth, as increase in optical density, shown for different initial concentrations (mM) of DMDS: 0.5 1-5 (O), 2.5 (A) and 5.0 (a).(b) Growth DMDS disappearance ( 0 )and sulphate formation ( 0 )shown for a culture growing on an initial concentration of 2-1 mM-DMDS.

u),

a),

observed following the complete consumption of 2 mM-DMDS or 5 a-thiosulphate. Weak anaerobic growth and denitrification was observed with thiosulphate and thiocyanate, but not with DMDS. Nitrate was reduced only as far as nitrite. Growth in liquid batch culture on DMDS commenced at initial concentrations up to 5 m~ (which is close to saturation of aqueous solution), but the initial rate of growth was depressed by increased concentrations of DMDS, indicating it to be an inhibitory substrate. Using eight concentrations of DMDS between 0-5 and 5 mM, initial specific growth rate (1 1-21 h after inoculation) decreased from 0.087 h-l (at 0-5 mM) to 0.07 (2 m~), 0.053 (4 mM) and 0.04 at 5 mM-DMDS. Final biomass levels were, however, proportional to the amount of DMDS supplied (Fig. la). The maximum specific growth rate (pmax) observed was about 0.09 h-l. Growth on 2 mM-DMDS showed sulphate formation to be correlated with DMDS disappearance and increase in biomass. Growth and sulphate formation continued at a slower rate following the disappearance of dissolved DMDS (Fig. 1b). Initially, about 87% of the added DMDS was in solution, the remainder being in the vapour phase in the flask. After 28 h (Fig. 1b) only 3% of this was left in solution, and biomass had increased to 80% of the final value achieved, but only 61% of the final concentration of sulphate had been produced. At 39 h, when dissolved DMDS was no longer detectable, increase in biomass was at 94% of the final level and sulphate production was at least 86%complete (Fig. 1b). This apparent continuation of growth after disappearance of dissolved DMDS could have been due to partition of DMDS between the liquid and vapour phases and its consumption in the final phase of growth at a rate determined by its solution rate. Additionally, there might have been some accumulation and subsequent use at a slower rate of intermediates from DMDS oxidation. Within the error of measurement, 2 mol sulphate were produced per mol DMDS supplied, as would be expected if both sulphur atoms of DMDS were oxidized to sulphate. The overall increase in biomass (Fig. 1b) for the complete consumption of DMDS was 27 mg TOC 1-l equivalent to a yield of 13.5 g cell carbon (mol DMDS)-l or 6.75 g cell carbon (mol sulphate produced)-'. From six determinations using cultures grown on 1-4 m~-DMDS,yield was 13.6 0.8 g cell carbon (mol DMDS)-l.


1412

N . A . SMITH A N D D. P. K E L L Y

Table 2. Growth and C 0 2fixation by strain E6 growing on 2 mM-DMDS in batch culture in the presence of 10 mM-NaH14C03 Time (h)

Optical density (440 nm)

0 8 24 32 48 56* 72 80* 96

0.027 0.030 0.052 0.101 0-213 0.227 0.348 0-352 0.483

Increase in biomass (mg TOC 1-l) 0 0.3 3.3

12-1

30.0 32.5

51.0

52.0 73.0

14C02fixed? (

~

c.p.m. ml-' 452 3 890 1 1 457 28 42 1 30 340 46 101 43 872 51 245

~

-

1

mg cellcarbon 1-l

Apparent proportion of TOC derived from 14C02

-

-

0.3 3.3 7.7 19.0 20.3 (28*4)$ 30.8 29.3 (52-8)$ 34.3 (75.3)$

93 79 63 63 62 @7)§ 59 56 (101)g 47 (103)§

* Indicates time of addition of a further 2 mM-DMDS. t Initial specific activity of 14C02, 17957 c.p.m. pmol-I.

$ Figures in parentheses are values for mg cell-carbon 1-l obtained from 14C02,calculated using specific activities based on the assumption that all the DMDS present was converted to C02. The specific activities at 56, 80 and 96 h were calculated as 12826, 9976 and 8162 c.p.m. (pmol C02)-l respectively. This overestimates 14C02 fixation, as specific activity obviously declines progressively rather than instantaneously for each quantity of DMDS oxidized, but does allow an estimate of the maximum amount of carbon derived directly from C02 fixation. 8 Proportions of total cell-carbon that could be derived from C 0 2 fixation, based on the above assumptions.

Fixation of 14C02during growth of strain E6 on DMDS Using cultures previously maintained through successive subculture on DMDS as sole substrate, simultaneous measurement of growth and 4C02 fixation showed that initially virtually all the cell carbon was derived from C 0 2 (Table 2), but the amount of 14Cfixed relative to increase in biomass decreased during growth. This phenomenon was due to dilution of the added 14C02by oxidation of the DMDS to C o t . Assuming complete oxidation to C 0 2of the 2, 4 and 6 mM-DMDS added at the times indicated in Table 2, the specific activity of the available 14C02could be recalculated and the actual C 0 2 fixation estimated. This showed that C 0 2 fixation was probably the only route for carbon assimilation during growth on DMDS (Table 2). Repeating this experiment with different initial amounts of 14C02provided further proof: the apparent fixation of 14C02was greatly decreased when 2.5 rnrather than 10 rnM-Hl4CO3was present, but calculating the true specific activities after the presumed complete oxidation of DMDS to C 0 2 indicated that at least 80% of the cell carbon was provided from C 0 2 fixation. These results and the demonstration that the proportion of cell carbon intially obtained from C 0 2was over 90% (Table 2), supports the assumption that the DMDS was completely oxidized giving 2 mol C 0 2 per mol DMDS consumed. Ribulose-I ,5-bisphosphate carboxylase in strain E6 Assay of cells taken from a DMDS-limited chemostat culture (see below) by the permeabilized whole cell procedure showed a specific activity of 33 nmol C 0 2 fixed min-l (mg dry wt)-l. Activity was linear for at least 30 min. This level of activity is adequate to acount for autotrophic growth at the dilution rate of 0.074 h-l. Growth of strain E6 in chemostat culture Steady state cultures were established on 2 mM-DMDS as the limiting substrate at dilution rates between 0.046 and 0.081 h-l, between which the steady state growth yield (Y) increased from 13.0 to 15.6 g cell carbon (mol DMDS)-l. The true growth yield (Y,,,) was estimated from plots of 1 / Y versus 1 /D and of the specific rate of DMDS consumption (qDMDS) versus D (Fig. 2a, b). DMDS consumption was estimated both from direct assay of the input medium and from the amount of sulphate produced in the steady state culture. Good correlation existed between the Downloaded from www.microbiologyresearch.org by IP: 86.151.64.96 On: Fri, 20 Nov 2015 11:36:41


1413

6 12 18 Reciprocal of dilution rate (h)

"

I

I

I

0.07 0.08 0.06 Dilution rate (h-') Fig. 2. Growth of strain E6 in chemostat culture. Steady state cultures on 2 m~-DMDSwere obtained at several dilution rates. (a) Specific rate of DMDS consumption [qDMDS = mmol DMDS consumed h-I (g cell-carbon)-l] versus dilution rate. DMDS consumption was estimated from both direct measurementof input DMDS concentration(0)and determinationof steady state production rates for sulphate (a).The line was drawn by linear regression fitting of the ( 0 )data points. Extrapolation to zero dilution rate gives a qDMDS value (the maintenance coefficient ) of 1.1. (b) Reciprocals of yield [g cell-carbon (mol DMDS)-l] versus dilution rate. The slope gives a qDMDS value (the maintenance coefficient) of 1.1. Symbols are as in (a).

two estimates of DMDS supplied, and also demonstrated that DMDS was quantitatively oxidized to sulphate in steady state cultures. No elemental sulphur was produced by the cultures from DMDS. Y, was about 17 g cell carbon (mol DMDS)-l. A steady state culture ( D = 0-078h-l) on 1-5mM-DMDS was simultaneously supplied with 2 mM-thiosulphate. Both substrates were completely oxidized. The steady state biomass rose from 22.5 to 27-5mg cell carbon 1-l, accompanied by a decrease from pH 6.7 to pH 6.3. This indicated a thiosulphate-dependent yield of 2.45 g cell-carbon per mol thiosulphate oxidized. Analysis of strain E6 cells removed from the chemostat culture showed it to contain carbon at 47% and protein at 62% of the dry wt, respectively. Kinetics of sulphur and carbon compound oxidation by strain E6 grown in the DMDS-limited chemostat culture Suspensions of strain E6 cells showed concentration-dependent oxidation of DMDS, MT, thiosulphate and sulphide, and could also oxidize formate and formaldehyde (Table 3). DMS and methylamine were not oxidized by DMDS-grown cells. Using different amounts of DMDS, the stoichiometry of oxygen uptake was found to be 6.5nmol O2 (nmol DMDS)-', which indicated the complete oxidation of DMDS as follows

(CH3)2S2

+ 6*502+2C02+ 2H2S04 + H20

Oxidation rates (t,) were measured using the following substrate concentration ranges (s) : DMDS (2-5-30 VM),MT (1-20 p ~ )formaldehyde , (10-500 p ~and ) formate (0.5-5 mM). K, and V, values estimated from Lineweaver-Burk plots (Fig. 3), and plots oft, versus t,/s and s/v versus s are summarized in Table 4. Strain E6 showed high affinities for the methylated sulphides (Fig. 3a, b), which were very rapidly oxidized, but lower affinity and oxidation rates for formaldehyde and formate (Fig. 3c, d ) . Downloaded from www.microbiologyresearch.org by IP: 86.151.64.96 On: Fri, 20 Nov 2015 11:36:41

1414

N. A. S M I T H A N D D . P . K E L L Y

m-

s

'c 200

-2.0-1.0 0 1.0 2.0 Reciprocal of substrate concn (mM-')

0.01 0.02 0.03 1 Substrate concn (mM) 0

I

I

0 250 500 750 1000 Reciprocal of substrate concn (mM-') -250

20 40 60 80 100 Reciprocal of substrate concn (mM-')

Fig. 3. Kinetics of oxidation of DMDS, MT, formaldehyde and formate by strain E6, taken from substrate-limited chemostat culture on DMDS. (a) Lineweaver-Burk plot of reciprocals of oxidation rates versus reciprocals of substrate concentration for DMDS ( 0 )and MT (0).(b) Oxidation rate versus substrate concentration for DMDS (0)and MT (0).(c) and (d) Lineweaver-Burk plots for formaldehyde ( 0 )and formate (0).

Table 3. Oxidation of one-carbon and sulphur compounds by strain E6 previously grown in chemostat culture on DMDS (2 mM ;0.074 h-' Substrate Sulphide Thiosulphate Methanethiol Dimethyl disulphide Formaldehyde Formate Dimethyl sulphide Methylamine Dimet hylamine

Concn. (p~)

Oxygen uptake rate* [nmol O2 min-1 (mg protein)-']

1000

588 410 581 30 1 51 33 0

500 7

20

500

5000 5-50 5-5000 5-5000

0 0

* Corrected for endogenous rate of oxygen uptake [6-7

nmol O2 min-' (mg protein)-'].

Table 4. Kinetic constants for the oxidation of four substrates by strain E6, previously grown in DMDS-limited chemostat culture (2 mM ;0.074 h- ) Substrate Methanethiol Dimethyl disulphide Formaldehyde Formate

Ks

vmax*

( p ~ ) [nmol O2 min-l (mg protein)-']

3-2

2.5

84 678

* Values are given as mean ( ~ S E for ) estimates using plots of

1003 60 348 f 18 76 f 10 35 f 0.5 l / v versus l/s, v versus u/s, and s/v versus s.



1415

DISCUSSION

This is the first reported demonstration of aerobic bacteria capable of the complete oxidation of DMDS as the sole source of energy for growth. The isolation of DMDS-degrading strains from a range of natural environments is of considerable interest, not only in showing how commonly this property occurs, but in indicating that DMDS is probably an important and ubiquitous intermediate in the biogeochemical cycling of sulphur. Because of the difficulty of obtaining colony development from enrichments using plates incubated in air containing DMDS (or DMS), all these DMDS-using strains have been obtained through their ability to grow as chemolithotrophic autotrophs with thiosulphate as the sole source of energy. No DMDS- or DMS-users were obtained by selection on methylamine medium. Consequently the isolation technique used exerted a strong selective bias towards the detection of thiosulphateoxidizing 'Thiobacillus-like' bacteria, able to use methyl sulphides, like T. thioparus TK-m (Kanagawa & Kelly, 1986) and Thiobacillus MS1 (Sivela, 1980). The results do not therefore preclude the possibility that DMDS (and DMS)-users, able to grow as methylotrophs but unable to grow autotrophically on thiosulphate, were also present in the samples tested. Such organisms might have resembled the DMSO and DMS- oxidizing hyphomicrobia described earlier (De Bont et al., 1981; Suylen & Kuenen, 1986; Suylen et al., 1986), but their presence in the enrichment cultures could only have been proved by an exhaustive screening of large numbers of bacteria able to produce colonies on agar media with DMDS, DMS, DMSO or, in some cases, methylamine as substrate (De Bont et al., 1981;Suylen & Kuenen, 1986). The inability of all our strains to use DMSO, and the inability of the hyphomicrobia to grow on thiosulphate, and in one case at least on methylamine or DMDS (De Bont et al., 1981), means that conclusive demonstration of the different types of bacteria that might degrade methyl sulphides in specific enrichments is not simple. We can tentatively conclude that the autotrophic strains described here were at least the major DMDS-users in their various habitats. The maximum specific growth rate exhibited by strain E6 on DMDS was about 0-09 h-l, which is at least equal to that of Hyphomicrobium EG on DMS, and considerably greater than that of Hyphomicrobium S (De Bont et al., 1981; Suylen & Kuenen, 1986). All the strains isolated exhibited a very narrow range of substrates used, although not as restricted as Hyphomicrobium S, which used only DMSO and DMS (De Bont et al., 1981). The ability to grow autotrophically using energy from oxidizing sulphide, sulphur, thiosulphate or tetrathionate under aerobic conditions is a characteristic of the Thiobacillus genus (Kelly & Harrison, 1988). All of the strains, except E3, also grew on thiocyanate, which is a property of T. thwpanrs (Kelly & Harrison, 1988) including the strain TK-m which also uses DMS or carbon disulphide, but not DMDS (Kanagawa & Kelly, 1986: N. A. Smith, unpublished). At this stage in their characterization these strains can probably be assigned to the genus Thiobacillus,with the distinctive feature that they are methyl sulphide oxidizers. This conclusion contradicts the statement by Sand (1987) that thiobacilli would not be able to produce sulphuric acid from DMDS, DMS and MT as a contributory factor in corrosion of concrete sewage pipelines. The mechanism of DMDS degradation can be inferred from our observations. DMS is excluded as an intermediate because of (a) the long lag before growth observed when cultures previously grown on DMS or DMDS were inoculated into media containing the other substrate; and (b)the failure of DMDS-grown organisms to oxidize DMS, although MT was oxidized very rapidly. This indicates that the initial step in DMDS metabolism is its reductive cleavage to MT. By analogy with DMS and MT oxidation by hyphomicrobia (De Bont et al., 1986; Suylen et al., 1986), MT is probably oxidized to H2S and formaldehyde, the latter being oxidized via formate to Cot, and the former to sulphate. The ability of DMDS-grown organisms to oxidize formaldehyde, formate and sulphide supports this view. MT oxidation by Hyphomicrobium EG was proposed to be by means of a H202-producingoxidase (Suylen et al., 1986) and inhibition of growth on DMSO, but not so severely on methylamine, by the catalase inhibitor 3-amino 1,2,4triazole (AT) provided indirect evidence for this (Suylen et al., 1986). Growth of strain E6 on DMDS was also strongly inhibited by AT, at lower concentrations than required to affect HyphomicrobiumEG, further supporting the proposed route for DMDS oxidation (N. A. Smith, unpublished data). Carbon assimilation by the DMDS-using strains appeared to be Downloaded from www.microbiologyresearch.org by IP: 86.151.64.96 On: Fri, 20 Nov 2015 11:36:41

1416

N. A. SMITH AND D . P . K E L L Y

predominantly by C02-fixation during growth on DMDS. The 14C02incorporation patterns during growth were consistent with all their carbon being derived from C 0 2 fixation by the Calvin cycle, the key enzyme of which (ribulose bisphosphate carboxylase) was present in strain E6 at an activity adequate to support growth at the observed growth rates. The growth yield of strain E6 in DMDS-limited chemostat culture (at a dilution rate of 0.078 h-l) was 14-6g cell-carbon (mol DMDS)-l, and increased in the presence of thiosulphate by the equivalent of 2.45 g cell-carbon (mol thiosulphate)-l. Assuming a yield equivalent to this to be due to each of the sulphide groups of the DMDS (Kelly & Kuenen, 1984), sulphide oxidation would support 4.9 of the observed 14.6 g cell-carbon (mol DMDS)-l. This would mean that oxidation of the methyl groups of DMDS to C 0 2 provided energy for production of 4-85 g cell-carbon per mol of C 0 2 produced. In view of the fact that the conversion of the methyl groups to formaldehydemight not result in energy conservation (Suylen et al., 1986), the energyyielding steps would be the oxidation of formaldehyde to formate and formate to C02. Previous work with Thiobacillus versutus (A2), which can grow autotrophically on formaldehyde and formate, showed yields (under growth conditions comparable to our work with strain E6) of about 5 g cell-carbon (mol formaldehyde)-' and 2 g cell-carbon (mol formate)-l (Kelly et al., 1979; Wood & Kelly, 1981; Kelly & Wood, 1984). The yields observed with strain E5 are thus wholly consistent with energy being derived simultaneously from the oxidation of formaldehyde and sulphide produced from DMDS, and being used to support autotrophic growth dependent on the Calvin cycle. Our failure to obtain growth on formate as sole substrate indicates that there may be some dependence for growth on the presence of an oxidizable sulphur compound. These observations firmly establish the importance of autotrophic Thiobacillus-like bacteria in the oxidative destruction of methylated sulphides in the natural environment. We thank the Natural Environment Research Council for support of this work. REFERENCES

BECHARD, M. J. & RAYBURN, W. R. (1979). Volatile organic sulfides from freshwater algae. Journal of Phycology 15, 379-383. BREMNER, J. M. & STEELE,C. G. (1978). Role of microorganisms in the atmospheric sulfur cycle. Advances in Microbial Ecology 2, 155-201. DACEY,J. W. H.& WAKEHAM, S. G. (1986). Oceanic dimethylsulfide: production during zooplankton grazing on phytoplankton. Science 233, 1314-1 3 16. DE BOW, J. A. M., VAN DIJKEN,J. P. & HARDER,W. (1981). Dimethyl sulphoxide and dimethyl sulphide as a carbon, sulphur and energy source for the growth of Hyphomicrobium S . Journal of General Microbiology 127, 315-323. DEPREZ,P. P.,FRANZMANN, P. D. & BURTON,H. R. (1986). Determination of reduced sulfur gases in Antarctic lakes and seawater by gas chromatography after solid absorbant preconcentration. Journal of Chromatography 362, 9-21. DROTAR, A., BURTON, G. A., TAVERNIER, J. E. & FALL, R. (1987) Widespread Occurrence of bacterial thiol methyltransferases and the biogenic emission of methylated sulfur gases. Applied and Environmental Microbiology 53, 1626163 1. FEREK, R. J., CHATFIELD,R. B. & ANDREAE,M. 0. (1986). Vertical distribution of dimethylsulphide in the marine atmosphere. Nature, London 320, 514516. FRANZMANN, P. D., DEPREZ,P. P., BURTON, H. R. & VAN DEN HOFF, J. (1987). Limnology of Organic Lake, Antarctica, a meromictic lake that contains high concentrations of dimethyl sulphide. Australian

Journal of Marine and Freshwater Research 38, 409417. KANAGAWA, T. & KELLY,D. P. (1986). Breakdown of dimethyl sulphide by mixed cultures and by Thiobacillus thwparus. FEMS Microbiology Letters 34, 1319. KANAGAWA, T., DAZAI, M. & FUKUOKA, S. (1982). Degradation of 0,O-dimethylphosphorodithioate by Thiobacillus thioparus TK-1 and Pseudomonas AK-2. Agricultural and Biological Chemistry 46, 2571-2578. KELLY,D. P. (1988). Oxidation of sulphur compounds. Symposia of the Society for General Microbiology 42, 65-98. KELLY,D. P. & WOOD,A. P. (1982). Autotrophic growth of Thiobacillus A2 on methanol. FEMS Microbiology Letters 15, 229-233. KELLY,D. P. & WOOD, A. P. (1984). Potential for methylotrophic autotrophy in Thiobacillus versutus (Thwbacillus sp. strain A2). In Microbial Growth on C 1 Compounds, pp. 324-329. Edited by R. L. Crawford & R. S. Hanson. Washington, DC : American Society for Microbiology. KELLY,D. P. & KUENEN,J. G. (1984). Ecology of the colourless sulphur bacteria. In Aspects of Microbial Metabolism and Ecology, pp. 21 1-240. Edited by G. A. Codd. London: Academic Press. KELLY,D. P. & HARRISON,A. P. (1988). Thwbacillus. In Bergey's Manualof Systematic Bacteriology, vol. 3. Edited by N. Pfenning & J. T. Staley. Baltimore: Williams & Wilkins (in the Press). KELLY,D. P., CHAMBERS, L. A. & TRUDINGER, P. A.


Oxidation of methylated sulphides (1969). Cyanolysis and spectrophotometric estimation of trithionate in mixture with thiosulfate and tetrathionate. Analytical Chemistry 41, 898-901. KELLY,D. P., WOOD, A. P., GOY-IXCHAL, J. C. & KUENEN,J. G. (1979). Autotrophic metabolism of formate by Thiobacillus strain A2. Journal of General Microbiology 114, 1-13. KIENE,R. P. & CAPONE,D. G. (1988). Microbial transformations of methylated sulfur compounds in anoxic salt marsh sediments. Microbial Ecology (in the Press). K ~ N I GW. , A., LUDWIG,K., SIEVERS, S., STOLTING, K. H. & G-R, W. (1980). Identificationof volatile organic sulfur compounds in municipal sewage systems by GC/MS. Journal of High Resolution Chromatography 3, 415-416. LEADBEATER, L., SIEBERT, K., WHOBERT, P. & BOWIEN, B. (1982). Relationship between activities and protein levels of ribulosebisphosphate carboxylase and phosphoribulokinase in Alcaligenes eutrophus. FEMS Microbiology Letters 14, 263-266. SAND, W. (1987). Importance of hydrogen sulfide, thiosulfate, and methylmercaptan for growth of thiobacilli during simulation of concrete corrosion. Applied and Environmental Microbiology 53, 16451648. SIVELA,S. K. (1980). Dimethyl sulphide as a growth substratefor an obligately chemolithotrophic Thiobacillus. Commentationes physico-mathematicae dissertationes, University of Helsinki.

1417

SUYLEN, G. M. H. & KUENEN,J. G. (1986). Chemostat enrichment and isolation of Hyphomicrobium EG. Antonie van Leeuwenhoek 52, 281-293. SUYLEN,G . M. H., STEFESS,G. C. & KUENEN,J. G. (1986). Chemolithotrophic potential of a Hyphomicrobium species, capable of growth on methylated sulphur compounds. Archives of Microbiology 146, 192-198. TOMITA,B., INOUE,H., CHAYA,K., NAKAMURA,A., HAMAMURA, N., UENO, K., WATANABE, K. &O~E, Y. (1987). Identification of dimethyl disulfide-forming bacteria isolated from activated sludge. Applied and Environmental Microbiology 53, 1541-1 547. TUOVINEN, 0.H. & KELLY,D. P. (1973). Studieson the growth of Thiobacillusferrooxidans. Archivfur Mikrobiologie 88, 285-298. WOOD,A. P. & KELLY,D. P. (1981). Mixotrophic growth of Thiobacillus A2 in chemostat culture on formate and glucose. Journal of General Microbiology 125, 55-62. ZEYER,J., EICHER,P., WAKEHAM, S. G. & SCHWARZENBACH, R. P. (1987). Oxidation of dimethyl sulfide to dimethyl sulfoxide by phototrophic purple bacteria. Applied and Environmental Microbiology 53, 2026-2032. ZINDER,S. H., DOEMEL, W. N . & BROCK,T. D. (1977). Production of volatile sulfur compounds during the decomposition of algal mats. Applied and Environmental Microbiology 34, 859-860.
