Proton Translocation Coupled to Dimethyl Sulfoxide Reduction in Anaerobically Grown Escherichia Coli HB101 PETER T

Home , Electron transport chain

JOURNAL OF BACTERIOLOGY, JUlY 1985, p. 369-375 Vol. 163, No. 1 0021-9193/85/070369-07$02.00/0 Copyright C 1985, American Society for Microbiology Proton Translocation Coupled to Dimethyl Sulfoxide Reduction in Anaerobically Grown Escherichia coli HB101 PETER T. BILOUS AND JOEL H. WEINER* Department ofBiochemistry, University ofAlberta, Edmonton, Alberta, Canada T6G 2H7 Received 27 February 1985/Accepted 22 April 1985

Proton translocation coupled to dimethyl sulfoxide (DMSO) reduction was examined in Escherichia coli HB101 grown anaerobically on glycerol and DMSO. Rapid acidification of the medium was observed when an anaerobic suspension of cells, preincubated with glycerol, was pulsed with DMSO, methionine sulfoxide, nitrate, or trimethylamine N-oxide. The DMSO-induced acidification was sensitive to the'uncoupler carbonyl cyanide p-trifluoromethoxyphenylhydrazone (60 ,uM) and was inhibited by the quinone analog 2-n-heptyl-4- hydroxy-quinoline-N-oxide (5.6 ,LM). Neither sodium azide nor potassium cyanide inhibited the DMSO response. An apparent -*H+/2e- ratio of 2.9 was obtained for DMSO reduction with glycerol as the reductant. Formate and H2(g), but not lactate, could serve as alternate electron donors for DMSO reduction. Cells grown anaerobically on glycerol and fumarate displayed a similar response to pulses of DMSO, methionine sulfoxide, nitrate, and trimethylamine N-oxide with either glycerol or H2(g) as the electron donor. However, fumarate pulses did not result in acidification of the suspension medium. Proton translocation coupled to DMSO reduction was also demonstrated in membrane vesicles by fluorescence quenching. The addition of DMSO to hydrogen-saturated everted membrane vesicles resulted in a carbonyl cyanide p-trifluoromethoxyphenylhydrazone-sensitive fluorescence quenching of quinacrine dihydrochloride. The data indicate that reduction of DMSO by E. coli is catalyzed by an anaerobic electron transport chain, resulting in the formation of a proton motive force.

Escherichia coli is a facultative anaerobe capable of growth. Cells were grown anaerobically for either 36 h deriving energy for aerobic growth by oxidative phosphory- (DMSO) or 24 h (FUM) at 37°C. lation or anaerobic growth by fermentation or anaerobic Preparation of everted envelopes. Membrane vesicles were respiration. Anaerobic respiration on fumarate (FUM), ni- prepared by French pressure cell disruption as described trate, and trimethylamine N-oxide (TMAO) have been well previously (1), except that the final membrane pellet was studied (5, 8). Several organisms, including E. coli, are suspended in 1 mM HEPES (N-2-hydroxyethylpiperazine- capable of reducing dimethyl sulfoxide (DMSO) to dimethyl N'-2- ethanesulfonic acid [pH 7.5]) containing 100 mM KCI. sulfide during growth (21), and in some cases it has been Preparation of whole cells for pH measurements. Cells were shown that DMSO can serve as the terminal electron accep- grown to late-log phase (optical density at 550 nm = 1.2 + tor during anaerobic growth (11, 20). We recently reported 0.1), harvested at 3,840 x g, and washed twice with 150 mM the ability of E. coli to grow anaerobically on a minimal KCI containing 0.5 mM dithiothreitol-0.1 mM sodium medium with DMSO as the terminal electron acceptor (1). pyrophosphate (one-fifth of medium volume per wash). The Under these conditions, an inducible and membrane-bound final cell pellet was suspended in 100 mM KCI containing 25 enzyme is synthesized which catalyzes DMSO reduction. mM KSCN, 0.1 mM sodium PP1, and 50 jxg of carbonic According to the chemiosmotic hypothesis, electron flow anhydrase per ml. The cell suspension, containing 2.8 x 109 through the respiratory chain of mitochondria, chloroplasts, cells per ml, was stored on ice in 4.0-ml volumes with either or bacteria is coupled to ATP synthesis by the formation of 7 mM GLY, 7 mM sodium formate, or 2.5 mM D-(-)-lactate a proton gradient across an energy-transducing membrane as energy sources. In some experiments, hydrogen gas (12). In the present paper, we demonstrate by pH and [H2(g)] was used as an energy source. For these experi- fluorescence quenching measurements with anaerobically ments, cells were bubbled slowly with H2(g) for 0.5 h before grown E. coli that proton translocation is coupled to DMSO use. reduction. Thus, anaerobic respiration on DMSO can serve pH measurements. For measurement of extracellular pH as an energy-yielding pathway for the growth of this orga- changes, 4.0 ml of prewarmed cells (23°C) were placed in an nism. oxygen electrode vessel (model OXSDIG, Rank Bros., Bot- tisham, Cambridge, England) and equilibrated under N2(g) MATERIALS AND METHODS or H2(g) with stirring until a stable pH was attained. Air was Growth conditions. E. coli HB101 (F- hsdR hsdM pro leu excluded by a Teflon stopper which held the pH electrode gal lac thi recA rpsL) was grown anaerobically on a minimal (model GK2321C, Radiometer, Copenhagen, Denmark). medium as described previously (1), with glycerol (GLY) as Gases, N2(g) or H2(g), were introduced over the surface of the carbon and energy source. DMSO (70 mM) or sodium the cells by syringe needle ports. The pH electrode was FUM (40 mM) were used as terminal electron acceptors for connected through a pH meter (PHM84, Radiometer) to a recorder. Dissolved oxygen was monitored to ensure anaerobic * Corresponding author. conditions were achieved and maintained throughout the 369 370 BILOUS AND WEINER J. BACTERIOL.

0 A 0cn a a. 0 C2, 0 U I I I I

min H+

to C a J- 0z B a -9 I-- U- I min I I min I H+ HV If

FIG. 1. Proton translocation in whole cells of E. coli HB101 grown anaerobically on GLY-DMSO medium. Late-log-phase cells were harvested, washed, and suspended in 100 MnM KCI containing 25 mM KSCN, 0.1 mM sodium PP,, 50 ,ug of carbonic anhydrase per ml, and 7 mM GLY under N2(g) until a stable pH was achieved. At the indicated points, GLY (2.7 ,umol), FCCP (60 ,uM, final concentration), and 500 nmol of DMSO, methionine sulfoxide, KNO3, TMAO, or FUM were added. The vertical arrows (H+) correspond to deflections resulting from a pulse of 500 nmol of HCI. experiment. Small additions of electron donors, acceptors, quinacrine dihydrochloride were all purchased from Sigma inhibitors, and standard solutions were introduced by Chemical Co. (St. Louis, Mo.). Carbonyl cyanide p- Hamilton microsyringes (type no. 701, Hamilton Co., Reno, trifluoromethoxyphenylhydrazone (FCCP)' was obtained Nev.). The buffering capacities of the suspensions were from Aldrich Chemical Co. (Milwaukee, Wis.). determined by injecting known quantities of 50 mM HCI in 100 mM KCI. RESULTS Estimation of -*H+/2e- stoichiometries. An estimation of Proton translocation coupled to DMSO reduction. An the number of protons translocated per electron pair trans- anaerobic suspension of E. coli HB101, prepared from cells ferred from donor to acceptor (--H+/2e-) was obtaiped as grown to late log phase in GLY-DMSO medium, was follows. The point of maximum medium acidification was preincubated in the presence of excess GLY. GLY was approximated'by extrapolating the traces of the initial rate of added as the electron donor in these experiments, as endog- proton extrusion and proton reentry to the point pf inter- enous energy was insufficient to yield proton translocation section. This maximum difference in pH was quantitated by with DMSO. A final, stable pH of 6.4 to 6.6 was achieved comparison with pH changes resulting from pulses of 500 with 02(g) saturation of less than 2%. When pulsed with a nmol of HCl (10 ,ul of 50 mM HCI containing 100 mM KCI). small quantity of DMSO, rapid acidification of the extracel- The resulting value was divided by the number of moles of lular medium was recorded which slowly, but not com- electron acceptor added to the suspension to yield an'esti- pletely, dissipated (Fig. 1A). A nearly equivalent response mate for -+H+/2e-. was obtained with the DMSO analog methionine sulfoxide as Fluorescence quenching measurements. Fluorescence the electron acceptor. quenching measurements were made on hydrogen-saturated The DMSO-induced acidification was sensitive to the everted membrane vesicles under conditions described by uncoupler FCCP (60 ,uM), resulting in an immediate deple- Jones (9). Fluorescence changes of quinacrine dihydro- tion of the proton gradient'and prevention of a further chloride were measured at an excitation wavelength of 449 response to DMSO (Fig. 1A). Although 10 p,M FCCP was nm and an emission wavelength of 510 nm with a Turner sufficient to deplete the proton gradient formed, 60 ,M was spectrofluorometer (model 430, G. K. Turner Associates, required to inhibit a further response to DMSO. The above Palo Alto, Calif.). Everted membrane vesicles (2.3 mg of data indicate that the reduction of DMSO or methionine protein) were suspended in 2.5 ml of an I-2(g)-saturated sulfoxide by E. coli is coupled to an uncoupler-sensitive buffer containing 300 mM KCl, 15 mM MgCl2, and 10 mM outward translocation of protons. HEPES (pH 7.5). Quinacrine dihydrochloride (4 ,uM), sub- It was previously reported that cells grown anaerobically strates, and inhibitors were added as indicated in the figure on GLY-DMSO medium synthesized nitrate, TMAO, and legends. FUM reductases in addition to DMSO reductase activity (1). Protein determination. Protein was estimated by a sodium As shown in Fig. 1B and C, pulses of nitrate and TMAO dodecyl sulfate Lowry procedure (10) with bovine serum resulted in proton translocation. However, we have been albumin as the standard. unable to demonstrate a respQnse to FUM with these cells Chemicals and reagents. Carbonic anhydrase, 2-n-heptyl- (Fig. 1C), despite the presence of FUM reductase activity. 4-hydroxy-quinoline-N-oxide (HOQNO), HEPES, and DMSO undergoes a two-electron reduction to form VOL. 163, 1985 PROTON TRANSLOCATION IN E. COLI 371

0 4) W, 0 o 0 A (f) -j ._ z 2 c.> E 0 < Yl a E-IL a (9 0f L

min ,H+

0 0 it) >-cI1) z cn 0 0 2 z II)(90 min i I I H+

FIG. 2. The effect of sodium azide and potassium cyanide on proton translocation coupled to DMSO reduction. Cells were prepared and incubated under N2(g) as indicated in the legend to Fig. 1. (A) Cells were given a test pulse of GLY (2.7 ,umol) and DMSO (500 nmol) before the addition of sodium azide (0.37 mM). (B) Potassium cyanide (5.0 mM) was added after test pulses of GLY and DMSO at the concentrations indicated above. Subsequent pulses of DMSO and KNO3 (500 nmol) were added where indicated. FCCP was added to a final concentration of 60 ,uM. The vertical arrows (Hi) correspond to pulses of 500 nmol of HCI.

dimethyl sulfide (17). The -*H+/2e- stoichiometry for conditions. Similar H+/2e values were obtained whether DMSO reduction was estimated by extrapolating the curve 500 nmol or 1 ,umol of DMSO was added. Due to a decrease of proton reentry to the point of maximum medium acidifica- in medium acidification with subsequent DMSO pulses (data tion. A value of 2.9 + 0.5 (mean ± standard deviation, n = not shown), only the first acidification profiles were used for 8 determinations) was obtained for DMSO reduction with these estimations. GLY as the electron donor, compared with a value of 3.3 + Effect of electron transport chain inhibitors. Several respi- 0.3 (n = 3) obtained for nitrate reduction under these ratory chain inhibitors were added to an anaerobic suspen-

0 0 z 0 0 a. 2 (9 0C]I0I A I I I I II I

I min H+

0 0 0 < ~~~a B I I II II I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

FIG. 3. Effect of HOQNO on proton translocation coupled to DMSO, KNO3, and TMAO reduction by whole cells. Cells were prepared and incubated under N2(g) as indicated in the legend to Fig. 1. At the indicated points, GLY (2.7 ,umol), FCCP (20 ,uM), HOQNO (5.6 p.M), and 500 nmol of DMSO, methionine sulfoxide, KNO3, or TMAO were added. Vertical arrows (H') correspond to pulses of 500 nmol of HCI. 372 BILOUS AND WEINER J. BACTERIOL.

4) analog HOQNO at 5.6 ,uM (Fig. 3A). Lower concentrations a-0~°J 2 c< Cf)o of HOQNO did not fully inhibit responses to DMSO. This C

I a I2 Y concentration of HOQNO (5.6,uM) had no effect on the nitrate (Fig. 3A) or TMAO responses (Fig. 3B). Both nitrate and FUM reduction are sensitive to HOQNO (13, 15), whereas TMAO reduction is reported to be insensitive to HOQNO (2). Our data suggests that DMSO reduction is more sensitive to HOQNO inhibition than the nitrate reductase pathway. No comparison to FUM reduction could be made. Alternate electron donors. In the previous sections, cells 0 which had been grown in GLY-DMSO medium were incu- 0 0o E Cl) e bated anaerobically with excess GLY for the proton 0 .) a Nc translocation experiments. To test alternate electron donors, U.- the harvested and washed cells were incubated with either I I II I D-(-)-lactate or potassium formate, or they were bubbled with H2(g). As shown in Fig. 4A, none of the indicated acceptors I min resulted in proton translocation when lactate was used as the electron donor. In contrast, a weak but measurable DMSO- Ha. induced, FCCP-sensitive proton translocation was observed with formate as the electron donor (Fig. 4B). Hydrogen gas has been used as a reductant for anaerobic growth on FUM, nitrate, or TMAO (18), a process requiring a membrane-bound hydrogenase. Pulses of DMSO resulted in an FCCP-sensitive acidification of the medium when H2(g) was used as an electron donor (Fig. 4C). In summary, several electron donors were shown to 0 o couple to DMSO reduction and result in a proton gradient. These include GLY, formate, and H2(g), but no response to 4 a L O lactate was recorded. I I DMSO reduction with GLY-FUM-grown cells. Due to the I I inability to demonstrate FUM-dependent proton transloca- I min tion with whole cells grown on GLY-DMSO medium, cells H+f'" were prepared from anaerobic growth on GLY-FUM medium, and proton translocation experiments were per- formed. It was previously reported that DMSO, nitrate, FIG. 4. Proton translocation in whole cells of E. coli HB101 with TMAO, and FUM reductase activities were present in cells lactate, formate, or H2(g) as the electron donor. Late-log-phase cells grown on GLY-FUM medium (1). grown anaerobically in GLY-DMSO medium were harvested, washed, and suspended in 100 mM KCI containing 25 mM KSCN, The addition of DMSO, TMAO, and nitrate to arnaerobic 0.1 mM sodium PP,, and 50 ,ug of carbonic anhydrase per ml, and cells preincubated with GLY as electron donor resulted in incubated with one of the following electron donors: 2.5 mM proton translocation in each case, but only a weak response D-(-)-lactate (A), 7 mM sodium formate (B), or incubation under to FUM was recorded (Fig. 5). When H2(g) was used as a H2(g) (C) after preconditioning by bubbling slowly for 0.5 h with reductant with these cells, DMSO-dependent as well as H2(g). Test pulses of 25 nmol of lactate (A) or 35 nmol of formate (B) methionine sulfoxide-, TMAO-, and nitrate-dependent pro- were made as indicated. FCCP was added to 10 ,uM final concentra- ton translocation could be demonstrated, but no response to tion, and 500 nmol of all electron acceptors was added. Vertical FUM was recorded (data not shown). arrows (H+) correspond to a pulse of 500 nmol of HCI. Fluorescence quenching measurement of DMSO reduction. Fluorescence quenching of the acridine dye quinacrine sion of E. coli to determine whether the DMSO-induced dihydrochloride has been used to demonstrate proton medium acidification requires a functional electron transport translocation in membrane preparations of E. coli grown chain. anaerobically on nitrate or FUM (6, 9). Quinacrine Azide, an inhibitor of cytochrome c oxidase in eucaryotes, dihydrochloride is a weak base which distributes according is known to be a competitive inhibitor of nitrate in the nitrate to the transmembrane ApH (14). Its accumulation within the reductase pathway (3, 7). The addition of 0.37 mM sodium vesicles, as a result of the inward translocation of protons, azide, a concentration which was necessary to completely gives rise to a quenching of fluorescence. Vesicles are inhibit a response to nitrate, had little or no effect on the employed to avoid problems associated with permeability, DMSO-induced proton translocation (Fig. 2A). transport, and contributions by cytoplasmic or periplasmic Cyanide is known to bind to cytochromes o and d and scalar reactions. It was of interest, therefore, to verify the inhibit aerobic electron transport in E. coli (8). It is also an results of the pH electrode study by this method. Hydrogen effective inhibitor of nitrate reductase in this organism (7). was used as a reductant for these experiments. The addition of 5.0 mM KCN, a concentration which se- Everted membrane vesicles, prepared from cells grown on verely inhibited responses to nitrate (Fig. 2B), had no GLY-DMSO medium, were suspended in an H2(g)-saturated apparent effect on the DMSO-induced proton translocation. buffer for experimentation. The addition of DMSO to hydro- In contrast to the CN- and azide results, DMSO and gen-saturated membrane vesicles (Fig. 6A) resulted in methionine sulfoxide reduction was inhibited by the quinone quenching of the acridine dye. The sudden return in fluores- VOL. 163, 1985 PROTON TRANSLOCATION IN E. COLI 373 cence is probably due to the depletion of hydrogen, as a 0'I) subsequent addition of DMSO was without effect. The A addition of NADH as an energy source restored quenching. The addition of ATP also resulted in fluorescence quenching, I mn due to hydrolysis of the substrate by the proton-translocating ATPase (5). The addition of FUM or nitrate to the same H+F preparation of H2(g)-saturated membrane vesicles also resulted in fluorescence quenching of quinacrine dihydrochloride (data not shown). The DMSO-dependent fluorescence quenching is sensitive to the action of the uncoupler FCCP (Fig. 6B), which equilibrates protons across the 0 membrane. The above results support the data obtained with - o whole cells (Fig. 4C) that hydrogen-coupled DMSO reduc- B tion results in the generation of a transmembrane pH gradi- I min ent. H+ DISCUSSION It was recently demonstrated that E. coli is capable of anaerobic growth with DMSO as the terminal electron acceptor (1). Under these growth conditions, a membrane- YI bound enzyme catalyzing DMSO reduction is synthesized. A I molybdenum cofactor requirement for DMSO reduction was I I min suggested by the inhibitory effects of sodium tungstate on H+ growth and by the lack of growth of chlorate-resistant F mutants chiA, chlB, chlE, and chlG on the GLY-DMSO medium. DMSO reductase activity was repressed by the presence of nitrate in the growth medium or by aerobic growth and appeared to be under the control of the fnr gene D product, a positive regulator for the expression of anaerobic terminal reductases (1). The data suggested that E. coli is I min capable of anaerobic respiration on DMSO. The coupling of DMSO reduction to the generation of a proton motive force was investigated in the present study. FIG. 5. Proton translocation in whole cells of E. coli HB101 According to the chemiosmotic hypothesis proposed by grown anaerobically on GLY-FUM medium. Cells were prepared Mitchell (12), electron flow through an electron transport and incubated under N2(g) as indicated in the legend to Fig. 1. At the chain results in the translocation of protons in a vectorial indicated points, 2.7 ,umol of GLY or 500 nmol of DMSO, TMAO, manner. This results in the generation of a gradient of pH KNO3, or FUM was added. Vertical arrows (H+) correspond to (ApH) and electrical potential (A4i) across the membrane. pulses of 500 nmol of HCI. These two components constitute the proton motive force, which can be a function of ApH or A*s exclusively or a whole cells of R. capsulata will generate a cytoplasmic combination of these two components. The resulting proton membrane potential (11). motive force is used for many energy-linked processes The stoichiometry for DMSO reduction, -H+/2e-, was across the membrane, including ATP synthesis via the estimated to be ca. 2.9 with GLY as an energy source. This reversible proton-translocating ATPase. value decreased with the time of anaerobic incubation, an Our results with whole cells and membrane vesicles ob- observation reported by Takagi et al. (16) during studies of tained from E. coli grown anaerobically on DMSO clearly TMAO reduction by E. coli. A stoichiometry of 2.9 suggests demonstrate the formation of a proton gradient coupled to the possibility of two energy-conserving sites during electron DMSO reduction. Electron flow from GLY, formate, and flow from GLY to DMSO. Under the same experimental H2(g) to DMSO resulted in the formation of a proton conditions, a value of 3.3 for nitrate reduction was obtained. gradient. Proton translocation was blocked by the un- By comparison a published stoichiometry of 2.1 for coupling action of FCCP or by the presence of the quinone -+H+/NO3- was obtained with GLY as an electron donor analog HOQNO. These results are consistent with electron with E. coli spheroplasts (3). It should be noted that the flow to DMSO through an electron transport chain, i.e., spheroplasts used for the latter experiments were obtained anaerobic respiration on DMSO. The midpoint oxidation- from cells grown anaerobically on GLY-KNO3 medium. reduction potential for DMSO/dimethyl sulfide has been Furthermore, the value reported by Garland et al. (3) was calculated by Wood (17) to be +160 mV. As this value lies obtained from the observed acidification peak, not the ex- between those of FUM/succinate and nitrate/nitrite (+30 mV trapolated peak as used in the present work. These condi- and +420 mV, respectively [8]), the coupling of DMSO tions differ from the experimental conditions employed in reduction to ATP synthesis is thermodynamically possible. this study and may account for differences in the --H+/NO3 E. coli is capable of anaerobic respiration on FUM, values. nitrate, and TMAO, and proton translocation coupled to We were not able to demonstrate any significant proton their reduction has been previously documented (4, 6, 16). translocation coupled to FUM reduction in whole cells The photosynthetic bacterium Rhodopseudomonas cap- obtained from growth on either GLY-DMSO or GLY-FUM sulata was shown to be capable of anaerobic growth in the medium, despite the presence of FUM reductase activity in dark with DMSO as the electron acceptor (19). It was these cells (1). Poor results with FUM may be due to the recently reported that the reduction of DMSO or TMAO by uptake of protons with FUM during transport as well as 374 BILOUS AND WEINER J. BACTERIOL.

0 0 Co cx a. 2 0 a z I I1I A

o X. 0 I I& B w zZ Z o zD U..IL 5 MINUTES FIG. 6. Fluorescence quenching ofhydrogen-saturated membrane vesicles ofE. coli HB101 grown anaerobically on GLY-DMSO medium. Everted membrane vesicles, prepared as outlined in the text, were suspended in an H2(g)-saturated HEPES buffer (10 mM, pH 7.5) containing 300 mM KCI and 15 mM MgCI2. Quinacrine dihydrochloride (4 ,uM) was added, and fluorescence was monitored until stable. At the indicated points, 2 ,umol (A) or 0.4 ,mol (B) of DMSO and 0.14 ,mol of NADH, 50 nmol of ATP, or FCCP (6 ,M) were added. incomplete or slow metabolism of FUM to succinate (4). rate in anaerobically grown Escherichia coli K12. Biochem. J. we have observed FUM-dependent fluorescence 164:265-267. However, 5. Haddock, B. A., and C. W. Jones. 1977. Bacterial respiration. quenching of quinacrine dihydrochloride in hydrogen- Bacteriol. Rev. 41:47-99. saturated membrane vesicles obtained from cells grown on 6. Haddock, B. A., and M. W. Kendall-Tobias. 1975. Functional GLY-DMSO medium (unpublished data). Proton transloca- anaerobic electron transport linked to the reduction of nitrate tion coupled to FUM reduction by endogenous substrates and fumarate in membranes from Escherichia coli as demon- has been previously reported (4), but we found it necessary strated by quenching of atebrin fluorescence. Biochem. J. to preincubate cells with various electron donors to demon- 152:655-659. strate proton translocation with DMSO, nitrate, or TMAO. 7. Hewitt, E. J., and B. A. Notton. 1980. Nitrate reductase systems The sensitivity of DMSO reduction to HOQNO inhibition in eukaryotic and prokaryotic organisms, p. 273-325. In M. en- is interest. has been shown to be an effective Coughlan (ed.), Molybdenum and molybdenum-containing of HOQNO zymes. Pergamon Press, Oxford. inhibitor of nitrate (13) and FUM (15) reduction, but appar- 8. Ingledew, W. J., and R. K. Poole. 1984. The respiratory chains ently not TMAO (2). As we were unable to demonstrate of Escherichia coli. Microbiol. Rev. 48:222-271. proton translocation coupled to FUM reduction, we could 9. Jones, R. W. 1979. Hydrogen-dependent proton translocation not compare the HOQNO sensitivity of DMSO reduction by membrane vesicles from Escherichia coli. Biochem. Soc. with that of FUM. However, our results do indicate a higher Trans. 7:1136-1137. sensitivity of DMSO reduction to HOQNO inhibition than 10. Markwell, M. A. K., S. M. Haas, L. L. Bieber, and N. E. shown for nitrate reduction. Neither azide nor CN-, inhibi- Tolbert. 1978. A modification of the Lowry procedure to sim- tors of nitrate reduction (3, 7), was effective in inhibiting plify protein determination in membrane and lipoprotein sam- proton translocation coupled to DMSO reduction. ples. Anal. Biochem. 87:206-210. 11. McEwan, A. G., S. J. Ferguson, and J. B. Jackson. 1983. ACKNOWLEDGMENTS Electron flow to dimethylsulphoxide or trimethylamine-N-oxide generates a membrane potential in Rhodopseudomonas This work was funded by grant MT5838 from the Medical Re- capsulata. Arch. Microbiol. 136:300-305. search Council of Canada to J.H.W. P.T.B. received support from 12. Mitchell, P. 1967. Proton-translocation phosphorylation in the Alberta Heritage Foundation for Medical Research. mitochondria, chloroplasts and bacteria: natural fuel cells and solar cells. Fed. Proc. 26:1370-1379. 13. Ruiz-Herrera, J., and J. A. DeMoss. 1969. Nitrate reductase LITERATURE CITED complex of Escherichia coli K-12: participation of specific 1. Bilous, P. T., and J. H. Weiner. 1985. Dimethyl sulfoxide formate dehydrogenase and cytochrome b, components in ni- reductase activity by anaerobically grown Escherichia coli trate reduction. J. Bacteriol. 99:720-729. HB101. J. Bacteriol. 162:1151-1155. 14. Schuldiner, S., and M. Avron. 1971. On the mechanism of the 2. Bragg, P. D., and N. R. Hackett. 1983. Cytochromes of the energy-dependent quenching of atebrin fluorescence in isolated trimethylamine N-oxide anaerobic respiratory pathway ofEsch- chloroplasts. FEBS Lett. 14:233-236. erichia coli. Biochim. Biophys. Acta 725:168-177. 15. Singh, A. P., and P. D. Bragg. 1975. Reduced nicotinamide 3. Garland, P. B., J. A. Downie, and B. A. Haddock. 1975. Proton adenine dinucleotide dependent reduction of fumarate coupled translocation and the respiratory nitrate reductase of Esch- to membrane energization in a cytochrome deficient mutant of erichia coli. Biochem. J. 152:547-559. Escherichia coli K12. Biochim. Biophys. Acta 396:229-241. 4. Gutowski, S. J., and H. Rosenberg. 1977. Proton translocation 16. Takagi, M., T. Tsuchiya, and M. Ishimoto. 1981. Proton coupled to electron flow from endogenous substrates to fuma- translocation coupled to trimethylamine N-oxide reduction in VOL. 163, 1985 PROTON TRANSLOCATION IN E. COLI 375

anaerobically grown Escherichia coli. J. Bacteriol. 148:762-768. 19. Yen, H.-C., and B. Marrs. 1977. Growth ofRhodopseudomonas 17. Wood, P. M. 1981. The redox potential for dimethyl sulphoxide capsulata under anaerobic dark conditions with dimethyl reduction to dimethyl sulphide. Evaluation and biochemical sulfoxide. Arch. Biochem. Biophys. 181:411-418. implications. FEBS Lett. 124:11-14. 20. Zinder, S. H., and T. D. Brock. 1978. Dimethyl sulfoxide as an 18. Yamamoto, I., and M. Ishimoto. 1978. Hydrogen-dependent electron acceptor for anaerobic growth. Arch. Microbiol. growth of Escherichia coli in anaerobic respiration and the 116:35-40. presence of hydrogenases with different functions. J. Biochem. 21. Zinder, S. H., and T. D. Brock. 1978. Dimethyl sulphoxide 84:673-679. reduction by micro-organisms. J. Gen. Microbiol. 105:335-342.