Vol. 142, No. 3JOURNAL OF BACTERIOLOGY, June 1980, p. 859-8680021-9193/80/06-0859/10$02.00/0

Influences of Growth Substrates and Oxygen on the ElectronTransport System in Acinetobacter sp. HO1-N

BURT D. ENSLEY, JR.,t AND W. R. FINNERTY*

Department ofMicrobiology, University of Georgia, Athens, Georgia 30602

The electron transport system of Acinetobacter sp. HO1-N was studied todetermine the specific cytochromes and to measure changes in the compositionof the respiratory system due to growth in various concentrations of oxygen or

types of growth substrates. Spectrophotometric analysis revealed that the quan-

tity and types of cytochromes changed in response to growth under variousconcentrations of oxygen. Growth on alkane and nonalkane substrates resulted inonly minor differences in cytochrome composition or oxidase activities. Mem-branes prepared from cells grown under oxygen-limiting conditions contained atleast one b-type cytochrome, cytochrome o, cytochrome d, and slight traces ofcytochrome a,, whereas membranes prepared from cells grown in the presence ofhigh oxygen concentrations contained only low levels of cytochromes b and o.

Polarographic measurements, electron transport inhibitor studies, and photoac-tion spectrum analyses indicated that cytochromes o, a,, and d were potentiallycapable of functioning as terminal oxidases in this organism. These experimentsalso revealed that all three cytochromes may be involved in the oxidation ofreduced nicotinamide adenine dinucleotide, succinate, or N,N,N',N'-tetramethyl-p-phenylenediamine.

Some bacteria are capable of altering the com-position of their respiratory systems in responseto changes in the availability of dissolved oxy-gen. The results of studies with Proteus vulgaris(6), Klebsiella aerogenes (16), Haemophilusparainfluenzae (45), Escherichia coli (31), andPseudomonas putida (43) have indicated thatcytochrome o is the major terminal oxidase incells grown in the presence of excess dissolvedoxygen, whereas growth under conditions of lim-ited aeration results in the production of cyto-chrome d (a2) as an additional terminal oxidase.Analyses of the oxygen uptake kinetics of E.

coli (37) and H. parainfluenzae (45) have sup-ported the proposal that cytochrome d is capableof efficiently reacting with low concentrations ofoxygen, permitting continued respiration underconditions of limited oxygen availability. Theadvantages of this type of respiratory systemmay be particularly applicable to hydrocarbon-oxidizing bacteria, since the growth of bacteriaon hydrocarbons has been associated with a highoxygen demand (26, 34).Acinetobacter calcoaceticus, a bacterium ca-

pable of growth on n-alkanes (42), contains res-piratory pigments which have been tentativelyidentified as cytochromes d and o (49). We havemeasured the influences of limited aeration andalkane growth on the composition of the cyto-

t Present address: Department of Microbiology, The Uni-versity of Texas at Austin, Austin, TX 78712.

chromes in Acinetobacter sp. HO1-N. Cellsgrown at the expense of acetate or hexadecanecontained cytochromes of similar composition.However, growth of the organism under oxygen-limiting conditions resulted in increased levelsof cytochromes b, o, and d and synthesis ofcytochrome a,. These changes were accompa-nied by differences in the photoaction spectraand an increased resistance to cyanide-mediatedinhibition of respiration.

MATERIALS AND METHODSBacterial strain. Acinetobacter sp. HO1-N (5, 20)

was used throughout these studies. The bacterium wasmaintained by serial transfers on agar slants of thefollowing composition (in grams per liter): nutrientbroth (Difco), 1.5; yeast extract (Difco), 2.0; peptone(Difco), 5.0; K2HPO4, 3.7; KH2PO4, 1.3; NaCl, 2.0; agar(Difco), 2.0; pH 7.8. Inoculated slants were incubatedat room temperature.

Culture conditions. The organism was grown in amineral salts medium described by Scott et al. (39).The minimal medium was supplemented with 0.8%nutrient broth (Difco) plus 0.5% yeast extract (Difco),with a sterilized solution of sodium acetate (Baker) toa final concentration of 0.4%, or with sterilized n-hexadecane (Humphrey Chemical Co.) to a final con-centration of 0.3%.

Starter cultures in 100 ml of nutrient broth-yeastextract were inoculated from the agar slants and usedto inoculate hexadecane or acetate starter cultures.Cultures were grown at 28°C in 3-liter volumes in a 4-liter Erlenmeyer flask, stirred at 300 rpm with a mag-

859

860 ENSLEY AND FINNERTY

netic stirrer, and either sparged with sterile air at 1 to2 liters/min for high aeration or not sparged for lowaeration. The organism was also grown in 10-liter lotsin a Chemapac fermentation vessel (14-liter workingvolume) with either stirring at 900 rpm without airsparging for low aeration or stirring at 600 rpm withair sparging at 18 liters/min for high aeration. Growthwas monitored with a Klett-Summerson colorimeterequipped with a 540-nm filter, and the cultures wereharvested during the midexponential phase of growthat 28°C. Oxygen concentrations in the fermentationvessel were monitored with a Chemapac oxygen elec-trode and meter.

Preparation ofmembranes. Acinetobacter mem-brane fractions were prepared by lysozyme treatmentof washed cells by the method of Scott et al. (39).

Cytoplasmic membranes were enriched from thecrude suspension by differential centrifugation (39).The membrane suspension was centrifuged at 65,000x g for 3 min, and the resulting translucent orangepellet was suspended in 50 mM potassium phosphatebuffer, pH 7.8, to a final protein concentration of 5 to10 mg/mnl.Enriched cytoplasmic membranes were either used

immediately or stored at 4°C in the presence of 20 ,ugof chloramphenicol per ml for not more than 3 days.Enzyme assays. NADH:cytochrome c reductase

(EC 1.6.99.3) and succinate:cytochrome c reductase(EC 1.3.99.1) were assayed by the method of King andDrews (24, 25).NADH dehydrogenase (EC 1.6.99.3) and succinate:

2,6-dichlorophenolindophenol (DCIP) reductase (EC1.3.99.1) were assayed by following the rate of reduc-tion of DCIP at 578 nm (24).

Succinate dehydrogenase (EC 1.3.99.1) activity wasmeasured by the phenazine methosulfate-mediatedreduction of DCIP as described by Arrigoni and Singer(3), except that a fixed phenazine methosulfate con-centration was used.

All activities were measured at 25°C in 3-ml glasscuvettes with a 1-cm light path. Specific activitieswere calculated by using millimolar extinction coeffi-cients of 20.6 mM-' cm-' for DCIP at 578 nm and 21mM-' cm-' for cytochrome c at 546 nm. Controlcuvettes contained all components except the sub-strate, and the low endogenous reduction rates meas-ured were subtracted from the experimental values.Oxygen uptake measurements. Rates of oxygen

utilization by membrane in the presence of varioussubstrates were measured polarographically with aYellow Springs Instrument model 53 oxygen monitorattached to a Perkin-Elmer model 159 strip chartrecorder. All rates were determined in a 3.4-ml jack-eted-glass reaction chamber maintained at 30°C withconstant stirimng. The reaction mixture contained 50mM potassium phosphate buffer, pH 7.0, and 1 to 2mg of membrane protein. After endogenous rates ofoxygen uptake were determined, the reactions werestarted by adding one of the following substrates:NADH to 1 mM; sodium succinate to 4 mM; DCIP to0.3 mM; or N,N,N',N'-tetramethyl-p-phenylenedia-mine (TMPD) (Sigma) to 0.3 mM. The redox dyesDCIP and TMPD were dissolved in water beforeaddition and were kept reduced in the reaction cham-ber with a twofold molar excess of ascorbic acid

(Baker). Succinic and ascorbic acids were dissolved inwater and neutralized to pH 7.0 with 0.1 M NaOH.The initial dissolved oxygen concentration in equilib-rium with atmospheric oxygen was assumed to be 233JIM.

Inhibition analysis. The effects of various elec-tron transport inhibitors on respiratory activities weredetermined by polarographic methods as describedabove. NaN3 (Baker) was dissolved in 50 mM potas-sium phosphate buffer (pH 7.0), and KCN was dis-solved in dilute KOH (pH 11). Antimycin A (Sigma),2-heptyl-4-hydroxyquinoline-N-oxide (HOQNO)(Sigma), and rotenone were dissolved in a minimalvolume of absolute ethanol. Various concentrations ofthe inhibitors were incubated with the membranepreparations for 3 min before the reactions werestarted. The influences of antimycin A, rotenone, andHOQNO were corrected for the inhibition caused bythe addition of absolute ethanol.

Optical spectra. Difference absorbance spectrawere recorded on an Aminco model DW-2 spectropho-tometer. A liquid nitrogen dewar and Pyrex cuvette(path length, 1 mm) were used to record spectra at77°K. The cytochromes in the respiratory particleswere readily oxidized by oxygen; therefore, shakingthe particles in air was usually sufficient to produceoxidized samples for the reference cuvette. Samplesoccasionally were oxidized by the addition of 2 to 12Mul of 50 mM K3Fe(CN)6. Reduced samples of respira-tory particles were prepared by the addition of solidsodium dithionite. When carbon monoxide was used,it was bubbled into the reduced sample for 2 min.

Estimates of amounts of cytochromes present weremade from room temperature, reduced-minus-oxidizeddifference spectra by using extinction coefficients andwavelength pairs as follows: cytochrome(s) b, 560 to575 nm, I = 22 mM-' cm-' (9); cytochrome d, 625 to650 nm, I = 15 mM- cm-' (17,47,48); and cytochromea,, 585 to 593 nm, I = 15 mM-' cm-' (38). Thecytochrome o concentration was estimated from thewavelength pair 417 to 490 nm with CO-reduced-mi-nus-reduced difference spectra and I = 80 mM-' cm-'(10, 41).Photoaction spectra. A photoaction spectrum,

the photoinduced reversal of carbon monoxide binding(7, 8), is the classical technique for identification ofterminal oxidases in electron transport chains. Theutilization of pulsed tunable dye lasers as a source ofradiation for photoaction spectra has been reportedpreviously (11, 40). This method has been adaptedhere for use with a laser producing a continuous beamof light.

Cytoplasmic membranes prepared from acetatecells grown under high and low aerations were mixedwith CO-saturated 50 mM phosphate buffer (pH 7.0)to give a final membrane protein concentration of 0.5to 1 mg/ml in the 3.5-ml reaction volume of a Gilsonwater-jacketed reaction vessel. The reaction vesselwas closed with a ground-glass stopper containing acapillary for addition of reagents. NADH, ascorbate-TMPD, or sodium succinate was added through thecapillary to give a final substrate concentration asdescribed for oxygen uptake. Respiration was allowedto proceed until the small amount of residual oxygenin the chamber had been exhausted. After anaerobic

J. BACTERIOL.

ELECTRON TRANSPORT IN ACINETOBACTER

incubation for 3 min, the reaction was started by theaddition of 10 to 20 pl of 02-saturated phosphatebuffer, and rates of oxygen uptake at 300C with con-stant stirring were monitored polarographically as de-scribed previously. Uninhibited membranes were in-cubated in the presence of N2-saturated, rather thanCO-saturated, phosphate buffer.

Photoaction spectra were measured by comparingoxygen uptake rates in unilluminated experimentswith those rates obtained when the reactions wereexposed to various wavelengths of laser radiation at afixed intensity. A Coherent Optics tunable dye laser(model 590) was operated at dutput wavelengths from542 to 650 nm, with Rhodamine 6-G dye used to emitfrom 650 to 750 nm and sodium fluorescein used forwavelengths from 542 to 569 nm. The tunable dyelaser was pumped with a Spectra Physics model 164argon ion laser at an output of 3 W, and the dye laseroutput was maintained at 60 mW. Wavelength outputof the laser was calibrated with a Spex 1401 doublemonochrometer, and radiation intensity incident onthe Gilson cell was measured with a joule meter.Output from the dye laser was reflected from a front-surfaced mirror to a beam splitter which divertedapproximately 4% of the beam to a Spectra Physics1401C power meter photometer for measurements ofbeam intensity. The majority of the radiation wasdirected to a convex lens, where the laser beam wasspread to a diameter of 1.4 cm at 14 cm. The radiationentered a window in an aluminum foil-covered Gilsonreaction vessel and passed through the water jacket tothe reaction chamber. A Yellow Springs Instrumentshigh-sensitivity oxygen electrode was used for meas-urements of oxygen uptake. Operation of the laser forthe 2- to 3-min period used in these experimentscaused no discernible temperature increase in the re-action chamber.

Oxygen uptake. rates were determined for radiationat intervals of 1 to 5 nm for wavelengths varyingbetween 645 and 542 nm. Each experiment measuredthe effect of one wavelength of radiation, and theresults obtained were compared with oxygen uptakerates in the presence of CO with no radiation todetermine the degree of inhibition relief at each wave-length. The results were expressed as relief relative torelief at an arbitrary wavelength (550 nm) (19).

Protein estimation. The protein content was es-timated by the method of Lowry et al. (27). Bovineserum albumin (Sigma) was used as a protein stan-dard.

RESULTSGrowth on acetate and hexadecane. Aci-

netobacter sp. HO1-N was grown under condi-tions of high and low aerations in a Chemapacfermentation vessel. It was observed that therate of growth could be controlled by regulatingthe flow of air into the fermentation vessel (Fig.1). Acetate-grown cells exhibited specific growthrates of 0.59 and 0.09 under conditions of highand low aerations, respectively. Hexadecane-grown cells showed specific growth rates of 0.62and 0.09 under conditions of high and low aera-

a0

z

s

2 4 6 S 10flours

FIG. 1. Growth curves of Acinetobacter sp. HOI-N. Symbols: 0, growth with high aeration; A, growthwith low aeration. Open symbols show growth at theexpense of acetate, and closed symbols show growthat the expense of hexadecane.

tions, respectively. Measurements of dissolvedoxygen concentrations after initiation of growthshowed a rapid decrease, and dissolved oxygenwas undetectable by the midexponential growthphase. An increase in the rate of aeration andstirring sufficient to maintain a dissolved oxygenconcentration of 4% caused no measurablechanges in the rate of growth. The maintenanceof dissolved oxygen concentrations above 16% ofsaturation resulted in the lower specific growthrate of 0.39. The growth rate of Acinetobactersp. HO1-N was similar with hexadecane or ace-tate and is in marked contrast to the growthresponse of Candida tropicalis to hydrocarbonsubstrates (12).Cytochrome composition. A comparative

analysis of the cytochrome composition of Aci-netobacter sp. HO1-N was undertaken to detectalterations in respiratory chain composition inresponse to growth in the presence of hexadec-ane or limited oxygen availability.

Cytoplasmic membranes prepared from cellsgrown at the expense of acetate or hexadecaneunder conditions ofhigh aeration were examinedfor the presence of cytochromes by reduced-minus-oxidized difference spectra taken at298°K (room temperature) and 77°K. A differ-

VOL. 142, 1980 861

862 ENSLEY AND FINNERTY

ence spectrum ofmembranes prepared from ace-tate-grown cells revealed the presence of at leasttwo b-type cytochromes, with a peaks at 564 and555 nm, ,8 peaks at 536 and 529 nm, and a broadSoret peak with a maximum at 425 nm (Fig. 2).In addition, the low-temperature spectrum re-vealed a small peak at 630 nm. Examination ofmembranes prepared from cells grown at theexpense of hexadecane under conditions of highaeration yielded a spectrum which was essen-tially superimposable on the spectrum obtainedfrom the membranes of acetate-grown cells.

In contrast, growth ofAcinetobacter sp. HO1-N at the expense of acetate or hexadecane underconditions of limited aeration caused easily de-tected changes in the cytochrome composition.The absorbance spectrum of membranes pre-pared from cells grown on acetate or hexadecaneunder conditions of limited aeration showed dis-tinct absorption maxima at 594 and 630 nm, inaddition to the double a and f8 peaks (Fig. 3).The new peaks were also easily discernible inroom temperature spectra. These data suggestedthat a significant production of new respiratorypigments resulted when cells were exposed tooxygen concentrations low enough to reduce therate of growth.Evidence regarding the identity of cyto-

chromes present in cells grown under low oxygentensions was obtained by carbon monoxide dif-ference spectra. The room temperature differ-ence spectra of membranes prepared from hex-adecane-grown cells were obtained for CO-

,.

I*I

5so 6o6i06B0WAV ELIENT (as)

FIG. 2. Difference spectra ofmembranes preparedfrom Acinetobacter sp. HOJ-N grown with acetateunder high aeration. The dithionite-reduced-minus-oxidized spectrum taken at room temperature (-)contained a membrane protein concentration of 6.2mg/ml. The difference spectrum obtained at 77°K( ) contained membrane protein at a concentra-tion of 3.8 mg/ml.

.4 .i i . .i .i .i si440 4810 520 560 600 640 680WAVELENGTI(nH )

FIG. 3. Difference spectra ofmembranes preparedfrom Acinetobacter sp. HO1-N grown with acetateunder low aeration. The spectrum obtained at roomtemperature (-----) contained a membrane proteinconcentration of 8.4 mg/ml, whereas the differencespectrum obtained at 77°K ( ) contained a mem-brane protein concentration of 4.7 mg/ml.

treated and dithionite-reduced membranes ver-sus dithionite-reduced membranes (Fig. 4). Ab-sorbance maxima at 417, 540, and 571 nm, whichare characteristic of cytochrome o (6), were ob-tained. In addition, the spectra of membranesprepared from cells grown under conditions oflow aeration showed peaks at 610 and 640 nm,which could represent CO complexes of a, andd, respectively (6). The inference is commonlymade that if a reduced cytochrome can form acomplex with CO to give a spectrally distinctspecies, it can also bind with oxygen, althoiugh ademonstration of participation in oxygen metab-olism requires additional evidence.The lack of an absorbance maximum in the

region of 450 nm indicated that the growth ofAcinetobacter sp. HO1-N at the expense of hex-adecane did not result in the production of de-tectable levels of cytochrome P-450. A CO-dif-ference spectrum with a pronounced peak at the450-nm region that appears after treatment withCO and reduction with dithionite represents adiagnostic characteristic for cytochrome P-450(21, 32, 33).A comparison of the concentrations of cyto-

chromes in membranes prepared from hexadec-ane- and acetate-grown cells revealed thatgrowth in hexadecane resulted in a doubling inthe concentration of cytochrome o (Table 1).Conversely, cultivation of the organism withgrowth-limiting concentrations ofoxygen causedsubstantial increases in the levels of membrane-

J. BACTERIOL.

ELECTRON TRANSPORT IN ACINETOBACTER 863

I%

aIt

9. 9a

4,40 fo sioi aoS00 oo *o0 0oWAVELEIITU (mm)

FIG. 4. Carbon monoxide difference spectrum ofmembranes prepared from Acinetobacter sp. HO1-Ngrown on hexadecane under conditions of low aera-tion. The sample cuvette contained dithionite-re-duced, CO-treated membranes, and the reference cu-vette contained dithionite-reduced membranes.

TABLE 1. Cytochrome composition ofmembranesCytochrome content (nmol/

Growth sub- Aeration mg of protein)strate rate'

b o d a,Hexadecane High 0.14 0.10 Tb NDc

Low 0.48 0.31 0.17 0.01

Acetate High 0.16 0.18 0.03 TLow 0.56 0.17 0.19 T

aAeration rates are as described in the text.bT, Traces.c ND, Not detectable.

bound cytochromes. A 3.5-fold increase in cyto-chrome b, a 3-fold increase in cytochrome o, anda 6- to 10-fold increase in cytochrome d andsynthesis of cytochrome a, were observed whengrowth rates were limited to one-sixth the max-imum rate by decreased aeration.

Soluble fractions prepared from Acinetobac-ter sp. H01-N did not contain detectable levelsof hemoprotein, even when the extracts hadbeen concentrated 20-fold by ultrafiltration.Photoaction spectra. Further evidence sup-

porting the identification of cytochromes o, a1,and d in Acinetobacter sp. H01-N was obtainedby the use of photochemical techniques. A pho-toaction spectrum of membrane-bound cyto-chromes prepared from Acinetobacter sp. H01-N grown at the expense of acetate with high andlow aerations is shown in Fig. 5. With NADH,succinate, or ascorbate-TMPD as electron do-nors, relief of CO-inhibited oxygen uptake wascharacterized by three peaks in the region of 540to 645 nm. The peaks at 570, 595, and 630 nm inthe spectrum of membranes prepared from cellsgrown with limited aeration corresponded to thea bands of photoaction spectra for cytochromes

o, a,, and d, respectively (6). This photochemicalevidence established the identity ofcytochromeso, a1, and d as potential terminal oxidases inAcinetobacter sp. HO1-N. A diminished contri-bution to respiration by cytochromes a1 and dwas observed in the photoaction spectrum ofmembranes prepared from cells grown underhigh aeration. Illumination of uninhibited mem-branes caused no measurable change in the rateof oxygen uptake.Oxygen uptake and dehydrogenase ac-

tivity. The oxidation of a number of electrondonors by membranes prepared from acetate- orhexadecane-grown cells under conditions of highor low aeration is summarized in Table 2. Acomparison of the rates of oxidation by mem-branes prepared from the various sources re-vealed that neither growth substrate nor aera-tion rate caused significant changes in 02 uptake.The most rapid oxidation ofTMPD and NADHoccurred when membranes prepared from 02-limited cells grown on acetate were used,whereas 02-limited growth on hexadecane pro-duced the highest levels of succinate oxidaseactivity. Reduced cytochrome c and DCIP werenot oxidized by membranes prepared from Aci-netobacter sp. H01-N. Soluble fractions con-tained no oxidase activities, and addition of sol-

1.2 -

a 1.0-

aaon

--Mo, I.

b-"a 0.6-

W 0.4'

0.2

al&

S60 S5o 600WAVELENCTH (mm)

620 640540

FIG. 5. Photoaction spectra of membranes pre-pared from acetate-grown Acinetobacter sp. HO1-N.Relief of CO-binding relative to relief at 550 nm wasplotted as a function of radiation wavelength. Mem-branes were prepared from cells grown under lowaeration (0) and high aeration (A).

A

.! . I

VOL. 142, 1980

9%

864 ENSLEY AND FINNERTY

TABLE 2. Oxidase activities ofmembranesActivity (umol of 02 consumed per min per mg of protein)

Electron donor Hexadecane-grown cells Acetate-grown cells

Aerated 02-limited Aerated 02-limitedSuccinate 0.05 ± 0.02" 0.10 ± 0.05 0.09 ± 0.04 0.07 ± 0.03NADH 0.15 ± 0.05 0.13 ± 0.06 0.10 ± 0.04 0.16 ± 0.06Ascorbate-TMPD 0.15 ± 0.1 0.07 ± 0.02 0.05 ± 0.02 0.11 ± 0.04Ascorbate-DCIP 0 0 0 0Ascorbate-cytochrome c 0 0 0 0

a Standard deviation of three determinations.

uble fractions to membrane preparations wouldnot stimulate oxidation rates by membranes.A report that growth of H. parainfluenzae

under oxygen-limited conditions resulted in in-creased levels of dehydrogenase activities (46)prompted an examination of dehydrogenase ac-tivities in Acinetobacter (Table 3). Membranesprepared from cells grown with high and lowaerations exhibited only small differences inNADH and succinate dehydrogenase activities.Inhibition analysis. The effects of inhibitors

on electron transport can provide data concern-ing the composition of the electron transportchain (13). Membranes prepared from Acineto-bacter sp. HO1-N were exposed to various con-centrations of selected inhibitors, and the ratesof oxidation of NADH, succinate, and TMPDwere determined (Table 4). The inhibitor con-centrations causing 50% inhibition were extrap-olated from plots of percent activity versus in-hibitor concentration. Antimycin A, rotenone,and HOQNO were not particularly effective in-hibitors of oxygen consumption. One unusualfeature of antimycin A inhibition was an effecton the oxidation of TMPD. TMPD oxidation byextracts from P. putida and Azotobacter vine-landii is not inhibited by similar concentrationsof antimycin A (19, 43). Succinate and NADHoxidase activities were inhibited by HOQNO,but TMPD oxidase activity was unaffected bythis inhibitor. Only small differences in sensitiv-ity to a particular inhibitor were noted whendata concerning membranes prepared from cellsgrown under high or low aeration were com-pared.A significant difference in the inhibitory effect

of cyanide was observed with membranes pre-pared from cells grown under high or low aera-tion (Fig. 6). Oxidations of NADH, succinate,and TMPD were inhibited to a similar extent inboth types of membranes. However, membranesprepared from cells grown with limited aerationwere more resistant to inhibition by cyanide.Oxygen uptake was inhibited 50% by 35 to 40,uM cyanide when extracts prepared from cells

TABLE 3. Dehydrogenase activities ofmembranesActivity (pmol of e- ac-ceptor reduced per min

Electron Electron acceptor per mg of protein)donor

Aerated 02-limitedcells cells

NADH DCIP 0.18 ± 0.06 0.22 ± 0.08NADH Cytochrome c 0.05 ± 0.02 0.06 ± 0.03Succinate DCIP 0.10 ± 0.05 0.11 ± 0.06Succinate DCIP + phenazine 0.07 ± 0.04 0.07 ± 0.04

methosulfateSuccinate Cytochrome c 0 0

TABLE 4. Effect of inhibitors on oxygenconsumption by membranes

InhibitorElectron .. concn (M)Aeration don Inhibitordonor causing 50%

inhibition

High NADH HOQNO 1.2 x 10-5NaN3 4.8 x 10-3Antimycin A 6.0 x 10-5Rotenone 5.3 x 10-4

Succinate HOQNO 6.6 x 10-5NaN3 4.2 x 10-3Antimycin A 1.0 x 10-4

Ascorbate- HOQNOTMPD Antimycin A 8.3 x 10-5

Rotenone 1.0 x 10-3

Low NADH HOQNO 2.5 x 10-5Antimycin A 6.1 x 10-5Rotenone 6.2 x 10-5

Succinate HOQNO 7.0 x 10-5Antimycin A 9.3 x 10-5

Ascorbate- HOQNOTMPD Antimycin A 1.3 x 10-4

grown with high aeration were examined,whereas greater than 300 ,M cyanide was re-quired for 50% inhibition of preparations fromcells grown with limited aeration. Concentra-

J. BACTERIOL.

ELECTRON TRANSPORT IN ACINETOBACTER

100

30

3 40

0.1 0.15 0.2CYANIEl CONCENTRATION, ml

FIG. 6. Effects of various concentrations of cyanide on membrane-associated oxidase activities of Acine-tobacter sp. HO1-N grown on acetate. Closed symbols refer to membranes prepared from cells grown underhigh aeration, and open symbols refer to membranes prepared from cells grown under low aeration. Oxygenuptake was measured in the presence of NADH (0, 0), succinate (A, A), and TMPD (5, *). Results areexpressed as a percentage of uninhibited activity.

tions of cyanide as high as 3 mM were necessaryfor complete inhibition of oxidase activities.

DISCUSSIONAbsorbance and photoaction spectra of Aci-

netobacter membranes gave evidence for thepresence of at least one b-type cytochrome inaddition to cytochrome o and small amounts ofcytochrome d. When growth was limited bydecreasing the dissolved oxygen concentration,increased levels of cytochromes b, o, and d andsynthesis of cytochrome a, were observed. Thisresponse is similar to that found in H. parainflu-enzae (45), E. coli (31), P. vulgaris (6), and K.aerogenes (16). For each of these species, thereis photochemical evidence that both cyto-chromes d and o can function as terminal oxi-dases. No significant differences in the cyto-chrome composition of membranes preparedfrom cells grown at the expense of acetate orhexadecane could be detected. This finding sup-ports the conclusions of Asperger et al. (4) intheir work with A. calcoaceticus.An examination of the data concerning oxi-

dase and dehydrogenase activities ofmembranesprepared from Acinetobacter sp. HO1-N indi-cated that neither hydrocarbon growth nor lim-ited aeration stimulated increases in enzymaticactivity. Dehydrogenase activities were similarin magnitude to those measured in other micro-organisms (24, 25), and the specific activitieswere not significantly altered by growth sub-strate or aeration rate. Measurements ofNADHand succinate dehydrogenase activities from

Paracoccus denitrificans have also shown thatoxygen concentration during growth had noclear effect on the specific activities of theseenzymes (36).

Oxidase activities were also unaffected by hy-drocarbon growth or low aeration rates. No sig-nificant differences were noted for the oxidationrates of NADH, succinate, or TMPD by mem-branes prepared from Acinetobacter grown un-der the various conditions. The oxidase rateswere similar to those measured in aerobicallygrown Rhodopseudomonaspalustris (25), P.pu-tida (43), and E. coli (37).The lack of an increase in oxidase activities

reflecting an increase in cytochrome composi-tion is not peculiar to Acinetobacter. Oxidaseactivities in E. coli (32), P. denitrificans (38),and Aerobacter aerogenes (16) have been re-ported to be unaffected by increases in the cy-tochrome composition of the cells. This has beenconsidered an indication that the dehydrogenasereactions are the rate-limiting step in respiration(1). The low dehydrogenase activities measuredin membranes prepared from Acinetobactertend to support the conclusion that respirationrates in this organism were also limited by thedehydrogenase.Data concerning the influence of hydrocarbon

growth on the levels of respiratory proteins inAcinetobacter should be interpreted with cau-tion. The relative amounts of the respiratoryproteins remained constant when measured perunit of membrane protein, but there is strongevidence that hydrocarbon growth causes a sig-

865VOL. 142, 1980

866 ENSLEY AND FINNERTY

nificant increase in the amount of membrane.Previous studies with this organism have docu-mented amounts of lipid phosphorous and mem-brane protein two to three times greater in cellsgrown at the expense of hexadecane (22, 28, 39).These experimental findings suggest that thetotal levels of respiratory proteins in Acineto-bacter, when calculated on the basis of units perwhole cell, are increased at least threefold bygrowth on hexadecane. This consideration re-lates further to the induction ofintracytoplasmicmembranes in Acinetobacter by growth on hex-adecane (22, 23, 39). If these specialized struc-tures do not contain respiratory enzymes, theircontamination of cytoplasmic membrane prep-arations would dilute the specific activities ofrespiratory proteins and mask hydrocarbon-in-duced increases in respiratory capacity.The flow of electrons through the respiratory

chain of Acinetobacter was analyzed with elec-tron transport inhibitors. Cyanide was the onlyinhibitor tested whose effects were significantlyinfluenced by the source of the respiratory par-ticles. Membranes prepared from cells grownwith limited aeration displayed less sensitivityto inhibition by cyanide. This effect is commonin a number of bacteria (31, 45, 48) and has beenattributed to increased synthesis of a cyanide-resistant cytochrome d (35). Since NADH andsuccinate oxidations were inhibited to a similarextent by the same concentration of cyanide,regardless of the source of respiratory particles,electrons abstracted from these substrates ap-pear to follow the same route through the elec-tron transport chain. This inference is supportedby data obtained with photorelief of carbonmonoxide inhibition. Identical photoaction spec-tra were measured with both NADH and succi-nate as electron donors, again indicating a simi-lar pattern of electron flow.The observation of TMPD oxidase activity in

membrane preparations derived from Acineto-bacter was surprising, since this organism hasbeen characterized as oxidase negative (5).Whole cells would not measurably oxidizeTMPD and could not be stimulated to do so bytreatment with toluene. However, membranepreparations would oxidize TMPD at one-halfto one-third the rate ofNADH oxidation, whichwas significantly greater than the autooxidationrate of ascorbate plus TMPD. Baumann et al.proposed that a positive oxidase test demon-strates that the organism possesses a cyto-chrome c component as part of its electron trans-port system (5). These workers also showed alack of spectrally detectable cytochrome c in sixtypically oxidase-negative strains ofAcinetobac-ter (5). Our studies also failed to detect the

presence of cytochrome c in Acinetobacter sp.HO1-N. Recent studies concerning the func-tional organization of electron transport chainshave placed cytochrome c at the outer surfaceof the cytoplasmic membrane (15, 18). If thismodel is correct, the requirement for cyto-chrome c in a positive oxidase reaction may berelated to the location of this protein near thecell surface.

Further analysis ofTMPD oxidation by mem-branes prepared from Acinetobacter revealed anumber of differences when compared to datafrom studies of oxidase-positive organisms. Cy-anide inhibited TMPD oxidation by Acineto-bacter membranes to approximately the sameextent as it did NADH and succinate oxidations.In addition, membranes containing elevated lev-els of cytochrome d demonstrated an increasedresistance to cyanide inhibition of TMPD oxi-dation. This is in direct contrast to findingsobtained with a number of other organisms.Potassium cyanide at concentrations sufficientto inhibit completely TMPD oxidation by ex-tracts of P. putida would only inhibit NADHand succinate oxidations by 30% (43). Mem-branes prepared from P. putida containing highlevels of cytochrome d exhibited increased re-sistance to cyanide inhibition of succinate andNADH oxidations, but not TMPD oxidation(18). The oxidation of TMPD is also confined toa cyanide-sensitive branch of respiratory sys-tems in Beneckea natriegens (44) and A. vine-landii (19). The results obtained in the presentstudy of Acinetobacter suggest that electronsfrom TMPD are not limited to a particularbranch of the respiratory pathway. Photoactionspectra of Acinetobacter membranes withTMPD as the electron donor showed reliefpeaksat wavelengths corresponding to all three ter-minal oxidase cytochromes, again suggestingthat TMPD may be oxidized by all of the ter-minal oxidase in Acinetobacter.Based on results of the present study, the

electron transport chain shown in Fig. 7 is pro-posed for Acinetobacter sp. HO1-N. NADH andsuccinate donate electrons via routes presentlyunknown to the respiratory system. The isola-tion of a ubiquinone from Acinetobacter (28),the ability ofcrude extracts to reduce menadionein the presence of NADH (Finnerty, unpub-lished data), and the inhibitory effects ofHOQNO on the oxidations of NADH and suc-cinate implicate ubiquinone as an essential in-termediate in the respiratory system. No orderof components in the chain has been demon-strated yet, but their arrangement probablyforms the basis of electron flow to the oxidaseso, d, and possibly a,.

J. BACTERIOL.

ELECTRON TRANSPORT IN ACINETOBACTER 867

TMPD

N AD H y-oUQbb\ d

S UCCINATEFIG. 7. Proposed electron transport pathway for

Acinetobacter sp. HOJ-N.

The multiple terminal oxidases of the electrontransport system proposed in Fig. 7 provide al-ternative routes for the flow of electrons to ox-

ygen. Little is known at present about factorswhich might favor respiration through one ter-minal oxidase over the others. Results from cy-anide inhibition studies led Pudek and Bragg toconclude that, when present in E. coli, bothcytochrome o and d function simultaneously asoxidases (36) and that the relative flow of elec-trons through these cytochromes is a function oftheir pool size in the membrane. However, arecent study on the kinetics of oxygen bindingby cytochromes o and d in E. coli indicated thatambient oxygen concentrations can regulate thecontribution to respiratory activity by either ofthese cytochromes (37).The respiratory function of cytochrome a, in

Acinetobacter is uncertain. Only trace amountsof this pigment were measured, although con-centrations did seem to increase in response togrowth of the organism under oxygen-limitingconditions. Photoaction spectra gave evidencethat cytochrome a1 in Acinetobacter is poten-tially a terminal oxidase, but photoaction spectraof the H. parainfluenzae respiratory chainshowed a peak of relief corresponding to cyto-chrome a,, even though Smith et al. demon-strated that this cytochrome al is not kineticallycompetent to support respiration (41). Cyto-chrome a1 in E. coli is also not considered tocontribute to observed respiration rates (14).However, in certain bacteria such asAcetobacterspecies, cytochrome al is thought to function asa terminal oxidase (6, 30).The electron transport system proposed here

for Acinetobacter is similar in many respects tothe respiratory system in Achromobacter, as

reported by Arima and Oka (2). Achromobacteralso lacks a c-type cytochrome, and electron flowdoes not appear to be restricted to a particularbranch ofthe respiratory pathway. In the Achro-mobacter model, terminal oxidases o, d, and alare all functional, but no photochemical evi-dence was given to support this contention.

ACKNOWLEDGMENTSWe thank L. Carriera for the use of his tunable dye laser

and D. DerVartanian, P. Hoffman, and V. Morgan for theirvaluable advice and technical assistance.

LITERATURE CITED1. Ackrell, B. A., and C. W. Jones. 1971. The respiratory

system of Azotobacter vinelandii. II. Oxygen effects.Eur. J. Biochem. 20:29-35

2. Arima, K., and T. Oka. 1965. Cyanide resistance inAchromobacter. I. Induced formation of cytochrome a2and its role in cyanide-resistant respiration. J. Bacteriol.90:734-743.

3. Arrigoni, O., and T. P. Singer. 1962. Limitations of thephenazine methosulphate assay for succinic and relateddehydrogenases. Nature (London) 193:1256-1258.

4. Asperger, O., H. P. Kleber, and H. Aurich. 1978.Cytochrome composition of Acinetobacter calcoaceti-cus. Acta Biol. Med. Ger. 37:191-198.

5. Baumann, P., M. Doudoroff, and R. Y. Stanier. 1968.A study of the moraxella group. II. Oxidative-negativespecies (genus Acinetobacter). J. Bacteriol. 95:1520-1541.

6. Castor, L N., and B. Chance. 1959. Photochemicaldeterminations of the oxidases of bacteria. J. Biol.Chem. 234:1587-1623.

7. Chance, B. 1952. The carbon monoxide compounds of thecytochrome oxidases. I. Difference spectra. J. Biol.Chem. 202:383-396.

8. Chance, B. 1952. The carbon monoxide compounds of thecytochrome oxidases. II. Photodissociation spectra. J.Biol. Chem. 202:397-406.

9. Chance, B. 1957. Techniques for the assay of the respi-ratory enzymes. Methods Enzymol. 4:273-329.

10. Chance, B. 1961. Cytochrome o, p. 433-435. In J. E. Falk,R. Lemberg, and R. K. Morton (ed.), Haematin en-zymes. Pergamon Press, Inc., Elmsford, N.Y.

11. Chance, B., and B. Erecinsa. 1971. Flow-flash kineticsof the cytochrome a3 reaction in coupled and uncoupledmitochondria using the liquid dye laser. Arch. Biochem.Biophys. 143:675-687.

12. Einsele, A., A. Fiechter, and H. P. Knopfel. 1972.Respiratory activity of Candida tropicalis duringgrowth on hexadecane and glucose. Arch. Mikrobiol.82:274-279.

13. Fernandez, R., M. Jones, and H. K. King. 1976. Puri-fication and properties of malate-NAD dehydrogenaseof Moraxella lwoffi. Biochem. Soc. Trans. 4:1080.

14. Haddock, B. A., J. A. Downie, and P. B. Garland.1975. Kinetic characterization of the membrane-boundcytochromes of Escherichia coli grown under a varietyof conditions by using a stopped-flow dual wavelengthspectrophotometer. Biochem. J. 154:285-294.

15. Haddock, B. A., and C. W. Jones. 1977. Bacterial res-piration. Bacteriol. Rev. 41:47-99.

16. Harrison, D. E. F. 1972. A study of the effect of growthconditions on chemostat-grown Klebsiella aerogenes.Biochim. Biophys. Acta 275:83-92.

17. Horio, T., T. Higashi, T. Yamanaka, H. Matsubara,and K. Okunukii. 1961. Purification and properties ofcytochrome oxidase from Pseudomonas aeruginosa. J.Biol. Chem. 236:944-951.

18. Jones, C. W. 1979. Energy metabolism in aerobes, p. 49-84. In J. R. Quayle (ed.), Microbial biochemistry. Uni-versity Park Press, Baltimore.

19. Jones, C. W., and E. R. Redfearn. 1967. The cyto-chrome system of Azotobacter vinelandii. Biochim.Biophys. Acta 143:304-353.

20. Juni, E. 1972. Interspecie$ transformation of Acinetobac-ter. genetic evidence for a ubiquitous genus. J. Bacteriol.112:917-931.

VOL. 142, 1980

868 ENSLEY AND FINNERTY

21. Jurtshuk, P., and G. E. Cardini. 1972. The mechanismof hydrocarbon oxidation by a Corynebacterium spe-cies. Crit. Rev. Microbiol. 7:239-289.

22. Kennedy, R. S., and W. R. Finnerty. 1975. Microbialassimilation of hydrocarbons. II. Intracytoplasmicmembrane induction in Acinetobacter sp. Arch. Micro-biol. 102:85-90.

23. Kennedy, R. S., W. R. Finnerty, K. Sudarsanan, andR. A. Young. 1975. Microbial assimilation of hydrocar-bons. I. The fine-structure of a hydrocarbon oxidizingAcinetobacter sp. Arch. Microbiol. 102:75-83.

24. King, M. T., and G. Drews. 1973. The function andlocalization of ubiquinone in the NADH and succinateoxidase system of Rhodopseudomonas palustris.Biochim. Biophys. Acta 305:230-248.

25. King, M. T., and G. Drews. 1975. The respiratory elec-tron transport system of heterotrophically-grown Rho-dopseudomonas palustris. Arch. Microbiol. 102:219-231.

26. Klug, M. J., and A. Markovitz. 1969. Fermentation of1-hexadecene by Candida lipolytica. Biotechnol.Bioeng. 11:427-440.

27. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1961. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

28. Makula, R. A., P. J. Lockwood, and W. R. Finnerty.1975. Comparative analysis of the lipids of Acinetobac-ter species grown on hexadecane. J. Bacteriol. 121:250-258.

29. McCaman, R. E., and W. R. Finnerty. 1968. Biosyn-thesis of cytidine diphosphate diglyceride by a particu-late fraction from Micrococcus cerificans. J. Biol.Chem. 243:5074-5080.

30. Meyer, D. J., and C. W. Jones. 1973. Reactivity withoxygen of bacterial cytochrome oxidases a,, aa3 and o.FEBS Lett. 33:101-105.

31. Moss, R. F. 1952. The influence of oxygen tension onrespiration and cytochrome a2 formation ofEscherichiacoli. Aust. J. Exp. Biol. Med. Sci. 30:531-540.

32. Omura, T., and R. Sato. 1964. The carbon monoxide-binding pigment of liver microsomes. I. Evidence for itshemoprotein nature. J. Biol. Chem. 239:2370-2378.

33. Omura, T., and R. Sato. 1964. The carbon monoxide-binding pigment of liver microsomes. II. Solubilization,purification and properties. J. Biol. Chem. 239:2379-2387.

34. Payne, W. J. 1970. Energy yields and growth of hetero-trophs. Annu. Rev. Microbiol. 24:17-52.

35. Pudek, M. R., and P. D. Bragg. 1974. Inhibition bycyanide of the respiratory chain oxidases ofEscherichia

coli. Arch. Biochem. Biophys. 164:682-693.36. Pudek, M. R., and P. D. Bragg. 1974. Reaction of

cyanide with cytochrome d in respiratory particles fromexponential phase Escherichia coli. FEBS Lett. 50:111-113.

37. Rice, C. W., and W. P. Hempfling. 1978. Oxygen-limitedcontinuous culture and respiratory energy conservationin Escherichia coli. J. Bacteriol. 134:115-124.

38. Sapshead, L. M., and J. W. Wimpenny. 1970. Theeffect of oxygen and nitrate on respiratory systems inMicrococcus denitrificans. J. Gen. Microbiol. 63:14.

39. Scott, C., R. A. Makula, and W. R. Finnerty. 1976.Isolation and characterization of membranes from ahydrocarbon-oxidizing Acinetobacter sp. J. Bacteriol.127:469-480.

40. Sewell, D. L., M. I. H. Aleem, and D. F. Wilson. 1972.The oxidation-reduction potentials and rates of oxida-tion of the cytochromes of Nitrobacter agilis. Arch.Biochem. Biophys. 153:312-319.

41. Smith, L., D. C. White, P. Sinclair, and B. Chance.1970. Rapid reactions of cytochromes of Hemophilusparainfluenzae on addition of substrates or oxygen. J.Biol. Chem. 245:5096-5100.

42. Stewart, J. E., R. E. Kallio, D. P. Stevenson, A. C.Jones, and D. 0. Schissler. 1959. Bacterial hydrocar-bon oxidation. I. Oxidation of n-hexadecane by a gram-negative coccus. J. Bacteriol. 78:441-448.

43. Sweet, W. J., and J. A. Peterson. 1978. Changes incytochrome content and electron transport patterns inPseudomonas putida as a function of growth phase. J.Bacteriol. 133:217-224.

44. Weston, J. A., and C. J. Knowles. 1974. The respiratorysystem of the marine bacterium Beneckea natriegens.II. Terminal branching to respiration of oxygen andresistance to inhibition of cyanide. Biochim. Biophys.Acta 368:148-157.

45. White, D. C. 1963. Factors affecting the affinity for oxygenof cytochrome oxidases in Hemophilusparainfluenzae.J. Biol. Chem. 238:3757-3761.

46. White, D. C. 1964. Differential synthesis of five primaryelectron transport dehydrogenases in Hemophilus par-ainfluenzae. J. Biol. Chem. 239:2055-2060.

47. White, D. C. 1965. Synthesis of 2-demethyl vitamin K2and the cytochrome system in Haemophilus. J. Bacte-riol. 89:299-305.

48. White, D. C., and P. R. Sinclair. 1971. Branched elec-tron transport systems in bacteria. Adv. Microb. Phys-iol. 5:173-209.

49. Whittaker, P. A. 1971. Terminal respiration in Morax-ella lwoffi. Microbios 4:65-70.

J. BACTERIOL.