JOURNAL OF BACTERIOLOGY, June 1973, p. 1045-1051Copyright i 1973 American Society for Microbiology

Vol. 114, No. 3Printed in U.S.A.

Genetic Mutations Affecting the RespiratoryElectron-Transport System of the Photosynthetic

Bacterium Rhodopseudomonas capsulataBARRY MARRS' AND HOWARD GEST

Department of Microbiology, Indiana University, Bloomington, Indiana 47401

Received for publication 3 November 1972

Alternative energy-converting systems permit the nonsulfur purple photosyn-thetic bacterium Rhodopseudomonas capsulata to grow either with light or(dark) respiration as the source of energy. Respiratory mutants, unable to growaerobically in darkness, can be readily isolated and the defective step(s) in theirrespiratory mechanisms can be identified by study of biochemical activities inmembrane fragments derived from photosynthetically grown cells. Such analysisof appropriate mutants and revertants permits construction of a model for therespiratory electron-transport system of the wild type. The results obtainedindicate differential channeling of electrons derived from succinate and reducednicotinamide adenine dinucleotide, and are interpreted in terms of a branchedelectron-transport scheme. The scheme provides a guide for further, more refinedanalysis of the respiratory mechanisms through biochemical genetic approaches,and several of the mutants isolated can be exploited for investigation of unsolvedproblems relating to interactions between respiratory and photosynthetic elec-tron transport and the mechanism of inhibition of bacteriochlorophyll synthesisby molecular oxygen.

Typical nonsulfur purple bacteria can obtainenergy for growth either by anaerobic photo-phosphorylation or dark aerobic (respiratory)phosphorylation. Both energy-conversion sys-tems appear to be localized in a commonmembranous framework, but, although theyare known to interact in various ways, themolecular details are still unclear. One ap-proach to elucidation of the alternative energy-yielding schemes and their relationships isthrough the study of appropriate mutants. Wechose to examine mutants (of Rhodopseu-domonas capsulata) blocked in respiration,which can still grow photosynthetically. In mu-tants of this kind, the residual components ofthe respiratory electron-transport system (ETS)present in photosynthetically grown cells can becharacterized, and this facilitates definition ofthe normal respiratory scheme. A previous re-port (9) from this laboratory described a mutantstrain of R. capsulata, respiratory-deficient byvirtue of a lesion in the reduced nicotinamideadenine dinucleotide (NADH) dehydrogenasecomplex, and in this paper we describe other

1 Present address: Department of Biochemistry, St. LouisUniversity School of Medicine, St. Louis, Mo. 63104.

types of electron-transport mutants. These in-clude strains with defects in the terminal por-tion of the ETS, detected through use of the"Nadi reaction" which has been employed instudies of mechanisms of biological respirationsince 1884 (5): a-naphthol + dimethyl-p-phenylenediamine + 02 - indophenol blue +H20. The mixed organic reactants (termed"Nadi" reagent, an acronym composed of thefirst two letters' of each chemical name) areoxidatively converted to the blue dye throughthe agency of cytochrome c plus cytochromeoxidase. Analysis of the various mutants iso-lated indicates that the ETS of R. capsulata isbranched and that electrons tend to be prefer-entially "channeled" through alternativeroutes, depending on their metabolic origins.Related studies, presented in an accompanyingreport (8), suggest that a flow of electrons froman intermediate catalyst of the ETS is essentialfor bacteriochlorophyll synthesis in the presenceof molecular oxygen.

MATERIALS AND METHODSBacterial strains and growth conditions. Unless

otherwise noted, the growth medium used was theso-called RCV synthetic medium (9; initial pH, 6.8),

1045

on March 27, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MARRS AND GEST

containing 0.4% DL-malic acid, 0.1% (NH,),SO.0.0001% thiamine hydrochloride, 0.01 M KPO4, andadditional inorganic salts as specified by Ormerod etal. (12). Illumination for photosynthetic growth wasprovided by Lumiline lamps (Sylvania) at an illumi-nance of approximately 500 foot-candles (5.38 x 10'lux). Such cultures were grown in vessels filled tocapacity with medium, or on agar plates (see below).

Strain Z-1 is an arsenate-resistant spontaneousmutant (16) of R. capsulata strain St. Louis (ATCC23782). Strain Ml (previously designated as Z-1[aer-13J; reference 9) is a respiratory-deficient deriva-tive of Z- 1 which lacks NADH dehydrogenase activityand, consequently, is unable to grow aerobically indarkness (i.e., with oxidative phosphorylation as theenergy supply). Strain M2 lacks both NADH andsuccinate dehydrogenase activities and was isolatedas follows. Ml was mutagenized with N-methyl-N'-nitro-N-nitrosoguanidine, and survivors were recov-ered as previously described (9). Mutagenized cellswere washed with 0.01 M KPO4 buffer (pH 6.8),containing 0.1% (NHJ),SO,, and resuspended in asuccinate-minimal medium (RCV-modified to con-tain 30 mM succinate instead of malic acid and onlyone-third as much CaCl,). The culture was incubatedanaerobically (argon atmosphere) in the light until itsturbidity had increased threefold, at which time 200 Uof penicillin G/ml were added. Photosynthetic condi-tions were maintained for 34 h, and surviving cellswere then washed free of penicillin. The washedsurvivors were spread on the surfaces of peptone plusyeast extract (0.3% each) agar plates, which wereincubated in an illuminated GasPack anaerobic sys-tem (BioQuest). Colonies of mutants unable to utilizesuccinate were subsequently identified by replicaplating to agar plates containing either succinate ormalate minimal medium. Strain M3 was isolatedfrom M2 as a spontaneous revertant which hadregained the ability to grow aerobically in darkness onmalate minimal medium (but not on succinate mini-mal).

Strain Z-1 was used as the parental type forisolation of mutants with defects in the terminaloxidase region of the respiratory system. Such mu-tants were identified by staining colonies on agarplates with the classical Nadi reagent (5) as follows.Plates are flooded for approximately 15 s with afreshly prepared 1:1 mixture of 1% a-naphthol in 95%ethanol and 1% N,N-dimethyl-p-phenylenediaminemonohydrochloride (DMPD) in water, then drained,and exposed to air. Colonies (grown either underanaerobic, photosynthetic or aerobic, dark conditions)composed of cells containing functional terminaloxidases stain a deep-blue color within 1 to 2 min.Lesions in terminal oxidase activities result in either aslow or virtual absence of staining. Survivors may berecovered from colonies on stained plates if the colonyis sufficiently large ( > 2 mm diameter) and if the cellsare restreaked on a fresh plate within a few minutes ofapplication of the stain. The Nadi-negative strainsM4 and M5 were identified as unstained coloniesformed by cells from cultures which had undergonepenicillin selection against aerobic growth as de-scribed earlier (9; we also isolated mutants similar to

M4 [see below] by staining colonies produced bymutagenized cells without prior selection).

Strains M6 and M7 are spontaneous revertants ofrespiratory-deficient strain M5; the revertants re-gained the capacity to grow aerobically in darkness.

Preparation of cell extracts and membrane frag-ments. Insofar as possible, membrane fragments(particles) were prepared from cells which had ex-perienced very similar culture conditions for manygenerations prior to disruption and extraction (withstrains unable to grow aerobically, however, themedium was supplemented with 0.5 mg of L-ascorbateper ml [9]; controls showed that such addition hadlittle or no effect on the activities measured). A"preculture" was inoculated with cells from frozenstorage stocks (photosynthetically grown in RCVmedium; glycerol was added to give a 10% concentra-tion just prior to storage at -70 C). The preculturewas incubated under anaerobic, photosynthetic con-ditions until early stationary phase was achieved(about 18 h at 30 C). A 5-ml amount of preculture wasused as inoculum for the "final" culture (180 ml),grown in the same fashion. Cells were harvested inearly stationary phase and disrupted in a Frenchpressure cell, and pigmented membrane fragmentswere isolated and stored essentially as describedearlier (9). In the current series of experiments, thesupematant fluid of centrifuged extracts was de-canted and the "fluffy" layer of relatively unpig-mented material, just above the deeply pigmentedmembrane sediment, was removed by vigorous "flick-of-the-wrist" snapping of the centrifuge tube. Themembrane fragments were suspended in 100 mMglycylglycine (pH 7.2) containing 4 mM MgCl,, andan equal volume of glycerol was added.

Analytical methods. Rates of 0, consumption bymembrane fragments suspended in 0.05 M KPO, (pH7.5; 1 mg of protein/ml) were measured polarograph-ically with a Yellow Springs Instrument Co. oxygenmeter, model 51, at 25 C; substrate concentrationswere: 4 mM NADH, 40 mM succinate, 0.2mM N, N, N', N'-tetramethyl-p-phenylenediamine(TMPD) plus 21 mM L-ascorbate, or 0.027 mM equinecytochrome c plus 21 mM L-ascorbate. NADH oxidaseactivity was also determined by spectrophotometricmeasurement of NADH oxidation rate in aerobicreaction mixtures as described earlier (9). Spectro-photometric determination of oxidation rate of re-duced cytochrome c was performed as follows. A 1mM solution of oxidized cytochrome c (equine, frac-tion Ill, Sigma Chemical Co.) was stoichiometricallyreduced (as determined by absorbancy measure-ments) by titration with 1 mM L-ascorbate. Thereduced cytochrome c was added to reaction mixtures[in 0.02 M tris(hydroxymethyl)aminomethane (Tris)-hydrochloride, pH 8.5; ca. 0.2 mg of membraneprotein per ml] to give an initial concentration of 0.04mM. Rate of absorbancy change at 550 nm wasmeasured with a Cary 14 recording spectrophotome-ter, and a reduced-minus-oxidized molar extinctioncoefficient of 2.1 x 104 cm'- was assumed (15).

Cytochrome c reductase activities were assayedspectrophotometrically as previously described (9).

Oxidative phosphorylation was assayed by meas-

1046 J. BACTERIOL.

on March 27, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

RESPIRATORY MUTANTS OF R. CAPSULATA

urement of esterification of 32P-labeled orthophos-phate as detailed by Melandri et al. (10).

Protein content of membrane fragments was meas-

ured by the method of Lowry et al. (7), as in an earlierstudy (9).

RESULTS

Nadi reaction and growth characteristicsof Z-1 and derived strains. Table 1 lists the R.capsulata strains considered in this investiga-tion and shows their facultative capacities fordark, aerobic growth, as well as response to theNadi reagent. Since use of the complete straindescriptions in discussions tends to be cumber-some, each isolate has been assigned a shortdesignation.

R. capsulata Z-1 is a typical nonsulfur purplebacterium in respect to ability to grow eitherphotosynthetically (anaerobically) or aerobi-cally in darkness. Strain Ml shows the normalNadi reaction, but is unable to grow as an

aerobe in the dark because of a severe block inan early step of the ETS, viz., the NADHdehydrogenase reaction (9). Strain M2, derivedfrom Ml, is additionally blocked in the succi-nate dehydrogenase reaction (see Table 3); thisadded alteration leads to inability to grow

photosynthetically with succinate as the source

of carbon and reducing power. M2 was con-

structed to obtain a cell type, blocked in both ofthe major pathways through which electronsenter the ETS, to study the mechanism bywhich 02 regulates bacteriochlorophyll synthe-sis (see reference 8). Strain M3 is a revertantobtained from M2, which has regained theability to grow aerobically in darkness with

malate as the substrate. Since the succinatedehydrogenase lesion persists in M3, succinatecannot support growth of this strain, eitherphotosynthetically or aerobically in darkness. Itis apparent that the succinate dehydrogenaselesion alone does not confer the respiratory-defi-cient phenotype. The latter, however, can resultfrom a severe block in the NADH dehydroge-nase step, since the bulk of the electrons usedfor dark oxidative phosphorylation are derivedfrom NADH.

Mutants with impaired ability to catalyzethe Nadi reaction can be readily obtained.Strains M4 and M5 were selected for furtherstudy because they showed particularly lowNadi reaction activities. Although M4 is Nadi-negative, it grows readily under aerobic condi-tions in darkness. To our initial surprise, we

observed that a large fraction of Nadi-negativestrains behaved in this way. M5, however, isNadi-negative and unable to grow aerobically(some similar isolates were found to have di-minished NADH dehydrogenase activity, andsuch presumed double mutants are not consid-ered in this communication). Under aerobicdark conditions in the liquid malate medium(RCV), revertants M6 and M7-derived fromM5-show the same growth rate and cell yieldas the parent Z-1 strain (also true for M4).From Table 1, it is evident that anaerobic

photosynthetic growth is independent of theNadi characteristic, as expected.Oxidase activities in membrane fragments.

To more precisely determine the loci of defectsin the Nadi-negative strains, membrane frag-ments were analyzed for their oxidase activ-

TABLE 1. Phenotypes of R. capsulata Z-1 and derivative strains

Strain Nadi Aerobic dark growthb Photosynthetic growth"designation Muainareaction Malate Succinate Malate Succinate

Z-1 asr-1 + + + + +M1 asr-1, aer-13 + _ _ + +M2 asr-1, aer-13, sdh-1 + _ _ +M3 asr-1, aer-13r4, sdh-1 + + _ +M4 asr-1, aer-414 _ + + + +M5 asr-1, aer-412-512 _ _ _ + +M6 asr-1, aer-412r20-512 + + + + +M7 asr-1, aer-412-512r34 + + + +

a asr, genes determining sensitivity to inorganic arsenate; mutation asr-1 permits growth in media containingequimolar concentrations of phosphate and arsenate (16). The designations 13r4, 412r20, and 512r34 aretentatively used to indicate genotypes of revertants, and reflect our present inability to distinguish among truereversions, second-site reversions, and suppressor mutations.

Growth on malate or succinate minimal agar plates (except for tests with M4-M7 on succinate). Symbols:±, colonies visible after 1 day of incubation (usually seVeral millimeters in diameter by 3 days); -, no visiblecolony formation after 1 week of incubation at 30 C. Tests for photosynthetic (anaerobic) and aerobic darkgrowth of M4, M5, M6, and M7 on succinate were conducted by using liquid medium; as indicated, growthoccurred in all instances except with M5 under aerobic, dark conditions.

1047VOL. 114, 1973

on March 27, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MARRS AND GEST

ities with several substrates (Table 2). As antic-ipated, the TMPD oxidase activities of theparticles correlated exactly with the in vivo(colony) DMPD staining reactions, and we

conclude that the Nadi-negative strains de-scribed are blocked in the ability to transportelectrons from the substituted phenylenedia-mine dyes to 0,. With exogenous, reducedcytochrome c (and reduced dichlorophenol-indophenol, data not shown), the same patternof activities was observed, indicating thatTMPD and reduced cytochrome c are bothoxidized via electron-transfer pathways withcommon elements. (Reduced cytochrome c oxi-dase activities measured spectrophotometri-cally corresponded closely to those determinedby the 0, consumption method.)

Antimycin A (20 gM) was found to have no

effect on the rate of TMPD oxidation by R.capsulata particles, and this is consistent withthe generally accepted conclusion (4, 5) thatendogenous cytochrome c is the point of entryfor electrons from donors such as TMPD (andexogenous reduced cytochrome c) in varioussystems. Attention is drawn to the fact thatcertain mutants (viz., M4 and M7), blocked inthe ability to transport electrons from exoge-nous reduced cytochrome c to 0, retain res-piratory activity with NADH and succinate(see below).

Table 3 shows that the cytochrome c reduc-

TABLE 2. Oxidase activities of membranepreparations from Z-1 and derivative strains

Oxidase activities (pmol of 02 cQnsumed orNADH oxidized per h per mg of protein)

Strain Reduced

NADH NADH Succi- cyto- TMPDnate chrome

Z-1 3.3 1.3 0.9 2.5 5.0M2 0.37 0.09 0.03 2.6 5.1M4 4.5 1.5 0.9 0.04 0.05M5 0.26 0.06 0.07 0.00 0.00M6 1.6 0.7 0.7 2.5 4.6M7 4.3 1.6 0.8 0.03 0.08

aExcept as noted, the values given were deter-mined by measuring 0° uptake; corrections were

applied for endogenous 0, consumption and autooxi-dation of substrates. Each value represents the aver-age of at least two determinations, except values forZ-1 which are the averages of about five determina-tions. Activities may vary as much as twofold; how-ever, in general, relative values obtained with parti-cles prepared and assayed in parallel do not showappreciable variability.

NADH - NAD+ determined spectrophotometri-cally, in the presence of air.

TABLE 3. Cytochrome c reductase activities ofmembrane fragments

Reductase activities (pmol of equinecytochrome c reduced per h per

Strain mg of membrane protein)

NADH Succinate

Z-l 12.2 4.5M2 0.7 0.2M3 14.7 0.1M4 14.5 4.2M5 16.1 4.7M6 20.6 6.0M7 23.7 4.9

tase activities of Nadi-negative strains are ashigh as those observed in the parental Z-1strain, and this provides additional evidencethat the mutational blocks in M4, M5, and M7must affect the oxidation of reduced, endoge-nous cytochrome c. As expected, antimycin Awas found to inhibit NADH-cytochrome c re-ductase activity (90% inhibition with 20 #Mantibiotic) as well as NADH oxidase activity(70% inhibition; cf. reference 6).

Oxidative phosphorylation by membranefragments. The data in Table 4 indicate thatmembranes from mutants M2 and M5, whichhave little or no ability to oxidize NADH andsuccinate with 02, show correspondingly lowcapacity to catalyze oxidative phosphorylation(in vitro) with these substrates. On the otherhand, membranes of strains M4 and M7,though blocked in transport of electrons fromreduced cytochrome c to 02, show appreciableoxidative phosphorylation activities, especiallywith NADH.

Channeling of electron flow from NADHand succinate. The data obtained with strainsM6 and M7 suggested that different mutationalchanges in the terminal portion of the ETS mayaffect the flow of electrons from NADH andsuccinate differentially. Although these strainsshow very similar cytochrome c reductase activ-ities (Table 3), their NADH and succinateoxidase activities differ relative to Z-1. Thus, inM6 the NADH oxidase activity is relativelymore depressed than that of succinate oxidase(Table 2); strain M7 shows a slight effect in theopposite direction. Separate tests of independ-ently arising revertants of M5 (capable ofaerobic dark growth and showing the Nadi-posi-tive phenotype) disclosed that all four isolatesmanifested comparable, depressed NADH oxi-dase activities, whereas their succinate oxidaseactivities were at the level observed in theparental strain (Z-1). Differential electron-flowrates can also be inferred from the oxidative

1048 J. BACTERIOL.

on March 27, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

RESPIRATORY MUTANTS OF R. CAPSULATA

TABLE 4. Oxidative phosphorylation activities ofmembrane fragmentsPhosphorylation activities (pmol of ATPsynthesized per h per mg of membrane

Strain protein)

Endogenousa NADHb Succinateb

Z-1 0.10 0.26 0.61M2 0.11 0.05 0.04M4 0.12 0.18 0.08M5 0.12 0.02 0.02M6 0.19 0.25 0.85M7 0.20 0.18 0.12

aRate without added substrate.b Rate with substrate added, corrected for endoge-

nous rate.

phosphorylation data of Table 4. Observationssuggesting channeling of electrons from the twosubstrates are summarized in Table 5. It is clearthat electrons originating from NADH andsuccinate do not necessarily share a commontransport pathway to oxygen, since differentlesions in the terminal portion of the ETS caninhibit electron flow from these sources differ-entially.The channeling possibility was further ex-

plored by spectrophotometric experiments asfollows. A suspension of membrane fragmentsfrom strain Z-1 (2 mg of protein/ml in 0.05 MKPO, buffer, pH 7.5) was supplemented with asaturating concentration of NADH (10 mM)and separated into two portions. These wereplaced in the reference and sample cuvettes,respectively, of a Cary 14 recording spectropho-tometer, and, as expected, no absorbancy differ-ence was observed in the region between 530and 580 nm. A precisely measured volume of asuccinate solution was then added to the samplecuvette (final succinate concentration, 100mM), and an equal volume of buffer was addedto the reference cell. This resulted in a clear-cutdifference spectrum with a peak at 560 nmaccounting for some 20% of the maximal 560-nmabsorbancy observable in a dithionite-minus-ferricyanide difference spectrum, in accord withthe hypothesis that there are differences inaccessibility of cytochrome b molecules within

TABLE 5. Evidence for channeling of oxidativeelectron flow in membrane fragments

Activity ratioaStrain

0° uptake ATP synthesis

M6 1.0 0.29Z-1 1.4 0.43M7 2.0 1.5

a NADH-dependent/succinate-dependent; calcu-lated from data of Tables 2 and 4.

the membrane to electrons derived from succi-nate as opposed to NADH.

DISCUSSIONThe data obtained with the various mutants

can be rationalized by the (minimal) schemeproposed in Fig. 1. In brief, the dark, aerobicETS is envisaged as consisting of two "parallel"sequences which can be cross-linked at theubiquinone-cytochrome b stage, and whichshare a common terminal oxidase. It is sug-gested that electrons from succinate preferen-tially (but not exclusively) follow the channelrepresented by the reaction sequence: 2, 3, 4, 6;and those from NADH, the alternative route: 1,5, 6. Crossover between the sequences presuma-bly can occur through the agency of ubiquinoneor by electron transfer between the cytochromeb components.

Several aspects of the reasoning leading tothe scheme warrant more detailed comment.The branched model accounts for the fact that

Light -GBChl mIBChl+ TMPD

I t DMPD, Soluble cyt c

etc.

Succinate dehyd. cyttbCt

Q0Fo 02

strain

M2M3M4M5M6M7

missing step(s)

t,2244,554

FIG. 1. Electron transport pathways responsi-ble for energy conversion in R. capsulata. Numberedarrows represent steps in respiratory electron-trans-port pathways; some steps may involve the functionof more than one electron carrier. BChl, bacterio-chlorophyll; cyt, cytochrome; UQ, ubiquinone; e,electron. The strains listed carry genetic defectswhich render the indicated steps inactive. The ar-rangement of cytochromes b and UQ in the diagram isintended to suggest the mechanism of electron chan-neling; see text for further details. TMPD, DMPD,etc. are thought to interact with the ETS via mem-brane-bound cytochrome c. The upper part of thediagram depicts, in simplified form, the light-drivencyclic electron-flow system responsible for anaerobicphotophosphorylation: heavy arrows indicate primaryevents consequent on absorption of light by BChl;cytochromes b and c are represented as being sharedby the photosynthetic and dark aerobic systems-sharing of certain catalysts could account for inhibi-tion of respiration by illumination and various otherinteractions between the alternative energy-transduc-ing mechanisms (see 2, 9, 14).

1049VOL. 114, 1973

on March 27, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MARRS AND GEST

certain mutant strains (M4 and M7) blocked inthe transport of electrons from exogenous do-nors such as TMPD, etc. to 02 retain thecapacity for dark aerobic growth. The possibil-ity that absence of oxidase activity with theexogenous electron donors noted could be duesimply to inaccessibility of such donors to theendogenous cytochrome c of the particles isconsidered unlikely on two grounds, namely (i)reductase activity with exogenous cytochrome cserving as acceptor is unaffected by the Nadi-negative type mutational blocks, and (ii) sepa-rate tests showed that treatment of M4 mem-brane fragments with a variety of detergents didnot elicit oxidase activity with TMPD as sub-strate.

Strain M5, which is devoid of oxidase activi-ties, is represented as a double mutant blockedin both branches (steps 4 and 5). In accord withthis interpretation is the observation that suchmutants were found to occur with lower fre-quency than mutants of the M4 type. Moreover,the model predicts, as observed, that M5 shouldgive rise to two distinct classes of revertantscapable of dark aerobic growth, namely, aNadi-positive class which has regained theability to oxidize TMPD but still shows de-pressed oxidase activity with NADH (M6), anda Nadi-negative class (M7) with propertiessimilar to those of M4.The branched scheme also lends itself to

explanation of the observations indicating dif-ferential channeling of electrons derived fromNADH as opposed to those from succinate.Current discussions (1, 3) on the fine structureof energy-converting membrane systems em-phasize the accumulating evidence for a precisetopological orientation of electron transfer cata-lysts, and on this basis a functional distinctionamong cytochrome b molecules within the samemembrane matrix can be entertained. This is,in fact, suggested by the spectrophotometricobservations described which indicate that asubstantial fraction of the total membrane-bound cytochrome b is reducible by succinate,but not by NADH. Similar findings for Rhodo-spirillum rubrum membranes have been re-ported by Horio et al. (4).The nature of the lesions which block electron

flow at steps 4 and 5 in the various mutantstrains isolated is still unknown. These blockscould be associated with still unidentified elec-tron carriers or with the terminal oxidase itself.The chemical identity of the latter in R. capsu-lata also remains unclear; an exhaustive study byKlemme and Schlegel (6), however, appears torule out a-type cytochromes and catalysts of thecytochromoid c variety. We suggest that the

branches of the ETS rejoin at a common termi-nal oxidase because the NADH oxidase activi-ties of membranes from strains Z-1, M4, andM6 were found to show identical sensitivity toKCN (90% inhibition with 1 mM KCN). Thepossible existence of two terminal oxidases withsimilar sensitivities to KCN, however, cannotbe excluded at present. A number of (nonphoto-synthetic) bacterial species, in fact, havebranched ETSs and two kinds of terminal oxi-dases (13). Even if the NADH and succinateoxidase activities of R. capsulata are mediatedby (two) specific terminal oxidases, interpreta-tion of the data presented would not be substan-tially altered. Thus, crossover between the twosequences prior to the oxidases-would still haveto be assumed, since (i) exogenous cytochrome cis reduced with both NADH and succinate assubstrates and (ii) succinate can be oxidized inthe absence of electron flow between cyto-chrome c and the terminal oxidase, as seen instrains M4 and M7 (Table 2; compare resultswith succinate, reduced exogenous cytochromec, and TMPD).Klemme and Schlegel (6) have compared the

oxidase activities of aerobic, dark and photo-synthetically grown wild-type R. capsulata andfound no qualitative differences between thesystems from cells grown in these two modes(see also reference 11, for a summary of litera-ture on this and related questions). In view ofthis and the internal consistency of our ownresults (especially with M5 and its revertants),we regard the study of membranes from photo-synthetically grown cells of respiratory-defi-cient mutants as a valid approach to analysis ofthe oxidase components of the ETS.

It can be anticipated that the mutants de-scribed will prove to be useful for study of avariety of problems relating to the energy me-tabolism of purple bacteria. The model ad-vanced for the ETS of R. capsulata provides abasis for further studies by the biochemicalgenetics approach, especially in that it suggeststhe types of mutants which should be sought toclarify unresolved questions (e.g., whether ornot there is more than one terminal oxidase).Certain of the mutants have been particularlyhelpful for study of the mechanism by which O2inhibits bacteriochlorophyll synthesis (see refer-ence 8), and it is likely that other mutants willpermit an incisive attack on the refractoryproblem of whether or not certain catalysts areshared in the photosynthetic and aerobic elec-tron-transport pathways.

ACKNOWLEDGMENTSWe thank Carol Stahl and Helen Ritchie for expert tech-

1050 J. BACTERIOL.

on March 27, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

RESPIRATORY MUTANTS OF R. CAPSULATA

nical assistance, and Clarke T. Gray (Dartmouth Medi-cal School) and Stephen Lien (Indiana University) for helpfuldiscussions. This research was supported by grant GB-7333Xfrom the National Science Foundation.

ADDENDUM IN PROOFAn investigation by Baccarini, Melandri, Zannoni,

and Melandri (personal communication) on the elec-tron transport system of aerobically grown R. cap-sulata indicates the presence of two terminal oxi-dases, and they propose a branched scheme verysimilar to that suggested by our study of electrontransport mutants.

LITERATURE CITED1. Fessenden-Raden, J. M., and E. Racker. 1971. Structural

and functional organization of mitochondrial mem-branes, p. 401-438. In L. I. Rothfield (ed.), Structureand function of biological membranes. Academic PressInc., New York.

2. Gest, H., and M. D. Kamen. 1960. The photosyntheticbacteria, p. 568-612. In W. Ruhland (ed.), Ency-clopedia of plant physiology, vol. V/2. Springer Verlag,Berlin.

3. Harold, F. M. 1972. Conservation and transformation ofenergy by bacterial membranes. Bacteriol. Rev.36:172-230.

4. Horio, T., K. Nishikawa, Y. Horiuti, and T. Kakuno.1968. Mode of coupling of the phosphorylation systemto the electron transport system in Rhodospirillumrubrum chromatophores, p. 408-424. In K. Shibata, A.Takamiya, A. T. Jagendorf, and R. C. Fuller (ed.),Comparative biochemistry and biophysics of photosyn-thesis. University Park Press, State College, Pa.

5. Keilin, D. 1966. The history of cell respiration andcytochrome. Cambridge University Press, Cambridge.

6. Klemme, J.-H., and H. G. Schlegel. 1969. Untersuchun-gen zum Cytochrom-Oxydase-System aus anaerob imLicht und aerob im Dunkeln gewachsenen Zellen von

Rhodopseudomonas capsulata. Arch. Mikrobiol.68:326-354.

7. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

8. Marrs, B., and H. Gest. 1973. Regulation of bacterio-chlorophyll synthesis by oxygen in respiratory mutants,of Rhodopseudomonas capsulata. J. Bacteriol.114:1052-1057.

9. Marrs, B., C. L. Stahl, S. Lien, and H. Gest. 1972.Biochemical physiology of a respiratory-deficient mu-tant of the photosynthetic bacterium Rhodopseudo-monas capsulata. Proc. Nat. Acad. Sci. U.S.A. 69:916-920.

10. Melandri, B. A., A. Baccarini-Melandri, A. San Pietro,and H. Gest. 1971. Interchangeability of phosphoryla-tion coupling factors in photosynthetic and respiratoryenergy conversion. Science 174:514-516.

11. Oelze, J., and G. Drews. 1972. Membranes of photosyn-thetic bacteria. Biochim. Biophys. Acta 265:209-239.

12. Ormerod, J. G., K. S. Ormerod, and H. Gest. 1961.Light-dependent utilization of organic compounds andphotoproduction of molecular hydrogen by photosyn-thetic bacteria; relationships with nitrogen metabo-lism. Arch. Biochem. Biophys. 94:449-463.

13. White, D. C., and P. R. Sinclair. 1971. Branched electron-transport systems in bacteria, p. 173-211. In A. H. Roseand J. F. Wilkinson (ed.), Advances in microbialphysiology, vol. 5, Academic Press Inc., New York.

14. Yamashita, J., S. Yoshimura, Y. Matuo, and T. Horio.1967. Relationship between photosynthetic and oxida-tive phosphorylations in chromatophores from light-grown cells of Rhodospirillum rubrum. Biochim. Bio-phys. Acta 143:154-172.

15. Yonetani, T. 1965. Studies on cytochrome c peroxidase.II. Stoichiometry between enzyme, H,O,, and fer-rocytochrome c and enzymic determination of extinc-tion coefficients of cytochrome c. J. Biol. Chem.240:4509-4514.

16. Zilinsky, J. W., G. A. Sojka, and H. Gest. 1971. Energycharge regulation in photosynthetic bacteria. Biochem.Biophys. Res. Commun. 42:955-961.

1051VOL. 114, 1973

on March 27, 2021 by guest

http://jb.asm.org/

Dow

nloaded from