Conservation and Transformation of Energy Bacterial Membranes · CONSERVATION ANDTRANSFORMATION OF ENERGY nical difficulties of studying largely insoluble proteins. Thetheme that

BACTERIOLOGICAL REviEws, June, 1972, p. 172-230Copyright 0 1972 American Society for Microbiology

Vol. 36, No. 2Printed in U.S.A

Conservation and Transformation of Energy byBacterial Membranes

F. M. HAROLDNational Jewish Hospital and Research Center and Department of Microbiology, University of Colorado

Medical Center, Denver, Colorado 80206

INTRODUCTION ............................................................

A NOTE ONTERMINOLOGY.ENERGY TRANSDUCTIONS IN MITOCHONDRIA .........................Theories of Energy Conservation ............................................Chemical coupling hnpothesis .............................................Conformational coupling ..................................................Chemiosmotic hypothesis ................................................

Point and Counterpoint .....................................................Permeability of the mitochondrial membrane to protons ....................Vectorial organization of respiratory catalysts .............................Proton extrusion and the generation of a membrane potential ...............The coupling device: ATPase and ion translocation ........................Uncoupling and proton conduction ........................................Fluorescent molecules as probes of the energized state ......................

Metabolite Transport by Mitochondria ......................................Accumulation of calcium ..................................................Accumulation of potassium ...............................................Transport of phosphate and substrate anions ..............................

Summary: Energy Transductions in Mitochondria ...........................ENERGY TRANSFORMATIONS IN BACTERIAL MEMBRANES............

Structural Basis ...........................................................Oxidative Phosphorylation ..................................................General features of respiration and phosphorylation ........................Coupling factors: the role of ATPase ......................................Nature of phosphorylating particles from bacterial membranes .............Coupling of respiration to phosphorylation .................................

Photosynthetic Phosphorylation .............................................Coupling of Metabolism to Transport ........................................Transport systems and carriers ...........................................Group translocation ......................................................Kinetic approach to energy coupling .......................................Coupling of transport to the respiratory chain in membrane vesicles.Ion gradients and energy coupling .........................................

Role of the Membrane in Motility ............................................Bacteriocins and the Energized State ........................................

SUMMARY AND PROSPECT ..............................................LITERATURE CITED .......................................................

172174175175176176177180180180181182183184185185186188189190191193193195196199200201201202205207210214215216216

"Disconcertingly few laymen-even few col-lege graduates-really understand what thescholar means by 'truth.' It is not a citadel ofcertainty to be defended against error; it is ashady spot where one eats lunch beforetramping on. The professional thinker enjoysbeing where he is, but he also looks forward tonew vistas around the next bend, over the nextcrest."-Lynn White, Machina ex Deo.

INTRODUCTIONMost of the triumps of biochemistry were

won by reducing the exquisite architecture of

the living cell to a homogenate. Now that theelectron microscope has revealed the sophisti-cation of cellular structure, it seems aston-ishing that our brutal methods should havebeen able to generate those intricate charts ofmetabolic pathways which adom laboratorywalls: catabolism, biosynthesis, even transcrip-tion and translation have been successfullyanalyzed by assuming that, for practical pur-poses, a cell is just a bag of enzymes. Processesassociated with membranes-oxidative phos-phorylation, transport, motility and cell divi-sion-have been more refractory, but onemight well attribute this merely to the tech-

172

on October 5, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

http://mmbr.asm.org/

CONSERVATION AND TRANSFORMATION OF ENERGY

nical difficulties of studying largely insolubleproteins.The theme that pervades the present article

is that membrane functions must be ap-proached from a fundamentally different pointof view. It seems evident that in bacteria, as inmitochondria and other organelles, certainenzyme systems are not just anchored to thecytoplasmic membrane but are organizedwithin and even across the osmotic barrier.Systems of this kind catalyze metabolic reac-tions that are oriented with respect to themembrane: vectorial metabolism. By virtue ofthe spatial organization of the catalysts, chem-ical reactions of this kind may be accompaniedby the translocation of molecules or groupsacross the membrane.The enzymes with which we are most fa-

miliar are those that are soluble, and conse-quently catalyze reactions that have no macro-scopic direction in space. However, at the levelof each individual enzyme molecule, many ifnot all enzymic processes must be thought ofas having a particular orientation relative tothe active site. Enzyme complexes, such asthose which carry out the oxidation of pyru-vate or of fatty acids (not to mention ribo-somes), clearly depend upon the precise articu-lation of successive molecular events. But onlyin the case of reactions which take place acrossa membrane would the implications of vec-torial metabolism be fully apparent, since suchreactions result in mass translocation from oneside of a barrier to the other: So long as thebarrier is intact, vectorial reactions may gen-erate differences in concentration across thebarrier and, if ions or electrons are translo-cated, differences in electrical potential. Con-versely, the rate and extent of a vectorial reac-tion will be influenced by the concentration ofreactants in the two compartments and pos-sibly by the potential difference. Finally, tworeactions can be coupled through gradients ofconcentration or of electrical potential, so as tomake one reaction drive the other even thoughthey do not share a common intermediate(chemiosmotic coupling). Since vectorial me-tabolism would require precise orientation ofthe components, structure and function be-come inextricably intertwined.Do vectorial reactions across membranes

exist in fact as well as in theory? Although theconcept is only now receiving widespread at-tention, it is by no means a novel one. Morethan forty years ago, Lund suggested that bioe-lectric phenomena may result from the separa-tion of charges during the redox reactions ofrespiration. Lundegardh subsequently pro-posed specifically that reduction of a cyto-

chrome by a hydrogen carrier, such as flavine,results in the release of a proton on one side ofa membrane. This was the germ of the ideathat the catalysts of the respiratory chain areso oriented as to separate protons and elec-trons across a membrane, and that many fun-damental membrane phenomena ultimatelydepend upon this separation of electricalcharges. Robertson (317) has recently surveyedthe genesis and subsequent evolution of thisidea, from which sprang such diverse conceptsas Conway's redox hypothesis of active trans-port, efforts to understand acid secretion bythe stomach, and Mitchell's chemiosmotichypothesis of oxidative phosphorylation. Re-search over the past decade has made it plau-sible, if not certain, that several enzyme sys-tems built into biological membranes do cata-lyze vectorial metabolism. Among these arethe oxidation chains of respiration and photo-synthesis, adenosine triphosphatase (ATPase)complexes which translocate Na+ and K+across mammalian cell membranes and per-haps protons across the membranes of mito-chondria and bacteria, the phosphotransferasesystem for sugar uptake by bacteria, and oth-ers, less familiar.The object of the present essay is to examine

the role of vectorial processes in the generationof biological energy and its utilization by bac-terial membranes: oxidative phosphorylationand photophosphorylation on the one hand,active transport and other work functions onthe other. The principles that underlie theseprocesses are presently more thoroughly under-stood in mitochondria and chloroplasts thanthey are in bacteria, and a major aim of thissurvey is to integrate knowledge from eukar-yotic organelles with the more fragmentarydata from prokaryotic cells. But insight flowsboth ways, and it is obvious that microorga-nisms have much to offer the student of basicmembrane physiology: the capacity for growthboth under aerobic and anaerobic conditions,in environments that are extreme with respectto pH, temperature, or ionic composition, andfinally the powerful technique of specific mu-tations. It may be true that in molecular ge-netics bacteria have had their day in the sun,but in membrane physiology it is not yet noon.

Like all reviews of the literature, this onedraws on the work of others for both insightand information. Not always do the referencesreflect this adequately since, in order to keepthe total within reasonable bounds, I haveelected to cite recent papers whenever possi-ble, sometimes to the neglect of earlier pi-oneering work. Coverage extends to 1972. But Ihave leaned most heavily on the writings of

173VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

Peter Mitchell (256-259, 261, 264, 265). Morethan any other contemporary investigator,Mitchell has explored the experimental andtheoretical implications of the idea that bio-chemical processes may be oriented in space.To him is also due much of the terminologyemployed to describe such processes, includingthe term vectorial metabolism. The extent ofmy indebtedness will be obvious to the reader.

A NOTE ON TERMINOLOGYMore than a little of the complexity of

membrane physiology is semantic in origin,but not therefore trivial: the multiplicity ofterms encountered in the literature reflects thedivergent viewpoints of several schools ofthought.There is little difficulty over the first group

of terms. Uptake is usually employed in apurely operational sense to designate removalof the substrate from the medium. The termcovers translocation and adsorption, the sub-strate may or may not be chemically alteredand nothing is implied regarding mechanisms.Transport and translocation refer in a generalsense to any movement from one side of amembrane to the other. The mechanism is notspecified, but in practice "transport" implies aprocess more complex than simple diffusion.Facilitated diffusion is distinguished fromsimple diffusion by signs pointing to interac-tion of the substrate with a component of themembrane; saturation kinetics and stereospec-ificity are observed, and the rate is greaterthan would be expected for passive diffusionacross a lipid phase. However, by definitionthe process involves no energy input and re-sults only in equilibration of the substrateacross the membrane in accord with the elec-trochemical potential.

Difficulties begin to arise when we come to"carrier"-mediated transport that results inthe accumulation of nutrients and metabolitesagainst a concentration gradient. Group trans-location is now generally used to describe theresult of a reaction catalyzed by an enzymesystem oriented across a membrane: chemicalmodification of the substrate occurs concur-rently with its translocation (256, 258, 259,321). Examples include the translocation ofprotons and electrons as well as the vectorialphosphorylation of sugars. Frequently, how-ever, no chemical change of the substrate isdemonstrable, and accumulation is attributedto active transport. Readers interested in theorigin and precise definitions of this termmust refer elsewhere (224, 259, 340, 358). Suf-fice it to mention here that the word is cur-

rently employed in two distinct senses. Rosen-berg originally defined active transport as aprocess which results in the movement of asubstance from a region of lower to one ofhigher electrochemical potential; movementagainst the electrochemical potential gradientrequires input of "energy," and the mechanismof energy coupling can thus be seen to lie atthe heart of the concept of active transport.Dependence upon metabolism is the most vis-ible hallmark of transport processes labeled"active," but the actual linkage between trans-location and the metabolic machinery canoccur by diverse molecular mechanisms. Someare exceedingly indirect, as in the accumula-tion of sugars by the mammalian intestine atthe expense of a sodium gradient which is intum maintained by a specialized ATPase. Forthis reason, some investigators prefer the defi-nition of Kedem who would restrict the usageof active transport to those translocationswhich, like the Na+, K+-ATPase, are directlylinked to metabolic reactions.

Mitchell's terminology starts from the con-cept of vectorial metabolism (259); it avoidsthe ambiguities inherent in the concepts of"active transport" and "energy coupling," andwill be employed in this article whenever pos-sible. Primary translocations are those inwhich translocation is directly linked to a bio-chemical reaction, and may be of two kinds.Group translocation, defined above, occurs atthe substrate level. Enzyme-linked solutetranslocation is a process in which the sub-strate itself does not participate in the ex-change of covalent bonds, but is translocatedas a result of such a reaction. This correspondsto Kedem's definition of active transport, andis exemplified by the Na+, K+, and the Ca2+transport ATPases of mammalian cells.Secondary translocations are by definition

not directly linked to a chemical or metabolicreaction, but may be secondarily coupled. Thesimplest is uniport, the translocation of asingle substrate by the carrier center, whichcorresponds to "facilitated diffusion." Uniportresults in equilibration of the substrate acrossthe membrane, in accord with its electrochem-ical potential. More complex situations arisewhen two substrates interact with the carrier(259). "Symport (cotransport) reactions arethose in which two solutes equilibrate acrossan osmotic barrier such that the translocationof one solute is coupled to the translocation ofthe other in the same direction." In this event,an electrochemical gradient which impels themovement of one substrate (Na+ or H+, forexample) can drive the movement of anothersubstrate which rides on the same carrier

174 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



(sugar, or an amino acid), even though thelatter may move against its own electrochem-ical potential. "Antiport (counter-transport)reactions are those in which two solutes equili-brate across a barrier such that translocation ofone solute is coupled to the translocation ofthe other in the opposite direction." Thus, anasymmetrical distribution of one substrate willdrive the movement of the other in the oppo-site direction. It is important to recognize thatthese are not merely hypothetical situations;uniport and antiport, at least, are establishedmodes of action of ion-conducting antibiotics.

Concentrative transport is a convenientterm to denote the familiar and ubiquitouscapacity of living things to move metabolitesor nutrients against apparent concentrationgradients, without specifying any mechanisms.The membrane components which mediate

any of the above translocations will be referredto as "transport systems," implying nothing asto their number, nature of mode of action. Theterm porter is employed by Mitchell (258, 259,264); strictly speaking, it applies only to thecatalysts of secondary translocations. "Per-mease" still has adherents among microbiolo-gists, but the meaning of the word has becomecloudy: some investigators use it to designatethe transport system as a whole, others referspecifically to that element which recognizesthe substrates (see references 206, 207, 259,301). These ambiguities render the term un-suitable for the present essay. The meaning ofcarrier and carrier center is self evident, eventhough their nature may be far different.

Finally, "energy" will be employed in theloose manner customary among biochemists todesignate the capacity to do work.

ENERGY TRANSDUCTIONS INMITOCHONDRIA

It is no longer heretical, or even novel, tosuggest that mitochondria and chloroplastsevolved from microbial symbionts. Their prob-able microbial ancestry would in itself justify asection on organelles in a review of bacterialmembranes, but there is a more compellingreason: Current concepts of the role of mem-branes in energy metabolism were developedand refined largely by studies with mitochon-dria and chloroplasts. The relatively detailedexamination of mitochondria which follows isintended to help narrow the gap between thesubcultures and to provide a point of depar-ture for the discussion of bacteria.

Theories of Energy ConservationThe oxidation of reduced nicotinamide ade-

nine dinucleotide (NADH), succinate, andother electron donors is catalyzed by a mul-tienzyme chain associated with the inner (cris-tae) membrane of the mitochondrion. The oxi-dation chain, to use Racker's noncommitalterm (310), includes both hydrogen carriers(flavoproteins, quinones) and electron carriers(cytochromes). As a result of the oxidation,adenosine triphosphate (ATP) is synthesizedwith what appears to be a fixed stoichiometryof 3 moles of ATP per mole of NADH, 2 permole of succinate. The coupling of ATP syn-thesis to the redox reactions takes place atspecific sites in the chain; it involves an as-sembly of proteins, one of which has ATPaseactivity, which are collectively referred to asthe coupling device.One thinks of the respiratory chain pri-

marily as a device for the synthesis of ATP,but in fact mitochondria carry out a variety ofenergy dependent processes: reversal of thedirection of oxidation, as in the reduction ofnicotinamide adenine dinucleotide (NAD) bysuccinate in the presence of ATP; transhydro-genation, the reduction of NAD by reducednicotinamide adenine dinucleotide phosphate(NADHP); and, of particular concern in thisarticle, the accumulation of ions and sub-strates against the concentration gradient.Oxidative phosphorylation proper, i.e., theproduction of ATP, is thus but one of severalmodes in which respiratory energy may be uti-lized.

Energy-linked functions can in general besupported either by oxidation of an electrondonor or (with the oxidation chain blocked bylack of substrate or with inhibitors) by exoge-nous ATP. Hydrolysis of ATP has been shownto support ion accumulation and transhydro-genation, as well as reversed electron trans-port. The three coupling sites of the oxidationchain appear to be equivalent, and energymade available at one site can be used at an-other. Finally, ATP synthesis and other func-tions can also be driven by a preexisting iongradient in the absence of other sources of en-ergy. There thus appears to be a network ofreversible reactions linking the redox chain,ATP, and ion concentration gradients (re-views: 42, 125, 150, 230, 308, 310, 350-352, 377)The network was mapped chiefly by means

of inhibitors of the "energy-transfer" reactionswhich result in the ultimate synthesis of ATP.Oligomycin, rutamycin, and dicyclohexyl-carbodiimide (DCCD) inhibit both A'T'P syn-thesis and the ATPase activity of mitochon-dria and of submitochondrial particles. Theseinhibitors also block transhydrogenation andion transport supported by ATP, yet do not

175VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

affect these reactions when an oxidizable sub-strate serves as energy donor (references to theoriginal literature will be found in the reviewslisted above). These findings are of the utmostimportance for the argument developed here:They point to the existence of an energizedentity or state, which can be generated at theexpense of oxidation, ATP, or even of ion gra-dients and is the link between the various en-ergy-dependent functions. Basically similarconclusions were reached from studies onsubmitochondrial particles depleted of AT-Pase; these can use oxidation but not ATP tosupport ion transport (350). The central posi-tion of this entity, variously referred to as"common factor" (125), "energy pressure"(352), or even plain "--" so as not to prejudgethe thorny issue of its nature, is diagrammati-cally shown in Fig. 1.

Respiration and phosphorylation in mito-chondria are normally coupled so that any-thing that retards phosphorylation retards res-piration as well: lack of adenosine diphosphate(ADP), for example, or oligomycin. This is thewell-known phenomenon called respiratorycontrol. However, many conditions and re-agents are known which dissociate the linkage:2, 4-dinitrophenol is only the best known ofmany "uncouplers" which allow respiration toproceed in the absence of phosphorylation.Many of these are believed to dissipate theenergized common factor and block not onlyATP synthesis but all the energy-linked func-tions.Thus far there is little disagreement, partic-

ularly not about the crucial postulate of anenergized state or entity preceding ATP. It isthe nature and origin of this energy-conservingentity that is the bone of contention.Chemical coupling hypothesis. The first

hypothesis to be formulated explicitly (see Sla-ter, 351, 352) has as its essential feature thepostulate that the free energy released by theoxidation/reduction of adjacent electron car-

NADH ± Respiratory Chain <--- 02

NADP __|X__I __ Cation, AnionTrans- o r Translocationhydrogenation or-Tanoatn

, OligomycinDCCD

Dissipation byUncouplers

FIG. 1. Pathways of energy conservation andtransformation in mitochondria.

riers is conserved in the form of a high-energyintermediate, for instance:

Ared + Box + I=; A0x - I + BredThis formulation is analogous to substrate-

level phosphorylations such as the oxidation of2-ketoglutarate, in which the energy released isconserved by formation of succinyl coenzymeA. Experimental findings which will not berecapitulated here made it necessary to postu-late additional intermediates, including thenonphosphorylated intermediate X I thatserves as the common energy donor of Fig. 1and a phosphorylated precursor of ATP, X -

P.There has been no lack of candidates for the

chemical links between oxidation and phos-phorylation, but thus far none have long with-stood critical examination. This should per-haps not be surprising: intermediates may wellbe stable only in the hydrophobic environmentof the lipid matrix in which the oxidationchain is embedded. This would surely be truefor the high-energy forms of cytochromeswhich have recently been inferred from carefulspectroscopic studies, especially in the regionof the second coupling site (see 352, 395; butalso 156). Storey (361) has formulated, in con-siderable chemical detail, a model in whichredox energy is conserved as a strained S-Sbond in a protein that participates in theredox reaction; energy transfer takes place bytransesterification with an adjacent protein toform an acyl thioester, which in turn can in-teract with additional proteins. Again, suchintermediates would not be expected to sur-vive outside the membrane.The failure to isolate intermediates of oxida-

tive phosphorylation was the impetus that ledto the formulation of alternative hypotheses.However, as we shall see below, the chemios-motic and conformational hypotheses also in-voke intermediates that have defied isolation. Iwould rather emphasize that, even in its con-temporary form, the chemical hypothesis re-quires the membrane only to act as an orga-nizer for the catalytic elements, and perhapsto supply a hydrophobic environment, but as-signs to it no intrinsic function in energy-linked processes.Conformational coupling. 'rhe most fa-

miliar transducer of biological energy is mus-cle, which converts the energy released by thehydrolysis of ATP into mechanical work; ananalogous process, but operating in the reversedirection, might account for the synthesis ofATP. In this spirit, several investigators (27,42, 124, 144, 293, 361, 403) have formulated

176 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



hypotheses in which the energy released by theoxidation chain is conserved in the form ofstrained, metastable and energy-rich confor-mations of elements of the cristae membrane.In the presence of ADP and Pi, the energizedstructure would relax with concomitant forma-tion of ATP. A mechanism has even been re-discovered (403)-albeit only in very schematicform-by which conformation changes couldinduce differences in pH and electrical poten-tial across the membrane, and secondarily ac-tivate ion transport. Relatively little is pres-ently known about high-energy conformationsof proteins, and the conformational couplinghypothesis suffers from the paucity of bothchemical detail and direct experimental sup-port. So far as I am aware, the only pointedevidence comes from electron micrographswhich illustrate drastic structural changes inmitochondria. The various energized and non-energized forms are closely correlated with themetabolic state of the mitochondria. The diffi-culty remains of proving that the conforma-tional changes are the cause, rather than theconsequence, of metabolic events such as ATPsynthesis or ion transport.

Formally, the conformational coupling hy-pothesis appears to be merely a variant of thechemical hypothesis, substituting a metastablestructure for an unstable intermediate. Butthis statement unfairly belittles the significantdifference in point of view. Green and his asso-ciates rightly stress the potential importanceof this concept in unifying a wide variety ofmembrane processes. Unlike the chemicalhypothesis, the conformational one assigns tothe structure of the system a profound, albeitill-defined role in phosphorylation and trans-port.Chemiosmotic hypothesis. A distinguished

colleague once described his attitude towardsthe chemiosmotic hypothesis as one of "ad-miring incomprehension," and it is my im-pression that his perplexity is widely shared inthe microbiological community. This is a pity,for the controversy that was sparked by theintroduction of the chemiosmotic hypothesishas transformed some of the basic premises ofmembrane research.

First formulated by Peter Mitchell in 1961(255), the chemiosmotic hypothesis has beenset forth in considerable detail. The intentionof the present section is only to outline theargument, as simply as may be. For detailedand rigorous exposition readers must refer toarticles by Mitchell (257, 261, 266) and espe-cially to the lucid scrutiny of the hypothesis bythe late G. D. Greville (125). The critical ap-

praisals by Racker (310), Slater (352), and Sku-lachev (350) should also be consulted.The chemiosmotic hypothesis rests upon the

following basic postulates. (i) The inner mito-chondrial membrane, in which the oxidationchain and coupling device are localized, is es-sentially impermeable to most ions, includingboth OH- and H+. In consequence the mem-brane, or at least the barrier portion, has a lowelectrical conductivity.

(ii) The respiratory chain is an alternatingsequence of hydrogen carriers and electron car-riers, arranged across the membranes in loops.Oxidation of a substrate results in the translo-cation of protons from one side of the mem-brane to the other: in any one loop, two pro-tons pass across. The particular arrangementshown in Fig. 2 is taken from Skulachev (350);it includes one loop corresponding to the trans-hydrogenase reaction, and three for the oxida-tion of NADH.

For illustration consider the oxidation ofNADH, corresponding to the first couplingsite. Flavine is reduced at the inner surfacebut reacts at the outer surface with nonhemeiron; protons are ejected while the electronsreturn to the inner surface. Here they are do-nated to coenzyme Q, together with a pair of

Cytoplasm Membrane Matrix

2H+ m N + H+

2H +

2H+ 2H+

e- ~~~~+

xH20

FIG. 2. Pathway of proton and electron transferduring oxidation of NADH, according to the chem-iosmotic hypothesis. After Mitchell (257), Greville(125), and Skulachev (350). Q, Coenzyme Q; Z, hy-pothetical hydrogen carrier; FeS, nonheme iron pro-teins; b, c, a, a, cytochromes.

VOL. 36, 1972 177


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

protons from the matrix fluid. In the secondloop, the reduced coenzyme Q reacts with oxi-dized cytochrome b at the outer surface; pro-tons are ejected, but the electrons return to theinner surface to enter the third loop. Overall,the passage of two reducing equivalents overeach loop results in the appearance of two pro-tons in the outer phase while two protons dis-appear from the inner phase. Oxidation ofNADH-linked substrate will translocate a totalof six protons.

Translocation of protons in one direction isequivalent, both in theory and in practice, tothe movement of OH- the other way. We cantherefore state the postulate thus: oxidation ofNADH results, not in the formation of water,but in the production of the elements of water,H+ and OH-, on opposite sides of an im-permeable membrane.The end result of substrate oxidation is the

generation across the membrane of a gradientof pH and of electrical potential, with the ma-trix phase alkaline and electrically negativerelative to the outer phase. Both gradientsexert a force on the protons extruded by therespiratory chain, tending to pull them backacross the membrane into the interior. This"proton-motive force" is the key element inenergy coupling. It is essential to recognizethat the proton-motive force is the sum of twocomponents which are related but not iden-tical: a chemical or osmotic component, due tothe pH difference; and an electrical compo-nent due to the membrane potential. These areinterconvertible, and it is convenient to ex-press the proton-motive force in electricalunits as the sum of the two components:

Ap = z I - ZApH

(Ap is the proton-motive force in electricalunits and can be taken as a measure of theelectrochemical potential of protons. A4 is theelectrical potential difference across the mem-brane. Z = 2.3 RT/F, where R, T, and F havetheir usual meanings, has a numerical valuenear 60 mv in the biological range. ApH is thepH difference between interior and exterior. Ifthe inner phase were more alkaline than theexterior by one unit, and electrically negativeto the extent of 180 mv, the total Ap would be- 240 mv).Obviously, a difference of pH or of electrical

potential can be maintained only so long asthe membrane forms a vesicle that is topologi-cally closed; any defect or leak would dissipatethe proton gradient.

(iii) The gradient of pH and of electrical

potential generated by the respiratory chainreverses the direction of an ATPase so as tobring about net synthesis of ATP.

Mitochondrial membranes contain an AT-Pase which is inhibitable by oligomycin andDCCD; it is attached by a stalk to the insideof the cristae membrane and gives the innersurface its characteristic, knobbed appearance.The studies of Racker and his associates (sum-marized in 310, 311) leave little doubt that thisenzyme catalyzes the terminal step in the bio-synthesis of ATP. Now, ATPase is assayed bythe hydrolysis of ATP, which normally pro-ceeds virtually to completion; if the enzyme isto catalyze net ATP synthesis, something mustdrive the reaction in the opposite direction.

According to the chemiosmotic hypothesis,the ATPase catalyzes the obligatory and re-versible translocation of protons: The reactioncatalyzed bv the enzyme is to be represented,not as is usually done, (disregarding ioniza-tion):

ATP + H2O = ADP + Pi

but by either of the following formulations:

ATP + H20 + H+in =ADP + Pi + H+out (ATPase I)

or ATP + H20 + 2H+in;=ADP + Pi + 2H+,,t (ATPase II)

If this is correct, the equilibrium constant ofthe ATPase should be written, not as it usuallydone:

(ADP) (Pi)Keq =

(ATP)

but rather:

Keq= x(ADP)(Pi) (H+)out

(ATP) Q-H+) i

(for ATPase I)

or

(ADP) (Pi)Keq= >

(ATP)

(H+) 2iut(for ATPase II)

It is evident that the poise of the equilibriumwould depend upon the proton activity onboth sides of the membrane and thus upon theproton motive force. Therefore, a disequili-

178 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



brium pf H+ activity could, in principle, re-verse the direction of the ATPase. Using cer-tain values for the free energy of hydrolysis ofATP and for the steady-state concentrations ofATP, ADP, and Pi (all of these, incidentally,are subject to dispute), Mitchell arrives at theconclusion that an ATP/ADP ratio of 1 couldbe maintained by a ApH of 3.5 units, a A' of-210 mv or any combination of lesser valuesadding up to the same total proton-motiveforce.There is another, intuitively easier way to

visualize the proposed function of the ATPase.We can say that the enzyme is so localized inthe membrane that ATP, ADP, Pi, and H+have access to the active site only from theinside, OH- only from the outside, and wateras such no access at all. Synthesis of ATP re-quires "extraction" of water from ADP and Pi,but it is in fact the elements of water that areextracted: H+ would be pulled inward, andOH- outward, by the gradient of pH and elec-trical potential established by the respiratorychain. By either formulation, synthesis of ATPis associated with movement of protons in-wards, completing the circuit of a proton cur-rent which can be regarded as the driving forcefor ATP synthesis (Fig. 3).

It is now necessary to specify in some chem-ical detail just how the hydrolysis of ATP isreversibly coupled to the translocation of pro-tons across the membrane. This is not pres-ently possible, nor is it known with certaintyfor any system whether ATP hydrolysis is ac-companied by the translocation of one or oftwo protons. Mitchell's original proposal (257)invoked an anhydride structure, X I, andthe translocation of ionizable groups belongingto a (protein?) component of the system.These have been critically discussed by Gre-ville (125), Racker (310), and Skulachev (350).A more recent formulation (266) dispenses withthe anhydride in favor of closely specifiedpathways for the conduction of protonsthrough the system. Detailed consideration ofwhat must still be regarded as a purely specu-lative mechanism seems out of place here: InGreville's words (125), "Until more is knownabout the structures and enzymatic mecha-nisms of the components, ... the postulatedproton-translocating ATPase system with itsunidentified groups XH and IOH will remainthe least objective part of the Mitchell hypoth-esis."

According to the chemiosmotic hypothesis,the proton-motive force is the common factorresponsible not only for the synthesis of ATP

H + H+ H+

Respiratory choin

ATPose

ApH and,&*

Reversible ATPasepoised by ApHand &#

OH

FIG. 3. Chemiosmotic hypothesis in principle:extrusion of protons by the respiratory chain, genera-tion of ApH and A At, and the poising of ATPase bythe proton-motive force.

but also for the other energy-linked functionsof mitochondria. In later sections we shall con-sider in detail the role of pH gradients andmembrane potentials in the transport of cat-ions, anions, and substrates.

It may be useful at this point to bring outthe essential differences between the chemicaland the chemiosmotic interpretations of mito-chondrial function. This is not primarily amatter of intermediates: unidentified and elu-sive intermediates have been invoked by both.There remain, however, two fundamental dif-ferences. The chemical hypothesis postulateschemical linkage between the oxidation chainand the ATPase; according to Mitchell, theseprocesses are coupled only via the proton-mo-tive force and do not share a chemical inter-mediate. For this reason, the demonstrationthat a gradient of pH or of electrical potential(or both) does in fact exist is a touchstone ofthe chemiosmotic hypothesis (without neces-sarily being incompatible with other explana-tions). At a more generalized level, theMitchell theory postulates vectorial metabo-lism both within and across the membrane andrequires that phosphorylation be dependent

179VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

upon the topological integrity of the system.

Point and CounterpointEfforts by the protagonists to rigorously

disprove one or another of the adversary posi-tions outlined above have generated hundredsof research papers during the past decade. Anobserver of pacific inclinations may derivesome satisfaction from the gradual melding ofviews which appeared utterly irreconcilable atfirst. Proposals have also been formulatedwhich embody elements of all three hypoth-eses. Williams, for example (393, 394),stresses the generation of protons as the pri-mary event, but within the membrane ratherthan across it; in the hydrophobic lipid phase,bare protons would serve as dehydratingagents to drive the synthesis of ATP. Thehistorical fact is, however, that it is the chemi-osmotic and chemical coupling hypothesesthat led to predictions verifiable by experi-ment, and it is principally in terms of thesehypotheses that we shall examine the experi-mental observations.The following survey, of necessity selective

and incomplete, deals chiefly with controver-sies that have arisen (or will arise shortly) inthe context of bacterial membranes. Many is-sues, including those generated by measure-ments of redox potentials and of other thermo-dynamic parameters, have been omitted alto-gether since no equivalent data from microor-ganisms are presently available. Thus I do notpurport to judge between the rival theories:Racker (310), Skulachev (350), and Slater (352)have complied "scoreboards" which reveal thedegree to which this matter is still sub judice.The intention is rather to outline the basis forthe belief that principles embodied in thechemiosmotic hypothesis offer valid insightsinto membrane function even if the presentformulation of the hypothesis should need tobe revised.The present section is chiefly concerned

with the nature of the energized state in rela-tion to oxidative phosphorylation. Membranetransport will be considered in a separate sec-tion, followed by a summary.Permeability of the mitochondrial mem-

brane to protons. Ever since the recognitionthat the slow oxidation of exogenous NADH bymitochondria is due to a permeability barrier,it has been generally admitted that the mito-chondrial membrane excludes a variety ofsmall molecules. Translocation of moleculeswhich participate in mitochondrial metabo-lism, such as orthophosphate (Pi), ADP, ATP,

and various substrate anions is known to bemediated by functionally specialized transportsystems which will be discussed below. Thechemiosmotic hypothesis requires, however,that the membrane be also impermeable to H+and OH-: were it permeable to these ubiqui-tous ions, gradients of pH and of electricalpotential could not be sustained. The effectiveproton conductance, i.e., the sum of thepermeability to H+ and OH-, was measuredby Mitchell and Moyle (267) by following therate of hydrogen ion titration across the mito-chondrial membrane and was found to be verylow indeed, 0.45 umho/cm2 or 0.11 ,ug of H+per sec per pH unit per g of mitochondrialprotein.

For a number of years, Chance and his asso-ciates denied that the mitochondrial mem-brane is impermeable to protons, for both the-oretical and experimental reasons (59, 65, 66).In a more recent paper (272), however, the es-sential ion impermeability of the membrane istaken for granted, and this may now be re-garded as generally accepted. Artificial phos-pholipid bilayer membranes are also exceed-ingly impermeable to protons, which is mostlikely a general property of lipid membranes.

Vectorial organization of respiratory cat-alysts. It is an essential postulate of the che-miosmotic hypothesis that membrane cata-lysts, including both the oxidation chain andthe ATPase, are organized across the mem-brane and catalyze vectorial metabolism.Some of the earliest evidence bearing on thispoint came from the recognition that themembrane as a whole has a definite polarity.When mitochondria are negatively stained

and examined by electron microscopy, thematrix side is seen to be lined with sphericalparticles attached to the membrane proper bya stalk. These were initially referred to as ele-mentary particles and believed to contain theelectron transport chain. More recent workfrom Racker's laboratory, however, makes itvirtually certain that the stalked particles rep-resent only the coupling device; the sphericalparticles are identical with coupling factor CF1and possess ATPase activity (310, 311).

Disruption of the mitochondria yields twokinds of submitochondrial particles, both ofwhich still carry out oxidative phosphoryla-tion. Particles prepared with the use of digi-tonin have the same polarity as do the originalmitochondria, but those produced by ultra-sonic disruption appear to be inside-out. Thisis the simplest explanation of the finding thatin sonic particles the direction of many mito-

180 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



chondrial functions is the reverse of that seenin intact organelles. Among these are the local-ization of the stalked particles, which now facethe medium; the direction of proton transloca-tion, which is inward in the particles (155, 257,260, 262); the polarity of ion movements,which indicates that the particles generate anelectrically positive interior whereas that ofthe parent mitochondria is negative (126, 272,349, 350); and the accessibility of cytochromes,ATPase, and other enzymes to substrates, in-hibitors, antibiotics, or solvent extraction.The latter characteristic has been exploited

in several laboratories, but especially byRacker and his associates, to dissociate thecomponents of the oxidation chain and cou-pling device and to reassemble them in theirproper order and topological orientation (re-viewed in 310, 311, 350, 377). This work leavesno doubt that the oxidation chain is, in fact,arranged across the membrane. To accomodatethe information presently available, a min-imum of one loop is required, which places thedehydrogenases and cytochrome oxidase nearthe matrix surface, cytochrome c near the sur-face facing the cytoplasm. The actual organi-zation may well prove to be more complex asmore data accumulate.One very striking conclusion to emerge is

that some of the components play a structuralas well as a catalytic role and that reconstitu-tion of oxidative phosphorylation is invariablyassociated with recovery of a topologicallyclosed vesicle (310, 311). In summary, there ismounting evidence that proper topological ori-entation of all the components is, in fact, es-sential to phosphorylation though not neces-sarily to oxidation. This, like the association ofphosphorylation with a vesicular structure,confirms general insights of the chemiosmotichypothesis. However, the actual sequence ofelectron and hydrogen carriers and their par-ticular orientation remain the subject of muchdispute.Proton extrusion and the generation of a

membrane potential. According to the chem-iosmotic hypothesis, the oxidation chain is soarranged as to catalyze the extrusion of pro-tons, thereby generating a difference of pHand of electrical potential which in turn poisesthe equilibrium of the ATPase. Since theproton-motive force is said to be the only linkbetween respiration and phosphorylation,demonstration of the existence of such gra-dients becomes crucial. This is especially sofor mitochondria respiring in the absence ofADP, since under these conditions the elec-

trical potential ought to be maximal.There is no doubt that respiring mitochon-

dria do eject protons, by a very rapid process(155, 263, 268, 304). Mitchell and Moyle re-ported that, upon admission of a pulse of ox-ygen to an anaerobic suspension of mitochon-dria, 6 protons were extruded per mole of anNADH-linked substrate, 4 protons per mole ofsuccinate (262, 263, 268); the extrusion of pro-tons, if not always the same stoichiometry, hasbeen confirmed in other laboratories (e.g., 155).The catch is that proton extrusion by mito-

chondria proved to be linked to concomitantaccumulation of Ca2 . It now appears clearthat Ca2+ leaks out of the organelles duringanaerobiosis and is reabsorbed when oxygen isadmitted (reviews: 125, 235, 377). It is there-fore uncertain which is primary, the extrusionof protons or the transport of Ca2+; we shallreturn to this problem below.Tupper and Tedeschi (370-372) made a he-

roic attempt to resolve the issue by direct de-termination of the electrical potential acrossthe mitochondrial membrane. They impaledthe giant mitochondria of Drosophila salivaryglands upon microelectrodes and recorded asmall potential, interior positive and inde-pendent of metabolism. These results flatlycontradicted the chemiosmotic hypothesis andwere, in fact, predicted by one version of thechemical hypothesis (143). However, consid-ering the enormous size of the microelectroderelative to its target and the surprisingly lowvalue reported for the electrical resistance ofthe membrane, the possibility of mechanicaldamage is all too real.

It seems to this reviewer that, despite theindirect approach, measurements of the elec-trical potential based on ion movements carrymore conviction. If the respiratory chain gen-erates an electrical potential, interior negative,mitochondria should tend to accumulate cat-ions and expel anions; the reverse shouldapply to sonic particles which are inside-out. Itis important to study ions which, unlike Ca2 ,are not normally transported. The first appli-cation of this principle was made by Mitchelland Moyle (260, 263, 270). Mitochondria donot ordinarily take up K+ but can be inducedto do so by addition of the antibiotic, valino-mycin. The explanation generally given is thatvalinomycin forms a lipid-soluble complexwith K+ and thus renders the membrane, to allpractical purposes, freely permeable to K+ (forreviews see 132, 150, 277, 307, 308). Therefore,K+ added to respiring mitochondria shoulddistribute itself in accordance with the electro-

VOL. 36, 1972 181


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

chemical potential. A negative potential wouldlead to K+ accumulation as required by theNernst equation:

RT [K+]0ln

F [K+],

(In this equation, A 4t is the membrane poten-tial, [K+ ]O and [K+ ]i refer to the potassiumconcentration on the outside and inside, re-spectively, and R, T, and F have their usualmeanings; activity coefficients are neglected.)In fact, valinomycin-treated mitochondriaaccumulate K+ to.an extent consistent with apotential of some -250 mv, interior negative(263, 270, 325); it is only fair to mention thatPressman and his associates gave the induc-tion of K+ accumulation by valinomycin quiteanother interpretation (307, 308).The most convincing series of experiments is

due to Skulachev, Liberman, and their asso-ciates (22, 126, 238, 241, 349, 350) who studiedthe translocation of synthetic, lipid-solublecations and anions. The structures of some ofthese are given in Fig. 4. Briefly, intact res-piring mitochondria were shown to accumulatethe cations and expel the anions, as would bepredicted for an electrically negative interior.Conversely, sonic submitochondrial particlesaccumulate the anions but expel the cations.The results can hardly be attributed to activetransport of these nonphysiological ions; gen-eration of a membrane potential by mito-chondrial membranes thus appears to havebeen demonstrated (238, 349, 350). The poten-tial could arise by a proton-translocating oxi-

CH3

C3FCH2-N-CH2

CHM3

DDA +

eMcPCE

B

TPMP+ TPB

FIG. 4. Lipid-soluble synthetic ions. DDA+, di-benzyldimethyl ammonium; TPMP+, triphenyl-methylphosphonium; PCB-, phenyldicarbaunde-caborane; TPB- tetraphenyl boron.

dation chain, as proposed by the chemiosmotichypothesis, or by some other process; the alter-natives are discussed at length in Skulachev'svaluable review (350).The coupling device: ATPase and ion

translocation. The coupling device, whichmediates the energy-transfer reactions thatculminate in the synthesis of ATP, has provento be a complex and sophisticated system. TheATPase proper, or F1 particle, is a large pro-tein (probably a hexamer) composed of sub-units of several kinds (62, 115, 343, 375). Atleast four additional proteins, and at least onephospholipid, are required to reconstitutenormal coupling function in submitochondrialparticles: one of the proteins inhibits, ormasks, the ATP hydrolase activity. Inhibitorsof the energy-transfer functions have provento be invaluable for analysis of the couplingdevice: oligomycin, rutamycin, and DCCDbind not to the ATPase itself but to othercomponents of the system. Inhibition of theATPase function is indirect and may reflecttransmitted effects on conformation. Otherinhibitors, including aurovertin and Dio 9,appear to interact with the F1 particle itself(60, 61, 218; reviews: 132, 151, 310).

It has long been apparent from studies withinhibitors that the coupling device is an ele-ment quite separate from the respiratorychain. Indeed, oxidation chain and couplingdevice can be physically dissociated: Groot etal. (127) found that promitochondria of anaer-obic yeast (which lack cytochromes as well asubiquinone) still carry out various energytransfer reactions involving ATPase. Con-versely, Kagawa and Racker (193) recombinedATPase, phospholipids, and an amorphousfraction devoid of respiratory activity to re-store vesicles which exhibited ATP-dependent,energy-linked functions.How this multienzyme system functions in

the synthesis of ATP is unknown. Various par-tial and exchange reactions, such as theADP/ATP exchange, are most easily under-stood in terms of a mechanism that invokesdiscrete steps in ATP synthesis, as does thechemical coupling hypothesis (42, 310). Thecase is strengthened by the detection (82) of aphosphorylated intermediate of the X-P type,which could be the elusive phosphoryl donor toADP. Fisher et al. (111) have reported a solubi-lized ATP synthetase complex which catalyzesvarious exchange reactions in a manner sensi-tive to uncouplers and oligomycin; they sug-gested that this preparation gives rise to atleast some of the intermediates of oxidativephosphorylation even though it is apparently

182 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



devoid of membranes.According to the chemiosmotic hypothesis,

the ATPase reversibly translocates protonsacross the membrane. Hydrolysis of ATPshould elicit electrogenic extrusion of protonsfrom mitochondria, and conversely, impositionof an appropriate gradient of pH or potentialshould bring about net ATP synthesis. Both ofthese predictions have been confirmed.

(i) The demonstration that a pH gradientimposed on chloroplasts in the dark resulted inATP synthesis (177) was an early and impres-sive vindication of the chemiosmotic hypoth-esis. Similar, but less dramatic, findings weremade with mitochondria (315). ATP synthesiscould also be supported by a K+ gradient: Ef-flux of K+ from mitochondria, induced by vali-nomycin, results in ATP synthesis with a stoi-chiometry so high as to be difficult to explainby the current version of the chemiosmotichypothesis (73, 322).

(ii) Addition of ATP to an anaerobic suspen-sion of mitochondria led to ejection of protons,whereas sonic particles absorbed protons underthese conditions. The stoichiometry is uncer-tain (257, 262, 269). This is accompanied byuptake of a cation, and little proton extrusionis seen unless a cation is present-Ca2 , orvalinomycin plus K+. Therefore, it is of crucialimportance to the interpretation of these ex-periments that ATP hydrolysis supports up-take of lipid-soluble cations by mitochondriaand of anions by the particles (22, 126, 238). Itfollows that ATP hydrolysis generates the pre-dicted electrical potential, presumably by elec-trogenic movements of protons.To the extent that the results were predicta-

ble, in principle if not in detail, from thechemiosmotic hypothesis, they must count inits favor. However, they are generally compat-ible with Fig. 1, regardless of the nature of"" so long as we take respiration, ion gra-dients, and ATP to be reversibly intercon-nected. Therefore, dramatic and important asthese experiments are, they do not permit anunambiguous choice between the mechanismsunder consideration.Uncoupling and proton conduction. Mito-

chondria are easily damaged by rough han-dling, detergents, and other treatments, all ofwhich dissociate respiration from phosphoryla-tion to a greater or lesser degree. Interest cen-ters, however, upon the chemical uncouplerswhich act at very low concentrations; 2,4 dini-trophenol was the first of these to be discov-ered and remains the most familiar. Typically,uncouplers block phosphorylation but stimu-late respiration; high concentrations of the

uncoupler may accelerate the dissipation of X- I, for instance by promoting the hydrolysisof an energy-rich intermediate (351).A major clue to the mode of action of uncou-

plers was supplied by the recognition thatmany uncouplers are lipid-soluble acids whosepK is such that at physiological pH valuesboth the protonated form and the anion willexist in substantial proportions (for structures,see Fig. 5). Mitchell (255-257, 260, 261, 264)proposed that uncouplers dissolve in themembrane and act as circulating carriers con-ducting protons across the barrier. Diffusion ofprotons would dissipate the proton gradient onwhich oxidative phosphorylation depends, andrelieve the restriction of respiration.During the past decade, much experimental

evidence has accumulated to support thethesis that many uncouplers are proton con-ductors. By use of their titration technique,Mitchell and Moyle (267) demonstrated thatcarbonylcyanide m-chlorophenylhydrazone(CCCP) and other uncouplers enhance the rateof proton diffusion across the mitochondrialcristae membrane by several orders of magni-tude. The rate of proton diffusion in the pres-ence of uncouplers is sufficient to account forthe uncoupling in terms of the chemiosmotichypothesis (155, 166, 241, 267, 268, 360). Com-pelling evidence for proton conduction by un-couplers came from studies with artificial lipidbilayers. Uncouplers increase the electricalconductivity and induce an electrical potentialacross a membrane separating two compart-ments which differ in pH (166, 239, 240). The

TCS

N--C-C-C_ N11N

NH

ICcCCCP

OH

f NO2

NO2

DNP

I

HC-CH\ /B10C 10

Decachlorobaren.

FIG. 5. Proton-conducting uncouplers. TCS, te-trachlorosalicylanilide; CCCP. carbonylcyanide m-chlorophenylhydrazone; DNP, dinitrophenol; dec-achlorobarene. Symbols: 0, CH; 0, BCI.

VOL. 36, 1972 183


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

results suggest that uncouplers enhance thediffusion of H+ (or OH-) specifically and havelittle effect on the diffusion of other ions. Theprecise mechanism of proton conduction isstill controversial. It has been suggested thatthe species which carries the current is thedimer of the undissociated and the dissociatedspecies, HA. A- (110). Most investigators,however, favor Mitchell's original proposalthat the uncoupler travels one way as the pro-tonated acid and the other way as the anion.The structures of the anions (Fig. 5) tend todelocalize the charge and enhance solubility ofthe anion in the lipid phase (165, 231, 239, 257,261, 264). b

It should be pointed out that the protonconductors shown in Fig. 5 carry out an elec-trogenic transport of protons. This stands incontrast to antibiotics, such as nigericin andmonensin, which catalyze antiport of H+ for K+or Na+ (reviews: 132. 150, 277, 307, 308). Thelatter antibiotics do not enhance the electricalconductivity of a membrane, nor are they un-couplers-presumably because they do not dis-sipate the electrical potential. On the otherhand, valinomycin, which conducts K+, is anuncoupler under certain conditions, becauseinflux of K+ dissipates the potential (261, 325).Despite the general acceptance of the phe-

nomenon of proton conduction, it is by nomeans universally accepted that uncoupling isthe consequence of diffusion of protons acrossthe membrane. It will be recalled that thechemical theory proposes that these two proc-esses are linked by energy-rich intermediateswhich normally exist in a lipid, hydrophobicenvironment. Proton conductors may well cat-alyze hydrolysis of such compounds by al-lowing access of protons to the active site (166,361). Wilson and his associates (367, 396) haveexamined in detail the pH profiles of uncou-pling and proton conduction. The two are notthe same, and the authors concluded that un-coupling is not the result of proton conductionbut rather is due to general acid or base catal-ysis of a hydrolytic reaction taking place in themembrane matrix (but see also 165). In thisconnection, it should be recalled that manyuncouplers bind to proteins of the mito-chondrial membrane (383), and this could alsoplay a role in uncoupling. In summary, whilethere is no question that many uncouplers doconduct protons across lipid membranes, itseems impossible at this time to say whetheruncoupling results from passage of protonsacross or into the coupling membrane.Fluorescent molecules as probes of the

energized state. The fluorescence of certainmolecules, such as ANS- (1-anilino-8-naph-thalene sulfonate), is greatly enhanced if thesubstance is localized in a hydrophobic envi-ronment. This property rendered ANS-, whichreadily binds to proteins, useful in detectingchanges in protein conformation. The recentapplication of fluorescent dyes of this kind tomitochondria provides a new tool with whichto probe the nature of energy coupling.

Briefly, binding of ANS- to submitochon-drial particles results in some enhancement offluorescence. Addition of substrate or of ATPproduces further enhancement, which can beblocked by uncouplers. The fluorescence re-sponse lags behind the change in oxidationstate of the respiratory carriers. From theseand other experiments, it was inferred thatANS- responds to the energized state of themitochondrion, in much the same way as doother energy-linked functions. The enhance-ment of fluorescence was traced partly to anincreased affinity for the dye, resulting in ad-ditional binding, partly to increased quantumyield of the fluorescence of bound dye. Clearly,ANS- fluorescence reports a change in thestate of the mitochondrial membrane-per-haps a change in the conformation of the mem-brane, or the advent of a more hydrophobicstate due to extrusion of water. The responseof ANS- is too slow to reflect the primary en-ergy-conserving event (19, 48, 64, 83).Of great importance to the interpretation of

ANS- fluorescence is the recognition that inthis, as in so many other respects, the effectsseen with whole mitochondria are opposite inpolarity to those of particles. The fluorescenceof ANS- associated with intact mitochondriawas decreased when the organelles were ener-gized. Moreover, the fluorescent cationic dyeAuramin-O was found to respond in a manneropposite to that of the anionic ANS-: fluores-cence and binding were decreased in particlesand increased in intact mitochondria (17).Thus, the fluorescence response is a function ofthe sidedness of the membrane and of the dis-tribution of electrical charges either within oracross the membrane. Indeed, ANS- fluores-cence responded to the artificial induction of amembrane potential in the sense that a poten-tial, interior negative led to a decrease in ANS-fluorescence (20, 179).

These very recent results led to two possibleinterpretations, both of which emphasize theamount of ANS- associated with the (hydro-phobic) mitochondrial membrane. Azzi et al.(20) suggested that when mitochondria are

184 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



energized the inner (matrix) side becomesmore positive and the outer side more nega-tive. In particles, it is the matrix side whichfaces the medium, resulting in increasedbinding of ANS-, an anion, and enhanced flu-orescence. Quite possibly the charge redistri-bution involves specific regions of the mem-brane related to the coupling sites, rather thanthe bulk of the membrane phase. An alterna-tive interpretation was favored by Jasaitis etal. (179): they argued that the amount of ANS-bound to the membrane is a function of theconcentration of dye in the mitochondrial ma-trix, and this in turn reflects the electricalpotential across the membrane. ANS- wouldbehave like the lipid-soluble anions and cat-ions discussed earlier: Submitochondrial parti-cles develop an electrically positive interior; asANS- migrates into the particles in responseto the potential, a larger fraction becomes as-sociated with the membrane and enhancedfluorescence ensues.

Metabolite Transport by MitochondriaTo one accustomed to the microbiological

transport literature, that on mitochondria con-veys quite another flavor since it gives pride ofplace to the relationship of metabolite translo-cation to oxidation and energy transduction.The mitochondrial literature is thus worthexamining precisely for the lessons that itsunique focus may hold for those concernedwith transport in cells.The cristae membrane appears to be freely

permeable to some major metabolites, in-cluding water, oxygen, CO2, and pyruvic acid,but is otherwise a barrier to diffusion. A va-riety of transport systems breach this barrier,both primary and secondary. The primary sys-tems for proton translocation were discussedin the preceding section. In addition, there isevidence that the uptake of glutamate, aspar-tate, and fatty acids occurs by group transloca-tion. Among secondary systems, the beststudied is the porter for adenine nucleotides,but since it appears to have no equivalent inmicroorganisms it will not be considered here.This section will focus instead on the accumu-lation of cations, anions, and substrates and itsrelationship to the energized state of the mito-chondrial membrane. The burgeoning litera-ture on metabolite transport in mitochondriahas been reviewed by Lehninger et al. (235),Pressman (308), Chappell (67), Klingenberg(215), and by Van Dam and Meyer (377).Accumulation of calcium. Most animal

mitochondria rapidly accumulate divalent cat-ions by an energy-linked process. Only that ofCa2+, which has been studied most thoroughly,will be considered here.There is overwhelming evidence that Ca2+

uptake is mediated by a specific carrier of highaffinity, which is under genetic control. Ca2+binding by liver mitochondria has an apparentdissociation constant of about 10-6 M and isspecifically inhibited by lanthanum and bypraesodymium (232, 234, 250, 376). Mitochon-dria of yeast (55) and of the blowfly (56), whichlack the high-affinity binding sites, also lackthe characteristic, rapid Ca2+ accumulation.Recently, Lehninger (233) reported the release,after osmotic shock, of a protein which bindsCa2+ in a manner identical with that of thewhole mitochondria. This protein, reminiscentof the binding proteins of bacterial cells but ofmuch greater molecular weight, may mediatethe initial step in Ca2+ uptake.

There is a voluminous literature on the rela-tionship of Ca2+ uptake to oxidation, whichcan only be summarized here (see also: 232,235, 308). Briefly, addition of a limitingamount of Ca2+ to respiring mitochondriaelicits a burst of respiration, which ceaseswhen all the Ca2+ has been taken up. Concom-itantly, protons are ejected and the internalpH of the mitochondria rises by a unit or more(7, 121). The precise stoichiometry depends onconditions, but a ratio of 2 Ca2+ per electronpair passing each coupling site is typical.These experiments are conducted in absence ofa permeant anion; the Ca2+ taken up, maxi-mally about 100 Amoles per g of mitochondrialprotein, remains associated with the mito-chondrial membrane.Much larger amounts of Ca2+ can be accu-

mulated in presence of an anion which tra-verses the barrier-phosphate or acetate, forexample. Alkalinization of the matrix andproton ejection are suppressed, and Ca2+ accu-mulates in the matrix in form of a salt: whenacetate is the anion, the mitochondria swell,even to the point of lysis. Phosphate, however,allows the precipitation of internal calciumphosphate, a surprisingly complex processwhich is still imperfectly understood.

Ca2+ uptake is supported by respiration andblocked both by inhibitors of the chain and byproton-conducting uncouplers; significantly,Ca2+ is discharged by these inhibitors. Oligo-mycin and DCCD do not inhibit uptake. How-ever, with the respiratory chain blocked, Ca2+uptake can be energized by ATP: for eachmolecule of ATP hydrolyzed, about 2 mole-

185VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

cules of Ca2+ and 1 molecule of Pi are ab-sorbed. The ATP-supported uptake, unlikethat supported by respiration, is blocked byoligomycin and DCCD. These observationsand earlier ones on the uptake of Mg2+ byheart mitochondria (see reference 235 for re-view) are of crucial importance to the mappingof energy transfer pathways: they point to anenergized state, or intermediate, other thanATP, as the driving force for Ca2+ uptake (Fig.1).When we inquire into the nature of the ener-

gizing event, we quickly find ourselves en-meshed in arguments related to the chemicaland chemiosmotic coupling theories. No fewerthan three schemes, summarized in Fig. 6,were formulated to account for the relationshipof proton efflux to cation uptake.

(i) According to Mitchell (257, 260, 262, 263),uptake of Ca2+ is electrophoretic, in responseto the membrane potential generated by therespiratory chain. Indeed, since the electro-genic extrusion of even a very few protons gen-erates a large potential, significant net H+ejection is to be expected only in presence of acation which can enter the mitochondrion andthus compensate for the electrical displace-ment.

(ii) Chance and his associates (65, 66, 308)championed scheme b, in which the hypothet-ical energy-rich intermediate, X -I, pumpsCa2+ inward while protons are expelled due tothe membrane potential, interior positive. Thisscheme assumes a proton-permeable mito-chondrial membrane.

(iii) Finally, scheme c introduces the conceptof a proton pump, actuated indirectly by theoxidation chain via an energy-rich X - I inter-

a. ChemiosmoticCoupling:

b. ChemicalCoupling:

Respiratory Chain=-H' translocation =-ATP

.*- (cations)

Respiratory Chain =X ~ I =ATP

9 cation pump

.* cations

(Protons)

c. Proton Pump: Respiratory Chain=X ~I =ATP

8 proton pump

Protons am

.4 (cations)

FIG. 6. Possible interrelationships of cation andproton translocations.

mediate. As written here, this pump extrudesprotons to generate an electrical potential, in-terior negative, which in turn drives the elec-trophoretic uptake of Ca2+ (68, 69, 238, 350).Alternatively, one might envisage obligatorylinkage between the movements of protons andof Ca2+ -i.e., an energized exchange of Ca2+for protons.With the realization that the mitochondrial

membrane is not readily permeable to protonsand that the interior is electrically negative,scheme b has been eliminated. It is more plau-sible that a chemically driven proton pumpmay be linked to the Ca2+ carrier (Fig. 6c) insuch a way that the exchange of H+ for Ca2+is, overall, electrically neutral. But there seemsto be a growing measure of agreement thatproton ejection is the primary event. The pro-tons may be translocated by a vectorial respi-ratory chain, or by a proton pump energizedby an X - I intermediate. By either scheme,we can account for the characteristics of Ca2+uptake on the assumption of electrophoreticCa2+ uniport in response to the electrical po-tential (232).Accumulation of potassium. Unlike the

rapid and extensive accumulation of K+ socharacteristic of bacteria, uptake of K+ bymitochondria is sluggish-limited by the lowpermeability of the mitochondrial membraneto K+. Addition of ion-conducting antibioticsof the valinomycin type, however, inducesmassive and rapid uptake of K+ (68, 306).The extensive literature concerning the na-

ture and mode of action of ionophores has beenrepeatedly reviewed in recent years (132, 150,277, 307, 308), so that a quick sketch will suf-fice here. Briefly, in addition to the protonconductors discussed above, we recognize twoclasses of alkali-metal ionophores (Table 1).

(i) Valinomycin is the type species of a K+-specific uniporter. The molecule is a cyclicdepsipeptide which forms a clathrate such thatK+ is encaged in the center of a shell whoseexterior is hydrophobic. The complex is conse-quently lipid-soluble and functions as a circu-lating carrier for K+. It must be noted that thecomplex as a whole bears a positive charge, sothat net K+ movement is electrogenic: it bothgenerates, and responds to, an electrical poten-tial. In other words, valinomycin moves K+ inaccord with the electrochemical potential.Valinomycin is exceedingly specific for K+.

Enniatins and the macrotetralide "nactins"are less selective but, like valinomycin, theyact as circulating carriers by formation oflipid-soluble clathrates. Gramicidin, which isquite promiscuous, apparently conducts cat-

186 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



ions by forming a cation-selective pore.(ii) Nigericin is the prototype of a second

class of ionophores, all of which are monocar-boxylic acids. They were originally recognizedby virtue of their capacity to reverse the actionof valinomycin-that is, to discharge K+ frommitochondria. Their mode of action is nowknown to depend again on formation of alipid-soluble clathrate. However, it is the anionof nigericin that complexes K+ to give an elec-trically neutral complex; the protonated niger-icin does not bind K+. Consequently, the anti-biotic tends to carry out exclhange of K+ for H+,or K+/H+ antiport. The related antibioticmonensin catalyzes Na+/H+ antiport.The characteristics of the K+ uptake in-

duced by valinomycin are qualitatively similarto those of Ca2+. Translocation, which occursagainst a large concentration gradient, is en-ergy linked and can be supported either byrespiration or by ATP. Uncouplers prevent K+

accumulation and discharge K+ already accu-mulated. Uptake of K+ is electrically compen-sated by ejection of protons or by concurrentaccumulation of anions (for summaries of theextensive studies see 69, 70, 125, 274, 307, 308).And the quest for the mechanism of energyinput leads us back to Fig. 6.The simplest interpretation now available is

based on the chemiosmotic hypothesis (68, 69,152, 176, 260, 270, 272): respiration and ATPhydrolysis both generate an electrical poten-tial, interior negative. Accumulation of K+ inpresence of valinomycin is due to the well es-tablished capacity of the antibiotic to conductK+ across lipid membranes. Nigericin medi-ates K+ efflux by exchange for protons fromthe medium; proton conductors dissipate theelectrical potential by allowing protons to flowin and thus elicit the same result.

Interpretations of K+ accumulation whichrely upon chemical coupling have also been

TABLE 1. A potpourri of inhibitors, antibiotics, and reagents which affect membrane processes

Metabolic region ] Inhibitor Mode of action

Respiratory chain CyanideAzide

ATPase

IonophoresHI

K+

K+, Na+ HI

K+/H+

Na+/H+

Lipid-soluble ions

Rotenone; piericidin

Antimycin; HOQNO

Oligomycin, rutamycin

DCCD

Dio 9

Dinitrophenol CCCP,

FCCP, TCS

Valinomycin, monactin

Gramicidin

Nigericin

Monensin

DDA+, TPMP+

TPB-, PCB-

Inhibits cytochrome oxidaseInhibits cytochrome oxidase, often ATPase as well; conducts

protonsSpecific inhibitors of first coupling site, probably on oxygen

side of coenzyme QSpecific inhibitors of second coupling site, between cyto-chromes b and c

Typically inhibits mitochondrial, but not bacterial ATPase;site of action, the "oligomycin-sensitivity-conferring-protein"

Inhibits both mitochondrial and bacterial ATPases; reactscovalently with a protein component of the membrane

Inhibits mitochondrial and bacterial ATPases; apparentlybinds to the ATPase itself

Conduct H+ very specifically across artificial and biologicalmembranes; uncouple oxidative phosphorylation; H+movement is electrogenic

Conduct K+ very specifically across artificial and biologicalmembranes. K+ movement is electrogenic; do not alwaysuncouple oxidative phosphorylation

Relatively nonspecific for monovalent cations; uncouplesoxidative phosphorylation

Mediates electrically neutral exchange of K+ for H+; notusually an uncoupler

Mediates electrically neutral exchange of Na+ for H+; notusually an uncoupler

Lipid-soluble cations (Fig. 4); accumulated by, and uncou-ple, intact mitochondria

Lipid-soluble anions (Fig. 4); accumulated by, and uncouple,submitochondrial particles

187VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

formulated. Massari and Azzone (246, 247,322) propose a chemically driven proton pumpwhich, however, carries out an obligatorilyneutral exchange of H+ for another cation (Ca2+, perhaps). Valinomycin renders this carrieraccessible to K+ but would not be actingacross the membrane, and a membrane poten-tial is not invoked. The model derives from asomewhat earlier one, developed in detail byPressman (308). Pressman postulated a carrierwhich forms a positively charged complex withcations-analogous to the valinomycin-K+complex. This carrier, whose function may beto drive uptake of anions against the electro-chemical potential (i.e., an "anion pump") isthought to be driven by an X - I, energy-richintermediate. Ordinarily, K+ is denied accessto the pump which is buried in the membranelipid, but valinomycin lets K+ pass to the ac-tive site. Carriers of this type can in principleaccount for the whole range of cation andanion translocations, but the mode of energycoupling must be specified more precisely thanhas yet been done.Transport of phosphate and substrate

anions. The recognition of specific carriers foranions resulted initially from application ofosmotic swelling techniques. Mitochondriawhose respiration is blocked are osmoticallystable in 0.15 M KCl, because the membraneis impermeable to both ions, but they swell inammoiiium phosphate or ammonium acetate.From studies on the effect of ionophorousagents, it was concluded that the entry ofphosphate is an electrically neutral processwhich can be formulated either as Pi-/OH-antiport or else as Pi-/H+ symport (67, 69, 70,271). Existence of a porter specific for phos-phate and arsenate was confirmed by the dis-covery that translocation of these metabolitesis specifically inhibited by certain mercurials(114, 373, 374).Most of the experiments on phosphate up-

take by mitochondria were done in the pres-ence of inhibitors both of respiration and ofATP utilization. Despite the lack of any en-ergy source, such mitochondria accumulate Piagainst a substantial concentration gradient.The accumulation is strongly dependent onthe external pH (for example, [Pi]J/[Pi]0 is 30at pH 6 but only 10 at pH 8 (298). Accumula-tion is inhibited by proton conductors and bynigericin, but not by cation conductors. Theseand other results suggest that the accumula-tion of phosphate by nonenergized mitochon-dria depends entirely on the establishment of apH gradient across the membrane. Measure-ments of proton movements and of the intra-

mitochondrial pH have confirmed that almosttwo protons accompany each phosphate ion:uptake of phosphate results in alkalinization ofthe medium; efflux of phosphate results inalkalinization of the mitochondrial interior(Fig. 7a). Whether we regard the process as Pi-H+ symport or as Pi-/OH- antiport, the in-ternal phosphate level is a function of the pHgradient across the membrane, according tothe relationship:

[Ann- n

[Ann-](where n is the valence of the anion and ApH= pHi - pHO; see references; 162, 215, 248,298, 300).Uptake of phosphate, and indeed of anions

generally, is enhanced if the mitochondria arepermitted to respire, and especially so when acation is provided: Ca2+, say, or K+ togetherwith valinomycin. How are we to envisage thecoupling of respiration to transport ofphosphate? Harris and Pressman (143) sug-gested that cation uptake is the primary eventand the anions follow passively. This scheme(Fig. 6b) predicts the generation of a positivepotential and is at variance with most of thedata currently available. Alternatively, activetransport of phosphate has been attributed toprimary anion pumps (see, for example, 308);this hypothesis is contradicted by the apparentelectroneutrality of phosphate uptake. What iscurrently known suggests, instead, that respi-ration leads to expulsion of protons and en-hancement of the pH gradient. This is limited,however, by the development of a membranepotential: only when a suitable cation is avail-able can net proton extrusion take place. Inother words, availability of the cation convertsA 4 to ApH; phosphate uptake is enhanced,and the salt accumulates in the mitochondrialmatrix (Fig. 7b; see references 215, 248, 309).

H+ H+ H+

~~~~~~~~~-~~~~~~~~~~~~~~~pi =-

2H+K NK+ ~ 2H +

FIG. 7. Transport of phosphate by mitochondria,via Pi-/H+ symport in response to a pH gradient.Left, Nonrespiring mitochondria; right, stimulationof phosphate uptake by respiration, in presence ofa cation. Val, Valinomycin.

188 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



The osmotic swelling technique led to thefirst recognition of not only the phosphate car-rier but of transport systems for substrate an-ions as well (67, 70). These interact in a re-markable manner. Transport of malate re-quires Pi for "activation" and is now known tooccur by exchange of malate for Pi; this is anelectrically neutral process which requires nometabolic coupling, since it occurs in mito-chondria whose respiration is blocked. Conse-quently, malate distribution across the mem-brane reflects that of phosphate and, secondar-ily, the pH gradient. Uptake of citrate, in turn,was found to require "activation" by malate.In fact, citrate uptake occurs by exchange formalate; one proton accompanies citrate, sothat once again the overall process is electri-cally neutral. Like that of other anions, citratedistribution is ultimately linked to the pHgradient (162, 248, 296-300, 309).

Overall, then, the uptake of anions by mito-chondria-either resting or respiring-isthought to depend upon a cascade of sec-ondary transport catalysts, as shown in Fig. 8.The key element is the phosphate porter, sincephosphate uptake is directly coupled to protonmovements. Net, uptake of malate must becoupled to prior Pi uptake, which thus actscyclically as a catalyst; citrate uptake, by ex-change for malate, is even more remotelylinked to the prime mover of the cascade-thepH gradient across the membrane. The greatlyenhanced uptake of anions, together with cat-ions, is at least qualitatively explained by anyscheme which assigns the respiratory chain thefunction of expelling protons with the genera-tion of a ApH and of a membrane potential.

FIG. 8. Interlocking systems for anion transportin mitochondria. (1) Phosphate-proton symport; (2)malate-phosphate antiport; (3) citrate-malate anti-port.

Summary: Energy Transductions inMitochondria

A reviewer feels most comfortable whenseated firmly on the fence. But it is clearlynecessary to conclude this torrent of experi-ment and argument by stating briefly the up-shot of the dispute. The following is inevitablyone man's view.

Figure 1 represents schematically the con-sensus regarding the general pattern of mito-chondrial energy transductions. The respira-tory chain supports not only the synthesis ofATP but a number of other energy-linkedfunctions, including particularly the transportof ions and substrates. The crucial insight isthat these functions do not require prior ATPsynthesis but are linked more directly to thechain. Conversely, exogenous ATP or even anion gradient can sustain other energy transfor-mations. Thus mitochondria, unlike the cell asa whole, do not use ATP as the universal en-ergy currency: ATP is but one of severalmodes of energy storage, all of which derivefrom a common pool of energized intermedi-ates or state.The nature of the energized state continues

to be, as it has for a decade, the heart of thedispute, and the devil finds no lack of scrip-tures to quote for his own purposes. But thereis a clearly discernible trend, in that elementsthat were originally derived from Mitchell'schemiosmotic hypothesis have been assimi-lated into all the major attempts to under-stand the molecular basis of energy coupling.Consequently, it no longer seems so urgent oreven meaningful to disprove one or another ofthe original hypotheses in toto; rather, theneed is for altemative and testable interpreta-tions of the data now at hand.

(i) The mitochondrial membrane is inher-ently impermeable to H+ and OH- and, in-deed, to ions generally. Functionally special-ized mechanisms exist for the translocation ofselected metabolites, which operate in such amanner as to maintain the osmotic equilib-rium and, apparently, the gradients of pH andof electrical potential.

(ii) Oxidative phosphorylation has neverbeen observed in a preparation that did notcontain topologically closed vesicles.

(iii) The respiratory chain contains bothhydrogen and electron carriers which appear tobe organized both within and across the mem-brane. The sequence of carriers is still debata-ble, and at present only one loop (rather thanthree) can be strung across the membrane. Butit is difficult to see what function the arrange-ment serves other than to separate protons and

VOL. 36, 1972 189


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

electrons across the membrane.(iv) The balance of the evidence is that the

oxidation chain does bring about a separationof electrical charges. Mitochondria can gen-erate an electrical potential of the order of -250 mv, interior negative. In chloroplasts andsonic submitochondrial particles, the polarityis reversed and the interior is positive. What-ever may be the mechanism of charge separa-tion, the very existence of large membranepotentials must be of profound significance tomitochondrial function.

(v) Is the ATPase of mitochondrial and chlo-roplast membranes reversibly coupled toproton translocation? The data on ion trans-port indicate strongly that ATP hydrolysisdoes bring about the electrogenic extrusion ofprotons, and conversely that ATP synthesiscan be driven by an ion gradient. The mecha-nism of proton extrusion by the ATPase seemsan area of darkness. Many data are in prin-ciple consistent with Mitchell's proposal thatthe proton gradient (ApH and A x,l) reversiblydetermine the poise of the ATP/ADP couple.Other findings, however, suggest that chemicalprecursors to ATP do exist.

(vi) Mitochondria accumulate a variety ofmetabolites-cations, anions, substrates, nu-cleotides. It is clear that specific transport sys-tems mediate the translocations, but there isno reason to invoke "active transport." Trans-port of ions and solutes against substantialconcentration gradients can be explained quitesatisfactorily as secondary responses to thegradient of pH and of the electrical potential.Evidence for the existence of the requisitesymport and antiport systems is mountingrapidly; for the present at least, there is noclear evidence for primary energy-linked anionor cation pumps.

(vii) The recognition that certain pharmaco-logical agents exert their effects by serving asartificial ion carriers is one of the most impor-tant results of recent membrane research. Theionophores are models for the association ofnautral membrane carriers with their sub-strates and provide tools with which to explorethe intimate relationship of energy metabolismto ion movements. Uncoupling seems to result,in many cases, from the conduction of protonsor other ions. It is not possible to state conclu-sively whether uncoupling results from theconduction of protons across the membrane, orinto its hydrophobic region, but a fundamentalrelationship of uncoupling to ion movementsseems to me undeniable.

(viii) And so we come at last to consider themolecular nature of energy conservation. It is

clearly possible, as Skulachev (350) has re-cently done, to formulate reasonable modelsconsistent with at least most of the data interms of both principal hypotheses. In thechemiosmotic view, the oxidation chain andthe ATPase are each seen as entirely separate,primary translocation systems for protons,which are coupled via ApH and A 4t. Accordingto the chemical hypothesis, the oxidation chainand ATPase each generate an unspecifiedhigh-energy intermediate; this in turn is theenergy donor for a proton pump that producesApH and A 46. There is much to recommend aconstructive synthesis, which includes chem-ical intermediates on the direct path to ATP,and ApH and A/v as the energy donors fortransport. Perhaps a counter-culture (210) de-serves the last word:

"Myself when young did eagerly frequentDoctor and saint, and heard great argumentAbout it and about, but evermoreCame out by the same door where in I went."

ENERGY TRANSFORMATION INBACTERIAL MEMBRANES

The cytoplasmic membrane of bacteria per-forms diverse functions which, in eukaryoticcells, are assigned to specialized organelles.The structure which regulates the metabolictraffic between medium and cytoplasm servesboth as osmotic barrier and as osmotic link. Itbears the pigments and catalysts of the respi-ratory chain and provides the structural frame-work for oxidative phosphorylation. Flagellaeare attached to the membrane, which may alsoprovide the energy for motility. And finally, itis the locus of enzymes for the biosynthesis ofstructures external to the osmotic barrier andan anchor point for the genophore.

Bacteria are profoundly different from mito-chondria in structure as well as in function,and it is obvious that one extrapolates frommitochondria to bacteria at some hazard. Yetthe fundamental unity of biochemistry makesit exceedingly likely that the basic principlesof membrane function will be more or less uni-versal, just as the principles of genetics areuniversal. In the remainder of this essay weshall examine energy-linked processes in bac-terial membranes-oxidative and photosyn-thetic phosphorylation, motility, and activetransport-in the hope that insights derivedfrom the study of organelles may illuminatesome of the mysteries of membrane function inbacteria.

190 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



Structural BasisThe envelope of bacterial cells consists of a

succession of layers, one within the other. Thatof gram-negative organisms is particularlycomplex, including two lipoprotein membranesand the periplasmic space between them. Theouter, lipopolysaccharide membrane, is usuallybelieved not to constitute a permeability bar-rier to small molecules. This may be oversim-plified, but it is nonetheless true that in allbacteria it is the cytoplasmic membraneproper which carries out the interconversion ofchemical, mechanical, and osmotic forms ofenergy. It appears in most electron micro-graphs as a plain "unit membrane," whoseunremarkable appearance gives no hint of itstrue complexity.

Autotrophic bacteria characteristically pos-

sess elaborate intracellular membranes, butthe more familiar heterotrophs do not. Here,the only morphologically differentiated regionsthat are regularly encountered are the vesic-ular bodies called mesosomes. These are now

known to occur in both gram-positive andgram-negative bacteria (reviewed in 318, 327),but after a decade of research their functionremains unclear. At one time it was quitewidely held that mesosomes are sites of oxida-tive metabolism, but this supposition can now

be discounted. Recent studies by Frehel et al.(118) demonstrate that deposits of tellurite,which are reliable cytological indicators of res-

piratory activity, are scattered over the plasmamembrane but never occur on the mesosomes.

Moreover, isolated mesosomes are deficient inNADH oxidase, succinic dehydrogenase, AT-Pase, and cytochromes (94, 107, 312, 365, 366).The evidence that mesosomes are involved incell division, circumstantial as it is, carriesmore conviction. Attachment of deoxyribonu-cleic acid to the membrane is often, but notalways, mediated by a mesosome. Moreover,the frequent appearance of mesosomes at thesite of septum formation points to a role inthis process, and perhaps in the synthesis ofcell wall or of membrane. The precise natureof this role remains uncertain. The membranefraction is known to contain enzymes for thebiosynthesis of cell wall peptidoglycan, but Iam not aware of any evidence as to their pres-ence in mesosomes. The C 55-polyisoprenoidalcohol which serves as a glycosyl carrier inwall biosynthesis is about equally distributedbetween the mesosome and plasma membranefractions of Lactobacilli (366). It also now ap-pears (94), contrary to earlier suggestions, thatmesosomes are not sites of preferential syn-thesis of phospholipids. An attractive specula-

tion is that mesosomes, like the Golgi appa-ratus of mammalian cells, are involved in theexport of macromolecules or building blocksacross the membrane. Lampen and his asso-ciates (123) have described the association ofan apparatus of tubules and vesicles with thesecretion of penicillinase by Bacillus licheni-formis.Another common morphological feature is

the occurrence of stalked particles on the innersurface of the plasma membrane of both gram-positive and gram-negative bacteria (1, 49, 122,278, 279). These structures contain the Ca2+-activated ATPase which is part of the appa-ratus of oxidative phosphorylation and are pre-sumably homologous with the stalked particlesso characteristic of the inner mitochondrialmembrane. They are scattered, apparently atrandom, over the entire membrane surface. Itis curious that the ATPase of the homofermen-tative organism Streptococcus faecalis, whichdoes not carry out oxidative phosphorylation,appears to lack a visible stalk (A. Abrams, per-sonnl communication).

Further insight into the organization of thecytoplasmic membrane has been obtained bymeans of chemical and physical methods. Itnow appears that the prolonged conflict overthe essential nature of membrane structure(see Hendler, 153, for an excellent summary) isbegining to subside. Davson and Danielli,some 40 years ago, proposed that large regionsof biological membranes are organized into abilayer of phospholipid molecules: The polarheadgroups face the medium and the fatty acylside chains make up a liquid, hydrophobicphase in the interior of the membrane whichconstitutes the permeability barrier to ionsand polar molecules. Despite widespread dis-sent just a few years ago, it now appears thatthis classical model provides an essentiallycorrect description of the membrane as a bar-rier. Much of the evidence derives from theexamination of Mycoplasma membranes by avariety of physical techniques (57, 95, 97, 357,368, 391). The properties of ion-conductingantibiotics are consistent with this model (277)as are various experiments which indicate thatthe lipid elements of the membrane exist in asemiliquid state (291, 369).On the other hand, students of membrane

functions have long insisted on the essentialrole of proteins in the organization of the cyto-plasmic membrane, a role not adequately ac-knowledged by the classical "picket fence."Here again the study of Mycoplasma mem-branes has been fruitful. Chemical dissection(276) suggests that the bulk of the membrane

191VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

proteins are associated with the surface, butthere is a fraction which is imbedded in theinterior. Freeze-etching of Mycoplasma mem-branes (368) reveals particles which may beprotein molecules buried in, or even pluggedthrough, the membrane in specific regions.The physical properties of the bulk lipids sug-gest that regions in which the lipids associatewith the proteins are quite limited in extent.

In this connection, it is worth stressing thatanalysis of the proteins of bacterial mem-branes by gel electrophoresis reveals severaldozen components but no single, major con-stituent to correspond to the expectation of a"structural protein" (71, 254, 324, 330, 331, butalso 302). The evidence on which the conceptof a structural protein in mitochondrial mem-branes was originally based is now known to befaulty (reviews: 197, 310); proteins whose solefunction is structural may well exist in mem-branes, but the idea now has little experi-mental foundation.

There emerges the impression that the bac-terial plasma membrane is a mosaic, composedof small functional regions that are predomi-nantly protein embedded in a relatively homo-geneous bilayer matrix. Hydrophobic interac-tions between these constituents predominateover ionic ones and the phospholipid head-groups may be largely exposed at the surface.The lipid bilayer must remain intact if themembrane is to serve as a permeability bar-rier: this inference, whose chief support wasresearch on antibiotics and drugs which in-teract with membranes (review: 132), wasgiven a firm foundation by Kaback in hisstudies on the role of lipids in bacterial mem-brane vesicles (reviews: 186-188). Bulk lipids,such as phosphatidylethanolamine in Esche-richia coli, are essential components of thebarrier. Functional membrane proteins areassociated with the lipid phase in varying de-grees of intimacy. Some, including the stalkedATPase, can be dissociated from the mem-brane by gentle washing and then reattachedto restore a functional complex (4, 6, 26). Atthe other extreme, the M protein of the 3-ga-lactoside transport system can be extractedonly with difficulty and detergents, and prob-ably requires lipid for its activity. Indeed,many membrane proteins are not known torequire lipids-in some cases specific lipids.Examples include NADH dehydrogenase (280)and the phosphotransferase system which cata-lyzes the vectorial phosphorylation of glucose(228, 253, 321). The role of lipids in the func-tioning of some bacterial membrane enzymes

has been reviewed (323). Virtually nothingappears to be presently known regarding thespatial relationship of functional centers toeach other and to the respiratory chains.One of the most intriguing consequences of

the essentially liquid nature of membranes istheir ability to reseal after mechanical injury.This fact, well known to cell physiologistssince the thirties, was first applied to thestudy of erythrocytes which can heal after lysisto form vesicles having the orientation of theoriginal cell. By judicious control of condi-tions, one may produce at will both normaland inverted vesicles (355, 356). By the sametoken, as was discussed above, disruption ofmitochondria can lead to the production ofboth kinds of vesicles.Membrane vesicles from bacteria have

begun to attract attention only recently, underthe impact of the work of H. R. Kaback, butare certain to assume increasing prominence inthe near future. To facilitate discussion of theorigin and orientation of such particles, Fig. 9illustrates the various vesicles that can be ex-pected to arise by disruption of the cyto-plasmic membrane; the stalked particles thatcontain the ATPase (see below) mark the cyto-plasmic side. The nature of "fragments," i.e.,topologically open structures, and of "right-side-out" and "inside-out" vesicles requires noelaboration. "Patchwork" vesicles could ariseby the annealing of fragments differing in ori-entation, so that the marker faces out in somepatches, inward in others. Tsukagoshi and Fox(369) recently reported the formation of hybrid

Fragments Right-sideoutvesicles

Inverted Hybrid orvesicles patchwork

vesicles

FIG. 9. Formation and orientation of membranevesicles. The stalked adenosine triphosphatase parti-cles are used as markers for the cytoplasmic side ofthe membrane.

192 HAROLD

.DV,


http://mm

br.asm.org/

Dow

nloaded from



vesicles when a mixture of membranes of dif-ferent origins was subjected to sonic treat-ment, and some of these may have the "patch-work" structure. It is also conceivable that amarker protein could flip over during prepara-tion of the vesicles. The last possibility raises adisturbing question: Is it conceivable thatmembrane proteins (apart from circulatingcarriers) could shift reversibly in a given vesi-cle, so as to face first one side and then theother, without impairing the topological integ-rity of the vesicle? If this possibility is admit-ted, the concept of sidedness becomes indeter-minate, at least for that particular protein.

Oxidative PhosphorylationGeneral features of respiration and phos-

phorylation. In the bacterial respiratorychains, as in that of mitochondria, reducingequivalents originating in NADH, succinate, orother electron donors pass over a cascade ofcarriers to oxygen. Various aspects of bacterialrespiratory chains are covered in recent re-views (49, 122, 168, 353, 390), all of whichstress features in which the respiratory chainsof bacteria differ from each other and fromthose of mitochondria. The essential point isthat bacterial respiratory chains are more flex-ible: They oxidize a wider range of substratesthan mitochondria do, and in some cases atleast can employ sulfate or nitrate as terminalelectron acceptors in place of oxygen. Bacterialrespiratory chains often appear to be branched(390), consisting of several distinct chainswhich communicate at particular sites. Oftenone finds constituents unique to bacteria, asamong the cytochromes. Some bacteria con-tain ubiquinones, others menaquinones (re-lated to vitamin K), and some contain both.Important as such differences are, there is

no reason to doubt that the coupling of respi-ration to phosphorylation is in principle thesame in bacteria as in mitochondria (49, 122).The present discussion, restricted to selectedorganisms which will be considered repeatedlyin this article, is intended only to summarizethe main features of the architecture of therespiratory chain and its overall metabolic per-formance.

(a) Escherichia coli inevitably heads the list.This is one of the most versatile of bacteria,capable of growing anaerobically either by gly-colysis or by the use of nitrate as terminalelectron acceptor. The enzymatic machinery ofoxidative phosphorylation is produced onlyunder aerobic conditions. In addition, it may

not always be appreciated that expression ofthe capacity for oxidative phosphorylation isrepressed by glucose (63, 149, 328).

Oxidative phosphorylation supplies at leasta large fraction of the ATP of aerobic cells ofE. coli and related organisms, since the leveldeclines if the cells are made anaerobic or ex-posed to uncouplers. Studies with intact cellssuggest the generation of 3 moles of adenosinetriphosphate for each NADH oxidized, for atotal of three coupling sites in this pathway(75, 145, 148).

E. coli contains both a benzoquinone (coen-zyme Q) and a naphthoquinone (vitamin K2).The isolation of mutants blocked in the syn-thesis of one or the other of these has contrib-uted greatly to the analysis of the role of qui-nones. It seems clear that coenzyme Q is thequinone chiefly involved in the respiratorychain-at least, in the oxidation of NADH,malate, and lactate. Oxidation of these sub-strates was impaired in a ubiquinone-deficientmutant, and glucose was converted to lacticacid (79). Studies with these mutants and withinhibitors of the respiratory chain (79, 354)suggest the chain shown in Fig. 10, with coen-zyme Q playing a vital but possibly indirectrole in electron transport. The role of mena-quinone appears to be much more restricted: itserves as a hydrogen carrier in a particularstep in uracil biosynthesis (283). (It shouldperhaps be mentioned that in Proteus, Kr6geret al. [227] place coenzyme Q in the directpathway of electron transport but suggest thatmenaquinone serves as an alternative carrier ofreducing equivalents in the oxidation ofNADH). There is no information at present asto the orientation of the respiratory chainwithin and across the membrane, but it is cu-rious that Hadjipetrou et al. (129) find inhibi-tion of respiration by ferricyanide; this wouldnot be expected to pass across the membraneand may perhaps be functioning as an unnat-ural electron acceptor at some exterior site.

(b) Micrococcus denitrificans is anotheradaptable organism. It grows heterotrophicallywith oxygen as terminal acceptor; anaerobi-cally, nitrate or nitrite can substitute for oxy-gen, and it can even be grown autotrophicallyon hydrogen gas under appropriate conditions.The respiratory chain, so far as it is known(Fig. 10), is remarkably similar to that ofmammalian mitochondria except for branchingat the oxygen terminus (15, 169, 170, 217, 339).The response of the chain to inhibitors is alsosimilar to that of mitochondria and definesthree possible coupling sites. Studies with in-

193VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


194 HAROLD

I xSubstrates NADH = FpD FeS CoQ Cyt.b =Cyt.c =Cyt. c l Cyt.aa3-02

NADPH FpS

SuccinateMi tochondri a

Substrates _ NAUH _ - Fp - ~Fe,/J/t Fp

iiADPH ,Lactate

CoQ--Cyt.b1,-. Fe, CoQ!° 02

1cicci nate

Esc

Substrates - NADH FFp-= Cyt.bl CoQ = Cyt.c 2/1r iiD- - -% Cyt.aja3!o*'

NADPH FpS

Succi nateMic

Substratee DADHFP FeS=Vit.K9=Ctt1 cyt.baa3 loz* 2

NADPH

Malate Cyt.bs

Succinate_ FpS ---,Fe - FSSuccinatecFp5~? ~FeS Mycobacterium phlei

FIG. 10. Respiratory chains of mitochondria and selected bacteria. Fp, Flavoproteins; CoQ, coenzyme Q;FeS, nonheme iron. For references see text.

tact cells, however, suggest that there may beas many as four (336). One of these may corre-spond to the transhydrogenase site first de-scribed by Asano et al. (15).

There is a hint that at least part of the chainmay be oriented across the membrane, sinceeven intact spheroplasts can oxidize exogenousferrocytochrome c (339). It is curious that cyto-chrome c can be partly released from intactcells by washing with salt solutions (334), butit remains to be seen whether this reflects theposition of cytochrome c at the extemal sur-face of the membrane, as appears to be thecase in mitochondria.

(c) Mycobacterium phlei is a strict aerobe.It has been studied in detail by Brodie and hisassociates, who have undertaken the arduoustask of dissociating and reassembling the com-

ponents of the oxidative phosphorylation appa-ratus. A detailed review has been prepared byBrodie and Gutnick (49).

Present knowledge of the respiratory proc-ess, with its three distinct chains leading fromNADH, succinate, and malate, is summarizedin Fig. 10 (after references 13, 49; and A. F.Brodie, personal communication.) The generalsimilarity of this chain, with three couplingsites and transhydrogenase, to that of mito-chondria is evident. There is, however, onenotable exception: the natural quinone in M.phlei, as in gram-positive organisms generally,is a naphthoquinone related to vitamin K.

It is clear that vitamin K is intimately andspecifically involved in respiration coupled tophosphorylation, but its precise role is uncer-tain. It has been proposed that a phosphoryl-

BACTERIOL. REV.

cherichia coli

crococcus deni tri fi cans

--I -1


http://mm

br.asm.org/

Dow

nloaded from



ated derivative of vitamin K is formed duringelectron transport and donates its phosphorylgroup to ADP; this, if true, would imply amechanism radically different from the mito-chondrial one, since the energy-rich interme-diate or state in these organelles is almost cer-tainly not phosphorylated. Gutnick and Brodie(128) detected incorporation of tritium into thenaphthoquinone of the respiratory chain underconditions when oxidative phosphorylation istaking place; this incorporation was sensitiveto uncouplers. Subsequently, Watanabe andBrodie (382) isolated a phosphorylated deriva-tive of vitamin K when membrane fragmentswere incubated with excess vitamin K as elec-tron sink; whether this, or a related compound,does indeed serve as the natural phosphoryldonor for ATP synthesis remains to be demon-strated.

In this context it should be mentioned thatin Bacillus megaterium, in which a vitamin Kis again the natural quinone, Kroger andDadak (226) find no evidence for participationof the naphthoquinone in phosphorylation.They suggest that it functions, as does ubiqui-none in mitochondria, to collect reducingequivalents from various dehydrogenases fordelivery to the cytochromes.

(d) Streptococcus faecalis. Until recently itseemed safe to regard streptococci as purelyfermentative organisms. Lacking cytochromesand oxidative phosphorylation, the only knownpathways for the generation of metabolic en-ergy were substrate-level phosphorylationssuch as glycolysis, the phosphoclastic cleavageof pyruvate, and the catabolism of arginine.The flavoprotein-linked NADH oxidase wasgenerally regarded as not coupled to energyproduction.

This simple pattem must now be qualified.Growth yields of aerobic cultures first pointedto the occurrence of limited oxidative phos-phorylation. This has now been confirmed withmembrane preparations which do appear tocouple NADH oxidation to ATP synthesis,albeit with low efficiency (105, 252). Resultsobtained with fumarate as an artificial electronacceptor suggest a chain corresponding to thefirst coupling site of mitochondria (Fig. 10).Oddly, the system was inhibited by antimycin,which is believed to act at the level of cyto-chrome b, but was unaffected by an uncoupler.To complicate matters further, it now appearsthat at least one strain of S. faecalis does formcytochromes when grown on a medium supple-mented with hematin. Phosphorylating mem-brane particles were again obtained (52).Coupling factors: the role of ATPase.

During the past decade, phosphorylatingmembrane particles have been prepared froma variety of bacterial species. The low effi-ciency of oxidative phosphorylation by suchparticles, whose P/O ratios are generally belowone, was at first taken to suggest that oxidativephosphorylation in bacteria is inherently lessefficient than in mitochondria. But now it isgenerally recognized that intact bacterial cellshave P/O ratios quite comparable to those oforganelles, and that the low efficiency of par-ticles reflects structural damage and the lossof components. Soluble "coupling factors"were found to be required for oxidation, phos-phorylation, or both; P/O ratios greater thanone, or even two, have been obtained from sev-eral systems including preparations from M.phlei, M. denitrificans, Azotobacter vinelandii,and Alkaligenes faecalis (review: 49).The nature, functions, and interrelationships

of these coupling factors are far from clear. M.phlei particles, for example, require severalfactors-some soluble, some easily dissociatedfrom the particles by urea (35, 154). However,brief heating of the particles or exposure totrypsin relieved the need for soluble factorsand increased the P/O ratio (11, 33, 34). Someof the factors may play a structural role, orelse exert their effects in an indirect manner.Alkaligenes faecalis extracts require two cou-pling factors whose proper attachment de-pends upon Mg2+ and K+. One of these, oncethought to be a nucleic acid, now appears to bea protein (8-10).One factor which appears to be common to

all phosphorylating preparations is one thatexhibits latent ATPase activity. The hydro-lytic activity can be unmasked either bytrypsin treatment or by heat (10, 32, 33, 172,173). The resulting ATPase characteristicallyrequires either Ca2+ or Mg2+ and is cold-labile.Activation results in loss of coupling activity.The soluble factor consists of 9-nm globules(10), thought to correspond to the stalked par-ticles that are so often seen on the inner sur-face of membranes from aerobic bacteria (1,10, 49, 122, 278, 279). It seems almost self evi-dent that this enzyme, like the analogous pro-tein in mitochondria and chloroplasts, cata-lyzes the terminal step in oxidative phosphoryl-ation. Indeed, Bogin et al. (32) have shownthat coupling factors of mammalian and bac-terial origin can replace each other. Direct evi-dence for the participation of ATPase in oxi-dative phosphorylation comes from the recentisolation of a mutant of E. coli which is unableto couple phosphorylation to oxidation; mem-brane fragments and whole extracts of this

195VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

mutant had only traces of ATPase activity.The authors tentatively propose that the genein question is the structural gene for ATPase(54).Thus far, the ATPase activity of prepara-

tions which exhibit coupling factor activity hasalways been found to be latent. Presumably, aswith the enzymes from mitochondria and chlo-roplasts, the hydrolytic activity is an artefact:in the native enzyme, the structure of thecomplex may exclude water and favor thetransphosphorylation step. The nature of theunmasking is not known; in mitochondria,removal of an inhibitory protein is required,whereas a relatively small configurationalchange suffices to reveal latent ATPase inchloroplast coupling factor (100, 101).ATPase activity has been found to be asso-

ciated with membranes from many bacterialspecies, and the presence of so potent an en-zyme catalyzing a "forbidden" and potentiallyharmful reaction raises a host of questions.Many preparations exhibit considerable activ-ity without the need for any unmasking (forexample 6, 98, 99, 127a, 127b, 137, 254a, 363a,and others cited in these references). Is suchactivity necessarily an artefact reflecting theunphysiological environment of a catalystwhose real function is transphosphorylation? Isthere only a single species of ATPase, or arethere several? Is ATPase, latent or exposed,necessarily associated with oxidative phos-phorylation? In this context, it should be re-called that ATPase occurs in E. coli grown ei-ther aerobically or anaerobically (54, 98, 99)and in S. faecalis which does not ordinarilyderive metabolic energy from oxidative phos-phorylation (4-6, 137, 332, 333). In both thesecases, and in others as well, there appears tobe only a single major ATPase species asjudged by purification, inhibition studies, andgenetic findings. It will be of great interest todetermine whether, and how, the ATPase ac-tivity is regulated and whether consistent dif-ferences can be found between enzymes fromglycolyzing and respiring cells.

It seems likely that ATPase enzymes serveas links between cytoplasmic and membranefunctions in both directions: on the one hand,to couple the membrane-bound catalysts ofoxidative phosphorylation to the synthesis ofATP and, on the other, to enable cells to uti-lize ATP as an energy source for membranefunctions. The mutant of E. coli, recently iso-lated by Butlin et al. (54), which is almostdevoid of ATPase activity will surely prove tobe of great value in defining the precise role ofATPase. It is already clear that this mutant

can neither carry out oxidative phosphoryla-tion (54) nor use ATP to derive the energy-linked transhydrogenation reaction (78). It cangrow on glucose as sole source of energy, butnot on succinate or lactate. There is at thistime no evidence that bears on Mitchell's view(265) that the equilibrium of bacterial ATPaseis poised by the proton-motive force. Indeed,evidence for participation of this enzyme inproton extrusion is as yet limited (141, 337).An important clue to the function of bac-

terial ATPase enzymes is their striking resem-blance to ATPase enzymes of mitochondrialand chloroplast membranes (62, 137, 333). Notonly are they alike in general size and shape,but they share such attributes as lability whenstored in the cold and sensitivity to DCCD.They require Mg2+ (and/or Ca2+) but are notparticularly stimulated by Na+ and K+ (Table2). These enzymes are evidently quite unlikethe ouabain-sensitive, Na+-K+ transport AT-Pase so characteristic of mammalian plasmamembranes. It appears that the latter enzymeis not present in bacteria, even though a traceof activity has been reported (130).Nature of phosphorylating particles from

bacterial membranes. Particles arising bycomminution of the bacterial membrane arefinding extensive use in the characterization ofoxidative phosphorylation, but little attentionhas yet been paid to their topology. If the con-clusions drawn from the work with mitochon-dria are correct, particles which carry out oxi-dative phosphorylation must, ipso facto, beclosed vesicles. Moreover, they cannot begrossly leaky to ions, since it appears thatmassive ion movements result in uncoupling.Now mitochondria are known to have anATP/ADP porter, but bacteria presumably donot: what evidence is currently available sug-gests that bacterial membranes are generallyvirtually impermeable to ATP, ADP, NADH,NADPH, and other central metabolites. How,then, can one obtain particles which are topo-logically closed yet accept NADH and ADP?The conclusion seems inescapable that suchparticles are inside-out, so that the couplingdevice faces the medium (260).Of course, membrane preparations may well

contain several kinds of vesicles as well asopen fragments, and interpretation of theiractivities must take this heterogeneity intoaccount. For example, if we assume thatNADH dehydrogenase is localized at the innersurface, then the rate of NADH oxidation willreflect nonphosphorylative oxidation by openfragments and coupled oxidation by invertedvesicles, but may not include that fraction of

196 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



._..-o

co

VOL. 36, 1972

0C*

c~i

:

co

I..

tsC)CQ:

a

197


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

the enzyme activity sequestered in structuresthat are right-side out. Thus, topological fac-tors may be responsible, at least in part, forthe apparent existence of both phosphorylatingand nonphosphorylating pathways of oxidationand other puzzling observations.The above is, admittedly, an a priori argu-

ment. However, a body of circumstantial evi-dence bearing on the orientation of membranevesicles is scattered through the literature.Some of the arguments tum on the propertiesof ionophores and lipid-soluble ions, which aresummarized in Table 1 for quick reference.

(a) Micrococcus denitrificans. Scholes andSmith (338) described the preparation ofmembrane vesicles by osmotic lysis of sphero-plasts. Like the original spheroplasts, themembranes oxidized succinate and NADHslowly, but both activities were enhanced byrepeated freezing and thawing. Oxidation ofsuccinate by membranes was also stimulatedby the nonionic detergent Triton X-100,whereas the activity of sonically treated mem-branes was not. The authors proposed thattheir membrane preparation consisted largelyof intact vesicles of the same orientation as theintact cell (i.e., right-side-out).Using a slightly different procedure of lysing

spheroplasts, John and Whatley (183) andJohn and Hamilton (181, 182) prepared parti-cles which couple oxidation of NADH to phos-phorylation with high efficiency, and exhibitrespiratory control: ADP, and also the uncou-pler CCCP, stimulated respiration. Like sub-mitochondrial particles, the M. denitrificansparticles could also be uncoupled by a combi-nation of valinomycin, K+, and nigericin. Weshall return to the significance of these obser-vations with ionophores in the following sec-tion. For the present we shall only note theconclusion (181, 182) that the bacterial parti-cles are closed vesicles and inside-out.

(b) Mycobacterium phlei. As mentionedabove, detailed studies have been carried outon particulate membrane preparations of M.phlei which perform oxidative phosphorylationwith relatively high efficiency (P/O ratiosabout two for NADH and succinate). Severalresults suggest that the process depends uponintact vesicles, especially the observation thatrepeated freezing and thawing abolished oxida-tive phosphorylation, but brief treatment at 50C restored it (11, 34).Very recently it has become apparent that at

least two kinds of particles can be generatedfrom M. phlei (14). Protoplast ghosts containall the components necessary but carried outvery little oxidative phosphorylation; pro-

longed preincubation with substrate and ADPwas required, and the ATP formed was at leastpartly retained. Oxidative phosphorylation bythese ghosts, like that by intact mitochondria,was uncoupled by valinomycin plus K+ and bydibenzyldimethyl ammonium (DDA+). Thusthe protoplast ghosts appear to have the sameorientations as do the parent cells. Sonic treat-ment produced much more active preparationswhich appear, by all available criteria, to beinverted. NADH oxidation and P/O ratioswere greatly increased; sonic particles were notuncoupled by valinomycin plus K+, or byDDA+, but only by valinomycin plus nigericin,or by the lipid-soluble anion tetraphenylboron;stalked particles corresponding to the latentATPase were observed facing the medium. Allthis strongly suggests that oxidative phospho-rylation by membrane fragments requires in-verted vesicles. By contrast, accumulation ofproline at the expense of respiration (157) isapparently characteristic of vesicles that areright-side-out.

(c) Escherichia coli. Of all bacterial mem-brane vesicles, those prepared from E. coli byKaback are by far the best characterized. Sev-eral lines of evidence indicate that vesicles,prepared by osmotic lysis of spheroplasts ac-cording to a carefully specified procedure, aretopologically closed and have the same orien-tation as do the original cells. Among the cri-teria employed are negative staining in elec-tron micrographs, the reduction in internalspace as a function of the osmotic pressure,retention of colloidal gold, and others (185,188). Such vesicles accumulate sugars, aminoacids, and other substances by a process linkedto the oxidation chain, which will be consid-ered in detail below. For the present, let usnote that such vesicles do not appear to carryout oxidative phosphorylation (43, 188, 213).Whether this is due to the loss of essential fac-tors or is simply a consequence of their orien-tation remains to be determined.

Particulate preparations of E. coli have beendescribed which do carry out oxidative phos-phorylation, albeit with low efficiency (44; seereference 49 for earlier preparations). Partic-ular attention has been devoted to the energy-linked transhydrogenase activity of these par-ticles (44, 112, 363). These particles, preparedby sonic treatment, must be inverted, but di-rect evidence on their polarity has not beenreported.

(d) Streptococcus faecalis. Gentle lysis ofprotoplasts by dilution in a medium of lowosmotic strength yields vesicles which havelow ATPase activity and appear to be right-

198 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



side-out. Sonic treatment reveals the ATPase,apparently by fragmentation (Altendorf andHarold, unpublished data).

In summary, the principle that bacterialmembrane preparations can reseal to give vesi-cles that are either right-side-out or inverted isan increasingly plausible one. To a first ap-proximation, it appears that gentle lysis ofprotoplasts is likely to yield right-side-out ves-icles, sonic treatment inside-out vesicles.However, there are already enough exceptionsto testify that factors other than mechanicalones play a role in determining the orientationof the vesicles. Some of the factors that mayapply to erythrocytes have been considered bySteck and his associates (355, 356).Coupling of respiration to phosphoryla-

tion. The labors of many investigators of mito-chondria lead eventually to formulation of thescheme shown in Fig. 1. According to this, theredox reactions of the oxidation chain are cou-pled to ATP synthesis indirectly, via one ormore nonphosphorylated, energy-rich interme-diates or states. The nature of this entity is, aswe saw in the first section, controversial-butits existence is not. It is a priori likely that sofundamental a relationship will be much thesame in bacteria as in mitochondria, but directevidence bearing on the validity of Fig. 1 forbacterial membranes is as yet meager.The burden of the argument is, in effect, to

demonstrate the essential identity of events inbacterial and mitochondrial membranes. Thusfar, the following energy-linked reactions havebeen studied, in addition to ATP synthesis byoxidative phosphorylation and, of course,transport.

(i) Reversed electron transport. Reduction ofNAD by succinate in presence of ATP hasbeen described in several organisms and exam-ined in some detail in M. denitrificans (16)and in E. coli (362). As in mitochondria, thisuncoupler-sensitive process is thought to implyreversal of the first coupling site. Its physiolog-ical significance in heterotrophs is uncertain,but in photosynthetic bacteria and in chemo-autotrophs it is likely to be a major sourceof reduced pyridine nucleotides (378, 381).

(ii) Transhydrogenase. The strongest evi-dence for a nonphosphorylated energy-rich in-termediate comes from the energy-dependentreduction of NADP by NADH by membraneparticles (15, 44, 78, 112, 363). Particles fromE. coli fully adapted to oxidative phosphoryla-tion could couple transhydrogenation either tothe oxidation of succinate or to the hydrolysisof ATP. When ATP served as the energysource, transhydrogenation was inhibited by

DCCD and also by uncouplers; when succinatewas the energy donor, the process was sensi-tive to uncouplers but not to inhibitors actingat the level of ATPase. Particles from cellsthat had been grown anaerobically could useATP as an energy donor, but not succinate(112). A mutant deficient in ATPase was un-able to couple transhydrogenase to the utiliza-tion of ATP, though activity with succinatewas normal (78). These results point squarelyto the existence of two distinct pathways,linked via a common, "energized" interme-diate as required by Fig. 1.As to the actual mechanism of coupling, i.e.,

the nature of the energized state or interme-diate designated "-," there is as yet little in-formation from bacterial systems. However, arecent series of studies with M. denitrificans(335, 336) suggests that, as in mitochondria,oxidative phosphorylation requires generationof a proton-motive force across the membrane.Intact cells are virtually impermeable to H+,but the proton conductance was greatly in-creased by FCCP (the fluorinated analogue ofCCCP [Fig. 5]); this also stimulated respira-tion. Respiration was associated with ejectionof protons from the cells. Maximal proton ejec-tion was observed in the presence of mobileions-either SCN- or K+ together with valino-mycin. Under these conditions, eight protonswere ejected per oxygen atom, and the rate ofrespiration (and proton ejection) was compar-able to that seen in presence of FCCP. Theuncoupler accelerated the return of protonsinto the cells, but did not affect the H+/Oratio. From these and other results, Scholesand Mitchell (335, 336) concluded that respira-tion of M. denitrificans is associated withtranslocation of protons outward, generating apH gradient and membrane potential. To ac-count for the results it is necessary to postu-late an oxidation chain of four loops; one ofthese may be the transhydrogenase. While nofinal claim was made, the results are compat-ible with a chemiosmotic mechanism of energycoupling.

Vesicles of M. denitrificans, which arethought to have an orientation opposite to thatof the intact cells, were examined by John andHamilton (181, 182). The particles had a P/Oratio of 1.3 and, unlike most bacterial prepara-tions described in the literature, exhibited res-piratory control. The crux of the argument isthat respiration was uncoupled by FCCP, byvalinomycin plus NH,+ and by the combina-tion of valinomycin K+, and nigericin. Theseparticular combinations characteristically un-couple submitochondrial particles (but not in-

199VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

tact mitochondria) and point to the impor-tance of an electrical potential, interior posi-tive, across these inverted membranes. (Letme reason out one of these, to clarify the basisof the conclusion. The respiratory chain inthese inverted vesicles would translocate pro-tons inward, generating a positive potential.Nigericin allows the H+ to exchange for K+,dissipating the ApH but not A A,. Valinomycinpermits the K+ to flow out, thus collapsing A4,as well. Thus, instead of a proton motive forcewhich could reverse the poise of the ATPase,there is a cyclic flux of K+ through the system)This conclusion is consistent with the obser-vation (64) that membrane particles from M.denitrificans, like submitochondrial particles,enhance the fluorescence of ANS- when al-lowed to respire.

Parallel observations are beginning to ap-pear from studies on M. phlei (14, 157) and onStaphylococcus (180, 286). Skulachev (350) re-fers to experiments which show that respiringvesicles of M. lysodeikticus accumulate lipid-soluble anions, indicating again the generationof a positive potential.Many questions must be raised concerning

the apparent relationship between proton-mo-tive force, membrane potential, and oxidativephosphorylation. It is not certain at this stagewhether the proton-motive force is a prerequi-site for ATP synthesis or an alternative formof energy utilization. It is not even establishedthat ATP as well as all coupling sites give riseto a common form of energy storage, nor thatall of the energy-linked functions of bacterialmembranes are reversibly interconnected asthe scheme of Fig. 1 suggests. The role andmechanism of action of the ATPase hasscarcely been explored. It would not be sur-prising if the organization of the processeswhich conserve and convert energy in bacterialmembranes were found to differ in some re-spects from that of mitochondria. But it isunmistakeably clear that microbiologists canno longer ignore the controversies that haveagitated their mitochondrial colleagues for thepast decade.

Photosynthetic PhosphorylationA small number of bacterial genera share

with the higher plants the capacity to utilizelight energy for growth. Perhaps for this rea-son, the investigation of bacterial photosyn-thesis has taken place largely within theframework of photosynthesis research ratherthan that of microbial physiology. To do jus-tice to this sophisticated topic within the con-fines of the present general survey seems to me

impossible; this section serves only to relatebacterial photosynthesis to other energy con-versions that occur in bacterial membranes.This cavalier treatment is excused by the ex-istence of several current and authoritativereviews (23, 81, 119, 150, 294. 342, 378, 380).

In intact bacterial cells, the photosyntheticapparatus appears as a system of closed sacs,vesicles, or lamellae. Whether the maturestructures are continuous with the plasmamembrane seems to be still uncertain. In anyevent, the chromatophores which can be iso-lated after comminution of the native appa-ratus are clearly a collection of closed vesicles.They contain bacterial chlorophyll and a va-riety of carriers for both hydrogen and elec-trons, and serve the same metabolic functionsas do the chloroplasts in plants: the generationof ATP and of reducing power for biosynthesis.The overall process of bacterial photosyn-

thesis as exemplified by Rhodospirillumrubrum is shown schematically in Fig. 11. Theinitial event is the photochemical excitation ofparticular chlorophyll molecules localized in alight-harvesting reaction center. In conse-quence, an electron is transferred to an uni-dentified primary acceptor, and the chloro-phyll accepts an electron from cytochrome c2.

Light

/ x

Cyt C2

Cyt b CoQcO

AnionTransport

Trans-hydrogenase I,

Ppi

ATP

ReverseElectronTransport

FIG. 11. Photosynthesis: electron transport andenergy transformations in Rhodospirillum rubrum.After Vernon (378) and Frenkel (190).

200 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



Between the unidentified acceptor and thecytochrome extends a chain of redox carrierswhose nature, number, and sequence are de-batable. This cyclic oxidation chain is coupledto phosphorylation at several sites-mostprobably two; the end product may be ATP orinorganic pyrophosphate.

Alternatively, light energy can be utilizedfor the generation of reducing power. Unlikegreen plants, which possess an additional pho-tochemical system, the bacteria are unable touse water as a hydrogen donor but require anexogenous reducing agent: H2, H2S, or an or-ganic reductant. Formation of NADH fromNAD and a reductant is envisaged to occur bya process analogous to the reversal of the mito-chondrial oxidation chain discussed above. Itshould be noted that either light or ATP canserve as the energy donor (174, 204, 216).

Apart from the unique photochemica'levents, the general arrangement of energytransductions appears to be identical with thatof oxidative phosphorylation (Fig. 1). The pas-sage of electrons over a cascade of redox car-riers is believed to generate an energized stateor intermediate, which in turn drives the en-ergy-linked functions: reduction of NAD,transhydrogenase (205), ANS- fluorescence(18), and the synthesis of ATP or pyrophos-phate. All these processes are inhibited by un-couplers which do not, however, interfere withelectron transport per se. The coupling to ATPsynthesis is mediated by a coupling devicewhich includes an ATPase sensitive to botholigomycin and Dio 9 and which requires phos-pholipids for activity. The ATPase is not, how-ever, involved in electron transport per se (21,40, 113, 214, 251, 337).A close relationship between photosynthetic

phosphorylation and the translocation of pro-tons was experimentally demonstrated by thediscovery (23, 379) that isolated chromato-phores accumulate protons from the mediumupon illumination. Addition of ATP likewiseresults in uptake of protons (260, 337). Con-versely, illumination of intact cells of R.rubrum results in extrusion of protons andacidification of the medium (91, 92, 337). Ittherefore appears that chromatophores are ves-icles that have inverted in the course of prepa-ration, so that their polarity is opposite to thatof intact cells.What is the nature of proton accumulation

by chromatophores? There is now much evi-dence to support the thesis that illuminationinduces rapid, electrogenic H+ influx, gener-ating across the membrane both a ApH (inte-rior acid) and a substantial membrane poten-

tial (+250 mv or more, interior positive). Thiswas initially inferred primarily from the effectsof ionophores on proton movements and pho-tophosphorylation (176, 202, 285). Recently,Isaev et al. (171) reported that illuminationsupports the accumulation of synthetic lipid-soluble anions, a clear indication of the devel-opment of a positive potential. In addition,Jackson and Crofts (175) showed that the shiftof carotenoid spectra upon illumination, whichhas been known for many years, is a conse-quence and a quantitative indicator of themembrane potential. Collapse of the gradientof pH and of electrical potential by certaincombinations of ionophores appears to accountfor the uncoupling of photophosphorylation(176, 273, 285).As is required by the scheme illustrated in

Fig. 1, ATP and even pyrophosphate serve asenergy donors to support membrane functionsin the dark: addition of ATP can elicit protonuptake, ANS- fluorescence, accumulation ofanions, transhydrogenase, and the shift in ca-rotenoid spectrum (18, 23, 171, 175, 205, 337).Inhibition of these effects by oligomycin impli-cates the ATPase of the chromatophore mem-brane in the utilization of ATP (like the mito-chondrial enzyme, chromatophore ATPase isinhibited by oligomycin, whereas other bac-terial ATP hydrolyzing enzymes are not). Inaddition, pyrophosphate can drive the genera-tion of ATP in the dark, again by an oligomy-cin-sensitive process (203).As with oxidative phosphorylation, the bone

of contention is the precise relationship be-tween proton movements and the primary en-ergy conservation step. According to the chem-iosmotic hypothesis, the primary event is thegeneration of an electrical potential upon illu-mination. If the acceptor of electrons fromchlorophyll and the electron donor are locatedon opposite sides of the chromatophore mem-brane, then illumination would not only reducethe one and oxidize the other but also producea separation of charges across the membrane.An appropriate arrangement of redox carrierswould transport H+ into the chromatophore,conserving the electrical energy and convertingpart of it into a pH gradient (81, 257). How-ever, it is not now possible to rigorously ex-clude the alternative view that proton uptakeand the membrane potential are due to a"proton pump," actuated by energy-rich inter-mediates or conformations of the membraneproduced during electron transport.

Coupling of Metabolism to TransportTransport systems and carriers. The basic

VOL. 36, 1972 201


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

principles of nutrient transport in microorga-nisms were set forth in a classic paper byCohen and Monod (74). Microbiologists eversince have held it to be virtually self-evidentthat the plasma membrane is largely imperme-able to the majority of polar molecules, in-cluding nutrients and metabolites. Uptake,when it occurs, is usually not a matter of"permeability" but of transport; that is, itdepends upon interaction of the substrate witha component of the membrane which recog-nizes the substrate with a high degree of speci-ficity, and translocates it across the membranein a catalytical manner. The neutral term"transport system" will be used here (in pref-erence to "permease") to designate in an oper-ational sense entities that are recognized onthe basis of genetic, kinetic, or other criteria tobe involved in the translocation of a substrateacross the membrane.

Kinetic and genetic analysis of many mi-crobial transport systems (reviews: 74, 186,206, 207, 242, 292) gave rise to the belief thatthe substrate combines reversibly with a "car-rier"; the resulting complex is often said to be"mobile in" or to "shuttle across" the osmoticbarrier. Molecules possessing the attributes ofcarriers in this sense are known: hemoglobin insaline solution, for instance, or valinomycin ina lipid phase. But microbial transport systemsappear to involve protein molecules firmlyembedded in the membrane, such as the Mprotein of the f3-galactoside transport systemof E. coli (206, 188). The Na+, K+, ATPase, ofmammalian cell membranes and mito-chondrial membrane porters (71) are likewiselarge protein molecules. Models based on theconcept of a carrier molecule freely diffusiblein a homogenous liquid are thus likely to beunrealistic. We must probably think in termsof conformational changes in a protein mole-cule, perhaps quite limited in magnitude,which modify the affinity of a binding site forits substrate and also alter the accessibility ofthe site from one side of the barrier to theother. (I have found the models in references178, 185, 186, 206, 242 particularly instructive.)Mitchell (258, 259) has stressed the analogy ofthe active site of an enzyme to the "carriercenter" of a porter and points out the impor-tant corollary that kinetic and energetic char-acteristics of the translocation in one directionneed not be identical with those in the oppo-site: carrier centers do not shuttle, they circu-late.At this point we must digress for a moment

to consider the relationship of these largelyhypothetical carriers to the concrete and ever

more numerous "binding proteins." (For re-cent reviews see references 186, 194, 242, 292,301.) Proteins which bind specific amino acids,sugars, vitamins, and ions with high affinityare easily released from gram-negative bacteriaby osmotic shock or by conversion to sphero-plasts. Genetic studies (12, 37, 146, 194, 236,290) provide persuasive evidence that thesebinding proteins are essential components oftransport systems, particularly of those withhigh affinity. Reports of changes in conforma-tion and in binding constants (36, 194, 292)lend weight to the hypothesis that binders rec-ognize the substrate at the external cell sur-face. There is, however, little to suggest thatthe binders serve to translocate substratesacross the plasma membrane: that may be theprovince of far more hydrophobic moleculessuch as the M protein, which remainembedded in the membrane. The persistenceof specific transport systems for a wide varietyof sugars and amino acids in membrane vesi-cles of E. coli (188, 208) and other organismsargues against obligatory involvement of theperiplasmic binding proteins in the transloca-tion step. Finally, dissociable binding proteinsare not found in gram-positive bacteria, whoseprotoplasts apparently retain normal transportcapacities. Perhaps the function of bindingproteins is connected with the complex, multi-layered envelope of gram-negative bacteria(but see reference 292).A striking feature of bacterial transport sys-

tems is their capacity to achieve large ap-parent concentration gradients; 1,000: 1 is notuncommon, and K+ can be transported by E.coli to a concentration gradient of 106:1. It isoften difficult to attach a precise meaning tothe concentration of a metabolite in the cyto-plasm, and small pools may well exist in abound state. But the large pools of sugars,amino acids, and other metabolites, which mayattain 0.1 to 0.2 M in the cytoplasm, simplyexceeded reasonable concentrations of anypossible macromolecular receptors. Moreover,many experiments suggest that such pools areosmotically active and hence in "free" solution(2, 3, 41, 45, 72, 96, 163, 185, 188, 219, 308, 347,401; but also 77 for an alternative view). Suchconcentrative uptake implies the performanceof work by the cells and interaction of the car-rier centers with the metabolic machinery.Group translocation. Group translocations

are processes in which passage of the substrateacross a membrane occurs concomitantly with,and as a consequence of, chemical transforma-tion of the substrate. It is thus a chemicalgroup, rather than an intact molecule, which

202 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



traverses the barrier (259, 321). Thirty yearsago it was thought that uptake of sugars by themammalian intestine may be linked to phos-phorylation, but after this was shown to be un-true the concept of group translocation fell intodisfavor. It is one of the lesser ironies of historythat today the transport system most fullyunderstood is precisely the vectorial phospho-rylation of sugars.

(i) Transport of sugars by vectorial phos-phorylation. In 1964, Kundig and Rosemandiscovered the existence of a novel system forthe phosphorylation of sugars in bacterial ex-tracts. The seqtience of reactions shown belowis now generally held to account for both thetransport of sugars and the initial step in sugardissimilation by many bacteria. The system iswidely, but not universally, distributed (319).

(1) PEP + HPr .-- HPr - P + pyruvate

(2) HPr - P + sugar sugar'- P + HPr

Sum: PEP + sugar M+

sugar - P + pyruvate

The extensive work on this system has beenrepeatedly and lucidly reviewed (185, 186, 242,321), so that a summary of the main pointswill suffice here. The first reaction is the phos-phorylation of a small protein, designatedHPr, by phosphoenolpyruvate (PEP); thephosphoryl group is linked to a histidine res-idue. This reaction is catalyzed by enzyme Iwhich4 like HPr, is constitutive and soluble.The second step, which is much more complex,consists of the transfer of the phosphoryl groupto any of a number of sugars. This reaction iscatalyzed by a family of membrane-boundenzymes, some constitutive and some induci-ble, having varying degrees of specificity.Among the sugars metabolized by the phos-photransferase system of E. coli are glucose(and its analogue, a-methylglucoside), man-nose, fructose, mannitol, and ,B-glucosides; aspecific enzyme II appears to exist for each.Recently Kundig and Roseman (228, 229) de-scribed the further resolution of enzyme II intoenzymes IIA (glucose, mannose, and fructoseeach required a particular IIA protein), andenzyme IIB (common to all three). In addition,the system requires phosphatidylglycerol foroptimal activity (228, 253). In Staphylococcusaureus, a sugar-specific soluble protein (factorIII) is needed in addition to enzymes I and II.

It was clear from the outset that the phos-photransferase system is required for the me-tabolism of many sugars since mutants defi-

cient in either enzyme I or HPr fail to grow onany of a large number of sugars; mutants thatlack a sugar-specific enzyme II fail to growonly on that particular sugar. But it was not atall self-evident that the phosphotransferasesystem is also required for translocation: itshould be recalled that glycerol, for example,passes across the membrane as such and issubsequently phosphorylated by a soluble ki-nase (see 242 for references). It is thereforenecessary to devise experiments to discrimi-nate between two alternative possibilities.

(i) The phosphotransferase system is re-quired only to phosphorylate sugars subse-quent to their translocation into the cytoplasmand serves to trap the sugar.

(ii) Phosphorylation occurs concomitantlywith, and is required for, the translocationprocess itself.The evidence, marshalled in detail in the

reviews cited above, overwhelmingly supportsthe thesis that transport of many sugars by E.coli and other bacteria occurs by "vectorialtransphosphorylation" (185). Among the prin-cipal findings with intact cells is that the sugarfirst appears in the cytoplasm as a phosphoryl-ated derivative, and the shorter the durationof uptake the greater is the proportion of phos-phate ester. In Staphylococcus aureus, phos-phorylation can be taken to be essential for theuptake of lactose because the f3-galactosidasefound in the cytoplasm attacks only lactosephosphate, not the free disaccharide. In mu-tants unable to phosphorylate the sugar, equi-librium is attained very slowly. Similarly, inmutants of Salmonella deficient in enzyme I,the maximal rate of a-methylglucoside uptakeis at least 50-fold less than in the wild type;qualitatively similar findings with other sugarssupport the contention that enzyme I is re-quired at least for transport at the normal rate(slow translocation of the sugars by diffusionor by other transport systems does occur). Theargument is greatly strengthened by studies onmembrane vesicles (see 185-187) which lackcytoplasmic enzymes; under appropriate con-ditions, all the glucose of a-methylglucosideaccumulated by vesicles is found as the phos-phorylated derivative in the lumen. Vesiclesprepared from mutant cells deficient in en-zyme I neither phosphorylated nor transportedit. In an ingenious double-label experiment,Kaback (184) found that intemal glucose-6-phosphate was almost entirely derived from3H-glucose added externally; "C-glucose al-ready present in the vesicles was not phos-phorylated. If all this falls short of absoluteproof, it does establish beyond reasonable

203VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

doubt that translocation of glucose and severalother sugars occurs by vectorial phosphoryla-tion.The focus of interest is now shifting to the

molecular events, particularly the precise rela-tionship of phosphorylation to translocation.Several recent findings pertain to this. InStaphylococcus, HPr - P phosphorylatesfactor III, and it is this latter protein which isthe immediate phosphoryl donor to the sugar(282). In E. coli, however, Rose and Fox (320)have obtained preliminary evidence that en-zyme II is phosphorylated and in turn donatesthe phosphoryl group to the sugar. The impli-cation is that the substrate-specific enzymes ofthe kind designated HA both phosphorylateand translocate the substrate. The balance ofavailable evidence suggests that translocationcan occur only in conjunction with phosphoryl-ation (184, 326), but it is perhaps not entirelyexcluded that enzyme II, in the absence ofphosphorylation, can mediate slow equilibra-tion or at least exchange of their substratesacross the membrane. Another question thatremains unsettled is the possible existence ofan additional sugar-specific element, distinctfrom enzyme II, whose role is to facilitate ac-cess of the substrate to the site of phosphoryla-tion. Thus, enzyme II would phosphorylate itsparticular sugar but not translocate it. Thispossibility has been suggested repeatedly,most explicitly so by Gachelin (120), whoseargument again turns on the ability of mutantsdefective in enzymes I or H to carry out atleast facilitated diffusion across the mem-brane. There is at this time no genetic evi-dence for the existence of a sugar-specifictranslocation.Since the phosphotransferase system is

unquestionably the main vehicle for transportand dissimilation of some sugars, could it be auniversal mechanism? Apparently the answeris no. In E. coli, the ,W-galactoside transportsystem can be differentiated from the phos-photransferase complex by two experimentalobservations. Accumulation of ,B-galactosidesis blocked by uncouplers, whereas that of a-methylglucoside is enhanced-perhaps becausethe uncouplers stimulate production of phos-phoenolpyruvate (reviewed in 186, 242). More-over, as Barnes and Kaback (24) demon-strated, uptake of ,B-galactosides by membranevesicles of E. coli can be coupled directly torespiration, under conditions in which partici-pation of phosphoenolpyruvate can be virtuallyexcluded (e.g., in a mutant deficient in en-zyme I of the phosphotransferase). The sugges-tion that the phosphotransferase may be the

catalyst of galactoside transport has recentlybeen revived by Koch (221), but for the presentthe weight of the evidence favors transport bytwo fundamentally distinct mechanisms.

(ii) Possible uptake of purines, pyrimidines,and fatty acids by group translocation. Theprinciple of group translocation is compatiblewith a wide range of metabolic transforma-tions, of which phosphorylation is but the mostfamiliar one. Whenever a substrate appears onthe inside rapidly and predominantly in theform of a derivative, group translocation maybe suspected. There is probably no need tocaution against facile application of what isonly a rule of thumb.

For some time evidence has been accumu-lating that the uptake of purines and pyrimi-dines by bacteria is intimately related to theirutilization. Especially when supplied at lowextracellular concentrations, these are incorpo-rated into nucleic acids without any lag andremoved from the medium at a rate controlledby the rate of nucleic acid synthesis (53, 284).In Bacillus subtilis, Berlin and Stadtman (28)found that 95% of the adenine taken up byresting cells was present as adenosine mono-phosphate (AMP). The specific activity of theAMP pool was higher than that of internaladenine itself. Such findings may indicatecompartmentation but are also compatiblewith the proposal that uptake of adenine oc-curs concurrently with its conversion to AMP.Thymine uptake by E. coli likewise appears tobe closely related to its conversion to deoxy-thymidine phosphate (195, 196). It is note-worthy that many enzymes catalyzing metabo-lism of nucleosides and bases are readily re-leased by osmotic shock and may be localizedin the periplasmic space.A thorough investigation of purine uptake by

E. coli has now been reported by Hochstadt-Ozer and Stadtman (159-161). Again, virtuallyall the adenine taken up was transferred toAMP and other nucleosides, apparently by theaction of purine phosphoribosyltransferase(adenine + PRPP = AMP + pyrophosphate).The initial rates of adenine uptake wereclosely correlated with the amount of this en-zyme present, which appears to be localizedeither in the membrane or else in the peri-plasmic space. In cells depleted of energy re-serves, exogenous phosphoribosyl pyrophos-phate (PRPP) greatly stimulated the uptake ofadenine; since the kinetic parameters weresimilar to those found for the isolated enzyme,the conversion of adenine to AMP appears toinvolve a site accessible to both adenine andPRPP from outside the osmotic barrier (161).

204 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



Even though most of the phosphoribosyl-transferase was lost from the cells upon os-motic shock, a fraction was retained whenmembrane vesicles were made. Uptake of ade-nine by the vesicles was again greatly stimu.lated by PRPP and resulted in conversion ofthe adenine to AMP which appeared inside thevesicles (160). It is thus likely that uptake ofadenine occurs by agency of the transphos-phorylase and represents a novel kind of grouptranslocation. However, the apparent involve-ment of PRPP from the external, rather thanthe internal, surface of the membrane is a puz-zling aspect of the proposed mechanism.Uptake of fatty acids by E. coli is closely

coupled to their degradation: free fatty acidsdo not appear in the cytoplasm, and mutantsgenetically defective in fatty acid catabolismhave correspondingly reduced rates of uptake.In particular, mutants lacking acyl coenzymeA synthetase do not take up fatty acids. Kleinet al. (212) therefore suggest that fatty aciduptake represents yet another case of grouptranslocation, by "vectorial acylation."Whether there exists another element whichfacilitates access of the fatty acid substrate tothe acyl synthetase is uncertain. In thisconnection it may be mentioned that thiamineuptake by E. coli is closely coupled to its phos-phorylation by an ATP-dependent kinase asso-ciated with the membrane. Here, however,studies with mutants suggest that it is thia-mine itself which may be translocated; thekinase is not thought to be required for trans-location, only for phosphorylation (199, 200).The distinction between group translocationand translocation followed by chemical trans-formation is not always easily made.

(iii) Vectorial biosynthesis of cell envelopepolymers. Peptidoglycan, lipopolysaccharides,and teichoic acids of the cell envelope are lo-calized outside the osmotic barrier. Since pre-cursors for their biosynthesis are produced inthe cytoplasm, some mechanism must exist forthe transport of building blocks from the cyto-plasm to the site of their assembly into poly-mers. The discovery that a C55 polyisoprenoidalcohol serves as glycosyl carrier in peptido-glycan synthesis proved to be of seminal im-portance: Not only are such compounds in-volved in the biosynthesis of envelope poly-mers generally, but they remain the only ex-amples of natural transport carriers of chemi-cally defined structure. The chemical struc-tures of the polyisoprenoid alcohols and theirrole in the biosynthesis of peptidoglycan, tei-choic acids, capsular polysaccharides, lipopoly-saccharides, and other polymers have been

comprehensively reviewed (93, 323) and willnot be further considered here. The biosyn-thesis of the cytoplasmic membrane itself alsofalls beyond the scope of this article.

Kinetic approach to energy coupling. Brit-ten's warning (45) that "the transport of smallmolecules into and within the bacterial celldoes not now appear to be an elementary proc-ess" is still timely. What insight has beenachieved stems primarily from the exhaustiveanalysis of the ,B-galactoside "permease" of E.coli and inevitably hypotheses formulatedfrom the study of this system have come todominate our thinking on microbial transportin general. It is neither possible nor necessaryto reiterate here material thoroughly coveredin the reviews of Kennedy (206), Lin (242), andKepes (207), except to set forth three conclu-sions which bear directly upon the coupling oftransport to energy-generating metabolism. (a)Exit of galactosides from E. coli is, like entry,a carrier-mediated process. Both geneticstudies and experiments with inhibitors sug-gest that the same carrier center mediatesboth entry and efflux (though it should be keptin mind that more than one pathway mayexist for each). (b) There is no evidence thatphosphorylation, acetylation, or any otherchemical modification of ,B-galactosides occursduring translocation. With the exception of arecent paper by Koch (221), current modelsagree that coupling of transport to metabolismoccurs at the level of the carrier rather thanthat of the substrate. (c) Recognition of the,B-galactosides and its translocation per se donot require metabolic energy; coupling to me-tabolism is necessary only for accumulationagainst a concentration gradient.The hypothesis that ,B-galactoside transport

occurs by facilitated diffusion with facultativeenergy coupling was first stated by Koch (220)and subsequently confirmed under a variety ofconditions (58, 198, 303, 399). It rests primarilyon experiments with metabolic inhibitors, es-pecially proton-conducting uncouplers (azide,dinitrophenol, CCCP). These stopped accumu-lation of thiomethylgalactoside (TMG) andother nonmetabolizable substrates of the trans-port system and caused rapid efflux of anyTMG previously accumulated by the cells.They did not, however, inhibit translocationper se: hydrolysis of o-nitrophenyl galactoside(ONPG) was only slightly inhibited and bothcounterflow and equilibration of the substrateacross the membrane were still observed.Winkler and Wilson (399) measured the ap-parent dissociation constants, Km, of the in-flux and efflux steps; in metabolizing cells, the

VOL. 36, 1972 205


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

Km for efflux was lower by a factor of 25 thanthat for influx, but in presence of an uncouplerthey were the same. It was proposed that en-ergy coupling, whatever its nature may be,occurs on the inner surface of the membraneand lowers the affinity of the carrier for itssubstrate which consequently accumulates inthe cytoplasm. Uncouplers interfere with en-ergy coupling and thus equalize the affinitiesat the two surfaces. Wong et al. (400) recentlyconfirmed this conclusion by detailed kineticanalysis of the effect of various inhibitors onthe uptake ofTMG and ONPG.Strong support for the view that energy cou-

pling converts a facilitated-diffusion systeminto one capable of active transport comesfrom the isolation of a mutant deficient in theenergy-coupling step (397, 402). This mutanthad largely lost the ability to accumulate f3-galactosides despite somewhat enhanced ca-pacity for carrier-mediated translocation. Themutation maps at the far end of the Y cistron(T. H. Wilson, personal communication).

Finally, Kashket and Wilson (198) showedthat Streptococcus lactis can carry out facili-tated diffusion of TMG even in the absence ofan exogenous energy source; addition of glu-cose supports concentrative transport of TMGand other galactosides.These observations, and the very large body

of kinetic measurements not mentioned here,are accounted for by models of the type illus-trated in Fig. 12 (206, 207, 399). This partic-ular model has been analyzed in detail byKepes (207). Briefly, the permease or carrier(identical with Kennedy's M protein) bothrecognizes and translocates the substrate.Pathways abcg and the reverse route cba re-quire no metabolic coupling and account forequilibration, counterflow, and exchange of

Medium

S .4-

Membrane

PS _ PS

a

b

P-9

..P

¢f edtP- A4 P--A

Cytoplasm

.- a S

-*o B

__.______--- ---S

FIG. 12. Kinetic model of f3-galactoside transport.After Kepes (207). P, "Permease" or carrier center;A - B, hypothetical energy donor; P - A, energizedfrom of permease; S, substrate; dashed line, variousunspecified exit processes.

galactosides in the presence of inhibitors.When cellular energy metabolism is intact, thecarrier can undergo a coupling reaction,pathway d, which lowers its affinity for galac-tosides. The original, high-affinity state of thecarrier is restored by the energy-dissipatingsequence ef.The view that energy is required only for

accumulation but not for translocation per seis not universally accepted. Manno andSchachter (244), for example, interpreted theirkinetic studies as indicating an energy-re-quiring entry process; the validity of theirmethods has, however, been questioned byKepes (207). The most serious challenge comesfrom Koch (221), whose earlier work has led tothe formulation of energy coupling as a facul-tative element of transport. Koch subjected E.coli to starvation in the presence of a-methyl-glucoside, in order to make the cells expendmetabolic energy on a useless transport cycleand thus deplete their reserves. Such "exer-cised" cells no longer hydrolyzed ONPG, eventhough they still possess both W-galactosidaseand the transport system. Addition of minuteamounts of glucose permitted rapid but lim-ited hydrolysis of ONPG, with an approximatestoichiometry of 25 to 30 moles of ONPG permole of glucose. Koch (221) concluded that in-flux of ONPG, even downhill, inherently re-quires expenditure of energy, and proposed amodel in which galactosides are phosphoryl-ated via the phosphotransferase system, fol-lowed by hydrolysis via a phosphatase. Pre-sumably, stoichiometric modification of thecarrier, rather than of the substrate, couldequally well account for his observations. I amnot persuaded that Koch's results require us todiscard the conventional view supported byevidence of many different kinds, that accu-mulation depends upon a manner of couplingto cellular metabolism which is not necessaryfor translocation as such. They may, however,suggest the need for at least minimal metabo-lism to keep the transport systems in a func-tional state.Even more debatable than the kinetic conse-

quences of energy coupling are the nature ofthe energy donor and of the energy-couplingreaction (d). Kepes favors the postulate that"energy coupling involves a covalent reactionwith the energy donor and permease protein(or a complex including permease protein). Itis believed that the result of the reaction is theestablishment of a covalent link between per-mease and one of the radicals of the energydonor according to the example of the phos-phorylation of the Na+-K+-dependent ATPase.

206 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



The derivative, symbolized P - A, will becalled the energized form of permease" (207).Uncouplers might either prevent formation ofthe donor A - B or act at one of the subse-quent steps.Could the energy donor, A - B, be ATP?

ATP does enhance translocation of B-galacto-sides under certain conditions (329, 387). How-ever, two lines of evidence suggest that therole of ATP is an indirect one: (i) E. coli accu-mulates TMG even under anaerobic condi-tions; accumulation is inhibited by proton-conducting uncouplers which do not signifi-cantly affect the ATP pool (303). (ii) Morecompelling is the finding that membrane vesi-cles of E. coli can couple respiration directly tothe accumulation of galactosides (24). Thusfar, the covalent energy donor A - B remainshypothetical and quite as elusive as the chem-ical intermediates of oxidative phosphoryla-tion.Kepes (207) leaves open the possibility that

A - B may, in fact, be a high-energy interme-diate of oxidative phosphorylation, but there isa plethora of attractive alternatives. Kaback,whose work will be discussed in the followingsection, suggested that the ,B-galactoside car-rier may itself be a member of a redox chain,and suffers cyclic changes of orientation andaffinity by oxidation and reduction of criticalsulfhydryl groups. But there is no compellingreason to assume any covalent modification ofthe carrier center at all. Lowered affinity ofthe carrier at the interior surface, or loweredmobility, could be due to an appropriate gra-dient of ion activity across the membrane (seebelow). Finally, Boyer and Klein (43) havedeveloped a model in which energy inputbrings about a conformational change in mem-brane structure, such as to block access of sub-strate from the inside to the carrier; thismodel does not invoke a lowered affinity of thecarrier for its substrate at all. Speculation isnot seriously hampered by facts, and transportworkers know for their sins that it is easier toconstruct models than to subject them to crit-ical test.How broadly applicable are generalizations

drawn from fB-galactoside transport-and espe-cially the hypothesis that accumulation againsta concentration gradient requires coupling tometabolism, but translocation per se does not?For many transport systems, little more isknown than that dinitrophenol and other in-hibitors block accumulation and, in somecases, discharge pools previously accumulated.Examples from E. coli include galactose (167),arabinose (289), sulfate (87), phosphate (249),

leucine (305), aromatic amino acids (50, 51),magnesium, and manganese (243, 345, 346).Citrate transport in B. subtilis (392), a-amino-isobutyrate in Bacillus megaterium (245), andproline uptake by Pseudomonas (201) are simi-larly inhibited. Rather more informative is thefinding that, in E. coli, exchange of aminoacids between the medium and the internalpool continues under conditions which preventnet accumulation of amino acids (46, 209).Similar observations were made in Pseudo-monas: an uncoupler, azide, inhibits accumu-lation of proline but not its entry into the cells(201). Recently, Hechtman and Scriver (147)described a mutant which can transport al-ala-nine across the membrane but does not accu-mulate it against a concentration gradient, andsuggested that the genetic lesion is in the en-ergy-coupling step. Fragmentary as it is, theevidence does suggest that in the case ofamino acids, translocation and accumulationhave different requirements for metabolic en-ergy.Streptococcus faecalis has many virtues,

among which the apparent absence of internalenergy reserves and of oxidative phosphoryla-tion are particularly pertinent to attempts toexplore the nature of energy coupling. Severaltransport systems in this organism appear tobe strictly dependent on an external energysource, including those for sucrose (3), glycyl-glycine (2, 3), phosphate (133), potassium (134),and glutamate (314). Some amino acids, how-ever, can enter the cells by exchange for othermembers of the cellular amino acid pool evenin the absence of glucose: glycine, threonine,alanine, serine, and their analogue, 2-amino-3-phosphonopropionic acid, fall into this cate-gory (47, 164, 275). Glucose stimulates accu-mulation of the latter amino acids; glucose-stimulated uptake, unlike the exchange, is in-hibited by proton-conducting uncouplers, yetboth appear to be mediated by the same trans-port system (136; Asghar and Harold, unpub-lished data). For these amino acids, then,translocation by exchange does not require anenergy source, but net accumulation beyondthis point does. The same conclusion wasreached by Kashket and Wilson (198) for TMGuptake by S. lactis.Coupling of transport to the respiratory

chain in membrane vesicles. A milestone inthe continuing analysis of bacterial membranefunction was passed with the description, byKaback and Stadtman (192), of a membranepreparation from E. coli which carried out ac-tive transport of proline. I propose to reviewthe results of this trial of research in some de-

VOL. 36, 1972 207


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

tail, both because of its importance as a tech-nique of broad applicability and because it hasled to iconoclastic views that deserve criticalscrutiny.

Vesicles are prepared by osmotic lysis ofspheroplasts under carefully specified condi-tions (184, 185, 187); they are usually closedsacs of relatively uniform size bounded by asingle unit membrane, but some strains yieldmulti-layered structures. Much effort was ex-pended on the difficult task of proving that thevesicles are topologically closed and possessthe same orientation as the parent cells, i.e.,right-side out (185, 188, 190). The vesicles arepractically devoid of cytoplasmic enzymes andcofactors. They also carry out no detectableATP synthesis by oxidative phosphorylation;whether this is due to loss of essential couplingfactors or simply reflects the orientation of theintact vesicles remains to be determined.However, when provided with a suitable en-ergy source, the vesicles accumulate a varietyof substrates against the concentration gra-dient and at a rate at least comparable to thatof the parent cells: ,B-galactosides (24, 25, 189),galactose, and other sugars (208), a variety ofamino acids (190, 191, 192), manganese (29),and, in presence of valinomycin only, K+ andRb+ (30, 188).Accumulation of these substrates requires

presence in the membrane of the appropriatetransport systems. Thus, vesicles derived frommutants deficient in transport of fB-galacto-sides, proline, or glycine share the deficiencyof the parent cells. However, vesicles appear tobe devoid of binding proteins, which are easilylost by osmotic shock. The binding proteinsthus appear to play no indispensable role intranslocation across the plasma membrane;however, the transport of galactose by vesiclesoccurred with relatively low affinity, sug-gesting that the binding proteins are involvedin conferring high affinity upon certain trans-port systems (208). Accumulation also dependsupon the integrity of the barrier functions ofthe membrane, and thus upon membranelipids: vesicles can be rendered leaky and re-sealed by various manipulations, includingchanges in temperature (185, 188, 189). Unlikethe uptake of glucose, which occurs by grouptranslocation, there is no evidence for anychemical transformation of the substratesduring concentrative uptake driven by respira-tion. Moreover, the accumulated substancesappeared to be in free solution in the lumen ofthe vesicles. Internal concentrations may be ofthe order of 15 mm, with concentration gra-dients of 100: 1 (189).

Most of the available information comesfrom vesicles of E. coli membranes. However,preparations capable of concentrative trans-port have also been obtained from S. aureus(344), B. subtilis (222, 223), Mycobacteriumphlei (157), and other organisms (222). Mem-brane vesicles of S. faecalis were unable tocouple respiration to active transport, in ac-cordance with the general belief that this orga-nism normally relies entirely upon substrate-level phosphorylations for metabolic energy.Accumulation of substrates is almost com-

pletely dependent upon respiration, and it isthe nature of the coupling between these proc-esses that has been the focus of recent papersby Kaback and his associates. (The publishedinformation refers predominantly to the up-take of ,B-galactosides, but qualitatively similarfindings were made with other metabolites[188, 208, 344]). The cornerstone of the argu-ment is the contention that ATP, derived fromoxidative phosphorylation, is not an interme-diate in the coupling of respiration to trans-port. The case rests partly upon the absence ofdetectable ATP from the vesicles, which alsodid not respond to ATP added externally.More convincing is the finding that even highconcentrations of arsenate, which is known touncouple oxidative phosphorylation in intactcells, did not inhibit respiration-linked trans-port by vesicles (24, 43, 157, 208, 213). More-over, DCCD, a standard inhibitor of ATPase,did not interfere with uptake of proline by ves-icles of M. phlei (157). The linkage betweenrespiration and transport thus appears to be afairly direct one.By no means do all substances oxidized by

the vesicle preparations also support trans-port. Vesicles of E. coli, for example, oxidizeNADH, D-lactate, and succinate, in that orderof preference. However, D-lactate was the mosteffective substrate for transport, followed bysuccinate, formate, a-glycerol phosphate, andL-lactate; NADH was a poor energy donor (25).Subsequently, the combination of ascorbateand phenazinemethosulfate was found to be anexcellent energy donor for f,-galactoside trans-port by vesicles of E. coli (and of amino acidsby vesicles from other sources [157, 222, 223]).In all cases, oxidation was mediated by anelectron transport chain. Since D-lactate, suc-cinate, and NADH all reduced the cytochromecomponents of the membrane, and to the sameextent, it appears that reducing equivalentsfrom all the substrates traverse the samecommon cytochrome chain (Fig. 10). D-Lactatewas the best energy donor for transport, andtherefore Barnes and Kaback (25) placed the

208 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



site of energy coupling of respiration to trans-port between D-lactate dehydrogenase and cy-tochrome b. Kerwar et al. (208) reached thesame conclusion with respect to galactose up-take.The localization of the coupling site was

supported by the application of inhibitors.Sulfhydryl reagents, especially p-chloromercu-ribenzoate and N-ethyl maleimide, inhibitedboth lactate oxidation and transport; both in-hibitors were reversed by dithiothreitol. D-Lactate dehydrogenase, when assayed sepa-rately, was completely indifferent to sulfhydrylinhibitors. Therefore the coupling site mustagain lie subsequent to the dehydrogenase, anda close relationship was inferred between oxi-dation and transport (189). By contrast, oxida-tion of NADH was not sensitive to sulfhydrylreagents and appears to involve a separatechain as far as cytochrome b.

Other inhibitors of respiration also blockboth oxidation of D-lactate and accumulationof ,B-galactosides. Of key importance to theargument is the observation that anaerobio-sis, KCN, and hydroxyquinoline N-oxide(HOQNO) stop accumulation and also induceefflux of galactosides previously accumulatedby the vesicles. By contrast, p-chloromercu-ribenzoate, and also oxamate (which inhibitsD-lactate dehydrogenase itself) were shown toblock accumulation without inducing efflux(189).These observations are the main pillars sup-

porting the tentative model put forth by Ka-back and Barnes (189). They suggest (Fig. 13)that a specialized electron transport chainconnects D-lactate with cytochrome b; NADHhas little or no access to this chain. The trans-port carrier for fB-galactosides, and indeed

every transport carrier, is either a member ofthis special oxidation chain or else closelylinked to its redox carriers. The transport car-rier is proposed to exist in two states, differingin orientation and in affinity for the substrate.The oxidized form, S-S, has high affinity for ,B-galactosides at the external surface. Upon re-duction of the disulfide bond, a reorientationoccurs; the substrate is translocated to theinner surface and concomitantly the affinity ofthe carrier center is lowered. Thus the sub-strate dissociates from the carrier and accumu-lates in the lumen of the vesicle. It is further-more suggested that the reduced, low-affinityform of the carrier can mediate efflux of thesubstrate across the membrane, as well as ex-change. The latter postulate is needed to ex-plain the remarkable observation that inhibi-tors which act beyond cytochrome b not onlyinhibit accumulation but also induce efflux ofany galactoside previously accumulated: Theelectron transport chain, including the carrier,would be expected to be reduced under theseconditions. Conversely, oxamate for example,inhibits accumulation but does not elicit ef-flux; according to the model, the carrier re-mains in the oxidized (high-affinity) state inthe presence of inhibitors that act before thecoupling site.

It seems to me quite solidly established thatthe vesicles couple respiratory energy to trans-port by a mechanism which does not involveATP. What is at issue is the precise nature ofthe coupling, and it is my opinion (pace Ka-back) that some of the data now availableadmit of alternative interpretations.The first doubts arise from topological con-

siderations. Right-side-out vesicles would notbe expected to oxidize NADH nor to hydrolyze

OUTPYR RED -.?r~RED\

D-LA XJioxXIN OXiDIZED

/ OW Km

OUT / OUTPYR RE E HIGH Km PY E E

Xfp to b ~ fp tobiD-LACO Xox oxDLC xo

IN REDUCED REDUCED

FIG. 13. Coupling of transport to D-lactic acid dehydrogenase, according to Kaback and Barnes (189). D-LAC, D-lactate; PYR, pyruvate; fp, flavoprotein; Cyto b, cytochrome b,; OX, oxidized; RED, reduced.OUT signifies the outside surface of the membrane; IN signifies the inside surface. The hemisphere locatedbetween fp and cyto b, represents the "carrier": Li, a high-affinity binding site; and W, a low-affinity bind-ing site. The remainder of cytochrome chain from cytochrome b, to oxygen has been omitted.

VOL. 36; 1972 209


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

ATP, since the requisite enzymes are almostcertainly localized at the inner surface. Thusthe reported synthesis of phospholipids fromATP (385), ATPase activity, and NADH oxi-dase may betray a fraction of vesicles whichare either open or, more likely, damaged insome more subtle way. It is at least arguablethat the failure of NADH or of ATP to supporttransport is due simply to their inability toreach their proper site of action at the innersurface of "competent" vesicles. If this be true,we would have to attribute the efficacy of D-lactate, succinate, and glycerol phosphate tothe retention by the vesicles of transport sys-tems mediating their entry. Furthermore, al-though D-lactate is the best energy source, sub-strates of other oxidation pathways do alsoserve: succinate, L-lactate, glycerolphosphate,and formate were half as effective as D-lactatein supporting the initial role of lactose uptakeand supported about one-third of,the maximalpool size (25). Kaback (personal communica-tion) suggests that these substrates are oxi-dized partly via the lactate dehydrogenasechain, partly by other routes-an ad hoc hy-pothesis which seems needlessly complex. Inci-dentally, the fact that many bacteria, in-cluding E. coli, can grow anaerobically at theexpense of glycolysis is an a priori argumentagainst redox intermediates as transport car-riers.The most glaring deficiency of the model of

Kaback and Barnes (189) is its failure to ac-count explicitly for the striking inhibition oftransport by uncouplers and by valinomycin inpresence of K+ (24, 25, 157, 208). The very factthat these compounds do not inhibit respira-tion, but only dissociate it from transport,implies that cyclic oxidation and reduction ofelectron carriers is not by itself sufficient todrive active transport.

It seems to me more plausible that transportis linked to respiration via energy-rich inter-mediates or an energized state, just as isthought to be the case for mitochondria (Fig.1). This energized state would be produced byany one of several branches of the respiratorychain, which vary in efflciency. A very closecorrelation between the redox state of thechain and concentrative transport would beexpected, but inhibition of both by sulfhydrylreagents would have to be explained in termsother than those suggested by Kaback. Thenature of the energized state and the mannerin which it drives concentrative transport is aseparate issue: Hypotheses can be framed interms of chemical, conformational, or chemios-motic coupling (e.g., 43, 189, 265, 361). But I

would stress that, as was earlier found to bethe case for mitochondria, it is necessary toaccount for the dissipation of the energizedstate by reagents which conduct protons orK+. A relationship between transport by cellsor vesicles and ion gradients has by no meansbeen "ruled out" (25, 188): on the contrary,there seems to be quite a lot to be said for it.Ion gradients and energy coupling. The

hypothesis that ion gradients are the link be-tween transport and metabolism rests uponthe presupposition that certain metabolicpathways catalyze reactions oriented acrossthe membrane, and thus generate primarygradients of metabolites. Because they bear anelectrical charge, gradients of ions are morelikely to serve in energy coupling than are gra-dients of neutral molecules. The primary iongradients support the transport of other nu-trients or metabolites against the electrochem-ical gradient by means of a hierarchy of sec-ondary carriers, which have affinity sites forboth the coupling ion and for particular nu-trients. In the case of symport, the substrateand the ion are translocated in the same direc-tion; the driving force on the coupling ion,which may have an electrical component aswell as a concentration component, is exertedalso on the substrate passenger and thus bringsabout "active" transport of the latter. Antiportdesignates the reverse situation, in which thecoupling ion moves in one direction, the pas-senger in the other, so that a previously estab-lished ion gradient can set up an opposed gra-dient of passenger. These mechanisms of "en-ergy coupling" are analogous to the more fa-miliar situation in which the product of oneenzymatic reaction is a substrate of a second:here it is the driving force of one translocationthat is balanced against a second translocation(256, 264, 265).The ions most likely to be involved in the

coupling of metabolism to transport in bac-teria are H+, Na+, and K+.

(i) Na+ and K+ as coupling ions. In mam-malian cells, extrusion of Na+ and accumula-tion of K+ by the Na+,K+-ATPase provides aprimary ion gradient. The evidence that trans-port of amino acids and sugars is coupled tothe Na+ gradient has been repeatedly reviewed(341, 358; but see 211 for an opposing view).Bacteria do not normally require Na+ in thegrowth medium, and a general role for Na+ incoupling transport to metabolism seems un-likely. Nonetheless, there are a number of re-ports of Na+-dependent transport processes:Na+ stimulates glutamate transport in E. coli(117) and is clearly involved in amino acid

210 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



transport by a halophilic Pseudomonas (364,401). Na+ specifically stimulates uptake ofTMG by "permease II" of Salmonella; con-versely, TMG stimulates 22Na uptake, sug-gesting symport of the cation and the sugar(359). These intriguing reports suggest thatNa+ may, after all, serve as coupling ion forparticular translocation processes, but at thistime there is no instance in which a Na+ gra-dient has been directly implicated as a sourceof energy for any bacterial transport process.

All bacteria accumulate K+, and it is apriori plausible that they might utilize some ofthe potential energy stored in this gradient todrive the accumulation of other metabolites.There is at present no evidence in support ofthis suggestion. Indeed, the finding that manybacteria grow, and presumably transport per-fectly well at external K+ levels close to theinternal one, argues against participation ofthe K+ gradient in concentrative transport. Onthe other hand, many investigators have notedan apparent K+ requirement for various trans-port processes (2, 3, 133, 135, 275, 295, 329,401). The function of K+ is not known. In onecursory study, cells of S. faecalis whose K+complement had been almost completely re-placed by Na+ were found to be still capable ofamino acid and phosphate transport (135), SOthat K+ is not obligatory in this system: K+may be required only because of its role in theextrusion of protons from the cells (140, 141).In yeast, however, evidence has been adducedby Eddy and his associates (88-90) that move-ments of K+ and of Na+ accompany the up-take of amino acids and that cation gradientssupply at least part of the driving force. Thepossible involvement of K+ and Na+ in en-ergy-linked transport by bacteria deservesmuch more rigorous investigation than it hasreceived so far.

(ii) Protons as coupling ions. The hypothesisthat the linkage between transport and metab-olism is effected by a gradient of proton ac-tivity is historically rooted in the familiar ob-servation that uncouplers of oxidative phos-phorylation block so very many transport proc-esses. It was the discovery that 2, 4-dinitro-phenol inhibits uptake of galactose and galac-tosides by E. coli and discharges preexistingpools that led Mitchell (256) to formulate thesetransport systems as H+-sugar symport, and toattribute the effect of uncouplers to theirability to facilitate diffusion of protons acrosslipid membranes.

Conventional wisdom, of course, offers quiteanother explanation for the inhibition of activetransport by uncouplers: Uncouplers interfere

with the generation of ATP by oxidative phos-phorylation and the inhibition of transport fol-lows secondarily from lack of ATP. Since thissyllogism is so often taken for granted, it isdesirable to restate explicitly the evidence thatrefutes it. (i) It has been found repeatedly (38,39, 225, 316, 398) that uncouplers inhibit trans-port even under anaerobic conditions. In S.faecalis and in anaerobically grown E. coli,both of which generate metabolic energy byglycolysis, a series of uncouplers was found toblock transport of K+, phosphate, sucrose, sev-eral amino acids, and TMG, even though un-couplers inhibited neither the generation northe utilization of glycolytic ATP (136, 198, 303).(ii) Membrane vesicles couple respiration toactive transport by a mechanism not involvingATP, which is again sensitive to uncouplers(see above). Clearly, the effect of uncouplerscannot be attributed to interference with ATPsynthesis but must be exerted directly on themembrane.

Uncouplers do, in fact, exert a direct effecton microbial membranes: They facilitate diffu-sion of protons across the plasma membrane,which is otherwise largely impermeable tothem (106, 131, 132, 136, 180, 256, 303, 335). Acorrelation between the rate of proton move-ments and the degree of inhibition of phos-phate uptake has been reported (136). It isthus a reasonable working hypothesis that theinhibition of active transport by uncouplersmay be due to their ability to collapse gra-dients of pH and of electrical potential acrossthe membrane.

In two recent articles, Mitchell (264, 265)has developed in some detail his views on themechanisms by which, in both mitochondriaand bacteria, protons serve to couple metabo-lism to the uptake of nutrients. As was dis-cussed earlier, the respiratory chain is thoughtto be arranged across the plasma membrane soas to extrude protons and generate a differenceof pH and of electrical potential. In anaerobicorganisms, proton extrusion may be accom-plished by oxidation/reduction of pairs of sub-strates (e.g., Stickland reaction) via an oxida-tion chain arranged across the membrane in aloop. Finally the Mg2+-dependent ATPase ofbacterial membranes is thought to catalyze thereversible translocation of protons outwardwhen ATP is hydrolyzed and would thus con-stitute an independent proton pump whichcan generate the proton gradient at the ex-pense of glycolytic ATP. The asymmetric dis-tribution of protons gives rise to a proton-mo-tive force which, as was discussed previouslyfor mitochondria, is the sum of two compo-

211VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

nents reflecting the pH and electrical gra-dients, respectively. The contributions of ApHand Ax to the total proton-motive force wouldvary with the organism and the circumstances.Aside from a limited number of group trans-

locations, the theory implies that most nu-trients are accumulated by secondary porterscoupled to metabolism only via the proton-motive force. K+ accumulation is attributed toan electrogenic porter specific for K+, and itsextent would be a function of the electricalpotential. Anion accumulation (e.g., sulfate orphosphate) is seen as anion-proton symportand, if electrically neutral overall, would occurto an extent determined largely by the ApH.Finally, accumulation of uncharged metabo-lites could be mediated by porters that cata-lyze symport with H+, as diagrammed for ,8-galactosides in Fig. 14; association of theporter with protons could determine either itsaffinity for the ligand or the "mobility" of thecomplex.A calculation may render this proposal more

concrete. The mechanism as written postulateselectrogenic translocation of one proton intothe cell with each molecule of f-galactoside.Let us assume, for the sake of illustration, a ApH (interior alkaline) of one unit and a A'I of- 120 mv (interior negative). The total proton-motive force, Ap, would then be -180 mv. Anequal force would be exerted on 0-galactosidemolecules; since a concentration factor of 10corresponds approximately to 60 mv, galacto-sides could in principle accumulate in the cellsto a concentration gradient of 1,000: 1.

In Mitchell's view, then, the metabolism ofbacterial cells (like that of mitochondria) isdominated by a circulation of protons betweencytoplasm and medium, shown in a greatlysimplified form in Fig. 15: Respiration andATPase translocate protons outward; ATPsynthesis and nutrient accumulation are, ingeneral, linked to translocation of protons

inward. Now, a steady-state electrical poten-tial across the membrane would cause externalcations, K+ and Na+, to accumulate in thecells; this would tend to collapse the potentialdifference and replace it by a rise in the in-ternal pH. To retain a significant potentialacross the membrane, Mitchell postulates theexistence of K+/H+ and Na+/H+ antiporterswhose function is to translocate these cationsoutward. The absolute difference of pH and ofelectrical potential will thus depend on therates of proton and other ion translocation,both inward and out. It is a network of mul-tiple interactions, whose complexity will becompounded if Na+-coupled transport systemsprove to be a reality. Quantitative analysis ofthis web is beyond both the scope of this ar-ticle and the competence of its author.

Until recently, proponents of the view thatactive transport is linked in some manner toproton movements had to rest their caselargely on the inhibition of transport by uncou-plers: not a compelling argument since, as wediscussed in the context of mitochondria, themode of action of uncouplers is still subject todispute. In particular, one should keep in mindthat it may not be possible to distinguishproton conduction across the membrane fromconduction to sensitive sites buried in the hy-drophobic regions of its interior. During thepast few years, however, evidence has begun toappear which points to the generation of ApHand A'I by bacteria and implicates these inactive transport.

In this laboratory, a systematic study wasinitiated of the relationship of proton move-ments to nutrient transport in S. faecalis-anorganism that is convenient precisely becauseof the absence of oxidative phosphorylation. Asfar as we have been able to determine, trans-port is energized entirely by glycolysis andother substrate-level phosphorylations; all theevidence points to ATP as the energy donor

H+-< ¢ ATPi+OH+

Hi(ADP+PIH)

v ~ H+

,a-Gal e-Ga\ctosideSym-

fl-Galactosde

FIG. 14. Accumulation of ti-galactosides in E. coli by a H+-symport mechanism. From Mitchell (265). a-Ga-lactoside translocation a, when the proton current is due to respiration; b, when the proton current is due toadenosine triphosphate (ATP) hydrolysis.

212 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



Respiratory chtain

HH+

Proton-translocating ATPase

H ADP,Pi

Proton-sugar symport ATP

H+

Proton-anion symport

An

Electrogenic cation uptake c+

Na

Sodium-proton antiport

Na

Sodium-nutrient symport S /_

Na

Leakage of Na+ and H+ H+

FIG. 15. Coupling of transport and metabolism bycirculation of protons and Na+. Proton circulationafter Mitchell (265).

for the accumulation of K+, phosphate, aminoacids, and many other substances. A potentMg2+-dependent ATPase activity (Table 2) isassociated with the plasma membrane. This isa multienzyme complex involving at leastthree components: the ATPase itself, a proteincalled nectin that is believed to link the en-

zyme to the membrane, and an unidentifiedcomponent of the membrane which confersupon the ATPase sensitivity to DCCD (4, 6,26, 137, 138, 332, 333). There is little doubtthat this enzyme is implicated in the couplingof glycolysis to membrane transport of many

nutrients. One exception is the uptake anddissimilation of glucose itself, which appar-

ently involves the phosphotransferase system.A possible role for the ATPase emerged from

studies on the linkage between glycolysis andthe accumulation of K+, in which extrusion ofprotons from the cells appears to play a cen-

tral role. Glycolyzing cells generate a pH gra-dient of up to one unit, interior alkaline, byextrusion of protons in exchange for K+ (139,142). The pH gradient is dissipated by exhaus-tion of the substrate, by proton-conductinguncouplers, and by antibiotics of the nigericintype, but not by valinomycin (142). More re-

cently, studies on the accumulation of lipid-

soluble cations (e.g., DDA+) provided con-vincing evidence that extrusion of protonsduring glycolysis is an electrogenic processwhich can generate a membrane potential ofsome - 150 to -180 mv (140, 141). The AT-Pase was clearly implicated in proton extru-sion and constitutes part of a proton pump,but its precise role remains to be explored.Accumulation of K+, quantitatively one of themajor transport processes carried out by S.faecalis, is ultimately dependent upon the ex-trusion of protons, since mutations and inhibi-tors which interfere with proton extrusion alsoimpair K+ uptake. Indeed, the proton pump isinvolved also in the expulsion of Na+ from thecells. There appears to be no Na+ pump, butrather an electroneutral antiport of Na+ forH+; the H+ is then ejected by the pump (141,271). As far as they go, these results are con-sistent with the scheme outlined in Fig. 15.However, unlike the case in mitochondria, theK+ complement of S. faecalis was not dis-charged by proton conductors and thus doesnot appear to be in equilibrium with an elec-trical potential across the membrane.

In S. faecalis, as in many other bacteria,transport of amino acids, phosphate, and cer-tain sugars is sensitive to proton-conductinguncouplers (136). It is also inhibited by niger-icin and by monensin, but not by valinomycinplus K+. Bearing in mind the known modes ofaction of these inhibitors (Table 1), it may bethat the pH gradient may play a role in thetransport process, but that a membrane poten-tial does not. Indeed, K+-replete cells of S.faecalis do not now appear to maintain a largesteady-state potential (141). Further work isclearly required to clarify the relationship of ApH and AI to nutrient transport, and espe-cially to search for co-transport of H+ into thecells together with amino acids and other li-gands.

Direct evidence for proton-substrate sym-port has, however, been presented by West(388): under certain conditions uptake of lac-tose by E. coli is accompanied by stoichio-metric influx of protons. More recent results(389) confirm that H+-lactose symport is anelectrogenic process, which would thus be con-trolled by a membrane potential. In staphylo-cocci, uptake of amino acids may likewise bedetermined by the membrane potential (286).Eddy and his associates (88-90), using yeast,reported influx of protons together with that ofamino acids; they suggested that H+, and per-haps K+ also, are co-substrates for amino acidaccumulation via an ion gradient.Do bacterial membrane vesicles extrude pro-

213VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

tons with the generation of a proton-motiveforce? This is an issue that is certain to behotly debated. Reeves (313) has just reportedexperiments which suggest that membranevesicles of E. coli respiring D-lactate extrudeprotons into the medium by a process sensitiveto uncouplers, and suggested the production ofa proton gradient. Subsequent experiments,however, led Kaback (188) to state that theextrusion of protons is observed even underconditions which destroy the permeability bar-rier. The experimental details will be awaitedwith interest.

In a number of vesicle preparations, activetransport of sugars and amino acids is inhib-ited by valinomycin plus K+ (25, 30, 157, 286),which suggests that transport depends insome manner upon an electrical potential, in-terior negative, that is dissipated by the elec-trophoretic influx of K+. It is pertinent herethat vesicles of E. coli, at least, have appar-ently lost the transport system for K+ andtake up K+ only in presence of valinomycin(30). Unlike valinomycin, nigericin did notinhibit the transport of proline by membranevesicles of M. phlei (157); this makes sensesince nigericin would not be expected to col-lapse a potential difference (Table 1). Finally,in the same system (157), transport of prolinewas uncoupled by lipid-soluble cations but notby anions: further evidence, however indirect,that transport is related to the generation bythe vesicles of an electrical potential, interiornegative. This conclusion is denied by Kaback(188), who feels that accumulation of Rb+ byE. coli vesicles in presence of valinomycin re-flects a pump coupled to the D-lactate dehy-drogenase rather than electrophoretic influx inresponse to the membrane potential. The ex-perimental findings on which Kaback basedthis view were not available at the time ofwriting, and further comment would be prema-ture.There seems to me little doubt that proton

translocation is somehow involved in the ac-tive transport of many nutrients. Whether theproton circuit (Fig. 15) correctly describes thenature of this relationship remains, however, amatter on which honest men may differ. Theconcept that metabolism and transport arelinked by a proton current depends for itssupport very heavily on what we believe to bethe modes of action of ionophores and otherreagents (Table 1). Thus the first requirementis clearly for direct and reliable measurementsof membrane potentials, pH gradients, and thestoichiometry of proton symport, to put thehypothesis upon firmer footing and to assess

its quantitative adequacy. Beyond this, onecan discern difficult questions to which an-swers will have to be sought. For example, therequirement that respiring cells maintain alarge electrical potential, interior negative, inthe steady state, raises a number of problems.In S. faecalis, at least, cells replete with K+ donot appear to maintain a large potential. In-deed, it is not easy to see how collapse of thepotential can be avoided in the presence ofhigh external K+-such as the 0.1 M buffersfrequently used. Mitchell and Moyle (265, 270,271) postulate the existence of a K+/H+ anti-porter to extrude such unwanted K+, but itmust be doubted whether such a system, thusfar unrecognized, would suffice. There are alsovarious data which suggest that E. coli com-pletely devoid of energy sources does nottranslocate metabolites at all (221). This maybe related to the many results from Kaback'slaboratory which point to the involvement ofsulfhydryl groups in the transport of sugarsand amino acids. These and many other find-ings must eventually find their place in an inte-grated view of transport, and clearly the timeis not yet.

Role of the Membrane in MotilityThe role of flagellae in bacterial motility is

universally recognized, but the mechanism bywhich the motion is produced is uncertain.Flagellae may be semirigid helices, activatedby a power source located at the base, whichimparts to them a rotary motion. Alterna-tively, the subunits may undergo periodicchanges of conformation such that helicalwaves are propagated from the base of the fla-gellum to its tip. The fact that removal of thecell wall immobilizes the organisms suggeststhe need for a rigid structure on which the fla-gellum exerts its thrust.No one doubts that motility requires a

source of energy, but the identity of the imme-diate energy donor is not at all established.Sherris and his associates, whose work hasbeen well reviewed by Doetsch and Hageage(86), found that a strain of Pseudomonas be-came immobile when the oxygen supply wasexhausted. However, arginine, which is catabo-lized by pathway that generates ATP by sub-strate-level phosphorylation, restored motilityunder anaerobiosis. These results gave rise tothe prevailing opinion that flagellae, like mus-cles, employ ATP as the immediate energydonor. The ATP would be supplied either byoxidative phosphorylation or by substrate-levelphosphorylation. However, there is nothing to

214 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



exclude the possibility, admittedly speculative,that the plasma membrane is more intimatelyinvolved in the generation of motive power.Perhaps the broadest hint in this direction isgiven by the superb electron micrographs pub-lished by De Pamphilis and Adler (84, 85).These reveal many details of the association offlagellae with both wall and membrane, com-plete with an engineer's view of the hook-up,and provide a structural basis for a role of themembrane in motility more direct than as asource of ATP.The only other evidence of which I am aware

is the finding (102-104) that motility was in-hibited by a number of antibiotics which exerttheir effects upon the membrane. Gramicidin,valinomycin, and monactin, all K+-conductingionophores, inhibited motility but neither ox-ygen uptake nor growth. The effects were atleast partly reversed by high K+. Motility wasalso inhibited by HOQNO and by a series ofdrugs known to depolarize nerve membranesand block impulse transmission. Colicins Eland K, which also exert their effects on en-ergy-linked functions of the membrane, areknown to inhibit motility of E. coli (109). Themeaning of these results is not obvious. Faustand Doetsch (103, 104) suggest that a mem-brane potential may be involved in motility, orthat conformational changes induced in themembrane at the base of the flagellum impartto it a rotary motion. These speculations lackclear definition at present, but they may con-tain the germ of a novel approach to thevexing problem of bacterial motility. It wouldbe of very great importance to study the ef-fects of ionophores in a more sensitive orga-nism and under conditions which lend them-selves more readily to interpretation.

Bacteriocins and the Energized StateThroughout this review, I have found it

useful to invoke the energized state of themembrane as a link between metabolism andvarious forms of work, recognizing that thisentity may share attributes of a membranepotential, a chemical compound, and of astrained conformational state. Uncouplers areinhibitors which exert their effects at, or closeto, this level, and others are found among thebacteriocins. Here we shall consider only coli-cins of types K, El, and I which interfere withenergy-linked membrane functions (for generalreviews see 132, 287).

Colicins K, El, and I evoke multiple effectsin sensitive cells. Synthesis of deoxyribonu-cleic acid, ribonucleic acid, and protein ceases

abruptly; the ATP level falls to about one-third, though respiration continues. Activetransport of TMG is blocked, and preexistingpools are discharged, but the permeability bar-rier in general remains intact. Indeed, thephosphotransferase system is not affected, andthe cells can still accumulate a-methylgluco-side by vectorial phosphorylation. Motility isinhibited (108, 109, 237). Thus far the effectsresemble those of azide and suggest that coli-cins uncouple oxidative phosphorylation. How-ever, not all the known metabolic effects maybe so explained. Fields and Luria (108, 109)found marked shifts in metabolic pathways,with excretion of glucose-6-phosphate, fructose1, 6 diphosphate, and other glycolytic interme-diates from the cells. Moreover, hemin-defi-cient mutants, which should be unable to carryout oxidative phosphorylation, are still colicin-sensitive. Conversely, strict anaerobiosis pro-tects the cells. Similar results have recentlybeen obtained with a bacteriocin from Serratia(116).The suggestion that colicins should not be

regarded primarily as uncouplers of oxidativephosphorylation is reinforced by the demon-stration that colicin El inhibited proline accu-mulation by membrane vesicles of E. coli (31).Since these vesicles do not carry out oxidativephosphorylation, it follows that the primaryeffect of the antibiotics is likely to be on someaspect of the energized state; the characteristicdrop in ATP level is probably a consequence ofthis primary action, rather than its cause (106).Confirmation of this suggestion may be

found in a paradoxical observation reported byCramer and Phillips (80): when colicin El wasadded to a suspension of E. coli cells con-taining ANS-, the fluorescence yield was en-hanced. The increased fluorescence of ANS-paralleled the decrease in the cellular ATPlevel. It is conceivable that this effect mirrorsthe collapse of the energized state, and espe-cially that of a membrane potential: like intactmitochondria, respiring E. coli may extrudeANS- by virtue of the electrical potential, in-terior negative.

Just what the nature of the primary colicineffect may be is presently not known. Colicinsbind to highly specific cellular receptors andare believed to remain bound to the receptorsites, external to the permeability barrier, ex-erting their effects from there. A single,bound-colicin molecule may be lethal. Mu-tants have been isolated which adsorb colicinsbut no longer respond to them, and the exist-ence of such colicin-tolerant mutants is oftencited as evidence for the existence of a "trans-

215VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

mission system:" such that the entire mem-brane responds as a unit to attachment of asingle colicin molecule (281, 287, 288). How-ever, a localized lesion or hole in the mem-brane (386) may equally account for the re-sults. Indeed, Feingold (106) reported the im-portant observation that colicin El facilitatesdiffusion of K+ across E. coli membranes, butnot that of protons. Rapid loss of K+ fromcells upon addition of colicin El was also de-scribed by Hirata et al. (158) and by Wendt(386). Whether K+ conduction will prove to bethe necessary and sufficient cause of the meta-bolic effects of colicin El remains to be seen.It seems, however, that colicins (unlike theclassical uncouplers) do not conduct protonsand thus make up a distinct class of probes forthe nature and physiological role of the ener-gized state.

SUMMARY AND PROSPECTThis essay is an attempt to integrate the

fragmentary information on energy transfor-mation by bacterial membranes with the moreextensive knowledge available from mitochon-dria.

It seems now to be generally admitted thatmitochondria utilize respiratory energy in anumber of fundamentally different ways. ATPsynthesis, transhydrogenation, ion transport,and other functions all appear to draw upon acommon energy donor whose nature remainselusive. At times it has the qualities of a chem-ical compound, at others those of a strainedconformational state; the transport data canbe rationalized only by accepting the existenceof a pH gradient and electrical potential acrossthe membrane. Efforts to assign a unique na-ture to this trinity continue, generating a liter-ature whose flavor is at times almost Byzan-tine.The main thesis of this article is that in

principle, though perhaps not in every detail,the nature and interrelationships of energy-conserving and energy-utilizing. reactions inbacteria are the same as those in mitochon-dria. Not for all regions of the metabolic map(Fig. 1) can this assertion be made with confi-dence. The case is best for oxidative phospho-rylation (and photosynthesis), where a strongargument can be made simply on the groundsof biochemical unity. It is weak indeed formotility, where hard evidence and persuasiveanalogy are alike lacking. Most intriguing isthe emerging inference that bacteria, too, gen-erate gradients of pH and of electrical poten-tial across the membrane, and that these are

intimately related both to oxidative phospho-rylation and to active transport.The coupling of transport to metabolism

poses special challenges. Historically, micro-biologists tended to approach transport fromthe kinetic and genetic side, to the neglect ofits metabolic connections. Students of mito-chondria achieved a better appreciation of theplace of transport in the overall economy oforganelles, because they encountered it as amode of energy utilization altemative to oxi-dative phosphorylation. It is obvious that thesetwo cultures must interact more than theyhave done so far, to the mutual enrichment ofboth. But it should not be taken for grantedthat models based on mitochondria can betransplanted intact to the bacteria. In nutrienttransport, if nowhere else, we can expect toencounter major differences, resulting from thedivergent life-styles of intracellular organelleand free-living organism.Today, when scientists are inundated by

what we genially refer to as The Literature, acall for more research should not be issuedlightly. It is difficult to see how the questionstouched upon in this article can be resolvedwithout more experimental data and criticalthought. But one has the uneasy feeling thatthe former is likely to be more plentiful thanthe latter.

ACKNOWLEDGMENTSI am indebted to Lynn White and the M.I.T. Press for

permission to reproduce the opening quotation from"Machina ex Deo" (1968); to V. P. Skulachev and AcademicPress for Fig. 2; to H. R. Kaback and the Journal of Biolog-ical Chemistry for Fig. 13; and to Peter Mitchell and Cam-bridge University Press for Figures 14 and 15. A. Abrams, K.H. Altendorf, P. Boyer, A. F. Brodie, F. Gibson, W. A. Ham-ilton, J. Hochstadt-Ozer, H. R. Kaback, E. Kashket, P.Mitchell, J. Moyle, E. C. C. Lin, D. Oxender, S. Silver, K.Thorne, I. C. West, and T. H. Wilson generously respondedto my requests for clarification, manuscripts, or constructivecriticism.

Original work from this laboratory was supported byPublic Health Service grant AI-03568 from the National In-stitute of Allergy and Infectious Diseases.

Special thanks are due to Ned Eig, Ethel Goren, EugeneLevin, and Nadia de Stackelburg for their help in preparingthis article for publication and to H. V. Rickenberg and myfamily for their patient forbearance.

LITERATURE CITED1. Abram, D. 1965. Electron microscope observa-

tions on intact cells, protoplasts, and the cy-toplasmic membrane of Bacillus stearother-mophilus. J. Bacteriol. 89:855-873.

2. Abrams, A. 1959. Reversible metabolic swellingof bacterial protoplasts. J. Biol. Chem. 234:383-388.

3. Abrams, A. 1960. Metabolically dependent pen-etration of oligosaccharides into bacterial

216 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



cells and protoplasts. J. Biol. Chem. 235:1281--1285.

4. Abrams, A., and C. Baron. 1967. The isolationand subunit structure of Streptococcal mem-brane adenosine triphosphatase. Biochem-istry 6:225-229.

5. Abrams, A., and C. Baron. 1970. Inhibitory ac-tion of carbodiimides on bacterial membraneATPase. Biochem. Biophys. Res. Commun.41:858-863.

6. Abrams, A., J. B. Smith, and C. Baron. 1972.Carbodiimide-resistant membrane ATPase I.Studies of the mechanism of resistance. J.Biol. Chem., in press.

7. Adanki, S., F. D. Cahill, and J. F. Sotos. 1968.Determination of intramitochondrial pH andintramitochondrial-extramitochondrial pHgradient of isolated heart mitochondria bythe use of 5,5-dimethyl-2,4-oxaxolidinedione.L. Changes during respiration and adenosinetriphosphate-dependent transport of Ca++,Mg++ and Zn++. J. Biol. Chem. 243:2337-2348.

8. Adolfsen, R., and E. N. Moudrianakis. 1971.Kinetic characterization of oxidative phos-phorylation in Alcaligenes faecalis. Biochem-istry 10:434-440.

9. Adolfsen, R., and E. N. Moudrianakis. 1971.Reconstitution of oxidative phosphorylationin Alcaligenes faecalis. Biochemistry 10:440-446.

10. Adolfsen, R., and E. N. Moudrianakis. 1971.Purification and properties of two solublecoupling factors of oxidative phosphorylationfrom Alcaligenes faecalis. Biochemistry 10:2247-2253.

11. Aithal, H. N., V. K. Kalra, and A. F. Brodie.1971. Reversal of the effects of freezing onoxidative phosphorylation in the Mycobac-terium phlei system. Biochem. Biophys. Res.Commun. 43:550-556.

12. Ames, G. F., and J. Lever. 1970. Compo-nents of histidine transport: histidine-bindingproteins and hisP protein. Proc. Nat. Acad.Sci. U.S.A. 66:1096-1103.

13. Asano, A., and A. F. Brodie. 1964. Oxidativephosphorylation in fractionated bacterial sys-tems. XIV. Respiratory chains of Mycobac-terium phlei. J. Biol. Chem. 239:4280-4291.

14. Asano, A., N. S. Cohen, and A. F. Brodie. 1971.Orientation of bacterial membranes and oxi-dative phosphorylation. Fed. Proc. Fed.Amer. Soc. Exp. Biol. 30:1285.

15. Asano, A., K. Imai, and R. Sato. 1967. Oxida-tive phosphorylation in Micrococcus denitrifi-cans. IIL The properties of pyridine nucleotidetranshydrogenase. Biochim. Biophys. Acta143:477-486.

16. Asano, A., K. Imai, and R. Sato. 1967. Oxida-tive phosphorylation in Micrococcus denitrifi-cans. III. ATP-supported reduction of NAD+by succinate. J. Biochem. (Tokyo) 62:210-214.

17. Azzi, A. 1969. Redistribution of the electrical

charge of the mitochondrial membraneduring energy conservation. Biochem. Bio-phys. Res. Commun. 37:254-260.

18. Azzi, A., M. Baltscheffsky, H. Baltscheffsky,and H. Vainio. 1971. Energy-linked changesof the membrane of Rhodospirillum rubrumchromatophores detected by the fluorescentprobe 8-anilinonaphthalene-1-sulfonic acid.FEBS Letters 17:49-52.

19. Azzi, A., B. Chance, G. K. Radda, and C. D.Lee. 1969. A fluorescence probe of energy-dependent structure changes in fragmentedmembranes. Proc. Nat. Acad. Sci. U.S.A. 62:612-619.

20. Azzi, A., P. Gherardini, and M. Santato. 1971.Fluorochrome interaction with the mito-chondrial membrane. The effect of energyconservation. J. Biol. Chem. 246:2035-2042.

21. Baccarini-Melandri, A., H. Gest, and A. SanPietro. 1970. A coupling factor in bacterialphotophosphorylation. J. Biol. Chem. 245:1224-1226.

22. Bakeeva, L. E., L. L. Grinius, A. A. Jasaitis, V.V. Kuliene, D. 0. Levitsky, E. A. Liberman,L. I. Severina, and V. P. Skulachev. 1970.Conversion of biomembrane-produced energyinto electric form. II. Intact mitochondria.Biochim. Biophys. Acta 216:13-21.

23. Baltscheffsky, H., M. Baltscheffsky, and A.Thore. 1971. Energy conversion reactions inbacterial photosynthesis. Curr. Top. Bio-energ. 4:273-325.

24. Barnes, E. M., Jr., and H. R. Kaback. 1970. ,-Galactoside transport in bacterial membranepreparations: energy coupling via membrane-bound D-lactic dehydrogenase. Proc. Nat.Acad. Sci. U.S.A. 66:1190-1198.

25. Barnes, E. M., Jr., and H. R. Kaback. 1971.Mechanisms of active transport in isolatedmembrane vesicles. I. The site of energy cou-pling between D-lactic dehydrogenase and #-galactoside transport in Escherichia colimembrane vesicles. J. Biol. Chem. 246:5518-5522.

26. Baron, C., and A. Abrams. 1971. Isolation of abacterial membrane protein, nectin, essentialfor the attachment of adenosine triphospha-tase. J. Biol. Chem. 246:1542-1544.

27. Bennun, A. 1971. Hypothesis for coupling en-ergy transduction with ATP synthesis or ATPhydrolysis. Nature New Biol. 233:5-8.

28. Berlin, R. D., and E. R. Stadtman. 1966. A pos-sible role of purine nucleotide pyrophos-phorylases in the regulation of purine uptakeby Bacillus subtilis. J. Biol. Chem. 241:2679-2686.

29. Bhattacharyya, P. 1970. Active transport ofmanganese in isolated membranes of Esche-richia coli. J. Bacteriol. 104:1307-1311.

30. Bhattacharyya, P., W. Epstein, and S. Silver.1971. Valinomycin-induced uptake of potas-sium in membrane vesicles from Escherichiacoli. Proc. Nat. Acad. Sci. U.S.A. 68:1488-1492.

217VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

31. Bhattacharyya, P., L. Wendt, E. Whitney, andS. Silver, 1970. Colicin-tolerant mutants ofEscherichia coli: resistance of membranes tocolicin El. Science 168:998-1000.

32. Bogin, E., T. Higashi, and A. F. Brodie. 1970.Interchangeability of coupling factors frombacterial and mammalian origin. Biochem.Biophys. Res. Commun. 38:478-483.

33. Bogin, E., T. Higashi, and A. F. Brodie. 1970.The effect of trypsin and heat treatment onoxidative phosphorylation in Mycobacteriumphlei. Biochem. Biophys. Res. Commun. 41:995-1001.

34. Bogin, E., T. Higashi, and A. F. Brodie. 1970.Effects of heat treatment of electron-trans-port particles on bacterial oxidative phos-phorylation. Proc. Nat. Acad. Sci. U.S.A. 67:1-6.

35. Bogin, E., T. Higashi, and A. F. Brodie. 1970.Oxidative phosphorylation in fractionatedbacterial systems. XLIII. Coupling factorsassociated with the NAD+-linked electrontransport pathway. Arch. Biochem. Biophys.136:337-351.

36. Boos, W., and A. S. Gordon. 1971. Transportproperties of the galactose-binding protein ofEscherichia coli. Occurrence of two conforma-tional states. J. Biol. Chem. 246:621-628.

37. Boos, W., and M. 0. Sarvas. 1970. Closelinkage between a galactose-binding proteinand the ,-methylgalactoside permease inEscherichia coli. Eur. J. Biochem. 13:526-533.

38. Borst-Pauwels, G. W. F. H., and S. Jager. 1969.Inhibition of phosphate and arsenate uptakein yeast by monoiodoacetate, fluoride, 2,4-dinitrophenol and acetate. Biochim. Biophys.Acta 172:399-406.

39. Borst-Pauwels, G. W. F. H., G. H. J. Wolters,and J. J. G. Henricks. 1971. The interactionof 2,4-dinitrophenol with anaerobic Rb+transport across the yeast cell membrane.Biochim. Biophys. Acta 225:269-276.

40. Bose, S. K., and H. Gest. 1965. Properties ofadenosine triphosphatase in a photosyntheticbacterium. Biochim. Biophys. Acta 96:159-162.

41. Bovell, C. R., L. Packer, and R. Helgerson.1963. Permeability of Escherichia coli to or-ganic compounds and inorganic salts meas-ured by light-scattering. Biochim. Biophys.Acta 75:257-266.

42. Boyer, P. D. 1968. Oxidative phosphorylation,p. 193-235. In T. P. Sanger (ed.), Biologicaloxidations. Interscience Publishers, Inc., NewYork.

43. Boyer, P. D., and W. L. Klein. 1972. Energycoupling mechanisms in transport. In C. F.Fox and A. D. Keith (ed.), Membrane molec-ular biology. Sinauer Associates, Stamford,Conn., to be published.

44. Bragg, P. D., and C. Hou. 1968. Oxidative phos-phorylation in Escherichia coli. Can. J.Biochem. 46:631-641.

45. Britten, R. J. 1965. The concentration of small

molecules within the microbial cell, p. 57-88.In Function and structures in microorga-nisms, XV Symp. Soc. Gen. Microbiol.,Cambridge University Press, Cambridge.

46. Britten, R. J., and F. T. McClure. 1962. Theamino acid pool in Escherichia coli. Bac-teriol. Rev. 26:292-335.

47. Brock, T. D., and G. Moo-Penn. 1962. Anamino acid transport system in Streptococcusfaecalis. Arch. Biochem. Biophys. 98:183-190.

48. Brocklehurst, J. R., R. B. Freedman, D. J.Hancock, and G. K. Radda. 1970. Membranestudies with polarity-dependent and excimer-forming fluorescent probes. Biochem. J. 116:721-731.

49. Brodie, A. F., and D. L. Gutnick. 1972. Elec-tron transport and oxidative phosphorylationin microbial systems. In T. E. King and M.Klingenberg (ed.), Electron and coupled en-ergy transfer in biological systems, vol. IB.Marcel Dekker, New York, in press.

50. Brown, K. D. 1970. Formation of aromaticamino acid pools in Escherichia coli K 12. J.Bacteriol. 104:177-188.

51. Brown, K. D. 1971. Maintenance and exchangeof the aromatic amino acid pool in Esche-richia coli. J. Bacteriol. 106:70-81.

52. Bryan-Jones, D. G., and R. H. Whittenbury.1969. Haematin-dependent oxidative phos-phorylation in Streptococcus faecalis. J. Gen.Microbiol. 58:247-260.

53. Buchwald, M., and R. J. Britten. 1963. Incorpo-ration of ribonucleic acid bases into the meta-bolic pool and RNA of Escherichia coli. Bio-phys. J. 3:155-166.

54. Butlin, J. D., G. B. Cox, and F. Gibson. 1971.Oxidative phosphorylation in Escherichia coliK 12. Mutations affecting magnesium ion orcalcium ion-stimulated adenosine triphospha-tase. Biochem. J. 124:75-81.

55. Carafoli, E., W. X. Balcavage, A. L. Lehninger,and J. R. Mattoon. 1970. Ca++ metabolism inyeast cells and mitochondria. Biochim. Bio-phys. Acta 205:18-26.

56. Carafoli, E., R. G. Hansford, B. Sacktor, and A.L. Lehninger. 1971. Interaction of Ca++ withblowfly flight muscle mitochondria. J. Biol.Chem. 246:964-972.

57. Carstensen, E. L., J. Maniloff, and C. W. Ei-nolf, Jr. 1971. Electrical properties and ultra-structure of Mycoplasma membranes. Bio-phys. J. 11:572-581.

58. Carter, J. R., C. F. Fox, and E. P. Kennedy.1968. Interaction of sugars with the mem-brane protein component of the lactose trans-port system of Escherichia coli. Proc. Nat.Acad. Sci. U.S.A. 60:725-732.

59. Caswell, A. H. 1969. Proton permeability andthe regulation of potassium permeability inmitochondria by uncoupling agents. J. Mem-brane Biol. 1:53-78.

60. Cattell, K. J., I. G. Knight, C. R. Lindop, andR. B. Beechey. 1970. The isolation of dicy-clohexylcarbodiimide-binding proteins frommitochondrial membranes. Biochem. J. 117:

218 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



1011-1013.61. Cattell, K. J., C. R. Lindop, I. G. Knight, and

R. B. Beechey. 1971. The identification of thesite of action of dicyclohexylcarbodiimide asa proteolipid in mitochondrial membranes.Biochem. J. 125:169-177.

62. Catterall, W. A., and P. L. Pedersen. 1971.Adenosine triphosphatase from rat liver mito-chondria. I. Purification, homogeneity, andphysical properties. J. Biol. Chem. 246:4987-4994.

63. Cavari, B. Z., Y. Avi-dor, and N. Grossowicz.1968. Induction by oxygen of respiration andphosphorylation of anaerobically grown Esch-erichia coli. J. Bacteriol. 96:751-759.

64. Chance, B. 1970. Fluorescent probe environ-ment and the structural and charge changesin energy coupling of mitochondrial mem-branes. Proc. Nat. Acad. Sci. U.S.A. 67:560-571.

65. Chance, B., C.-P. Lee, and L. Mela. 1967. Con-trol and conservation of energy in the cyto-chrome chain. Fed. Proc. Fed. Amer. Soc.Exp. Biol. 26:1341-1354.

66. Chance, B., and L. Mela. 1967. Energy-linkedchanges of hydrogen ion concentration insubmitochondrial particles. J. Biol. Chem.242:830-844.

67. Chappell, J. B. 1969. Transport and exchangeof anions in mitochondria, p. 335-350. In Th.Biicher and H. Sies (ed.), Inhibitors-tools incell research. Springer-Verlag, New York.

68. Chappell, J. B., and A. R. Crofts. 1965. Grami-cidin and ion transport in isolated liver mito-chondria. Biochem. J. 95:393-402.

69. Chappell, J. B., and A. R. Crofts. 1966. Iontransport and reversible volume changes ofisolated mitochondria, p. 293-314. In J. M.Tager, S. Papa, E. Quagliariello, and E. C.Slater (ed.), Regulation of metabolic proc-esses in mitochondria. Elsevier, Amsterdam.

70. Chappell, J. B., and K. N. Haarhoff. 1967. Thepenetration of the motochondrial membraneby anions and cations, p. 75-92. In E. C. Sla-ter, Z. Kaniuga, and L. Wojtczak (ed.), Bio-chemistry of mitochondria. Academic PressInc., New York.

71. Chavin, S. I. 1971. Isolation and study of func-tional membrane proteins. Present status andfuture prospects. FEBS Letters 14:269-282.

72. Cirillo, V. P. 1966. Symposium on bioelectro-chemistry of microorganisms. I. Membranepotentials and permeability. Bacteriol. Rev.30:68-79.

73. Cockrell, R. S., E. J. Harris, and B. C.Pressman. 1967. Synthesis of ATP driven bya potassium gradient in mitochondria. Nature(London) 215:1487-1488.

74. Cohen, G. N., and J. Monod. 1957. Bacterialpermeases. Bacteriol. Rev. 21:169-194.

75. Cole, H. A., J. W. T. Wimpenny, and D. E.Hughes. 1967. The ATP pool in Escherichiacoli. I. Measurement of the pool using a mod-ified luciferase assay. Biochim. Biophys.Acta 143:445-453.

76. Contessa, A. R., and A. Bruni. 1971. Differenteffects of oligomycin and dicyclohexyl-carbodimide on ATPases from mammaliancells. Biochim. Biophys. Acta 241:334-340.

77. Cope, F. W., and R. Damadian. 1970. Cell po-tassium by 39K spin echo nuclear magneticresonance. Nature (London) 228:76-77.

78. Cox, G. B., N. A. Newton, J. D. Butlin, and F.Gibson. 1971. The energy-linked transhydro-genase reaction in respiratory mutants ofEscherichia coli K 12. Biochem. J. 125:489-493.

79. Cox, G. B., N. A. Newton, F. Gibson, A. M.Snoswell, and J. A. Hamilton. 1970. Thefunction of ubiquinone in Escherichia coli.Biochem. J. 117:551-562.

80. Cramer, W. A., and S. K. Phillips. 1970. Re-sponse of an Escherichia coli-bound fluores-cent probe to colicin El. J. Bacteriol. 104:819-825.

81. Crofts, A. R., C. A. Wraight, and D. E.Fleischman. 1971. Energy conservation in thephotochemical reactions of photosynthesisand its relation to delayed fluorescence.FEBS Letters 15:89-100.

82. Cross, R. L., B. A. Cross, and J. H. Wang. 1970.Detection of a phosphorylated intermediatein mitochondrial oxidative phosphorylation.Biochem. Biophys. Res. Commun. 40:1155-1161.

83. Datta, A., and H. S. Penefsky. 1970. Interactionof fluorescent probes with submitochondrialparticles during oxidative phosphorylation. J.Biol. Chem. 245:1537-1544.

84. DePamphilis, M. L., and J. Adler. 1971. Finestructure and isolation of the hook-basal bodycomplex of flagellae from Escherichia coliand Bacillus subtilis. J. Bacteriol. 105:384-395.

85. DePamphilis, M. L., and J. Adler. 1971. At-tachment of flagellar basal bodies to the cellenvelope: specific attachment to the outer,lipopolysaccharide membrane and the cyto-plasmic membrane. J. Bacteriol. 105:396-407.

86. Doetsch, R. N., and G. J. Hageage. 1968. Mo-tility in bacteria. Biol. Rev. (Cambridge) 43:317-362.

87. Dreyfuss, J., and A. B. Pardee. 1966. Regulationof sulfate transport in Salmonella typhimu-rium. J. Bacteriol. 91:2275-2280.

88. Eddy, A. A., K. Backen, and G. Watson. 1970.The concentration of amino acids by yeastcells depleted of adenosine triphosphate.Biochem. J. 120:853-858.

89. Eddy, A. A., K. J. Indge, K. Backen, and J. A.Nowacki. 1970. Interactions between potas-sium ions and glycine transport in the yeastSaccharomyces carisbergensis. Biochem. J.120:845-852.

90. Eddy, A. A., and J. A. Nowacki. 1971. Stoi-chiometrical proton and potassium ionmovements accompanying the absorption ofamino acids by the yeast Saccharomycesc.rlsbergensis. Biochem. J. 122:701-711.

91. Edwards, G. E., and C. R. Bovell. 1969. Charac-

219VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

teristics of a light-dependent proton trans-port in cells of Rhodospirillum rubrum.Biochim. Biophys. Acta 172:126-133.

92. Edwards, G. E., and C. R. Bovell. 1970. Actionof phenazine methyl sulfate, inhibitors, anduncouplers on the light-induced proton trans-port by cells of Rhodospirillum rubrum. J.Membrane Biol. 2:95-107.

93. Ellar, D. 1970. The biosynthesis of protectivesurface structures of prokaryotic and eukar-yotic cells. In Organization and control inprokaryotic and eukaryotic cells, XXthSymp. Soc. Gen. Microbiol., p. 167-202.

94. Ellar, D. J., T. D. Thomas, and J. A. Posgate.1971. Properties of mesosomal membranesisolated from Micrococcus lysodeikticus andBacillus megaterium. Biochem. J. 122:44P.-45P.

95. Engelman, D. M. 1970. Lipid bilayer structurein the membrane of Mycoplasma laidlawii. J.Mol. Biol. 58:153-165.

96. Epstein, W., and S. G. Schultz. 1967. Ion trans-port and osmoregulation in bacteria, p. 186-193. In L. B. Guze (ed.), Microbial proto-plasts, spheroplasts and I-forms. The Wil-liams & Wilkins Co., Baltimore.

97. Esfahani, M., A. R. Limbrick, S. Knutton, andS. J. Wakil. 1971. The molecular organizationof lipids in the membrane of Escherichia coli:phase transitions. Proc. Nat. Acad. Sci.U.S.A.68:3180-3184.

98. Evans, D. J., Jr. 1969. Membrane adenosinetriphosphatase of Escherichia coli: activationby calcium ion and inhibition by monovalentcations. J. Bacteriol. 100:914-922.

99. Evans, D. J., Jr. 1970. Membrane Mg++-(Ca++)-activated adenosine triphosphatase of Esche-richia coli: characterization in the membrane-bound and solubilized states. J. Bacteriol.104:1203-1212.

100. Farron, F. 1970. Isolation and properties of a

chloroplast coupling factor and heat activatedadenosine triphosphatase. Biochemistry 9:3823-3828.

101. Farron, F., and E. Racker. 1970. Studies on themechanism of the conversion of couplingfactor 1 from chloroplasts to an active adeno-sine triphosphatase. Biochemistry 9:3829-3836.

102. Faust, M. A., and R. N. Doetsch. 1969. Effectof respiratory inhibitors on the motility ofPseudomonas fluorescens. J. Bacteriol. 97:806--811.

103. Faust, M. A., and R. N. Doetsch. 1971. Effectof membrane-active antibiotics on motilityand K'2 permeability of Pseudomonas flu-orescens. Can. J. Microbiol. 17:183-189.

104. Faust, M. A., and R. N. Doetsch. 1971. Effectof drugs that alter excitable membranes or

motility of Rhodospirillum rubrum andThiospirillum jevense. Can. J. Microbiol. 17:191-196.

105. Faust, D. J., and P. J. Vandemark. 1970. Phos-phorylation coupled to NADH oxidation with

fumarate in Streptococcus faecalis 10CL.Arch. Biochem. Biophys. 137:393-398.

106. Feingold, D. S. 1970. The mechanism of colicinEl action. J. Membrane Biol. 3:272-386.

107. Ferrandes, B., C. Frehel, and P. Chaix. 1970.Fractionnement et purification des systemesmembranaires cytoplasmiques etmesosomiques de Bacillus subtilis. etude dequelques-unes de leurs proprietes oxydo-reductrices associees a la chaine respiratoire.Biochim. Biophys. Acta 223:292-308.

108. Fields, K. L., and S. E. Luria. 1969. Effects ofcolicins El and K on transport systems. J.Bacteriol. 97:57-63.

109. Fields, K. L., and S. E. Luria. 1969. Effects ofcolicins El and K on cellular metabolism. J.Bacteriol. 97:64-77.

110. Finkelstein, A. 1970. Weak-acid uncouplers ofoxidative phosphorylation. Mechanism of ac-tion on thin lipid membranes. Biochim. Bio-phys. Acta 205:1-6.

111. Fisher, R. J., J. C. Chen, B. P. Sani, S. S. Ka-play, and D. R. Sanadi. 1971. A soluble mito-chondrial ATP synthetase complex catalyzingATP-phosphate and ATP-ADP exchange.Proc. Nat. Acad. Sci. U.S.A. 68:2181-2184.

112. Fisher, R. J., and D. R. Sanadi. 1971. Energy-linked nicotinamide adenine dinucleotidetranshydrogenase in membrane particles fromEscherichia coli. Biochim. Biophys. Acta 245:34-41.

113. Fisher, R. R., and R. J. Guillory. 1967. Inhibi-tion of the energy conservation reactions ofRhodospirillum rubrum by Dio-9. Biochim.Biophys. Acta 143:654-656.

114. Fonyo, A. 1968. Phosphate carrier of rat-livermitochondria: its role in phosphate outflow.Biochem. Biophys. Res. Commun. 32:624-628.

115. Forrest, G., and S. J. Edelstein. 1970. On thesubunit structure of the cold labile adenosinetriphosphatase of mitochondria. J. Biol.Chem. 245:6468-6470.

116. Foulds, J. 1971. Mode of action of a bacteriocinfrom Serratia marcescens. J. Bacteriol. 107:833-839.

117. Frank, L., and I. Hopkins. 1969. Sodium-stim-ulated transport of glutamate in Escherichiacoli. J. Bacteriol. 100:329-336.

118. Frehel, C., B. Ferrandes, and A. Ryter. 1971.Reactions de'oxydo-reduction au niveau desmembranes cytoplasmiques et mesosomiquesde Bacillus subtilis. Biochim. Biophys. Acta234:226-241.

119. Frenkel, A. W. 1970. Multiplicity of electrontransport reactions in bacterial photosyn-thesis. Biol. Rev. (Cambridge) 45:569-616.

120. Gachelin, G. 1970. Studies on the a-methylglu-coside permease of Escherichia coli. A two-step mechanism for the accumulation of a-

methylglucoside 6-phosphate. Eur. J.Biochem. 16:342-357.

121. Gear, A. R. L., C. S. Rossi, B. Reynafarje, A. L.Lehninger. 1967. Acid-base exchanges in mi-

220 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



tochondria and suspending medium duringrespiration-linked accumulation of bivalentcations. J. Biol. Chem. 242:3403-3413.

122. Gel'man, N. S., M. A. Lukoyanova, and D. N.Ostrovskii. 1967. Respiration and phosphory-lation of bacteria. English translation,Plenum Press, New York.

123. Ghosh, B. K., M. G. Sargent, and J. 0.Lampen. 1968. Morphological phenomenaassociated with penicillinase induction andsecretion in Bacillus licheniformis. J. Bac-teriol. 96:1314-1328.

124. Green, D. E., J. Asai, R. A. Harris, and J. T.Penniston. 1968. Conformational basis of en-ergy transformations in membrane systems.III. Configurational changes in the mito-chondrial inner membrane induced bychanges in functional states. Arch. Biochem.Biophys. 125:684-705.

125. Greville, G. D. 1969. A scrutiny of Mitchell'schemiosmotic hypothesis. Curr. Topics Bioe-nerg. 3:1-78.

126. Grinius, L. L., A. A. Jasaitis, Y. P. Kadziaus-kas, E. A. Liberman, V. P. Skulachev, V. P.Topali, L. M. Tsofina, and M. A. Vladimi-rova. 1970. Conversion of biomembrane-pro-duced energy into electric form. I. Submito-chondrial particles. Biochim. Biophys. Acta216:1-12.

127. Groot, G. S. P., L. Kovac, and G. Schatz. 1971.Promitochondria of anaerobically grownyeast. V. Energy transfer in the absence of anelectron transfer chain. Proc. Nat. Acad. Sci.U.S.A. 68:308-311.

127a. Gross, R., and N. W. Coles. 1968. Adenosinetriphosphatase in isolated membranes ofStaphylococcus aureus. J. Bacteriol. 95:1322-1326.

127b. Guarraia, L. J., and H. D. Peck, Jr. 1971. Din-itrophenol-stimulated adenosine triphospha-tase activity in extracts of Desulfovibriogigas. J. Bacteriol. 106:890-895.

128. Gutnick, D. L., and A. F. Brodie. 1965. Phos-phate-dependent incorporation of tritiuminto a naphthoquinone during oxidative phos-phorylation. J. Biol. Chem. 240:3698-3699.

129. Hadjipitrou, L., M. D. Lilly, and P. Kouroun-akis. 1970. Effect of ferricyanide on Esche-richia coli. Antonie van Leeuwenhoek J. Mi-crobiol. Serol. 36:531-540.

130. Hafkenscheid, J. C. M., and S. L. Bonting.1969. Studies on (Na+-K+)-aactivated AT-Pase. XXmI. A Mg++-ATPase in Escherichiacoli, activated by monovalent cations.Biochim. Biophys. Acta 178:128-136.

131. Hamilton, W. A., and R. E. Jeacocke. 1971. Theion-specific increases in membrane permea-bility with a group of membrane-active anti-bacterial agents. Biochem. J. 127:56P.

132. Harold, F. M. 1970. Antimicrobial agents andmembrane function. Advan. MicrobialPhysiol. 4:45-104.

133. Harold, F. M., R. L. Harold, and A. Abrams.1965. A mutant of Streptococcus faecalis

defective in phosphate uptake. J. Biol. Chem.240:3145-3153.

134. Harold, F. M., R. L. Harold, J. R. Baarda, andA. Abrams. 1967. A genetic defect in reten-tion of potassium by Streptococcus faecalis.Biochemistry 6:1777-1784.

135. Harold, F. M., and J. R. Baarda. 1968. Effectsof nigericin and monactin on cation permea-bility of Streptococcus faecalis and metaboliccapacities of potassium-depleted cells. J.Bacteriol. 95:816-823.

136. Harold, F. M., and J. R. Baarda. 1968. Inhibi-tion of membrane transport in Streptococcusfaecalis by uncouplers of oxidative phospho-rylation and its relationship to proton conduc-tion. J. Bacteriol. 96:2025-2034.

137. Harold, F. M., J. R. Baarda, C. Baron, and A.Abrams. 1969. Inhibition of membrane-boundadenosine triphosphatase and of cation trans-port in Streptococcus faecalis by N, N1-dicy-clohexylcarbodiimide. J. Biol. Chem. 244:2261-2268.

138. Harold, F. M., J. R. Baarda, C. Baron, and A.Abrams. 1969. Dio-9 and chlorhexidine: in-hibitors of membrane-bound ATPase and ofcation transport in Streptococcus faecalis.Biochim. Biophys. Acta 183:129-136.

139. Harold, F. M., J. R. Baarda, and E. Pavlasova.1970. Extrusion of sodium and hydrogen ionsas the primary process in potassium ion accu-mulation by Streptococcus faecalis. J. Bac-teriol. 101:152-159.

140. Harold, F. M., and D. Papineau. 1972. Cationtransport and electrogenesis by Streptococcusfaecalis. I. The membrane potential. J. Mem-brane Biol. 8:27-44.

141. Harold, F. M., and D. Papineau. 1972. Cationtransport and electrogenesis by Streptococcusfaecalis. II. Proton and sodium extrusion. J.Membrane Biol. 8:45-62.

142. Harold, F. M., E. Pavlasova, and J. R. Baarda.1970. A transmembrane pH gradient inStreptococcus faecalis: origin, and dissipationby proton conductors and N,N1-dicyclohex-ylcarbodiimide. Biochim. Biophys. Acta 196:235-244.

143. Harris, E. J., and B. C. Pressman. 1969. Thedirection of polarity of the mitochondrialtransmembrane potential. Biochim. Biophys.Acta 172:66-70.

144. Harris, R. A., J. T. Penniston, J. Asai, and D.E. Green. 1968. The conformational basis ofenergy conservation in membrane systems. II.Correlation between conformational changeand functional states. Proc. Nat. Acad. Sci.U.S.A. 59:830-837.

145. Harrison, D. E. F., and P. K. Maitra. 1969.Control of respiration and metabolism ingrowing Klebsiella aerogenes. The role ofadenine nucleotides. Biochem. J. 112:647-656.

146. Hazelbauer, G. L., and J. Adler. 1971. Role ofthe galactose binding protein in chemotaxisof Escherichia coli toward galactose. Nature

221VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACrERIOL. REV.

New Biol. 230:101-104.147. Hechtman, P., and C. R. Scriver. 1970. The iso-

lation and properties of a fB-alanine per-measeless mutant of Pseudomonas fluores-cens. Biochim. Biophys. Acta 219:428-436.

148. Hempfling, W. P. 1970. Studies of the effi-ciency of oxidative phosphorylation in intactEscherichia coli B. Biochim. Biophys. Acta205:169-182.

149. Hempfling, W. P. 1970. Repression of oxidativephosphorylation in Escherichia coli B bygrowth in glucose and other carbohydrates.Biochem. Biophys. Res. Commun. 41:9-15.

150. Henderson, P. J. F. 1971. Ion transport by en-ergy-conserving biological membranes. Annu.Rev. Microbiol. 25:394-428.

151. Henderson, P. J. F., and H. A. Lardy. 1969.Antibiotic inhibition of mitochondrial energy-transfer reactions. Antimicrob. Ag. Chem-other. 1969., p. 18-27.

152. Henderson, P. J. F., J. D. McGivan, and J. B.Chappell. 1969. The action of certain anti-biotics on mitochondrial, erythrocyte and ar-tificial phospholipid membranes. The role ofinduced proton permeability. Biochem. J.111:521-535.

153. Hendler, R. W. 1971. Biological membrane ul-trastructure. Physiol. Rev. 51:66-97.

154. Higashi, T., E. Bogin, and A. F. Brodie. 1970.Oxidative phosphorylation irr fractionatedbacterial systems. XLII. The effect of cou-pling factors on urea-treated particles fromM. phlei. Arch. Biochem. Biophys. 136:331-336.

155. Hinkle, P. C., and L. L. Horstman. 1971. Respi-ration-driven proton transport in submito-chondrial particles. J. Biol. Chem. 246:6024-6028.

156. Hinkle, P., and P. Mitchell. 1970. Effect ofmembrane potential on equilibrium poisebetween cytochrome a and cytochrome c inrat liver mitochondria. J. Bioenerg. 1:45-60.

157. Hirata, H., A. Asano, and A. F. Brodie. 1971.Respiration dependent transport of prolineby electron transport particles from Myco-bacterium phlei. Biochem. Biophys. Res.Commun. 44:368-374.

158. Hirata, H., S. Fukui, and S. Ishikawa. 1969. Ini-tial events caused by colicin K infection-cation movement and depletion of ATP pool.J. Biochem (Tokyo) 65:843-847.

159. Hochstadt-Ozer, J., and E. R. Stadtman. 1971.The regulation of purine utilization in bac-teria. I. Purification of adenine phosphori-bosyltransferase from Escherichua coli K 12and control of activity by nucleotides. J. Biol.Chem. 246:5294-5303.

160. Hochstadt-Ozer, J., and E. R. Stadtman. 1971.The regulation of purine utilization in bac-teria. H. Adenine phosphoribosyltransferasein isolated membrane preparations and itsrole in transport of adenine across the mem-brane. J. Biol. Chem. 246:5304-5311.

161. Hochstadt-Ozer, J., and E. R. Stadtman. 1971.The regulation of purine utilization in bac-

teria. III. The involvement of purine phos-phoribosyltransferases in the uptake of ade-nine and other nucleic acid precursors by in-tact resting cells. J. Biol. Chem. 246:5312-5320.

162. Hoek, J. B., N. E. Lofrumento, A. J. Meyer,and J. M. Tager. 1971. Phosphate transportin rat-liver mitochondria. Biochim. Biophys.Acta 226:297-308.

163. Holden, J. T. (ed.) 1962. Amino acid pools.American Elsevier, New York.

164. Holden, J. T., J. N. A. van Balgooy, and J. S.Kittredge. 1968. Transport of aminophos-phonic acids in Lactobacillus plantarum andStreptococcus faecalis. J. Bacteriol. 96:950-957.

165. Hopfer, U., A. L. Lehninger, and W. J. Len-narz. 1970. The effects of the polar moiety oflipids on bilayer conductance induced byuncouplers of oxidative phosphorylation. J.Membrane Biol. 3:142-155.

166. Hopfer, U., A. L. Lehninger, and T. E.Thompson. 1968. Protonic conductanceacross phospholipid bilayer membranes in-duced by uncoupling agents for oxidativephosphorylation. Proc. Nat. Acad. Sci. U.S.A.59:484-490.

167. Horecker, B. L., J. Thomas, and J. Monod.1960. Galactose transport in Escherichia coli.II. Characteristics of the exit process. J. Biol.Chem. 235:1586-1590.

168. Horio, T., and M. D. Kamen. 1970. Bacterialcytochromes. II. Functional aspects. Annu.Rev. Microbiol. 24:399-428.

169. Imai, K., A. Asano, and R. Sato. 1967. Oxida-tive phosphorylation in Micrococcus denitrifi-cans. I. Preparation and properties of phos-phorylating membrane fragments. Biochim.Biophys. Acta 143:462-476.

170. Imai, K., A. Asano, and R. Sato. 1968. Oxida-tive phosphorylation in Micrococcus denitrifi-cans. IV. Further characterization of electron-transfer pathway and phosphorylation ac-tivity in NADH oxidation. J. Biochem. (To-kyo) 63:207-218.

171. Isaev, P. I., E. A. Liberman, V. D. Samuilov, V.P. Skulachev, and L. M. Tsofina. 1970. Con-version of biomembrane-produced energy intoelectric form. III. Chromatophores of Rhodos-pirillum rubrum. Biochim. Biophys. Acta216:22-29.

172. Ishikawa, S. 1966. Identity of a bacterial cou-pling factor and ATPase. J. Biochem. (To-kyo) 60:598-601.

173. Ishikawa, S. 1970. Properties of an oxidativephosphorylation system reconstituted fromcoupling factors in Micrococcus Iysodeikticus.J. Biochem. (Tokyo) 67:297-312.

174. Jackson, J. B., and A. R. Crofts. 1968. Energy-linked reduction of nicotinamide adeninedinucleotides in cells of Rhodospirillumrubrum. Biochem. Biophys. Res. Commun.32:908-915.

175. Jackson, J. B., and A. R. Crofts. 1971. The ki-netics of light induced carotenoid changes in

222 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



Rhodopseudomonas spheroides and their rela-tion to electrical field generation across thechromatophore membrane. Eur. J. Biochem.18:120-130.

176. Jackson, J. B., A. R. Crofts, and L.-V. von Ste-dingk. 1968. Ion transport induced by lightand antibiotics in chromatophores from Rho-dospirillum rubrum. Eur. J. Biochem. 6:41-54.

177. Jagendorf, A. T., and E. Uribe. 1966. ATP for-mation caused by acid-base transition ofspinach chloroplasts. Proc. Nat. Acad. Sci.U.S.A. 55:170-177.

178. Jardetzky, O. 1966. Simple allosteric model formembrane pumps. Nature (London) 211:969-970.

179. Jasaitis, A. A., V. V. Kuliene, and V. P. Sku-lachev. 1971. Anilinonaphthalene-sulfonatefluorescence changes induced by non-enzy-matic generation of membrane potential inmitochondria and submitochondrial particles.Biochim. Biophys. Acta 234:177-181.

180. Jeacocke, R. E., D. F. Niven, and W. A. Ham-ilton. 1972. The proton motive force inStaphylococcus aureus. Biochem. J. 127:57P.

181. John, P., and W. A. Hamilton. 1970. Respira-tory control in membrane particles from Mi-crococcus denitrificans. FEBS Letters10:246-248.

182. John, P., and W. A. Hamilton. 1971. Release ofrespiratory control in particles from Micro-coccus denitrificans by ion-translocating anti-biotics. Eur. J. Biochem. 23:529-532.

183. John, P., and F. R. Whatley. 1970. Oxidativephosphorylation coupled to oxygen uptakeand nitrate reduction in Micrococcus denitri-ficans. Biochim. Biophys. Acta 216:342-352.

184. Kaback, H. R. 1968. The role of the phosphoe-nolpyruvate-phosphotransferase system inthe transport of sugars by isolated membranepreparations of Escherichia coli. J. Biol.Chem. 243:3711-3724.

185. Kaback, H. R. 1970. Transport of sugars acrossisolated bacterial membranes Curr. Top.Membranes Transport 1:35-99.

186. Kaback, H. R. 1970. Transport. Annu. Rev.Biochem. 39:561-598.

187. Kaback, H. R. 1971. p. 99. In W. B. Jakoby (ed.),Methods in enzymology, vol 22. AcademicPress Inc., New York.

188. Kaback, H. R. 1972. Transport across isolatedbacterial cytoplasmic membranes. Biochim.Biophys. Acta, in press.

189. Kaback, H. R., and E. M. Barnes, Jr. 1971.Mechanisms of active transport in isolatedmembrane vesicles. II. The mechanism ofenergy coupling between D-lactic dehydro-genase and jl-galactoside transport in mem-brane preparations from Escherichia coli. J.Biol. Chem. 246:5523-5531.

190. Kaback, H. R., and T. F. Deuel. 1969. Prolineuptake by disrupted membrane preparationsfrom Escherichia coli. Arch. Biochem. Bio-phys. 132:118-129.

191. Kaback, H. R., and L. S. Milner. 1970. Rela-

tionship of a membrane-bound D-(-)-lacticdehydrogenase to amino acid transport in iso-lated bacterial membrane preparations. Proc.Nat. Acad. Sci. U.S.A. 66:1008-1015.

192. Kaback, H. R., and E. R. Stadtman. 1966. Pro-line uptake by an isolated cytoplasmic mem-brane preparation of Escherichia coli. Proc.Nat. Acad. Sci. U.S.A. 55:920-927.

193. Kagawa, Y., and E. Racker. 1971. Partial reso-lution of the enzymes catalyzing oxidativephosphorylation. XXV. Reconstitution of ves-icles catalyzing 32Pi-adenosine triphosphateexchange. J. Biol. Chem. 246:5477-5487.

194. Kalckar, H. M. 1971. The periplasmic galactosebinding protein of Escherichia coli. Science174:557-565.

195. Kammen, H. 0. 1967. Thymine metabolism inEscherichia coli. I. Factors involved in utili-zation of exogenous thymine. Biochim. Bio-phys. Acta 134:301-311.

196. Kammen, H. O., and M. Strand. 1967. Thy-mine metabolism in Escherichia coli. H. Al-tered uptake of thymine after bacteriophageinfection. J. Biol. Chem. 242:1854-1863.

197. Kaplan, D. M., and R. S. Criddle. 1971. Mem-brane structural proteins. Physiol. Rev. 51:249-272.

198. Kashket, E. R., and T. H. Wilson. 1972. Role ofmetabolic energy in the transport of ,-galac-tosides by Streptococcus lactis. J. Bacteriol.109:784-789.

199. Kawasaki, T., I. Miyata, K. Esaki, and Y. Nose.1969. Thiamine uptake in Escherichia coli. I.General properties of thiamine uptake systemin Escherichia coli. Arch. Biochem. Biophys.131:223-230.

200. Kawasaki, T., I. Miyata, and Y. Nose. 1969.Thiamine uptake in Escherichia coli. HI.The isolation and properties of a mutant ofEscherichia coli defective in thiamine up-take. Arch. Biochem. Biophys. 131:231-237.

201. Kay, W. W., and A. F. Gronlund. 1969. Prolinetransport by Pseudomonas aeruginosa.Biochim. Biophys. Acta 193:444-455.

202. Keister, D. L., and N. J. Minton. 1969. Energy-linked reactions in photosynthetic bacteria.V. Relation of the light-induced proton up-take to photophosphorylation in R. rubrumchromatophores. Proc. Nat. Acad. Sci. U.S.A.63:489-495.

203. Keister, D. L., and N. J. Minton. 1971. ATPsynthesis driven by inorganic pyrophosphatein Rhodospirillum rubrum chromatophores.Biochem. Biophys. Res. Commun. 42:932-939.

204. Keister, D. L., and N. J. Yike. 1967. Energy-linked reactions in photosynthetic bacteria. I.Succinate-linked ATP-driven NAD+ reduc-tion by Rhodospirillum rubrum chromato-phores. Arch. Biochem. Biophys. 122:415-422.

205. Keister, D. L., and N. J. Yike. 1967. Energy-linked reactions in photosynthetic bacteria.II. The energy-dependent reduction of oxi-dized nicotinamide-adenine dinucleotidephosphate by reduced nicotinamide-adenine

VOL. 36, 1972 223


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

dinucleotide in chromatophores of Rhodospi-rillum rubrum. Biochemistry 6:3847-3857.

206. Kennedy, E. P. 1970. The lactose permeasesystem of Escherichia coli, p. 49-92. In J. R.Beckwith and D. Zipser (ed.), The lactoseoperon. Cold Spring Harbor, New York.

207. Kepes, A. 1971. The ,8-galactoside permease ofEscherichia coli. J. Membrane Biol. 4:87-112.

208. Kerwar, G. K., A. S. Gordon, and H. R. Ka-back. 1972. Mechanism of active transport inisolated membrane vesicles. IV. Galactosetransport by isolated membrane vesicles fromEscherichia coli. J. Biol. Chem. 247:291-297.

209. Kessel, D., and M. Lubin. 1962. Transport ofproline in Escherichia coli. Biochim. Biophys.Acta 57:32-34.

210. Khayyam, 0. 1080 (approx.). The Rubaiyat (E.Fitzgerald, translation).

211. Kimmich, G. A. 1970. Active sugar accumula-tion by isolated intestinal epithelial cells. Anew model for sodium-dependent metabolitetransport. Biochemistry 9:3669-3677.

212. Klein, K., R. Steinberg, B. Fiethen, and P.Overath. 1971. Fatty acid degradation inEscherichia coli. An inducible system for theuptake of fatty acids and further characteri-zation of old mutants. Eur. J. Biochem. 19:442-450.

213. Klein, W. L., A. S. Dahms, and P. D. Boyer.1970. The nature of the coupling of oxidativeenergy to amino acid transport. Fed. Proc.Fed. Amer. Soc. Exp. Biol. 29:341.

214. Klemme, B., J. H. Klemme, and A. San Pietro.1971. PPase, ATPase and photophosphoryl-ation in chromatophores of Rhodospirillumrubrum: inactivation by phospholipase A;reconstitution by phospholipids. Arch.Biochem. Biophys. 144:339-342.

215. Klingenberg, M. 1970. Metabolite transport inmitochondria: an example for intracellularmembrane function. Essays Biochem. 6:119-159.

216. Knobloch, K., J. H. Eley, and M. I. H. Aleem.1971. Generation of reducing power in bac-terial photosynthesis. Rhodopseudomonaspalustris. Biochem. Biophys. Res. Commun.43:834-839.

217. Knobloch, K., M. Ishaque, and M. I. H. Aleem.1971. Oxidative phosphorylation in Micro-coccus denitrificans under autotrophic growthconditions. Arch. Mikrobiol. 76:114-124.

218. Knowles, A. F., R. J. Guillory, and E. Racker.1971. Partial resolution of the enzymes cata-lyzing oxidative phosphorylation. XXIV. Afactor required for the binding of mito-chondrial adenosine triphosphatase to theinner mitochondrial membrane. J. Biol.Chem. 246:2672-2679.

219. Knowles, L., and L. Smith. 1971. The relation-ship between substrate-induced respirationand swelling in Azotobacter vinelandii.Biochim. Biophys. Acta 234:153-161.

220. Koch, A. L. 1964. The role of permease intransport. Biochim. Biophys. Acta 79:177-200.

221. Koch, A. L. 1971. Energy expenditure is obliga-tory for the downhill transport of galacto-sides. J. Mol. Biol. 59:447-459.

222. Konings, W. N., E. M. Barnes, Jr., and H. R.Kaback. 1971. Mechanisms of active trans-port in isolated membrane vesicles. III. Thecoupling of reduced phenazine methosulfateto the concentrative uptake of fB-galactosidesand amino acids. J. Biol. Chem. 246:5857-5861.

223. Konings, W. N., and E. Freese. 1971. L-serinetransport in membrane vesicles of Bacillussubtilis energized by NADH or reduced phen-azine methosulfate. FEBS Letters 14:65-68.

224. Kotyk, A., and K. Janacek. 1970. Cell mem-brane transport. Plenum Press, New York.

225. Kovac, L., and S. Kuzela. 1966. Effect of un-coupling agents and azide on the synthesis of,B-galactosidase in aerobically and anaerobi-cally grown Escherichia coli. Biochim. Bio-phys. Acta 127:355-365.

226. Kr6ger, A., and V. Dadak. 1969. On the role ofquinones in bacterial electron transport. Therespiratory system of Bacillus megaterium.Eur. J. Biochem. 11:328-340.

227. Kr6ger, A., V. Dadak, M. Klingenberg, and F.Diemer. 1971. On the role of quinones in bac-terial electron transport. Differential roles ofubiquinone and menaquinone in Proteus rett-geri. Eur. J. Biochem. 21:322-333.

228. Kundig, W., and S. Roseman. 1971. Sugartransport. I. Isolation of a phosphotransferasesystem from Escherichia coli. J. Biol. Chem.246:1393-1406.

229. Kundig, W., and S. Roseman. 1971. Sugar trans-port. II. Characterization of constitutivemembrane-bound enzymes II of the Esche-richia coli phosphotransferase system. J. Biol.Chem. 246:1407-1418.

230. Lardy, H. A., and S. M. Ferguson. 1969. Oxida-tive phosphorylation in mitochondria. Annu.Rev. Biochem. 38:991-1034.

231. LeBlanc, 0. H., Jr. 1971. The effect of uncou-plers of oxidative phosphorylation on lipidbilayer membranes: carbonylcyanide m-chlo-rophenylhydrazone. J. Membrane Biol. 4:227-251.

232. Lehninger, A. L. 1970. Mitochondria and cal-cium ion transport. Biochem. J. 119:129-138.

233. Lehninger, A. L. 1971. A soluble, heat-labile,high affinity Ca++-binding factor extractedfrom rat liver mitochondria. Biochem. Bio-phys. Res. Commun. 42:312-318.

234. Lehninger, A. L., and E. Carafoli. 1971. Theinteraction of La3+ with mitochondria in rela-tion to respiration-coupled Ca"+ transport.Arch. Biochem. Biophys. 143:506-515.

235. Lehninger, A. L., E. Carafoli, and C. S. Rossi.1967. Energy-linked ion movements in mito-chondrial systems. Adv. Enzymol. 29:259-320.

236. Lengeler, J., K. 0. Hermann, H. J. Unsold, andW. Boos. 1971. The regulation of the ,8-meth-ylgalactoside transport system and of the ga-lactose binding protein of Escherichia coli K

224 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



12. Eur. J. Biochem. 19:457-470.237. Levisohn, R., J. Konisky, and M. Nomura.

1967. Interaction of colicins with bacterialcells. IV. Immunity breakdown studied withcolicins IA and lb. J. Bacteriol. 96:811-821.

238. Liberman, E. A., and V. P. Skulachev. 1970.Conversion of biomembrane-produced energyinto electric form. IV. General discussion.Biochim. Biophys. Acta 216:30-42.

239. Liberman, E. A., and V. P. Topaly. 1968. Selec-tive transport of ions through bimolecularphospholipid membranes. Biochim. Biophys.Acta 163:125-136.

240. Liberman, E. A., V. P. Topaly, and A. Y. Sil-berstein. 1970. Charged and neutral ion car-

riers through bimolecular plgspholipid mem-

branes. Biochim. Biophys. Acta 196:221-234.241. Liberman, E. A., V. P. Topaly, L. M. Tsofina,

A. A. Jasaitis, and V. P. Skulachev. 1969.Mechanism of coupling of oxidative phos-phorylation and the membrane potential ofmitochondria. Nature (London) 222:1076-1078.

242. Lin, E. C. C. 1972. The molecular basis ofmembrane transport systems. Curr. Top.Membranes Transport, in press.

243. Lusk, J. E., and E. P. Kennedy. 1969. Magne-sium transport in Escherichia coli. J. Biol.Chem. 244:1653-1655.

244. Manno, J. A., and D. Schachter. 1970. Energy-coupled influx of thiomethylgalactoside intoEscherichia coli. J. Biol. Chem. 245:1217-1223.

245. Marquis, R. E., and P. Gerhardt. 1964. Respira-tion-coupled and passive uptake of a-amino-isobutyric acid, a metabolically inert trans-port analogue, by Bacillus megaterium. J.Biol. Chem. 239:3361-3371.

246. Massari, S., and G. F. Azzone. 1970. The mech-anism of ion translocation in mitochondria. 1.Coupling of K+ and H+ fluxes. Eur. J.Biochem. 12:301-309.

247. Massari, S., and G. F. Azzone. 1970. The mech-anism of ion translocation in mitochondria. 2.Active transport and proton pump. Eur. J.Biochem. 12:310-318.

248. McGivan, J. D., and M. Klingenberg. 1971.Correlation between H+ and anion movementin mitochondria and the key role of the phos-phate carrier. Eur. J. Biochem. 20:392-399.

249. Medveczky, N., and H. Rosenberg. 1971. Phos-phate transport in Escherichia coli. Biochim.Biophys. Acta 241:494-506.

250. Mela, L. 1969. Inhibition and activation of cal-cium transport in mitochondria. Effect oflanthanides and local anesthetic drugs. Bio-chemistry 8:2481-2486.

251. Melandri, B. A., A. Baccarini-Melandri, A. SanPietro, and H. Gest. 1970. Role of phos-phorylation coupling factor in light-depend-ent proton translocation by Rhodopseudo-monas capsulata membrane preparations.Proc. Nat. Acad. Sci. U.S.A. 67:477-484.

252. Mickelson, M. N. 1969. Phosphorylation andthe reduced nicotinamide adenine dinucleo-

tide oxidase reaction in Streptococcus agalac-tiae. J. Bacteriol. 100:895-901.

253. Milner, L. S., and H. R. Kaback. 1970. The roleof phosphatidylglycerol in the vectorial phos-phorylation of sugar by isolated bacterialmembrane preparations. Proc. Nat. Acad.Sci. U.S.A. 65:683-690.

254. Mirsky, R. 1969. Membrane proteins from Ba-cillus megaterium KM. Biochemistry 8:1164-1169.

254a. Mirsky, R., and V. Barlow. 1971. Purificationand properties of ATPase from the cyto-plasmic membrane of Bacillus megateriumKM. Biochim. Biophys. Acta 241:835-845.

255. Mitchell, P. 1961. Coupling of phosphorylationto electron and hydrogen transfer by a chem-iosmotic type of mechanism. Nature (Lon-don) 191:144-148.

256. Mitchell, P. 1963. Molecule, group and electrontranslocation through natural membranes.Biochem. Soc. Symp. 22:142-168.

257. Mitchell, P. 1966. Chemiosmotic coupling inoxidative and photosynthetic phosphoryla-tion. Biol. Rev. (Cambridge) 41:445-502.

258. Mitchell, P. 1967. Active transport and ion ac-cumulation, p. 167-197. In M. Florkin and E.H. Stotz (ed.), Comprehensive biochemistry,vol. 22. Elsevier Publishing Co., Amsterdam.

259. Mitchell, P. 1967. Translocations through nat-ural membranes. Adv. Enzymol. 29:33-87.

260. Mitchell, P. 1967. Proton translocation phos-phorylation in mitochondria, chloroplasts andbacteria: natural fuel cells and solar cells.Fed. Proc. Fed. Amer. Soc. Exp. Biol. 26:1370-1379.

261. Mitchell, P. 1968. Chemiosmotic coupling andenergy transduction. Glynn Research Ltd.,Bodmin, United Kingdom.

262. Mitchell, P. 1969. Orientated chemical reac-tions and ion movements in membranes, p.483-518. In D. C. Tostoson (ed.), The molec-ular basis of membrane functions. PrenticeHall Inc., Englewood Cliffs, N.J.

263. Mitchell, P. 1969. The chemical and electricalcomponents of the electrochemical potentialof H+ ions across the mitochondrial cristaemembrane. FEBS Symp. 17:219-232.

264. Mitchell, P. 1970. Reversible coupling betweentransport and chemical reactions, p. 192-256.In E. E. Bittar (ed.), Membranes and iontransport, vol. 1. Wiley-Interscience, NewYork.

265. Mitchell, P. 1970. Membranes of cells and or-ganelles: morphology, transport and metabo-lism. In: Organization and control in prokar-yotic and eukaryotic cells. XXth Symp. Soc.Gen. Microbiol., p. 121-166.

266. Mitchell, P. 1972. Chemiosmotic coupling inenergy transduction: a logical development ofbiochemical knowledge. J. Bioenerg., in press.

267. Mitchell, P., and J. Moyle. 1967. Acid-base ti-tration across the membrane system of rat-liver mitochondria. Biochem. J. 104:588-600.

268. Mitchell, P., and J. Moyle. 1967. Respirationdriven proton translocation in rat-liver mito-

225VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACrERIOL. REV.

chondria. Biochem. J. 105:1147-1162.269. Mitchell, P., and J. Moyle. 1968. Proton trans-

location coupled to ATP hydrolysis in rat-liver mitochondria. Eur. J. Biochem. 4:530-539.

270. Mitchell, P., and J. Moyle. 1969. Estimation ofmembrane potential and pH difference acrossthe cristae membrane of rat-liver mitochon-dria. Eur. J. Biochem. 7:471-489.

271. Mitchell, P., and J. Moyle. 1969. Translocationof some anions, cations and acids in rat-livermitochondria. Eur. J. Biochem. 9:149-155.

272. Montal, M., B. Chance, and C.-P. Lee. 1970.Ion transport and energy conservation insubmitochondrial particles. J. MembraneBiol. 2:201-234.

273. Montal, M., M. Nishimura, and B. Chance.1970. Uncoupling and charge transfer in bac-terial chromatophores. Biochim. Biophys.Acta 223:183-188.

274. Moore, C. L. 1971. Profiles of mitochondrialmonovalent ion transport. Curr. Top. Bioe-nerg. 4:191-236.

275. Mora, J., and E. E. Snell. 1963. The uptake ofamino acids by cells and protoplasts of S.faecalis. Biochemistry 2:136-141.

276. Morowitz, H. J., and T. M. Terry. 1969. Char-acterization of the plasma membrane of Mi-coplasma laidlawii. V. Effects of selectiveremoval of protein and lipid. Biochim. Bio-phys. Acta 183:276-303.

277. Mueller, P., and D. 0. Rudin. 1970. Transloca-ters in biomolecular lipid membranes: theirrole in dissipative and conservative bioenergytransductions. Curr. Top. Bioenerg. 3:157-249.

278. Mufioz, E., J. H. Freer, D. J. Ellar, and M. R.J. Salton. 1968. Membrane-associatedATPase activity from Micrococcus lyso-deikticus. Biochem. Biophys. Acta 150:531-533.

279. Mufioz, E., M. R. J. Salton, M. H. Ng, and M.T. Schor. 1969. Membrane adenosine triphos-phatase of Micrococcus lysodeikticus. Purifi-cation, properties of the "soluble" enzymeand properties of the membrane-bound en-zyme. Eur. J. Biochem. 7:490-501.

280. Nachbar, M. S., and M. R. J. Salton. 1970.Characteristics of a lipid-rich NADH dehy-drogenase-containing particulate fractionobtained from Micrococcus lysodeikticusmembranes. Biochim. Biophys. Acta 223:309-320.

281. Nagel de Zwaig, R., and S. E. Lauria. 1967.Genetics and physiology of colicin-tolerantmutants of Escherichia coli. J. Bacteriol. 94:1112-1123.

282. Nakazawa, T., R. D. Simon, J. B. Hays, and S.Roseman. 1971. Phosphorylation of a sugar-specific protein component of the lactosetransport system in Staphylococcus aureus.Biochem. Biophys. Res. Commun. 42:836-843.

283. Newton, N. A., G. B. Cox, and F. Gibson. 1971.The function of menaquinone (vitamin K2) in

Escherichia coli K 12. Biochim. Biophys.Acta 244:155-166.

284. Nierlich, D. P. 1967. Radioisotope uptake as ameasure of synthesis of messenger RNA. Sci-ence 158:1186-1188.

285. Nishimura, M., and B. C. Pressman. 1969. Ef-fects of ionophorous antibiotics on the light-induced internal and external hydrogen ionchanges and phosphorylation in bacterialchromatophores. Biochemistry 8:1360-1370.

286. Niven, D. F., and W. A. Hamilton. 1972. Themechanism of energy coupling in the activetransport of amino acids by Staphylococcusaureus. Biochem. J. 127:58P.

287. Nomura, M. 1967. Colicins and related bacteri-ocins. Annu. Rev. Microbiol. 21:257-284.

288. Nomura, M., and C. Witten. 1967. Interactionof colicins with bacterial cells. III. Colicin-tolerant mutations in Escherichia coli. J.Bacteriol. 94:1093-1111.

289. Novotny, C. P., and E. Englesberg. 1966. TheL-arabinose permease systems in Escherichiacoli B/r. Biochim. Biophys. Acta 117:217-230.

290. Ohta, N., P. R. Galsworthy, and A. R. Pardee.1971. Genetics of sulfate transport by Salmo-nella typhimurium. J. Bacteriol. 105:1053-1062.

291. Overath, P., H. U. Schairer, and W. Stoffel.1970. Correlation of in vivo and in vitro phasetransitions of membrane lipids in Escherichiacoli. Proc. Nat. Acad. Sci. U.S.A. 67:606-612.

292. Oxender, D. L. 1972. Membrane transport.Annu. Rev. Biochem. 41, in press.

293. Packer, L. 1970. Relation of structure to energycoupling in rat-liver mitochondria. Fed. Proc.Fed. Amer. Soc. Exp. Biol. 29:1533-1540.

294. Packer, L., S. Murakami, and C. W. Mehard.1970. Ion transport in chloroplasts andplantmitochondria. Annu. Rev. Plant Physiol. 21:271-304.

295. Packer, L., and M. Perry. 1961. Energy-linkedlight-scattering changes in Escherichia coli.Arch. Biochem. Biophys. 95:379-388.

296. Palmieri, F., G. Prezioso, E. Quagliariello, andM. Klingenberg. 1971. Kinetic study of thedicarboxylate carrier in rat-liver mitochon-dria. Eur. J. Biochem. 22:66-74.

297. Palmieri, F., and E. Quagliariello. 1969. Corre-lation between anion uptake and the move-ment of K+ and H+ across the mitochondrialmembrane. Eur. J. Biochem. 8:473-481.

298. Palmieri, F., E. Quagliariello, and M. Klingen-berg. 1970. Quantitative correlation betweenthe distribution of anions and the pH differ-ence across the mitochondrial membrane.Eur. J. Biochem. 17:230-238.

299. Papa, S., N. E. Lofrumento, D. Kanduc, G.Paradies, and E. Quagliariello. 1971. Thetransport of citric-acid-cycle intermediates inrat-liver mitochondria. Electrical nature andcoupling of the exchange-diffusion reactionswith proton translocation. Eur. J. Biochem.22:134-143.

300. Papa, S., N. E. Lofrumento, M. Loglisci, and E.Quagliariello. 1969. On the transport of inor-

226 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



ganic phosphate and malate in rat-liver mito-chondria. Biochim. Biophys. Acta 189:311-314.

301. Pardee, A. B. 1968. Membrane transport pro-teins. Science 162:632-637.

302. Patterson, P. H., and W. J. Lennarz. 1970.Novel protein composition of a bacterialmembrane. Biochem. Biophys. Res.Commun. 40:408-415.

303. Pavlasova, E., and F. M. Harold. 1969. Energycoupling in the transport of ,B-galactosides byEscherichta coli: effect of proton conductors.J. Bacteriol. 98:198-204.

304. Penniston, J. T., J. H. Southard, D. E. Green,and M. Luzzana. 1971. Speed of respiration-dependent proton ejection by mitochondria.Application of a pH-measuring system with10-msec resolution. Arch. Biochem. Biophys.142:638-644.

305. Pipemo, J. R., and D. L. Oxender. 1968. Aminoacid transport systems in Escherichia coli K12. J. Biol. Chem. 243:5914-5920.

306. Pressman, B. C. 1965. Induced active transportof ions in mitochondria. Proc. Nat. Acad. Sci.U.S.A. 53:1076-1083.

307. Pressman, B. C. 1969. Mechanism of action oftransport mediating antibiotics. Ann. N.Y.Acad. Sci. 147:829-841.

308. Pressman, B. C. 1970. Energy-linked transportin mitochondria, p. 213-250. In E. Racker(ed.), Membranes of mitochondria and chlo-roplasts. Van Nostrand Reinhold Co., NewYork.

309. Quagliariello, E., G. Genchi, and F. Palmieri.1971. Respiration-dependent anion uptake byrat-liver mitochondria. FEBS Letters 13:253-257.

310. Racker, E. 1970. Function and structure of theinner membrane of mitochondria and chloro-plasts, p. 127-171. In E. Racker (ed.), Mem-branes of mitochondria and chloroplasts. VanNostrand Reinhold Co., New York.

311. Racker, E. 1970. The two faces of the innermitochondrial membrane. Essays Biochem.6:1-22.

312. Reaveley, D. A., and H. J. Rogers. 1969. Someenzymic activities and chemical properties ofthe mesosomes and cytoplasmic membranesof Bacillus licheniformis 6346. Biochem. J.113:67-79.

313. Reeves, J. P. 1971. Transient pH changesduring D-lactate oxidation by membrane ves-icles. Biochem. Biophys. Res. Commun. 45:931-936.

314. Reid, K. G., N. M. Utech, and J. T. Holden.1970. Multiple transport components for di-carboxylic amino acids in Streptococcus fae-calis. J. Biol. Chem. 245:5261-5272.

315. Reid, R. A. 1970. The synthesis of adenosinetriphosphate by transmembrane ionic gra-dients. Biochem. J. 116:12P.

316. Riemersma, J. C. 1968. Effects of sodium azideand 2, 4-dinitrophenol on phosphorylationreactions and ion fluxes in Saccharomycescerevisiae. Biochim. Biophys. Acta 153:80-87.

317. Robertson, R. N. 1968. Protons, electrons, phos-phorylation and active transport. CambridgeUniversity Press, Cambridge.

318. Rogers, H. J. 1970. Bacterial growth and thecell envelope. Bacteriol. Rev. 34:194-214.

319. Romano, A. H., S. J. Eberhard, S. L. Dingle,and T. D. McDowell. 1970. Distribution ofthe phosphoenolpyruvate: glucose phospho-transferase system in bacteria. J. Bacteriol.104:808-813.

320. Rose, S. P., and C. F. Fox. 1971. The ,B-gluco-side system of Escherichia coli. II. Kineticevidence for a phosphoryl-enzyme II interme-diate. Biochem. Biophys. Res. Commun. 45:376-380.

321. Roseman, S. 1969. The transport of carbohy-drates by a bacterial phosphotransferasesystem. J. Gen. Physiol. 59:1385-1405.

322. Rossi, E., and G. F. Azzone. 1970. The mecha-nism of ion translocation in mitochondria. 3.Coupling of K+ efflux with ATP synthesis.Eur. J. Biochem. 12:319-327.

323. Rothfield, L., and D. Romeo. 1971. Role oflipids in the biosynthesis of the bacterial cellenvelope. Bacteriol. Rev. 35:14-38.

324. Rottem, S., and S. Razin. 1967. Electrophoreticpatterns of membrane proteins of Myco-plasma. J. Bacteriol. 94:359-364.

325. Rottenberg, H. 1970. ATP synthesis and elec-trical membrane potential in mitochondria.Eur. J. Biochem. 15:22-28.

326. Saier, M. H., Jr., W. S. Young III, and S.Roseman. 1971. Utilization and transport ofhexoses by mutant strains of Salmonella ty-phimurium lacking enzyme I of the phos-phoenolpyruvate-dependent phospho-transferase system. J. Biol. Chem. 246:5838-5840.

327. Salton, M. R. J. 1967. Structure and function ofbacterial cell membranes. Annu. Rev. Micro-biol. 21:417-442.

328. Sanwal, B. D. 1970. Allosteric control of amphi-bolic pathways in bacteria. Bacteriol. Rev.34:20-39.

329. Scarborough, G. A., N. K. Rumley, and E. P.Kennedy. 1968. The function of adenosine 5'-triphosphate in the lactose transport systemof Escherichia coli. Proc. Nat. Acad. Sci.U.S.A. 60:951-958.

330. Schnaitman, C. A. 1970. Examination of theprotein composition of the cell envelope ofEscherichia coli by polyacrylamide gel elec-trophoresis. J. Bacteriol. 104:882-889.

331. Schnaitman, C. A. 1970. Protein composition ofthe cell wall and cytoplasmic membrane ofEscherichia coli. J. Bacteriol. 104:890-901.

332. Schnebli, H. P., and A. Abrams. 1970. Mem-brane adenosine triphosphatase from Strepto-coccus faecalis. Preparation and homoge-neity. J. Biol. Chem. 245:1115-1121.

333. Schnebli, H. P., A. E. Vatter, and A. Abrams.1970. Membrane adenosine triphosphatasefrom Streptococcus faecalis. Molecularweight, subunit structure, and amino acidcomposition. J. Biol. Chem. 245:1122-1127.

VOL. 36, 1972 227


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

334. Scholes, P. B., G. McLain, and L. Smith. 1971.Purification and properties of a c-type cyto-chrome from Micrococcus denitrificans. Bio-chemistry 10:2072-2076.

335. Scholes, P., and P. Mitchell. 1970. Acid-basetitration across the plasma membrane ofMicrococcus denitrificans: factors affectingthe effective proton conductance and the res-piratory rate. J. Bioenerg. 1:61-72.

336. Scholes, P., and P. Mitchell. 1970. Respiration-driven proton translocation in Micrococcusdenitrificans. J. Bioenerg. 1:309-323.

337. Scholes, P., P. Mitchell, and J. Moyle. 1969.The polarity of proton translocation in somephotosynthetic microorganisms. Eur. J.Biochem. 8:450-454.

338. Scholes, P. B., and L. Smith. 1968. The isola-tion and properties of the cytoplasmic mem-brane of Micrococcus denitrificans. Biochim.Biophys. Acta 153:350-362.

339. Scholes, P. B., and L. Smith. 1968. Composi-tion and properties of the membrane-boundrespiratory chain system of Micrococcus de-nitrificans. Biochim. Biophys. Acta 153:363-375.

340. Schultz, S. G. 1969. Mechanisms of absorption,p. 59-108. In R. M. Dowben (ed.), Biologicalmembranes. Little, Brown and Co., Boston.

341. Schultz, S. G., and P. F. Curran. 1970. Coupledtransport of sodium and organic solutes.Physiol. Rev. 50:637-718.

342. Schwartz, M. 1971. The relation of ion trans-port to phosphorylation. Annu. Rev. PlantPhysiol. 22:469-484.

343. Senior, A. E., and J. C. Brooks. 1971. The sub-unit composition of the mitochondrial oligo-mycin-insensitive ATPase. FEBS Letters 17:327-329.

344. Short, S. A., D. C. White, and H. R. Kaback.1972. Active transport in isolated bacterialmembrane vesicles. V. The transport ofamino acids by membrane vesicles preparedfrom Staphylococcus aureus. J. Biol. Chem.247:298-304.

345. Silver, S. 1969. Active transport of magnesiumin Escherichia coli. Proc. Nat. Acad. Sci.U.S.A. 62:764-771.

346. Silver, S., P. Johnseine, and K. King. 1970.Manganese active transport in Escherichiacoli. J. Bacteriol. 104:1299-1306.

347. Sistrom, W. R. 1958. On the physical state ofthe intracellularly accumulated substrates of,B-galactoside-permease in Escherichia coli.Biochim. Biophys. Acta 29:579-587.

348. Skou, J. C. 1971. Sequence of steps in the (Na++ K+)-activated enzyme system in relationto sodium and potassium transport. Curr.Top. Bioenerg. 4:357-398.

349. Skulachev, V. P. 1970. Electric fields in cou-pling membranes. FEBS Letters 11:301-308.

350. Skulachev, V. P. 1971. Energy tranformationsin the respiratory chain. Curr. Top. Bioenerg.4:127-190.

351. Slater, E. C. 1966. Oxidative phosphorylation,p. 327-396. In M. Florkin and E. H. Stotz

(ed.), Comprehensive biochemistry, vol. 14.Elsevier, Amsterdam.

352. Slater, E. C. 1971. The coupling between en-ergy-yielding and energy-utilizing reactionsin mitochondria. Quant. Rev. Biophys. 4:35-72.

353. Smith, L. 1968. The respiratory chain of bac-teria, p. 55-122. In T. P. Singer (ed.), Biolog-ical oxidations. Interscience Publishers, NewYork.

354. Snoswell, A. M., and G. B. Cox. 1968. Pieri-cidin A and inhibition of respiratory chainactivity in Escherichia coli K 12. Biochim.Biophys. Acta 162:455-458.

355. Steck, T. L., G. Fairbanks, and D. F. H. Wal-lach. 1971. Disposition of the major proteinsin the isolated erythrocyte membrane. Pro-teolytic dissection. Biochemistry 10:2617-2624.

356. Steck, T. L., R. S. Weinstein, J. H. Straus, andD. F. H. Wallach. 1970. Inside-out red cellmembrane vesicles: preparation and purifica-tion. Science 168:255-257.

357. Steim, J. M., M. E. Tourtellotte, J. C. Reinert,R. N. McElhaney, and R. L. Rader. 1969.Calorimetric evidence of the liquid-crystal-line state of lipids in a biomembrane. Proc.Nat. Acad. Sci. U.S.A. 63:104-109.

358. Stein, W. D. 1967. The movement of moleculesacross cell membranes. Academic Press Inc.,New York.

359. Stock, J., and S. Roseman. 1971. A sodium-dependent sugar co-transport system in bac-teria. Biochem. Biophys. Res. Commun. 44:132-138.

360. Stockdale, M., and M. J. Selwyn. 1971. Effectsof ring substituents on the activity of phenolsas inhibitors and uncouplers of mitochondrialrespiration. Eur. J. Biochem. 21:565-574.

361. Storey, B. T. 1970. Chemical hypothesis forenergy conservation in the mitochondrial res-piratory chain. J. Theoret. Biol. 28:233-259.

362. Sweetman, A. J., and D. E. Griffiths. 1971.Studies on energy-linked reactions. Energy-linked reduction of oxidized nicotinamide-adenine dinucleotide by succinate in Esche-richia coli. Biochem. J. 121:117-124.

363. Sweetman, A. J., and D. E. Griffiths. 1971.Studies on energy-linked reactions. Energy-linked transhydrogenase reaction in Esche-richia coli. Biochem. J. 121:125-130.

363a. Thompson, J., M. L. Green, and F. C. Hap-pold. 1969. Cation-activated nucleotidase incell envelopes of a marine bacterium. J. Bac-teriol. 99:834-841.

364. Thompson, J., and R. A. MacLeod. 1971. Func-tions of Na+ and K+ in the active transportof a a-aminoisobutyric acid in a marine pseu-domonad. J. Biol. Chem. 246:4066-4074.

365. Thome, K. J. I., and D. C. Barker. 1969. Bacto-prenol, ATPase and acetate activating en-zymes of a vesicular fraction from Lactoba-cillus casei. Eur. J. Biochem. 11:582-591.

366. Thorne, K. J. I., and D. C. Barker. 1972. Theoccurrence of bactoprenol in the mesosome

228 HAROLD


http://mm

br.asm.org/

Dow

nloaded from



and plasma membranes of Lactobacillus caseiand Lactobacillus plantarum. J. Gen. Micro-biol. 70:87-98.

367. Ting, H. P., D. F. Wilson, and B. Chance. 1970.Effects of uncouplers of oxidative phosphory-lation on the specific conductance of bimolec-ular lipid membranes. Arch. Biochem. Bio-phys. 141:141-146.

368. Tourtellotte, M. E., D. Branton, and A. Keith.1970. Membrane structure: spin labeling andfreeze etching of Mycoplasma laidlawii. Proc.Nat. Acad. Sci. U.S.A. 66:909-916.

369. Tsukagoshi, N., and C. F. Fox. 1971. Hybridiza-tion of membranes by sonic irradiation. Bio-chemistry 10:3309-3313.

370. Tupper, J. T., and H. Tedeschi. 1969. Mito-chondrial membrane potentials measuredwith microelectrodes: probable ionic basis.Science 166:1539-1540.

371. Tupper, J. T., and H. Tedeschi. 1969. Microe-lectrode studies on the membrane propertiesof isolated mitochondria. Proc. Nat. Acad.Sci. U.S.A. 63:370-377.

372. Tupper, J. T., and H. Tedeschi. 1969. Microe-lectrode studies on the membrane propertiesof isolated mitochondria. II. Absence of ametabolic dependence. Proc. Nat. Acad. Sci.U.S.A. 63:713-717.

373. Tyler, D. D. 1968. The inhibition of phosphateentry into rat liver mitochondria by organicmercurials and by formaldehyde. Biochem. J.107:121-123.

374. Tyler, D. D. 1969. Evidence of a phosphate-transporter system in the inner membrane ofisolated mitochondria. Biochem. J. 111:665-678.

375. Tzagoloff, A., and P. Meagher. 1971. Assemblyof the mitochondrial membrane system. V.Properties of a dispersed preparation of therutamycin-sensitive adenosine triphosphataseof yeast mitochondria. J. Biol. Chem. 246:7328-7336.

376. Vainio, H., L. Mela, and B. Chance. 1970. En-ergy dependent bivalent cation translocationin rat liver mitochondria. Eur. J. Biochem.12:387-391.

377. Van Dam, K., and A. J. Meyer. 1971. Oxidationand energy conservation by mitochondria.Annu. Rev. Biochem. 40:115-160.

378. Vernon, L. P. 1968. Photochemical and electrontransport reactions of bacterial photosyn-thesis. Bacteriol. Rev. 32:243-261.

379. Von Stedingk, L.-V., and H. Baltscheffsky.1966. The light-induced, reversible pHchange in chromatophores from Rhodospi-rillum rubrum. Arch. Biochem. Biophys. 117:400-404.

380. Walker, D. A., and A. R. Crofts. 1970. Photo-synthesis. Annu. Rev. Biochem. 39:389-428.

381. Wallace, W., and D. J. P. Nicholas. 1969. Thebiochemistry of nitrifying microorganisms.Biol. Rev. (Cambridge) 44:359-391.

382. Watanabe, T., and A. F. Brodie. 1966. Enzy-matic formation of a phosphorylated deriva-tive of vitamin Kl. Proc. Nat. Acad. Sci.

U.S.A. 56:940-945.383. Weinbach, E. C., and J. Garbus. 1968. Struc-

tural changes in mitochondria induced byuncoupling reagents. The response to proteo-lytic enzymes. Biochem. J. 106:711-717.

384. Weiss, D. E. 1969. Energy-transducing reac-tions in biological membranes. IV. A model ofoxidative phosphorylation and ion accumula-tion in mitochondria. Aust. J. Biol. Sci. 22:1389-1412.

385. Weissbach, H., E. L. Thomas, and H. R. Ka-bach. 1971. Studies on the metabolism ofATP by isolated bacterial membranes: forma-tion and metabolism of membrane-boundphosphatidic acid. Arch. Biochem. Biophys.147:249-254.

386. Wendt, L. 1970. Mechanism of colicin action:early events. J. Bacteriol. 104:1236-1241.

387. West, I. C. 1969. The site of action of adeno-sine-5'-triphosphate on ,B-galactoside trans-port in Escherichia coli. FEBS Letters 4:69-71.

388. West, I. C. 1970. Lactose transport coupled toproton movements in Escherichia coli.Biochem. Biophys. Res. Commun. 41:655-661.

389. West, I. C., and P. Mitchell. 1972. Protonmovements coupled to the transport of ,B-galactosides into Escherichia coli. Biochem.J. 127:56P.

390. White, D. C., and P. R. Sinclair. 1971.Branched electron-transport systems in bac-teria. Adv. Microbiol. Physiol. 5:173-211.

391. Wilkins, M. H. F., A. E. Blaurock, and D. M.Engelman. 1971. Bilayer structure in mem-branes. Nature New Biol. 230:72-78.

392. Willecke, K., and A. B. Pardee. 1971. Inducibletransport of citrate in a gram-positive bacte-rium, Bacillus subtilis. J. Biol. Chem. 246:1032-1040.

393. Williams, R. J. P. 1969. Electron transfer andenergy conservation. Curr. Top. Bioenerg. 3:79-156.

394. Williams, R. J. P. 1970. Electron transfer, con-formation changes and energy conservation,p. 7-23. In M. J. Tager, S. Papa, E. Quagli-ariello, and E. C. Slater (ed.), Electron trans-port and energy conservation. Adriatica Edi-trice, Bari.

395. Wilson, D. F., M. Koppelman, M. Erecinska,and P. L. Dutton. 1971. Energy conservationin detergent-treated mitochondria and puri-fied succinate-cytochrome c reductase.Biochem. Biophys. Res. Commun. 44:759-766.

396. Wilson, D. F., H. P. Ting, and M. S. Kop-pelman. 1971. Mechanism of action of uncou-plers of oxidative phosphorylation. Biochem-istry 10:2897-2902.

397. Wilson, T. H., M. Kusch, and E. R. Kashket.1970. A mutant in Escherichia coli energy-uncoupled for lactose transport; a defect inthe lactose-operon. Biochem. Biophys. Res.Commun. 40:1409-1414.

398. Windisch, F., and W. Heumann. 1960. "Ent-

229VOL. 36, 1972


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

kopplung" von Spaltungsstoffwechsel undanoxygener Energie (ATP) Bildung. Natur-wissenschaften 47:209.

399. Winkler, H. H., and T. H. Wilson. 1966. Therole of energy coupling in the transport of ,B-galactosides by Escherichia coli. J. Biol.Chem. 241:2200-2211.

400. Wong, J. T.-F., A. Pincock, and P. M. Bron-skill. 1971. Site of energy coupling in the car-rier mechanism for ,B-galactoside transport.Biochim. Biophys. Acta 233:176-188.

401. Wong, P. T. S., J. Thompson, and R. A.MacLeod. 1969. Nutrition and metabolism of

a marine bacteria. XVII. Ion-dependent re-tention of a-aminoisobutyric acid and its re-lation to Na+-dependent transport in a ma-rine pseudomonad. J. Biol. Chem. 244:1016-1025.

402. Wong, P. T. S., E. R. Kashket, and T. H.Wilson. 1970. Energy coupling in the lactosetransport system of Escherichia coli. Proc.Nat. Acad. Sci. U.S.A. 65:63-69.

403. Young, J. H., G. A. Blondin, and D. E. Green.1971. Conformational model of active trans-port: role of protons. Proc. Nat. Acad. Sci.U.S.A. 68:1364-1368.

230 HAROLD


http://mm

br.asm.org/

Dow

nloaded from


Documents

Conservation and Transformation of Energy Bacterial Membranes · CONSERVATION ANDTRANSFORMATION OF ENERGY nical difficulties of studying largely insoluble proteins. Thetheme that