Leghemoglobin and Rhizobium Respiration - Semantic Scholar · It is misleading to describe the...

Ann. Rev. Plant Physiol. 1984.35:443-78Copyright © 1984 by Annual Reviews Inc. All rights reserved

LEGHEMOGLOBIN ANDRHIZOBIUM RESPIRATION

Cyril A. Appleby

Division of Plant Industry, Commonwealth Scientific and Industrial Research Organiza-tion, Canberra, ACT 2601, Australia

CONTENTS

INTRODUCTION ..................................................................................... 444LEGHEMOGLOBIN ................................................................................. 445

Occurrence, Purification, Sequence Analysis ................................................ 445Biosynthesis and its Control; Genetic Origin ................................................ 446Maintenance, Turnover, and Degradation ................................................... 448Leghemoglobin Structure in Relation to Oxygen Reactivity ............................... 449lntracellular Location of Leghemoglobin ..................................................... 451

LEGUME BACTEROIDS AND THE OXYGEN PARADOX ................................ 452Nitrogenase Location and Oxygen Lability: Possible Protective Mechanisms ......... 452Cytochrome Variation Among Free-living Rhizobia and Bacteroids ..................... 453Periplasmic Location of Bacteroid Oxidases? ............................................... 455

LEGHEMOGLOBIN AND RHIZOBIUM RESPIRATION EFFICIENCY ................. 457Oxygen Delivery and Other Possible Leghemoglobin Functions ......................... 457The Soybean Symbiosis: Recognition of Efficient and Inefficient Phases of Bacteroid

Respiration ........................................................................... 459Other Legume Symbioses: Evidence for Increasing Oxygen Tolerance ................. 461Respiration Substrate in Relation to Symbiotic Efficiency ................................. 466Hemoglobin in Nonlegume Symbioses ......................................................... 468Rhizobium Nitrogen Fixation Without Leghemoglobin ..................................... 470

CONCLUSIONS ...................................................................................... 471

0066-4294/84/0601-0443 $02.00

443

www.annualreviews.org/aronlineAnnual Reviews

Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

http://www.annualreviews.org/aronline

444 APPLEBY

INTRODUCTION

It is misleading to describe the principal function of leghemoglobin (Lb)1 as abarrier or scavenger which, by binding 02, prevents free O2 from reachingnitrogenase within the N2-fixing Rhizobium bacteroids of legume root nodules(39, 50). By 1974, after the first comprehensive reviews of Lb structure andproperties were published (4, 58), its function was generally accepted as thefacilitation of Oz flux to the vigorously respiring, phosphorylating, N2-fixingRhizobiurn bacteroids, albeit at a stabilized 02 tension (- 10 nM in soybeannodules) presumed to be too low to damage the O2-intolerant nitrogenaseenzyme complex. As late as 1980, however, the reported absence of Lb fromthe Nz-fixing symbiosis between Rhizobium and a nonleguminous plant Para-sponia (originally misnamed Trema) (45), and the demonstration of nitrogenrase activity in certain free living rhizobia grown under microaerobic condi-tions, led Robson and Postgate (130) to the reasonable conclusion that "leghe-moglobin is a sophistication rather than essential." Since then the perspectivehas changed dramatically. A hemoglobin (Hb) which resembles Lb in some(but not all) of its properties has been purified from Parasponia nodules (12,13), Hb has been extracted from Casuarina nodules (145; A. I. Fleming and C.A. Appleby, unpublished observations), and Hb spectra have been observed inslices of actinorhizal nodules from other plant families (145).

A principal objective of this review will be the reexamination of Lb prop-erties and function, particularly in relation to an efficient phase of Rhizobiumbacteroid respiration which undoubtedly occurs in legume root nodules atextremely low free O2 tension, and also in relation to the apparently uncoupled

~Abbreviations and definitions: The term leghemoglobin, and occasionally the abbreviationsLb, LbOz (oxyferrous Lb), LbCO (carboxyferrous Lb), are used only to describe the O2-carryinghemoprotein found in legume root nodules. Myoglobin (Mb) refers only to the monomeric proteinfound in animal muscle tissue. The term hemoglobin (Hb), with descriptive prefix, e.g. ParasponiaHb, is used to describe the proteins found in all other animal and plant tissues. "Oxygenation"describes the reversible combination of ferrous Lb, Mb, or Hb with 02; "oxidation" describes theconversion of ferrous Lb, Mb, or Hb to the corresponding ferric species, thereby rendering theseproteins incapable of oxygenation until they have been re-reduced to the ferrous species. "Bacter-oid" is used to describe the N2-fixing rhizobia found in both legume and Parasponia nodules.cccP is used to describe the uncoupling agent carbonyl cyanide m-chlorophenylhydrazone. Withsome reluctance, the author uses the recently popularized term "peribacteroid membrane" (129)instead of the classical term "membrane envelope" (25) to describe the plant membrane whichsurrounds each bacteroid or group of bacteroids in N2-fixing legume root nodules; hence the term"peribacteroid space" rather than "envelope space" is used to describe the space between theperibacteroid membrane and its contained bacteroid(s). The term "periplasmic space" is used describe the space beiween the bacteroid inner (plasma) membrane and outer membrane (cell wall).It is critically important to distinguish between the "periplasmic" and "peribacteroid" spaces; theymay have very different functions (see below). In describing membrane energetic phenomena, represents the electrical potential (or membrane potential), A IxI-I + the electrochemical potential (orproton motive force), and Aplt the pH difference, all between the bulk phases on either side of membrane; at 25°C, A~H+ = A~ - 60 ApI-I.


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


LEGHEMOGLOBIN FUNCTION 445

(protective?) respiration which is demonstrable in certain bacteroids at artifi-cially induced high O2 tension. The genetic origin of Lb, the mechanism andcontrol of its biosynthesis, and relationships among Lb and other plant hemo-globins will also be considered. If Hb does indeed seem to be a necessary part ofall natural symbioses involving rhizobia, then knowledge of the genetic originand control of biosynthesis of present plant hemoglobins may be critical for thesuccess of new genetically engineered symbioses. On the other hand, if thereexist natural or mutated Rhizobium strains which can be shown to fix N2vigorously in aerated cultures in the absence of an O2 carrier, will these be theorganisms favored for the establishment of new, simpler symbioses? Or willthe otherwise advantageous containment of endophyte at high local density in anodule or similar structure (69) cause such constraint on O~ supply as to requirethe presence of an O2 carrier similar to Lb of legume nodules?

Almost all early studies on the physiology of symbiotic N2 fixation, includ-ing those on Lb and Rhizobium respiration, were made with the experimentallyconvenient soybean nodule which, with hindsight, appears to be one of themost demanding symbioses, at least with respect to O2 supply (see below). It probable that some recent controversies, notably those on Lb localization (25,109, 129, 161) and the nature of the efficient and inefficient phases of nodulerespiration (22, 158), may have arisen because of the unrecognized but greaterO2 tolerance of other symbioses such as that between the pea and R. legumino-sarum. Except where it would cause historical injustice, the overall plan of thisreview will be to discuss the soybean symbiosis first, then consider in turn whatseem to be increasingly O2-tolerant systems. This review will assume know-ledge of the Lb literature to 1974, summarized elsewhere (4, 58). Because space limitations, some topics will be covered only by a single recent referencefrom each laboratory active in the area. As far as possible these will be to papersin which authors have summarized their own work in relation to that of others.All cited Russian papers are available in English translations.

Leghemoglobin, with very high O2 affinity resulting from an unusual com-bination of extremely fast O2 association rate and rather slow 02 dissociationrate (4), has been much investigated by those wishing to answer fundamentalquestions about Hb structure-function relationships. This subject, of necessityonly superficially treated in the present review, will be adequately coveredelsewhere (P. E. Wright, manuscript in preparation). General references con-sulted during the preparation of this review are (4, 6, 23, 52, 57, 58, 65, 69, 79,l13a, 114, 125, 130, 159, 162, 169).

LEGHEMOGLOBINOccurrence, Purification, Sequence AnalysisIt is well documented (e.g. 4) that the Lb content of most legume root nodules correlated with their N2-fixing ability. Although some mutant strains of rhizo-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

446 APPLEBY

bia have the ability to form nodules containing Lb or its apoprotein but unableto fix N2 (32, 104, 133), the complementary situation, of N2-fixing nodulescontaining no Lb, has not been recorded. The purification of Lb is generallyachieved by extraction of nodules with aqueous buffer, ammonium sulfateprecipitation, oxidation of LbO2 to ferric Lb with ferricyanide, removal ofanionic ligands such as nicotinic acid by gel filtration at slightly alkaline pH,and separation of ferric Lb isomers by anion exchange chromatography. In thisway, soybean Lb (8, 51, 58), kidney bean and yellow lupin Lb (51), broad Lb (51, 102), pea Lb (102, 156), alfalfa Lb (78), subclover Lb (143), Sesbania rostrata root and stem nodule Lb (36) have been characterized multiple molecular species. The most discriminating procedure for separatingLb species is isoelectric focusing (63,107), and in this way soybean ferric Lb shown to consist of four major components (a, cl, cz, c3) and four minorsatellite components (b, dl, d2, d3) which are N-terminal acetylation productsof the major components (165). The amino acid sequences of kidney bean Lb (100), pea Lb I (101), broad bean Lb (128), lupin Lb I and II (55), and Lb a (58) have been determined by protein sequencing procedures. The aminoacid sequences of soybean Lb cl, c2, c3, and a corrected sequence ofLb a havebeen deduced following gene sequencing (75, 166). The differences in aminoacid sequence among Lb components from a particular plant species arerelatively slight. Thus, between yellow Lupin Lb I and II (both 153 residues)the sequence difference is 13%; between soybean Lb a (143 residues) and Lb (144 residues), the most unlike of the soybean leghemoglobins, the differenceis only 8% (75,166). This compares with sequence differences of 48% betweensoybean Lb a and yellow lupin Lb I, and - 85% between soybean Lb a and theanimal globins (74). Despite these large differences, sequence alignmentstudies (74) reveal that all leghemoglobins and animal hemoglobins studiedhave conserved sequences in critical regions, and make it appear likely thatthey share the same genetic origin. This has been dramatically confirmed byrecent analysis of the soybean Lb genes and their intron structures (77; seebelow).

Biosynthesis and its Control; Genetic Origin

The first indication that Lb apoprotein might be a plant product came fromobservations that the pattern of Lb isomers produced in legume nodules wascontrolled by the plant host and not the Rhizobium symbiont (4, 52). Thisconclusion was reinforced by isolation of poly(A)-containing mRNA fromsoybean nodules, its translation in a wheat germ system to produce Lb apopro-teins (163), and the demonstration that purified Lb-cDNA prepared from suchmRNA would hybridize with soybean DNA but not with Rhizobium DNA (19,135). Since then, the laboratories of K. Marcker and D. P. S. Verma havemoved rapidly in the characterization of Lb genes and their control mecha-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


nisms. Detailed analysis of the structure of a soybean Lb gene (75) shows theamino acid coding region to be interruped by two intervening sequences(introns) at exactly the same positions found for animal globin genes, Sinceintrons have never been detected in bacterial genes, this finding denies anearlier suggestion (4) that Lb might have arisen by globin gene transfer fromRhizobium to a primitive legume host. Furthermore, since soybean Lb genes allcontain a third, central intron (40, 75, 77, 166, 167) missing from modemanimal globins but in the exact position predicted from computer analysis ofanimal globin protein structure, "the Lb gene has all the appearance of aprimitve globin gene" (75).

Besides the four genes responsible for the four major soybean Lb species a,cl, c2, c3 there have been recognized one or more Lb pseudogenes which do notappear to be transcribed. An analysis of gene organization (37, 99) suggeststhat the four Lb genes and two pseudogenes are arranged in two independentclusters in the soybean genome. On one cluster the order is 5’ Lb a, Lb Cl(pseudo Lb?), Lb c3 3’, and on the other the order is 5’ (second pseudo Lb?),Lb c2 31. Since isoelectric focusing analysis of Lb from nodule extracts (63)and in vitro translation of nodule polysomes (160) show that Lb c2 and ca areexpressed earlier than Lb a during nodule development, "it would thus appearthat soybean Lb genes are activated in the opposite order in which they arearranged on the chromosome" (37). Uninfected plant cells do not contain anytrace of Lb (12, 32, 162), so derepression of Lb genes appears to be under thecontrol of Rhizobium or of nodule tissue formed after infection of the plant byRhizobium.

The protoheme moiety of Lb is made by the Rhizobium symbiont (4, 52);since heme synthesis in isolated R. japonicum is greatly stimulated by mic-roaerobic conditions (83) similar to those expected to exist in the developingnodule, it might be supposed that heme export from nascent bacteroids to plantcytoplasm [where Lb apoprotein is known to be made (160)] could be principal stimulus for such protein synthesis. But in ineffective nodules formedby some R. japonicum mutant strains having a deficiency in heme synthesis(115), the plant cytoplasm contains almost a normal amount of Lb polypeptide.Thus, its synthesis may not be controlled by heme. Nor is the synthesis of Lbdependent on nitrogenase. Ineffective nodules formed by certain mutant rhizo-bia are reported to contain Lb (104, 133,162), and in normal, effective nodulesformed by both slow and fast-growing rhizobia, the apoprotein can be detectedprior to the appearance of nitrogenase (4, 33, 52, 160).

It seems that the synthesis of functional nitrogenase by legume nodulebacteroids is dependent on the prior appearance of Lb, with events occurring inthis probable order: increased O2 demand by the proliferating prebacteroidrhizobia causing microaerophilic conditions and consequent stimulus of rhizo-bial heme synthesis (83); export of heme through the bacterial plasma mem-brane, periplasmic space, bacterial cell wall, peribacteroid space ~ and peribac-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

43,8 APPLEBY

teroid membrane~ to the host cytoplasm (perhaps by an active process yet to bedescribed), and assembly of Lb by combination of protoheme with apoproteinsynthesized on the host endoplasmic reticulum (160); the O2 buffering andtransporting capacity of Lb then allowing the vigorous flux of 02 necessary forbacteroid oxidative phosphorylation and synthesis of nitrogenase protein at alow free 02 tension which does not damage nitrogenase (4, 52; see below).

Maintenance, Turnover, and Degradation

The presumed principal function of Lb as an O2 carrier demands that it bemaintained in vivo as ferrous Lb,~ and this presumption is confirmed byexperimental observation (4, 52, 113). In animal erythrocytes the principal reductase is an NADH-specific soluble flavoprotein, donating electrons via anintermediate carrier, cytochrome bs, to ferric Hb (73). While it is possible thatone of the soluble cytochromes b demonstrable at low concentration in extractsof unnodulated soybean roots (C. A. Appleby and Y. P. Chen, unpublishedobservations) might be similarly involved in Lb reduction, this remains un-proved.

Anaerobic suspensions of washed soybean bacteroids can reduce ferric Lb toferrous Lb, although slowly, in the absence of added "natural" carrier such ascytochrome b5 or low molecular weight carrier such as ferrocyanide or redoxdye (2). Electron microscope examination of these bacteroids (D. J. Goodchildand C. A. Appleby, unpublished observations) showed that the peribacteroidmembranes, and presumably proteins from the peribacteroid space, had beenlost during isolation. Hence it is likely that the "enzyme" responsible for suchLb reduction originated in the periplasmic space~ between the bacteroid plasmamembrane and cell wall. In contrast, Kretovich et al (90) have claimed that Lb reductase leached during 5 hr incubation of washed lupin bacteroids origi-nated in the peribacteroid space. But their incubation conditions were probablyharsh enough to leach periplasmic space contents, and no proof was offered thatthe peribacteroid membranes had remained substantially intact during thewashing procedures which preceded leaching. The liberated enzyme has beenpurified and characterized as a 30,000 mol wt NADH-specific flavoprotein (90,144), but its physiological significance as Lb reductase must remain in doubt.This is because of a very low rate of reaction with ferric Lb, only 3% of the ratewith ferric Mb and 2% of the rate with ferric cytochrome c (144). Furthermore,in lupin nodules at least (see below), all or most of the Lb is located in the hostcytoplasm and therefore not accessible to a "reductase" located in the peribac-teroid or periplasmic space. On the other hand, Tikhomirova et al (144),Borodenko et al (38), and Puppo et al (124) have extracted from legume nodulecytosol a variety of NAD(P)H-dependent enzymes, some of which are fla-voproteins or metalloflavoproteins, and all of which have very low activity asdirect ferric Lb reductase but higher activity when dyes such as 2,6-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



dichlorophenol indophenol are used as mediators. Saari (132) has purified fromsoybean nodule cytosol an interesting NADH-dependent Lb reductase whichhas reasonable activity in the absence of dye co-factors but which (L. Saari,personal communication) requires microaerobic conditions for activity andwhich may involve the formation of a "peroxide-level" intermediate complexof NADH, or enzyme or Lb. Puppo et al (123) have recently reviewed the workof their laboratory on the involvement of NADH reductase, nitrite reductase,superoxide dismutase, and indole-3-acetic acid oxidation in Lb reductionmechanisms, but it is the reviewer’s judgment that the natural mechanismwhich keeps Lb completely reduced in effective nodules (4) remains to elucidated.

The turnover rate of the two major components of Lb in pea nodules, namedLb I and Lb II, is about 2 days (34),2 much shorter than the lifetime of the zoneof effective tissue in pea nodules (85). Also, control mechanisms not yetunderstood must be responsible for the observed predominance of Lb I over LbIV in young pea nodules (156)2 but not in older nodules, and the predominanceof Lb c2, c3 in young but not older soybean nodules (55). The time scale of thesechanges in relative concentration of Lb isomers, which possibly reflects theirslightly different functional properties, is slower than the protein turnover timesrecorded by 35S incorporation experiments (34).

It is probable that very many factors, including loss of photosyntheticsubstrate and consequent failure of Lb reductase mechanisms, the appearanceof nicotinic acid or other ligands which might prevent the rereduction of ferricLb (124, 132), the accumulation of nitrite or nitric oxide which might causeheme degradation, and the appearance of active endopeptidases in senescingnodules, are responsible for the observed loss of functional Lb from suchnodules. Recent papers by Puppo et al (123), Klucas & Arp (87), and Pfeiffer al (121) include useful key references to the voluminous studies in this area.The work by Pfeiffer et al (121) is especially interesting in showing thatsoybean nodules can completely recover from stress-induced prematuresenescence. In general, it seems clear that the recognition and elimination ofgenetic factors which predispose nodules to early senescence and loss of Lbcould be responsible for a considerable increase in the efficiency of symbioticN2 fixation in legume nodules.

Leghemoglobin Structure in Relation to Oxygen Reactivity

The extremely high 02 affinity of Lb, critical for its biological function (seebelow), is the consequence of an extraordinarily fast O2 combination rateconstant and moderately slow 02 dissociation rate constant (4, 52). At pH 7-9

2Note that pea Lb IV (156, 157) is almost certainly the component named Lb II in (102) and (34).Since it represents the fourth in a series of five Lb isomers, the name Lb IV must be preferred.


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


450 APPLEBY

(20°C) the combination rate constant for soybean Lb a is 1.18 × 108M- is-1and the dissociation rate is 5.55 s-1 (11), allowing the calculation of equilib-rium dissociation constant (K’) as 47 nM dissolved 02. This is in goodagreement with a recalculated (cf 4) value of K’ for soybean Lb a (thepredominant component of mature nodules) as 44 nM dissolved 02, determinedby a direct equilibrium procedure (157). This direct equilibrium study (157)also yielded the values K’ = 66 nM 02 for soybean Lb c (the predominantcomponent of young nodules), K’ = 149 nM for pea Lb IV (the predominantcomponent of mature pea nodules), and K’ = 204 nM for pea Lb I (thepredominant component of young pea nodules). These comparative values mayprove to be very important. On the one hand, they show that the predominantLb components of very young soybean and pea nodules [Lb c and Lb Irespectively (63, 157)] have lower 02 affinities than do the predominantcomponents (Lb a and Lb IV) of mature nodules. This may be related to themore complex structure and greater 02 constraints within larger nodules. Onthe other hand, the principal Lb components of pea nodules have only one-thirdof the 02 affinity of the corresponding components of soybean nodules. Thismight be correlated directly with the greater O2 tolerance exhibited by thenitrogen-fixing Rhizobium bacteroids of pea nodules compared with soybeanbacteroids (see below).

X-ray crystal structure analyses of lupin Lb II (15) and soybean Lb a (117)have confirmed earlier suggestions (4, 52) that the heme pocket of Lb might larger than in other monomeric hemoglobins, although such structural studiescannot directly prove that a large distal pocket is principally responsible for thevery fast 02 combination rate of Lb. The extraordinary mobility of Lb distalhistidine, proved by a number of spectroscopic studies (summarized in 10, 11,15), is almost certainly one of the factors which allows the rapid approach andcombination of O2 with the heme iron of ferrous Lb. But a suggestion (10) thatthis distal histidine, by moving close over the bound 02, might be principallyresponsible for the slow 02 dissociation rate, has not been substantiated. It isnow known (11) that in soybean Lb aO2, at neutral pH, the distal histidineremains relatively free. Only as distal histidine is protonated (with pK 5.5) is stabilizing hydrogen bond formed between it and the ligated O2 (11). Forma-tion of this hydrogen bond has, as might be expected, no effect on the O2combination rate but causes a fivefold decrease of the 02 dissociation rate withconsequent fivefold increase in 02 affinity of Lb at acid pH. The physiologicalsignificance of this proton-dependent 02 binding reaction of Lb is unclear, aspH and its variation in the Lb domain of legume nodules remains unknown.Under equivalent conditions (i.e. when distal histidine is hydrogen bonded tothe ligated O2 in both LbO2 and MbO2) O2 dissociates ten times more slowlyfrom LbO2. It has been proposed that peripheral interactions between heme andnearby protein residues, by causing changes in heme electronic structure, areresponsible for the very stable Fe-Oz bond in LbO2 (10, 11).


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



Intracellular Location of Leghemoglobin

This remains one of the most controversial aspects of Lb physiology. By 1974,evidence summarized in (4) suggested that Lb might occur both inside andoutside the membrane envelopes (now renamed peribacteroid membranes1)

which surround each bacteroid or small group of bacteroids in effective legumenodules. More recently, Verma et al (summarized in 162) have concluded,from experiments involving Lb antibody interaction with the cut surfaces ofprefixed soybean nodule tissue and Lb antibody interaction with the contents ofsoybean peribacteroid membrane vesicles prepared by 150,000 × g centrifuga-tion through discontinuous hypertonic sucrose gradients, that Lb exists only inthe host cell cytoplasm and not within the peribacteroid membranes. Fromexperiments involving medium speed (16,000 × g) centrifugation throughdiscontinuous hypertonic sucrose gradients to prepare membrane-enclosedbacteroids from lupin nodules, Robertson et al (129) concluded that in lupinnodules also, Lb exists only in the host cell cytoplasm. Two recent Russianstudies (103, 109), both involving the suspect use of frozen-thawed nodulematerial, used simple centrifugation in 0.3 M sucrose, phosphate buffer toprepare lupin bacteroids apparently retaining their peribacteroid membranes.One study (103) reported the presence of Lb in peribacteroid space, the other(109) its absence.

By a procedure involving the gentle crushing of sliced fresh soybean nodulesin an osmotically balanced medium followed by gentle (160 × g) centrifuga-tion and washing in similar media without the use of hypertonic sucrosegradients, Bergersen & Appleby (25) isolated apparently intact peribacteroidmembranes. These could be ruptured, releasing Lb, merely by recentrifugationin the same osmoticum at 10,000-20,000 × g for 10 min or more efficiently byresuspending them in hypotonic buffer before 10,000 × g recentrifugation.When ~31I-labeled Lb had been added to the isolation medium just beforecrushing the sliced nodules, the observation of a parallel decline in radioactiv-ity and Lb content during five washes in iso-osmotic medium, followed by asubstantial release of unlabeled Lb after peribacteroid membrane rupture,confirmed the compartmentation of Lb within soybean nodules. The concentra-tion of Lb in the free space within soybean peribacteroid membranes wascalculated as 200-500 IxM, and the Lb undoubtedly present in the relativelysparse host cell cytoplasm of soybean symbiotic tissue was calculated as - 3mM (25). From the estimated relative volumes of plant cytoplasm and peribac-teroid space in these soybean nodules it was calculated that about one-third ofthe total Lb was present within the peribacteroid space (25). It is the reviewer’sopinion that the overall findings of this study (25) cannot be dismissed asserting that "minor quantities of Lb observed inside the peribacteroid mem-branes appear to be due to some damage in the membrane and artifacts ofdetection techniques (162)."

It is, of course, entirely possible that the relatively large peribacteroid


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


452 APPLEBY

membrane structures of soybean symbiotic tissue, each enclosing severalbacteroids in abundant free space (69), do contain Lb, and that lupin peribacte-roid membranes, enclosing only one bacteroid in very little free space (69), not "require" Lb presence. This topic, the subject of recent vigorous discussion(6), will be considered further below.

With the low molecular weight protoheme prosthetic group of Lb beingsynthesized in the bacteroid (4, 52), and the - 16,000 mol wt apoproteinsynthesized on the rough endoplasmic reticulum of the host cytoplasm (160), is reasonable to assume that heme export to the cytoplasm precedes theassembly and substantial residence of Lb there. It does not appear that theacetylated minor species of soybean Lb represent the portions of the majorcomponents which had been modified for transport across the peribacteroidmembrane (165). Isoelectric focusing analysis (25) showed that soybean present in the host cytoplasm and the peribacteroid space had the same relativeproportion of each major to minor acetylated component.

LEGUME BACTEROIDS AND THE OXYGEN PARADOX

Nitrogenase Location and Oxygen Lability: Possible ProtectiveMechanisms

The incompatibility between free O2 and the Fe and MoFe protein componentsof nitrogenase is well documented (130). Not only does O2 cause rapid,irreversible oxidation of nitrogenase metal-S centers, it also represses thesynthesis of nitrogenase proteins (134). Until recently, with nitrogenaseapparently located in the cytoplasm of rhizobia (122), resolution of the 02-requirement paradox of this obligately aerobic N2 fixer seemed relativelysimple. First, mechanical 02 diffusion barriers were seen as contributing to O2restriction in nodules (130), although there was disagreement with respect the presence (26) or absence (150) of continuous passages for air conductionthrough the cortex to the well defined network of air tubules existing within thesymbiotic tissue (26). Second, the presence of Lb seemed to permit the rapidtransport and even distribution throughout the nodule of this mechanicallyrestricted 02 supply, with final delivery to the bacteroid surface at a very lowfree O2 tension stabilized in the O2-unloading region of LbO2 (4, 52, 141, 169).Thereafter it has been supposed that efficient bacteroid respiration coupled toATP production at low 02, and inefficient (uncoupled) protective respiration higher 02, by oxidases located on or in the bacteroid plasma membrane (5, 22)might permit no penetration of 02 to the bacteroid cytoplasm, the apparentnitrogenase domain.

The 02 paradox, however, has been recently revived, with suggestions ofstructural or functional interaction between nitrogenase and plasma membrane


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



in Azotobacter and Rhodospirillum (72, 158) and perhaps also in Rhizobium(158). In aerobic bacteria the principal terminal oxidases, the O2-consumingenzymes, are thought to be located on the inner (cytoplasmic) surface of theplasma membrane (79, 114), the newly postulated site of functional nitrogen-ase.

There are several possible ways of resolving the paradox. First and mostsatisfactory would be direct demonstration that purified nitrogenase is notdestroyed when incubated, in the presence of partly oxygenated Lb, at 3-10 nMfree, dissolved 02, the probable maximal concentration (4, 24, 29) in theperibacteroid space or host cytoplasm of soybean nodules. With the likelihoodthat final diffusion barriers would result in an even lower concentration on theinside surface of the bacteroid plasma membrane, it would then not matter ifOz-consuming oxidases and O2-intolerant nitrogenase were located side by sideon this inner surface. A second resolution of the paradox would involvedemonstration that in efficient bacterial respiration, the first steps of oxygenreduction occurred in the periplasmic space~ or on the outer surface of thebacteroid plasma membrane, with the result that OE did not penetrate to theplasma membrane inner surface. The third resolution would involve demon-stration that if uncoupled or coupled respiration did occur via oxidases conven-tionally located on the inner surface of the bacteroid plasma membrane, the 02combination and turnover rates of the oxidases might be so much faster than theO2 combination and degradation rates of nitrogenase and its accessory proteinsthat the latter would be infrequent events. Consequently, the remainder of thissection will summarize the voluminous evidence concerning cytochromes andother putative respiratory enzymes in rhizobia and the scanty evidence abouttheir location and function. Later sections of the review will explore the ideathat efficient, Lb-facilitated bacteroid respiration, at extremely low free 02, is acharacteristic feature of every natural symbiosis involving Rhizobium.

Cytochrome Variation Among Free-living Rhizobia andBacteroids

Perhaps the earliest significant biochemical difference recognized betweenair-grown rhizobia and the bacteroids from N2-fixing soybean root nodules wasthe presence of the classical oxidase cytochrome aa3 (3, 7, 52, 155) andcytochrome o (3, 52) in air-grown cells and their absence from bacteroids. japonicum bacteroids were characterized by an approximate doubling of theamount of cytochrome c found in air-grown cells and appearance of theCO-reactive (and therefore putative oxidase) cytochromes c-552 and c-554,and cytochromes P-450, P-428 (cytochrome ax?), and P-420 (2, 47,155). whereas the function of cytochromes aa3 and o as terminal oxidases in air-grown rhizobia was easily proved by photochemical action spectra for the reliefof CO-inhibited respiration (3) at - 70 IxM dissolved 02, bacteroid


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


454 APPLEBY

respiration was insensitive to CO under the same (70 ixM 02) experimen-tal conditions (2). As will be discussed below, it is now realized that bacte-roid respiration is almost completely inhibited by CO at the very low O2tensions where LbO2 would act as an 02 buffer if it were present (28, 29).Hence, these alternative (oxidase?) cytochromes might play a very significantpart in R. japonicum bacteroid respiration at low (physiological) 02 concentra-tions.

Kretovich et al (89, 105) observed similar changes in cytochrome composi-tion between air-grown R. lupini and the bacteroids from effective lupinnodules, namely increases in bacteroid cytochromes c-552 and P-450 asnodules matured, and the absence of cytochrome P-450 and retention ofcytochromes aa3 in bacteria from ineffective nodules which lacked Lb. More-over, since Ching et al (44) observed a parallel increase in nitrogenase activityand cytochrome P-450 content and a corresponding decline in cytochrome aa3in maturing soybean bacteroids isolated by density gradient centrifugation, itseemed that a common pattern had emerged, with loss of the (unwanted?)cytochrome aa3 and o oxidases which functioned at high 02, and the appear-ance in bacteroids of new hemoproteins which might have a specific function ineffective nodules in the presence of leghemoglobin. More recently, however,Keister et al (81) have found that the oxidase cytochromes aa3 and o persist inthe bacteroids of effective soybean nodules formed by R. japonicum strain61 A76. Also, some R. japonicum strain 61 A76 mutants which lack cytochromeaa3 grow about as well as the wild type in aerobic culture but are mostlysymbiotically ineffective (59).

In the bacteroids ofR. leguminosarum strain 96 isolated from effective broadbean (Vicia faba) nodules, cytochrome P-450 seems to be entirely absent,being replaced by a high concentration of CO-reactive protohemoprotein withSoret peak near 424 nm (92). In the R. leguminosarum strain PRE bacteroidsisolated from effective pea (Pisum sativum) nodules, cytochrome P-450 canstill be detected at low concentration, but a new CO-reactive protohemoproteinresembling that found in broad bean bacteroids also appears (9). Although thisR. leguminosarum bacteroid protohemoprotein is nominally classified with thecytochrome o found in other air-grown rhizobia (9), C. A. Appleby (unpub-lished observations) finds that it very easily forms an 02 complex and may haveother properties between those of a hemoglobin and oxidase. Later in thisreview it will be proposed that the greater O2 tolerance of pea bacteroids (incomparison with soybean bacteroids) might be related to the presence of thishemoprotein. This "cytochrome o-like" protohemoprotein does not, however,seem to be a ubiquitous component of the more O2-tolerant N2-fixing rhizobia.When the cowpea Rhizobium strain 32H1 is grown in pure culture in amicroaerobic environment to produce N2-fixing cells, the cytochrome aa3


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



oxidase is retained, cytochrome o becomes low, and cytochrome P-450 isinduced (9, 62).

The study of de Hollander & Stouthamer (49,142), although relating only the non-Nz-fixing air-grown form ofR. trifolii, is extremely interesting insofaras it shows a branching point for respiration located at cytochrome b-555.Apparently, electrons can flow directly from cytochrome b-555 to cytochromeb-561 (cytochrome o?), an oxidase with moderately high affinity for Oz, butthis pathway lacks the site III of oxidative phosphorylation. Electron flow tocytochrome a-597 (cytochrome aa3?), an oxidase with lower O2 affinity,proceeds via cytochrome c-549 and seems to involve three sites of oxidativephosphorylation. These two branches of the R. trifolii respiratory chain, withdiffering phosphorylating efficiency, also have differing inhibitor sensitivity.

Periplasmic Location of Bacteroid Oxidases?

With the possibility that Rhizobium nitrogenase is associated with the innersurface of the bacteroid plasma membrane (72, 158), evidence from inhibitionand other studies of branched electron transport pathways of varying phos-phorylation efficiency in bacteroids (2, 3, 5, 14, 28, 49, 52, 126, 142), anddemonstration of a richly varied pattern of putative O2-reactive hemoproteins inrhizobia according to Oa environment (see above), it becomes important know their spatial relationships. An attractive possibility, mentioned above, isthat the terminal oxidase of the principal, efficiently phosphorylating respira-tion pathway of bacteroids might be located in the bacteroid periplasmic spaceor on the outer periplasmic surface of the plasma membrane, remote from theO2-intolerant nitrogenase enzyme complex. Periplasmic enzymes are bestdefined operationally on the basis of their release during spheroplast formation(20). Unfortunately, although spheroplasts have been prepared from free-livingR. japonicum (42) and periplasmic enzymes liberated from free-living leguminosarum by lysosyme/EDTA treatment of osmotically protected cells(68), there has yet been no report of successful spheroplast formation fromRhizobium bacteroids. Consequently, one must rely on inferential evidence forbacteroid oxidase location. The principal CO-reactive (and hence putativeoxidase) hemoproteins of R. japonicum are cytochrome P-450, cytochromec-552, and cytochrome c-554 (2, 47). From bacteroids disrupted by increasing-ly harsh sonic oscillation or French press treatment (2; C. A. Appleby, unpub-lished observations) cytochrome P-450 is released into solution a little morereadily than is the major, CO-unreactive cytochrome c-550, and cytochromec-552 is released a little less readily than cytochrome c-550. Since the majorcytochrome c of other aerobic bacteria is usually loosely attached to theperiplasmic side of the bacteroid plasma membrane (79), these observations


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


456 APPLEBY

are compatible with, but by no means prove, the concept that the bacteroidcytochromes c-552 and P-450 are located in the periplasmic space, withcytochrome c-552 being more firmly attached to the periplasmic membranethan is P-450.

The efficient, phosphorylating phase of R. japonicum bacteroid respirationwhich occurs at low 02, but not the uncoupled respiration at high 02, isinhibited by N-phenylimidazole, a specific inhibitor of cytochrome P-450 (5,14, 28), and it has been suggested (14) that bacteroid cytochrome P-450 mightreduce O2 to the level of peroxide, which becomes further reduced by a specificperoxidase. While there is no precedent for P-450 involvement in a phosphory-lating respiration pathway, in yeast a cytochrome c peroxidase located in themitochondrial intermembrane space [the equivalent functional space to bacte-rial periplasmic space (79, 114)] has been shown to be linked with the sites and II of oxidative phosphorylation (61). Since lupin bacteroids contain readily solubilized cytochrome c peroxidase (126), and the CO-reactivecytochromes c-552 and c-554 of soybean bacteroids have spectral propertiesresembling those of a known bacterial cytochrome c peroxidase (138), Appleby(5) tried, but without success, to demonstrate cytochrome c peroxidase activityin purified preparations of bacteroid cytochromes c-552 and c-554.

Soybean bacteroid plasma membranes also contain an unusual cytochrome coxidase which may have high-potential flavin as a prosthetic group (5, 116).This oxidase can be partly solubilized by bacteroid rupture in a French press inthe absence of detergent (5), suggesting that it is an extrinsic protein, perhapsfrom the peribacteroid space, rather than an intrinsic membrane protein. Thisoxidase is insensitive to CO (5) and has other inhibition characteristics (5, 116)which seemingly relate it to the oxidase that mediates the uncoupled (protec-tive?) phase of R. japonicum bacteroid respiration occurring at high 02 (5, 14,28). In disrupted cells of air-grown, non-N2-fixing R. japonicum (3) and strainswhich retain the oxidase cytochromes aa3 and o when developing symbioticallyas effective bacteroids (80, 81), these two cytochromes remain firmly bound plasma membrane fragments. They may, therefore, be assumed to be intrinsicproteins, probably exposed on the interior, cytoplasmic side of the plasmamembrane as in other aerobic bacteria (79, 114). In contrast, the cytochromeo-like hemoprotein which persists in the Oz-tolerant (see below) bacteroids R. leguminosarum (9, 92) is very readily solubilized (92) and exhibits Hb-likeproperties including apparent ability to form an 02 complex (C. A. Appleby,unpublished observations). The concept of an 02 cartier in the periplasmicspace of bacteroids is an attractive one (169); hence it becomes important know, from successful spheroplast formation, or procedures such as antibodylabeling, the location of this R. leguminosarum bacteroid "Hb-like" proteinand, indeed, of all bacteroid putative oxidases, on the periplasmic or cytoplas-mic sides of the bacteroid plasma membrane.


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



LEGHEMOGLOBIN AND RHIZOBIUM RESPIRATIONEFFICIENCY

Oxygen Delivery and Other Possible Leghemoglobin Functions

The demonstration of extremely fast kinetics for the association reactionbetween O2 and ferrous Lb and moderately fast dissociation kinetics, bysatisfying some formal requirements of the process, led to the suggestion thatLb facilitated the diffusion of O2 to respiting Rhizobium bacteroids in anenvironment of very low mean O2 pressure (170). At the same time, it wassuggested that Lb-facilitated O2 diffusion might serve to flatten out the gradientof nodule O2 pressure to make it everywhere nearly the same (170). In modelexperiments which followed directly from this kinetic study, the N2-fixingefficiency of respiring R. japonicum bacteroids (30, 171) and R. lupini bacter-oids (18, 108) was dramatically increased by the presence of part-oxygenatedferrous Lb or other O2-carrier proteins having fast kinetics and high 02 affinity.Since the addition of catalase, peroxidase, or superoxide dismutase (171) didnot reverse this stimulatory effect, it seemed not to involve the formation ofperoxide or superoxide as active intermediates, at least in the extra-bacteroidalenvironment accessible to these soluble proteins. Since the substitution of ferricLb for part-oxygenated ferrous Lb virtually abolished the stimulatory effect onR. japonicum bacteroids (171), it was then concluded that a cyclic electrontransfer process involving ferric and ferrous Lb was not a significant physiolog-ical role of this hemoprotein.

It is possible that the presence of superoxide dismutase and peroxidase innodule cytosol (123) is significant for the protection of Lb against accidentalautoxidation. Nevertheless, it is the reviewer’s prejudice that the demonstra-tions of pseudoperoxidatic activity (123, 136) and indoleacetic acid oxidaseactivity (118, 123) of purified Lb do not reflect significant physiologicalactivities of this hemoprotein. Nor is it likely that the ability of Lb to combinewith CO, or NO produced by nodule nitrite metabolism, represents an "in-tended" function (67) of Lb, any more than the intended function of animal is to protect the body against inhaled or ingested or metabolically producedtoxins.

Wittenberg (169) has found in soybean nodule cytosol an iron protein familyin concentration equal to that of Lb, and suggested that it might have a role inelectron transfer between Lb and the bacteroid surface. Since ferric Lb added torespiting bacteroids did not significantly stimulate respiration or N2-fixingactivity (171), the ferric valence state of Lb is unlikely to be part of the electrontransfer process proposed by Wittenberg (169). Following the demonstration direct and reversible formation of Lb IV from Lb II (ferrous Lb) (17), J. Wittenberg (personal communication) has put forward the idea that this newiron protein might be involved in an Lb II~--Lb IV valence change process.


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


458 APPLEBY

Although the model experiments which seemingly establish O2 transport asthe dominant function of Lb were made with soybean bacteroids (stripped freeof their peribacteroid membranes) in intimate contact with purified part-oxygenated Lb (30, 171), the most reliable recent evidence suggests that lupin nodules (with one bacteroid in each small peribacteroid membrane sac)all Lb is in the plant cytoplasm and none in the peribacteroid space (129). Evenin soybean nodules (25), with several bacteroids in each rather large peribacter-oid membrane sac, a tenfold higher concentration of Lb is demonstrated inplant cytoplasm compared with peribacteroid space. One has to enquirewhether this substantial separation of Lb from the bacteroid surface invalidatesthe whole concept of facilitated O2 diffusion as the function of Lb; the availableevidence suggests that it does not.

By using some of the equations worked out for theoretical and model systemsand inserting measured physical quantities of the R. japonicum/soybean sym-biosis, including 0.35 mM Lb in the peribacteroid space and 3.76 mM Lb in thehost cytoplasm, Bergersen (24) calculated that with 20% oxygenation cytoplasmic Lb (4) and N2-fixing bacteroids respiting at the rate observed model experiments, the steady state concentration of free, dissolved 02 at thebacteroid surface would be about 7 nM. This is very close to the measured Ozdissociation constant for the principal, efficient oxidase of washed R. japoni-cum bacteroids respiting in part-oxygenated purified Lb (29). In earlier calcula-tions (22), Bergersen had estimated that if Lb were present only in hostcytoplasm and not in the peribacteroid space, a slightly greater fractionaloxygenation of Lb would have been needed to sustain the same bacteroidrespiration rate as calculated with Lb in both spaces. If Lb were present only inthe peribacteroid space, its fractional oxygenation would need be a little lessthan if present in both spaces (22). But this last calculation did not take intoaccount the very large problem of Oz transfer from the network of nodule airtubules (26, 150) via unfacilitated diffusion through host cytoplasm if Lb werenot present, to the outside surfaces of the many peribacteroid membranes ineach cell of the symbiotic tissue. From the data presented in table 15 and figure24 of (24) it can be calculated that, with 10 nM free, dissolved Oz at theperibacteroid membrane surface, there would be 70 IxM Lb-bound 02 (aconcentration 7000 times greater) in the peribacteroid space, and 752 ~MLb-bound O2 (a 750,000 times greater concentration than of free, dissolved 02)in the host cytoplasm. These figures give simple meaning to the concept ofLbO2 as an 02 store, transporter, buffer, and stabilizer of delivered 02 concen-tration in the O2-unloading region of this hemoprotein (4, 22, 24, 52, 141,169).

Implicit in the above discussion is an assumption that O2 supplied by Lb isconsumed principally by the N2-fixing bacteroids, but Robertson et al (128a)have recently suggested that Lb is concerned also with 02 supply to the (very


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



much less abundant) host cell mitochondria. Before this suggestion can betaken seriously it will be necessary to demonstrate that mitochondria isolatedfrom legume nodule symbiotic tissue can respire efficiently in closed systemscontaining part-oxygenated Lb as sole O2 source. It is the use of such ex-perimental systems (8, 27, 29) which has established that the terminal ox-idase(s) ofR. japonicum bacteroids can function efficiently at 5-10 nM free O2,the lower end of the O2-unloading region of soybean Lb. By luminescenceprocedures, Km(02) for the oxidase of isolated rat liver mitochondria has beenestimated as 380 nM and Km(O2) for pigeon heart mitochondria as 30-270 (79a). These higher values, and the repeated observation that the mitochondriaof mature symbiotic tissue have all migrated to the cell periphery (e.g. 128a),especially to regions adjacent to air tubules and interstitial cells where O2concentration is supposed to be highest (24, 26), suggest that such mitochon-dria might not respire efficiently in the O2-unloading region of Lb.

The Soybean Symbiosis: Recognition of Efficient andInefficient Phases of Bacteroid Respiration

The first positive evidence of a role for Lb in legume nodule respiration, whichemphasized the importance of defined 02 environment, came from Tjepkema& Yocum (149). With very thin slices of soybean nodules suspended in medium containing only 0.5 I~M free, dissolved 02, respiration was halvedwhen Lb was inactivated as LbCO. Later, in contrast to experiments where thegross respiration of clover nodules in air (137) or soybean nodules in 20% 02,80% argon was unaffected by CO, Bergersen et al (30) found that the nitrogen-ase (H2-evolving) activity of these same soybean nodules in 20% 02, 80%argon was almost completely inhibited by 1% CO. Controls showed thenitrogenase (H2-evolving) activity of washed, Lb-free bacteroids to be un-affected by CO, so the inhibitory effect of CO on whole nodule nitrogenaseactivity was ascribed to blockage of a specific Lb-mediated process (30).Furthermore, with R. japonicum bacteroid suspensions shaken under 02-limiting conditions, the addition of Lb (and its maintenance in part-oxygenatedform by control of shaking or stirring rate) caused a very great increase innitrogenase activity for a relatively modest increase in total respiration (30).Since other high-affinity O2-carder proteins (not all of them hemoglobins)could be substituted for Lb to cause a similar stimulatory effect (171), original interpretation in terms of a specific terminal oxidase in the bacteroidmembrane which reacted directly with LbO2 (30) was replaced by the hypoth-esis that Lb exerted its effect by facilitating the diffusion of free 02 to thebacteroids (171). In the absence of LbO2, stimulation of total respiration increased gaseous 02, or increased shaking rate, was largely ineffective insupporting nitrogenase activity (171).

To accommodate these experimental findings it was proposed that R. japoni-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

460 APPLEBY

cum bacteroids contained two terminal oxidase pathways of differing O2affinity (171). The one which operated at Lb-buffered 02 tension was supposedto be efficient in terms of ATP production and consequent nitrogenase activity;the oxidase operating at unbuffered 02 was supposed to be inefficient. It wasconfirmed experimentally that cellular ATP level, ATP/ADP ratio, and nitro-genase activity all rose when LbO2 was added to respiring suspensions of R.japonicum bacteroids, and that increased gaseous 02 tension was much lessefficient than LbO2 in stimulation of bacteroid ATP production (14). More-over, since N-phenylimidazole, an inhibitor of cytochrome P-450, had aspecific effect on the efficient phase of bacteroid respiration it was supposedthat this R. japonicum bacteroid cytochrome (see above) might be a componentof the efficient, but not of the inefficient, respiration pathway. It is possible thata cyanide-sensitive metalloflavoprotein (5, 116) is the terminal oxidase of theinefficient respiration pathway.

The development of closed experimental systems involving spectrophoto-meter cuvettes in which LbO~, MbO2, or other oxygenated Hb acted as O2carrier and also as reporter of free O2 concentration (8, 27, 29), and of electrode systems with low background currents (28, 29) enabled the quantita-tive measurement of free O2 levels during the efficient and inefficient phases ofR. japonicum bacteroid respiration. A high-affinity oxidase system, sensitiveto N-phenylimidazole and CO, was most active when dissolved O2 was be-tween 10 and 100 nM. At > 100 nM 02, this oxidase had little activity and O2was consumed by a low-affinity system insensitive to these inhibitors (28).Measurements made under non-steady-state conditions suggested that the high-affinity pathway produced up to five times greater bacteroid ATP levels (andconsequently higher ATP/ADP ratio and nitrogenase activity) than did theoxidase operating above 1000 nM free O2 in the presence of fully oxygenatedLb, part oxygenated Mb, or in the absence of O2 carrier (28). It has beensuggested (5, 28) that this uncoupled phase of R. japonicum bacteroid respira-tion at high Oz represents a protective respiration capacity similar to that whichprotects Azotobacter nitrogenase at high external 02 concentration (79). Sincethe average oxygenation of Lb in mature soybean nodules in air is 0.2 or less,corresponding to a free O2 concentration of 10 nM or less (4, 24), perhaps it only in very young nodules or in the vicinity of air tubules in mature nodules(see above) that local O2 concentration might rise above Lb-saturating levels(~ 500 nMOz for soybean Lb) and provoke protective respiration. But it known that intact mature soybean nodules can adapt to altered rhizophere pO2(46), and it has been suggested (46, 169) that adaptation results from return optimal internal O~ pressure by adjustment of the protective respiration rate.

A recent measurement of R. japonicum bacteroid respiration (29) revealedfour separate O2 affinity states or four separate terminal oxidases. The one oflowest affinity (K,~ - 1400 nM 02) was insensitive to CO and almost insensio


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



tive to N-phenylimidazole. The one of highest affinity (Kin -- 5 nM 02) was

most sensitive to these inhibitors and the other oxidases (Km ~ 20 nM; Km ~200 nM) showed intermediate inhibition characteristics. The highest affinityoxidase (Km ~ 5 nM 02) was most efficient in terms of nitrogenase activity perunit of 02 consumed, and calculations of O2 gradients during Lb-facilitated 02diffusion (24; see above) suggest that this might be the dominant oxidase undernormal, physiological conditions, where O2 concentration at the bacteroidsurface is no more than 7-8 nM. As will be described in more detail below, theisolated bacteroids from other legume symbioses also seem to be most effi-cient, in terms of nitrogenase activity, at O2 concentrations approximatelycorresponding to the O2-unloading region of Lb. What seems to be in dispute isthe extent to which uncoupling of respiration from phosphorylation accompa-nies and/or controls the cessation of nitrogenase activity invariably observed athigher O2.

In a brief study which measured respiration-driven proton translocation(rather than ATP, ADP, and AMP levels) of aerobically isolated R. japonicumbacteroids, Ratcliffe et al (127) reported H+/O ratios of 5-7 for succinate-stimulated or endogenous respiration at unstated but probably high (20-90 ~M)dissolved O2 in the absence of Lb. In accordance with the now questionable(168) conventions which rigidly equate proton translocation and oxidativephosphorylation in aerobic organisms (114), these authors calculated (127) P/O ratio of about 3 for R. japonicum bacteroid respiration at high O2, a resultin stark contrast with those of the Canberra group (14, 28). Also, aftermeasurements which showed the maintenance of high ATP/ADP ratios at highO2 in R. leguminosarum bacteroids (see below), Laane et al (94) claimed the apparent lowering of ATP/ADP in R. japonicum bacteroids must have beenthe result either of insufficient time for establishment of equilibrium conditionsor of artifactual uncoupling of phosphorylation by unsaturated fatty acidperoxidation in R. japonicum membranes in the absence of bovine serumalbumin as protective agent. These claims were vigorously refuted byBergersen (22), who had obtained the same low ATP/ADP ratios in long-termand short-term experiments with R. japonicum bacteroids at high O2, as will bediscussed below.

Other Legume Symbioses: Evidence for Increasing OxygenToleranceTHE LUPIN SYMBIOSIS The overall cytochrome pattern of R. lupini bacter-oids from lupin nodules is very similar to that ofR. japonicum (soybean nodule)bacteroids, with the cytochromes aa3 and o of air-grown cells being replaced bycytochromes c-552 and P-450 (89, 105). Moreover, since a number of inhibi-tors of cytochrome P-450 suppressed R. lupini bacteroid nitrogenase activity toa greater extent than O2 uptake (91), it appeared that this cytochrome


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


462 APPLEBY

might be part of an efficient respiratory pathway in R. lupini bacteroids similarto that proposed for R. japonicum bacteroids (5, 14, 28). There are, on the otherhand, a number of indications that R. lupini bacteroids are more O2-tolerantthan are R. japonicum bacteroids. First, the Kretovich group routinely isolatesN2-fixing R. lupini bacteroids under aerobic conditions (18) which completelydestroy the Nz-fixing activity of soybean bacteroids (24). Second, whereas theaddition of MbO2 (which buffers 02 supply at a higher free O2 concentrationthan does LbO2) to respiring R. japonicum bacteroids stimulates nitrogenaseless than does addition of LbO2 (27,171), the nitrogenase activity ofR. lupinibacteroids was stimulated to a greater extent by MbO2 than by LbO2 (18,108).Third, whereas the addition of specific respiration pathway inhibitors to N2-fixing R. japonicum bacteroids allows the facile delineation of efficient,high-O2-affinity (LbO~-facilitated) and inefficient, low-affinity, unfacilitatedrespiration pathways (5, 14, 28), similar experiments with R. lupini bacteroids(76, 108) could not distinguish between LbO2-facilitated and unfacilitatedrespiration. In the particular case of experiments designed to prove that LbO2 isnot involved in 02 supply to aR. lupini bacteroid respiration pathway includingcytochrome P-450 (76), the results may be criticized on the grounds that thequite high concentrations of inhibitors used, including amines to 29 mM, mighthave had unspecific effects.

THE PEA AND FRENCH BEAN SYMBIOSES: IS MEMBRANE POTENTIAL OR

ATP/ADP RATIO THE PRINCIPAL FACTOR CONTROLLING NITROGENASEACTIVITY? As indicated in the Introduction, the convenience of the soybeannodule as an experimental object has led to the dominance of this symbiosis inmetabolic studies of legume N2 fixation with perhaps a too-easy assumptionthat all other symbioses would have similar characteristics. A vivid demonstra-tion that they do not comes from the studies by Veeger et al (94-97, 158) energy transduction in N2-fixing R. leguminosarurn bacteroids. After initialdifficulties, resulting from the uncoupling effects of contaminating unsaturatedfatty acids on membrane-associated energetic,processes of washed bacteroids,were overcome by the addition of bovine serum albumin, this group was able toshow that ATP/ADP ratios remained high at dissolved O2 concentrations (>1000 riM) which caused the switch-off of nitrogenase activity (94). In leguminosarum bacteroids, then, the nitrogenase switch-off at high 02 is notprimarily a result of declining ATP/ADP ratio, as has been claimed for R.japonicurn bacteroids (14, 28). But it is not correct, as imputed by Laane et (94), that the apparent decline in R. japonicum bacteroid ATP/ADP at high 02is an artifact, traceable either to fatty-acid-induced membrane energetic uncou-pling at high 02 or to non-steady-state conditions preventing the attainment ofmeaningful ATP/ADP ratios at the beginning (high 02 region) of short-termexperiments. Bergersen (22) points out that constant, low ATP/ADP ratios


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



were found at 1 and 3 min sampling times when R. japonicum bacteroids wereexposed to 8000-2000 nM 02 with LbO2 present as 02 buffer. Furthermore,under comparable experimental conditions, in the absence of LbO2 or MbO2and of bovine serum albumin as fatty acid adsorbent, Bergersen found that R.japonicum bacteroids had about four times the nigrogenase activity of R.leguminosarum bacteroids (22). Then, whereas the addition of fatty-acid-freebovine serum albumin caused a threefold stimulation of R. leguminosarumbacteroid nitrogenase activity, it did not increase (in fact it somewhat de-creased) the nitrogenase activity ofR. japonicum bacteroids. From these resultsit appears that the declining ATP/ADP ratio in R. japonicum bacteroids at high02 is not an artifact, so the maintenance of a high ratio in protected R.leguminosarum bacteroids at high 02 might reflect different energy conserva-tion strategies between these two species.

What the R. leguminosarum bacteroid study (94) did confirm was theparallel increase in nitrogenase activity and ATP/ADP ratio in the zero to low02 region already observed for R. japonicum bacteroids (14, 27, 28), and thestimulatory effects of LbO2 and MbO2 of both processes. But there weresignificant differences in 02 concentrations for peak activities. Under semi-equilibrium conditions in the absence of LbO2 or MbO2, R. leguminosarumbacteroid nitrogenase activity was maximal at 1000-2000 nM dissolved O2(22, 94) and R. japonicum bacteroid nitrogenase activity maximal (and some-what higher) at 100-200 nM 02 (22). In the presence of LbOz or MbO2, maximal nitrogenase activity of R. leguminosarum bacteroids (94) was againobserved at higher O2 than with R. japonicum bacteroids (27), For R. legumino-sarum bacteroids the overall "best" nitrogenase activity occurred in the pres-ence of MbO2 at - 800 nM free dissolved 02 (94); forR.japonicum bacteriods"best" nitrogenase activity occurred in the presence of LbO2 at - 100 nM free02 (27). Wt~. en allowance is made for the use of bacteroid protein as the basisfor specific activity calculation in the R. leguminosarum study (94) and bacter-oid dry weight in the R. japonicum study (27), the "best" isolated-bacteroidnitrogenase activity appears to be much the same for both species. Takentogether, these results show that R. japonicum bacteroids do best at lower 02than do R. legurninosarum bacteroids. In this respect it is interesting to notethat recent measurements (11, 156, 157) show the 02 affinity of the majorcomponent of soybean Lb to be about three times higher than is the 02 affin-ity of the major pea Lb component. Also, the putative oxidase pattern of R.leguminosarum bacteroids, compared with that of soybean bacteroids (seeabove; also 9, 92) suggest that the former may be more O2-tolerant. Thus,one begins to see a correlation between O2 affinity of Lb, bacteroid respira-tory enzyme pattern, and O2-dependence of N2-fixing activity in these twosymbioses.

In both R. japonicum (14) and R. leguminosarum (94) bacteroids the un-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

464 APPLEBY

coupling agent CCCP was shown to have a greater effect in suppression ofnitrogenase activity than on the lowering of whole cell ATP/ADP ratio.Whereas Appleby et al (14) thought that this might indicate the existence more than one domain to ATP formation, with lowered ATP/ADP in one ofthem being responsible for diminished nitrogenase activity, Laane et al (94)favored the explanation that a lowering of the energized state of the bacteroidmembrane by CCCP action reduced the supply of electrons to nitrogenase morethan it inhibited ATP formation. Addition of increasing amounts of theionophore valinomycin to R. leguminosarum bacteroids resulted in decreasedmembrane potential (A~) and nitrogenase activity and increased ATP/ADPratio. At low concentration valinomycin hardly affected the electrochemicalpotential (A~H÷) since the A~ was compensated by an increase in ApH (96).The ionophore nigericin, at < 0.2 p,M, stimulated nitrogenase activity of R.leguminosarum bacteroids at the same time as it increased A~. In this situa-tion, again, ATP/ADP stayed almost constant as did A~H+ (96).

Having established this correlation between A~ and nitrogenase activity,Laane et al (96) assumed a direct relationship between A~ and the generationoI’reducing equivalents for nitrogenase. An imaginative diagram was presented(97) which showed high A~ (membrane potential) causing a conformationalchange in a membrane-bound flavodoxin reductase which in turn allowed thereduction of flavodoxin or low-potential bacteroid ferredoxin (41) not possiblein the relaxed system. In usual formulations of the Mitchell chemiosmotichypothesis (79, 114) the common currency for membrane-associated energeticreactions is the electrochemical potential or protonmotive force (AttH+), andthe Mitchell hypothesis does not allow A~, one component of AIxH+, to haveindependent status in the promotion of membrane-associated energetic reac-tions. It is the reviewer’s prejudice that the embracing of Williams’ localizedproton hypothesis (79), which allows localized microcircuits of protons be-tween individual respiratory chains (114) and hence the "discriminating cou-pling of energy to different processes" (168), would give logical sense to theVeeger group’s experimental findings (94, 95) without necessarily invokingspecific conformational change. Also, it might make easier the understandingof how the differently organized respiratory chains of R. japonicum and R.leguminosarum bacteroids, with their differing O2 sensitivities (see above),could allow ATP/ADP ratio to control nitrogenase activity in the R. japonicumsymbiosis (14, 28) and membrane potential to control it in the R. leguminosar-um symbiosis (94, 96).

The shutoff of nitrogenase activity at high 02, without diminution of ATPratio, is not a unique characteristic of R. leguminosarurn bacteroids. Whenanaerobically isolated R. phaseoli bacteroids from French bean (Phaseolusvulgaris) nodules were shaken under restricted aeration in the absence of LbOz,with glucose as substrate (151), maximal nitrogenase activity, a high steady


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



state of cytochrome c reduction, and an overall cellular ATP/ADP ratio of only1.3, was achieved at 25 mm gaseous 02 tension (3 I~M dissolved 02). At higher(60 mm) gaseous 02, nitrogenase activity was suppressed, cytochrome ~c wasmostly oxidized, and the ATP/ADP ratio had risen to 4.4! Trinchant et al (151)see a correlation between the redox state of cytochrome c, bacteroid reducingpower, and the allocation of electrons to nitrogenase via bacteroid ferredoxin.As suggested by the R. leguminosarum studies (see above), the generation nitrogenase reducing power is supposed to be a property of the bacteroidplasma membrane. Since nitrogenase itself may be associated with this mem-brane (see above; see also 72, 158), it is relevant to note that the turnover rate purified nitrogenase and the efficiency of ATP utilization is much increased asredox potential becomes more negative (71).

Because purified nitrogenase is most active at high ATP/ADP ratios andstrongly inhibited as this ratio approaches 1 (112), the low ratio of 1.3measured for R. phaseoli bacteroids under optimal Nz-fixing conditions (151)illustrates the hazards of attempting to understand events in the nitrogenasedomain by measuring whole-cell adenine nucleotide levels. Nevertheless, if thestartling differences in whole-cell ATP/ADP ratio among R. japonicum (14,28), R. leguminosarum (94), and R. phaseoli (151) bacteroids at high 02 are notartifactual (22, 94), they must reflect the very different responses of thesebacteroid species to 02 stress. For R. japonicum bacteroids the low ATP/ADPratio had been seen as indicating the operation of uncoupled, protective respira-tion (5, 14, 28) via a terminal oxidase distinguished from the efficient, high-Oz-affinity terminal oxidase on the basis of altered inhibitor sensitivity. Although abranched respiration pathway (with the two sides having different inhibitorsensitivities) has been demonstrated in nonsymbiotic R. trifolii (49, 142), theanalogy with R. japonicum bacteroids is not good since it is the low-O2 affinitybranch ofR. trifolii respiration which is more efficiently coupled to phosphory-lation.

A branched respiratory chain with terminals of varying phosphorylationefficiency is one of the more obvious ways in which aerobic N2-fixing organ-isms could protect their O2-sensitive N2 fixation apparatus (79). But H. Haak-er, quoted in (6), favors the explanation that A. vinelandii proton translocation(and hence ATP generation) is controlled by the reduction state of the respira-tion chain. For instance, the reduction state of the ubiquinone-cytochrome bcomplex of Azotobacter modulates the efficiency of proton translocation at siteI, although there is as yet no indication that the same system operates in R.leguminosarum bacteroids (97). In plant host tissues, including legume roots(98), there is evidence for electron spillover to inefficiently phosphorylatingpathways at the site of NAD(P)H oxidation as well as at the ubiquinone site.Also, in A. vinelandii, an NADH dehydrogenase whose activity is enhanced byhigh ATP/ADP ratio, may direct electron flow through an inefficiently phos-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

466 APPLEBY

phorylating respiration pathway (1). Again, there is no direct evidence for theexistence of similar spillover points in Rhizobium bacteroids.

Respiration Substrate in Relation to Symbiotic Efficiency

The foregoing discussion has been concerned with the low- and high-efficiencyphases of Rhizobium bacteroid respiration and nitrogenase activity revealed byvariation of O2 concentration and the presence of inhibitors, but it is necessaryto consider also the effects of various respiration substrates. With R. japonicumbacteroids, Bergersen and Turner (21, 29) found the highest affinity oxidase(K,~ -- 4 nM 02) operating in the presence of LbO2 as 02 carrier to be mostefficient in support of nitrogenase when endogenous substrate was beingrespired. Even though the addition of succinate caused an increase in respira-tion and nitrogenase activity, the mean efficiency (02 uptake/acetylene re-duced) was lower (21,29). At higher gaseous or bulk-liquid 02 concentrations,the large stimulation of nitrogenase activity in bacteroid suspensions by succin-ate was probably the result of vigorous respiration of this substrate causing alowered (but favorable) 02 concentration at the bacteroid surface (21). Similar-ly, Trinchant et al (152) found that whereas the vigorous respiration of succin-ate could support acetylene reduction by R. phaseoli (French bean), R. japoni-cum, and R. leguminosarum bacteroids at high (> 40 mm) gas phase pressure in the absence of LbO2, the less vigorous respiration of sucrose orglucose could not. Only at low (1-5 nM) dissolved 02, in the presence part-oxygenated Lb, could sucrose or glucose support maximum rates ofacetylene reduction (152). Since these two sugars are principal photosyntheticproducts translocated to root nodules, and succinate is not, they could well bethe most significant respiration substrates under the normal physiologicalcondition of low free 02. But N2-fixing bacteroids contain large amounts ofpoly-13-hydroxybutyrate (24), and it is not known what proportion of translo-cated photosynthate is respired directly rather than transformed to and stored aspoly-13-hydroxybutyrate, perhaps with eventual depolymerization to yield 6-hydroxybutyrate or other endogenous substrate for the efficient phase ofbacteroid respiration at low 02.

Exogenous 13-hydroxybutyrate is a poor substrate for respiration (93) andrespiration-supported N2 fixation (119) in Rhizobium bacteroids compared withsuccinate, but this difference may reflect a permeability problem. Cell-freeextracts of pea nodule bacteroids have [3-hydroxybutyrate dehydrogenase acti-vities 2-3 times higher than those of succinate or malate dehydrogenases (84),and a 13-hydroxybutyrate dehydrogenase has been purified from R. lupinibacteroids (173). Acetylene reduction could be stimulated by the addition B-hydroxybutyrate to a crude extract of R. japonicum bacteroids (88), but was necessary to have NADH, benzylviologen, and an ATpogenerating systempresent also, and the total activity was only 8% of that obtained when sodium


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



dithionite was supplied instead of 13-hydroxybutyrate and NADH as reductants.For two varieties of soybean nodule a moderate rise in bacteroid poly-13-hydroxybutyrate was reported at the onset of senescence at the same time asnitrogenase activity was lost and Lb disappeared (86), and it was suggested thatthis rise might be related to the loss of Lb-facilitated respiration.

An exhaustive study by Kretovich et al (93) involving R. lupini and R.leguminosarum (both pea and broad bean) nodules and bacteroids found reverse correlation between the intensity of nitrogen fixation on the one hand,and the content of poly-13-hydroxybutyrate on the other." Most interesting wasthe demonstration that the poly-13-hydroxybutyrate content of effective noduleswas significantly higher at midnight (low nodule N2-fixing activity) than midday (high N2-fixing activity) (93). This diurnal variation of poly-13-hydroxybutyrate level had not been observed in an earlier study using soybeannodules (172) which found only a slow decline in polymer level of detachednodules or of nodules from plants kept many days in the dark. The rapid declinein nitrogenase activity of soybean nodules maintained under the same condi-tions (172) showed that the presence of poly-13-hydroxybutyrate was notsufficient for maintenance of high nitrogenase activity. These two studies (93,172) are in agreement to the extent of suggesting that some directly utilizedphotosynthetic product, as well as a product of poly-13-hydroxybutyrate hydrol-ysis, might be needed to support the energetic processes leading to bacteroid N2fixation. Perhaps this is another manifestation of the phenomenon revealed byVeeger et al (see above; also 94-97, 158), namely the separate maintenance membrane potential and high ATP/ADP ratios as requirements ofR. legumino-sarum bacteroid N2 fixation.

Respiration substrates besides sucrose, glucose, and succinate, which havebeen shown to supplement the endogenous microaerobic nitrogenase activity ofisolated Rhizobium bacteroids, include gluconate (21,29), malate, other orga-nic acids, alcohols, and aldehydes (119, 120). It is claimed (119, 120) alcohols and aldehydes as potential physiological substrates might arise byfermentative reactions in near-anaerobic plant cell cytosol.

A substrate of particular importance to the internal economy of N2-fixingrhizobia is H2. Hydrogen evolution, an inherent property of the nitrogenasereaction, results in the "wasteful" use of ATP and reductant (24), and not allRhizobium bacteroid species possess an uptake hydrogenase capable of cou-pling the reoxidation of this substrate to ATP formation. Since the topic hasbeen reviewed authoritatively in the preceding volume of this series (57), onlythe most relevant questions will be considered here: Is there an exclusiverespiration pathway for H2? Can H2 participate in uncoupled, protective re-spiration as well as in phosphorylating respiration, as suggested by Dixon (53)?Early measurements of H2 respiration by whole R. japonicum bacteroids, in theabsence of Lb as 02 carrier, showed an apparent K,~ for 02 of 1300 nM (107).


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


468 APPLEBY

As this phase of unfacilitated H2 respiration is insensitive to CO (131) and mayinvolve a flavoprotein terminal oxidase (116) [probable characteristics of theuncoupled respiration of other substrates by R. japonicum bacteroids at high 02(5, 14, 28)], it too might be seen as "protective" respiration. On the other hand,when part-oxygenated Lb was used as 02 carder, I-t 2 respiration by R. japoni-cure bacteroids was mediated by a terminal oxidase with K,~ of l0 nM O2 (60).This phase of high-O2-affinity HE respiration caused a stimulation of acetylenereduction, lowered the rate of endogenous substrate respiration, and main-tained the ATP content of bacteroids (60). Hence, it seems no different from thepreviously recognized phase of efficient R. japonicum bacteroid respiration ofother substrates at low O2 which supports nitrogenase activity (5, 14, 28).There is, however, good evidence for the presence of a H2-reduciblecytochrome b 559 in hydrogenase-positive strains of R. japonicum (56, 116)but not in hydrogenase-negative wild strains or mutants (56), so one mustconsider that I-I 2 oxidation is mediated by an exclusive electron transport chain,at least at the substrate end.

Hemoglobin in Nonlegume Symbioses

THE RHIZOBIUM-PARASPONIA SYMBIOSIS The N2-fixing symbiosis withParasponia (a member of the Ulmaceae) is the only one known to involve theassociation of Rhizobium and a nonleguminous plant (154). In this symbiosisthe rhizobia remain within infection threads in the host cells rather thanbecoming enclosed in peribacteroid membranes, and the narrower 02 range(near 0.2 atm) for nitrogen fixation by Parasponia nodules (153, 154), com-pared with legume nodules, suggested the operation of a less flexible 02 supplysystem. The apparent absence of Lb from field-grown Parasponia nodules (45)and apparem retention of the low-O2-affinity cytochromes aa3 and o as oxidasein Rhizobium bacteroids from Parasponia nodules led to a suggestion that thissymbiosis might function, without assistance from an 02 carder, at relativelyhigh free O2 (9). But Parasponia nodules were shown to have the same lowinternal O2 tension as had legume nodules (147), and a reexamination glasshouse-grown Parasponia nodules with better extraction procedures re-vealed the presence of a dimeric Hb with volume-averaged concentration abouthalf that of cowpea (leguminous) nodules formed with the same Rhizobiumstrain (13).

Parasponia Hb was distinguished from plant peroxidase by its reversiblereaction with O2 to form HbO2 and reversible displacement of 02 by CO.Kinetic measurements (C. A. Appleby, B. A. Wittenberg, J. B. Wittenberg,manuscript in preparation) revealed fast 02 association and dissociation ratesand an equilibrium dissociation constant of about 180 nM dissolved O2 at 20°,

which is about one third that of soybean Lb a (4, 11) and about equal that of themajor component of pea Lb (156, 157). Since a limited study of Parasponia


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



nodule bacteria separated by gradient fractionation suggests that mature bacter-oids may lose cytochrome aa3 but retain cytochromes c 552 and o (110), situation very similar to that in R. leguminosarum bacteroids (9), it seems thatthe Parasponia symbiosis may resemble that in the pea (Pisum s~tivum), atleast in respect to bacteroid 02 tolerance and Hb-buffered O2 supply. Prelimi-nary evidence (F. J. Bergersen and C. A. Appleby, unpublished observations)shows that isolated bacteroids from P. andersonii nodules have maximalnitrogenase activity in the O2-unloading region of Lb. The intracellular locationof Parasponia Hb is not known, although in a gentle extraction procedure itremains adsorbed to the bacteroid surface (13). Even so, the efficient transla-tion of the total or poly(A)-enriched RNA from Parasponia nodules in a wheatgerm protein-synthesizing system, yielding Parasponia Hb as principal labeledproduct, establishes this protein as a plant not a bacterial gene product (12).

Although Parasponia Hb and soybean Lb a have no immunological cross-reactivity, and there is no cross-hybridization between a full length eDNA forsoybean Lb a and Parasponia nodule RNA (12), the similarity betweenParasponia Hb and soybean Lb a is emphasized by the close amino acidhomology between many regions of the two proteins (12; also A. A. Kortt, J. E.Burns, A. A. Inglis, M. J. Trinick, C. A. Appleby, manuscript in preparation).This homology, being much greater than between soybean Lb a and animalMb, proteins already thought to have the same genetic origin (35, 75, 77), ledAppleby et al (12) to conclude that legume nodule Lb and Parasponia Hb mightlie close to each other in evolutionary history.

ACTINORHIZAL SYMBIOSES In the N2-fixing actinorhizal root nodulesformed by symbiotic assocition between Frankia endophytes and nonlegumi-nous hosts there appears to be a lesser mechanical restriction against 02penetration to infected cells than in legume nodules (146, 164). Such observa-tions, the ability of Frankia spp. to form vesicles in nodules and pure cultureand the ability of vesiculated Frankia to fix N2 in pure culture at air O2 tension(146, 148), suggested that O2-carrier proteins might have no function actinorhizal symbioses. But energy usage for N2 fixation is similar to that inlegume nodules (146), so it is unlikely that uncoupled, low-phosphorylating-efficiency respiration operates as a significant protective mechanism in acti-norhizal nodules. In 1960, Davenport (48) reported observations of functional,membrane-bound Hb in Casuarina cunninghamiana nodules by visual spec-troscopy, but others (145) could see nothing. It was not until 1983 thatTjepkema, by sensitive spectrophotometry (145), confirmed the occurrence high concentrations of Hb in nodule slices of C. cunninghamiana and Myricagale. Lower concentrations of Hb were observed in the nodules of Comptoniaperegrina, Alnus rubra, and Eleagnus angustifolia; zero to trace amounts werefound in Ceanothus americanus and Datisca glomerata (145). The Hb from


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


470 APPLEBY

Casuarina nodules could be partially solubilized (145; also A. I. Fleming andC. A. Appleby, unpublished observations) and shown to undergo reversiblereactions with 02 and CO, thereby distinguishing it from plant peroxidase (cf13).

Rhizobial Nitrogen Fixation without Leghemoglobin

In 1975, the detection in several laboratories of nitrogenase activity in purecultures of a number of slow-growing and one fast-growing Rhizobium speciesin the absence of Lb or other plant material (see review in 66) led to optimisticexpectations that the further study of these apparently undemanding Rhizobiumstrains might be crucial for the extension of the symbiotic N2 fixation toagriculturally important nonleguminous plants. But many subsequent liquidculture and chemostat studies, including those on Rhizobium sp. 32H1, R.japonicum 3Ilb83, R. japonicum 61A76 (82), Rhizobium sp. CB756 (43), leguminosarum strain 128C30 (140), and the Parasponia-Rhizobium strainANU 289 (111), showed that none of these strains were self-sufficient in termsof N: fixation. They required carefully controlled low concentrations of fixednitrogen sufficient to support (especially initial) growth, but then insufficient prevent nitrogenase expression. More importantly, however, for the substanceof this review, was the common experience that free O2 in solution had to bemaintained at a concentration too low to cause nitrogenase repression ordestruction but sufficient to allow respiration and oxidative phosphorylation.

In one of these studies (43) it was shown that for Rhizobium sp. CB756 theoptimal dissolved O2 concentration for nitrogenase expression was only 30 nM.Free-living N2-fixing cells of Rhizobium sp. CB756 and the closely relatedRhizobium sp. 32H1 were shown (9, 62; also C. A. Appleby and F. J.Bergersen, unpublished observations) to retain the characteristic oxidasecytochromes aa3 and o of air-grown R. japonicurn (3), and also to have presentthe characateristic cytochromes c-552 and P-450 of R. japonicum N2-fixingbacteroids (2). Whether or not the cytochromes aa3 and o permitted protectiverespiration at high 02, the N2-fixing, chemostat-cultured cells ofRhizobium sp.CB756 and 32H1 were more efficient, in terms of nitrogenase activity, duringLb-facilitated respiration at ~ l0 nM free, dissolved O2 (29, 31).

In contrast with the feeble N2-fixing activities of the above Rhizobium spp. inpure culture are recent forthright demonstrations of the ability of Rhizobiumstrain ORS571 (isolated from Sesbania rostrata nodules) to grow in pureculture with gaseous N2 as sole nitrogen source (54, 64, 139). In chemostatcultures, when dissolved Oz concentration was kept between 2-9 ixM (64,139), the nitrogenase activity of culture samples was the highest yet recordedfor Rhizobium (free-living or bacteroids), and nitrogenase activity was notstimulated by the addition of LbO2 to assay vessels. Spectral examinationshowed the oxidase cytochrome aa3 to be absent, but the oxidase cytochrome oto be present in N2-fixing cells grown at 2 I~M free dissolved 02 (139). In view


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



of these startling, O2-tolerant properties of N2-fixing Rhizobium strain ORS571in pure culture, it is all the more remarkable that the N2-fixing root and stemnodules formed by this strain of Sesbania rostrata, a tropical legume, containhigh concentrations of Lb (36).

CONCLUSIONS

The overwhelming conclusion from this review is that all known natural

N2-fixing symbioses involving Rhizobium require the presence of a hemoglo-bin. The recent demonstration of Hb (apparently closely related to Lb) Parasponia nodules (12, 13) and of Hb in several nonlegume-actinorhizalsymbioses (145) leads to the prediction that Hb genes will be found to widespread in the plant kingdom and may not have to be inserted to helpachieve new symbioses.

The function of Lb seems confirmed as facilitation of 02 flux to respiringRhizobium at extremely low, nontoxic free 02 concentration, although confu-sion remains about the nature of, or even the necessity for, Rhizobium protec-tive respiration at higher 02. In this respect the Rhizobium strain ORS571,which fixes N2 vigorously in aerated culture (54, 64, 139), may be worthy intensive study. Even though its natural, N2-fixing association with Sesbaniaincludes Lb (36), will the O2-tolerant features of this organism permit theestablishment of engineered symbioses of lesser complexity than presentlegume symbioses?

Literature Cited

1. Ackrell, B. A. C., Erickson, S. K.,Jones, C. W. 1972. The respiratory chainNADPH dehydrogenase of Azotobactervinelandii. Eur. J. Biochem. 26:387-92

2. Appleby, C. A. 1969. Electron-transportsystems of Rhizobium japonicum. I.Hemoprotein P-450, other CO-reactivepigments, cytochromes, and oxidases inbacteroids from N2-fixing root nodules.Biochim. Biophys. Acta 172:71-87

3. Appleby, C. A. 1969. Electron transportsystems of Rhizobium japonicurn. II.Rhizobium haemoglobin, cytochromes,and oxidases in free-living (cultured)cells. Biochim. Biophys. Acta 172:88-105

4. Appleby, C. A. 1974. Leghemoglobin.See Ref. 125, pp, 521-54

5. Appleby, C. A. 1978. Function of P-450and other cytochromes in Rhizobium res-piration. InFEBS, Colloq. B6, Functionsof Alternative Terminal Oxidases, ed. H.Degn, D. Lloyd, G. C. Hill, 49:11-20.Oxford:Pergamon. 196 pp.

6. Appleby, C. A. 1981. Leghemoglobinand other oxygen control systems. SeeRef. 65, pp. 265-70 (Discussion sum-mary)

7. Appleby, C. A., Bergersen, F. J. 1958.Cytochromes of Rhizobium. Nature 182:1174

8. Appleby, C. A., Bergersen, F. J. 1980.Preparation and experimental use ofleghaemoglobin. In Methods of Evaluat-ing Biological Nitrogen Fixation, ed. F.J. Bergersen, pp. 315-35. Chichester,UK:Wiley. 702 pp.

9. Appleby, C. A., Bergersen, F. J., Ching,T. M., Gibson, A. H., Gresshoff, P. M.,Trinick, M. J. 1981. Cytochromes of rhi-zobia from Parasponia, pea and soybeannodules, and from nitrogen fixing con-tinuous cultures. See Ref. 65, p. 369(Abstr.)

10. Appleby, C. A., Blumberg, W. E., Brad-bury, J. H., Fuchsman, W. H., Peisach,J., et al. 1982. The mobile distal histidineof leghemoglobin. Does it control oxy-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

472 APPLEBY

gen binding kinetics? In Hemoglobin andOxygen Binding, ed. C. Ho, pp. 435-41.New York: Elsevier/North Holland. 486PP

11. Appleby, C. A., Bradbury, J. H., Mor-ris, R. J., Wittenberg, B. A., Witten-berg, J. B., Wright, P. E. 1983. Leghe-moglobin. Kinetic, nuclear magnetic res-onance, and optical studies of pH de-pendence of oxygen and carbon monox-ide binding. J. Biol. Chem. 258:225zl-59

12. Appleby, C. A., Chen, Y. P., Dudman,W. F., Higgins, T. J., Inglis, A. S., etal.1984. Parasponia hemoglobin: Implica-tions for the origin of plant hemoglobins.See Ref. 159, p. 493

13. Appleby, C. A., Tjepkema, J. D., Tri-nick, M. J. 1983. Hemoglobin in a nonle-guminous plant, Parasponia: Possiblegenetic origin and function in nitrogenfixation. Science 220:951-53

14. Appleby, C. A., Turner, G. L., Macni-col, P. K. 1975. Involvement of oxy-leghaemoglobin and cytochrome P-450in an efficient oxidative phosphorylationpathway which supports nitrogen fixationin Rhizobium. Biochim. Biophys. Acta387:461-74

15. Arutyunyan, E. G. 1981. The structure ofleghemoglobin. Mol. Biol. (USSR) 15:27-44

16. Deleted in proof17. Aviram, I., Wittenberg, B. A., Witten-

berg, J. B. 1978. The reaction of ferrousleghemoglobin with hydrogen peroxideto form leghemoglobin IV. J. Biol.Chem. 253:5685-89

18. Bashirova, N. F., Inozemtseva, I. A.,Melik-Sarkisyan, S. S., Kretovich, V. L.1978. Effect of hemoproteins on nitro-gen-fixing activity of the bacteroids ofRhizobium lupini. Fiziol. Rast. 25:18-24

19. Banlcombe, D., Verma, D. P. S. 1978.Preparation of a complementary DNA forleghaemoglobin and direct demonstra-tion that leghaemoglobin is encoded bythe soybean genome. Nucleic Acids Res.5:4141-53

20. Beacham, 1. R. 1979. Periplasmic en-zymes in gram-negative bacteria. Int. J.Biochem. 10:877-83

21. Bergersen, F. J. 1978. Leghaemoglobin,oxygen supply and nitrogen fixation:Studies with soybean nodules. In Limita-tions and Potentials for Biological Nit-rogen Fixation in the Tropics, ed. J.Dtbereiner et al, pp. 247-61. New York.Plenum. 398 pp.

22. Bergersen, F. J. 1980. Leghaemoglobin,oxygen supply and nitrogen fixation:Studies with soybean nodules. Ann.Proc. Phytochem. Soc. Eur. 18:139-60

23. Bergersen, F. J. 1981. Energy produc-tion and utilization in nitrogen fixation.See Ref. 65, pp. 125-30 (Discussionsummary)

24. Bergersen, F. J. 1982. Root Nodules ofLegumes: Structure and Functions.Letchworth, Herts: Res. Stud. Press. 164PP.

25. Bergersen, F. J., Appleby, C. A. 1981.Leghaemoglobin within bacteroid-enclosing membrane envelopes fromsoybean root nodules. Planta 152:534-43

26. Bergersen, F. J., Goodchild, D. J. 1973.Aeration pathways in soybean rootnodules. Aust. J. Biol. Sci. 26:729-40

27. Bergersen, F. J., Turner, G. L. 1975.Leghaemoglobin and the supply of O3 tonitrogen-fixing root nodule bacteroids:Studies of an experimental system withno gas phase. J. Gen. Microbiol. 89:31-47

28. Bergersen, F. J., Turner, G. L. 1975.Leghaemoglobin and the supply of ox-ygen to nitrogen-fixing root nodule bac-teroids: Presence of two oxidase systemsand ATP production at low free O2 con-centration. J. Gen. Microbiol. 91:345-54

29. Bergersen, F. J., Turner, G. L. 1980.Properties of terminal oxidase systems ofbacteroids from root nodules of soybeanand cowpea and of N2-fixing bacteriagrown in continuous culture. J. Gen.Microbiol. 118:235-52

30. Bergersen, F. J., Turner, G. L., Ap-pleby, C. A. 1973. Studies on the phys-iological role of leghemoglobin in soy-bean root nodules. Biochim. Biophys.Acta 292:271-82

31. Bergersen, F. J., Turner, G. L., Gibson,A. H., Dudman, W. F. 1976. Nitrogen-ase activity and respiration of cultures ofRhizobium spp. with special reference toconcentration of dissolved oxygen.Biochim. Biophys. Acta 444:16zl-74

32. Bisseling, T., Been, J., Klugkist, J., vanKammen, A., Nadler, K. 1983. Nodule-specific host proteins in effective and in-effective root nodules of Pisum sativum.EMBO J. 2:961-66

33. Bisseling, T., Moen, A. A., van denBos, R. C., van Kammen, A. 1980. Thesequence of appearance of leghaemoglo-bin and nitrogenase components I and IIin root nodules of Pisum sativum. J. Gen.Microbiol. 118:377-81

34. Bisseling, T., van Straten, J., Houwaard,F. 1980. Turnover of nitrogenase andleghemoglobin in root nodules of Pisumsativum, Biochim. Biophys. Acta 610:360-70


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



35. Blanchetot, A., Wilson, V., Wood, D.,Jeffreys, A. J. 1983. The seal myoglobingene: an unusually long globin gene. Na-ture 310:732-34

36. Bogusz, D. 1984. Leghaemoglobin fromroot and stem nodules of Sesbania rostra-m. See Ref. 159, p. 534 (Abstr.)

37. Bojsen, K., Abildsten, D., Jensen, E.~J., Paludan, K., Marcker, K. A. 1983.The chromosomal arrangement of sixsoybean legbemoglobin genes. EMBO J.2:1165-68

38. Borodenko, L. I., Zhiznevskaya, G. Ya.,Appleby, C. A. 1978. The oxidation-reduction reaction between leghemoglo-bin and a non-heine iron protein fromlupine root nodules. In IUPAClnt. Syrup.Chem. Nat. Prod. l lth, ed. N. Marekovet al, 1:141 (Abstr.) Sofia Bulg. Acad.Sci. 274 pp.

39. Brill, W. J. 1980. Biochemical geneticsof nitrogen fixation. Microbiol. Rev.44:449-67

40. Brisson, N., Verma, D. P. S. 1982. Soy-bean leghemoglobin gene family: Nor-mal, pseudo, and truncated genes. Proc.Natl. Acad. Sci. USA 79:4055-59

41. Carter, K. R., Rawlings, J., Orme-Johnson, W. H., Becker, R. H., Evans,I-I. J. 1980. Purification and characteriza-tion of a ferredoxin from Rhizobiumjapo-nicum bacteroids. J. Biol. Chem. 255:4213-23

42. Child, J. J., Sietsma, J. H. 1975. Sphero-plasts from Rhizobium japonicum. PlantSci. Lett. 4:267-71

43. Ching, T. M., Bergersen, F. J., Turner,G. L. 1981. Energy status, growth andnitrogenase activity in continuous cul-tures ofRhizobium sp. strain CB756 sup-plied with NH4+ and various rates ofaeration. Biochim. Biophys. Acta636:82-90

44. Ching, T. M., I-Iedtke, S., Newcomb,W. 1977. Isolation of bacteria, trans-forming bacteria, and bacteroids fromsoybean nodules. Plant Physiol. 60:771-74

45. Coventry, D. R., Trinick, M. J., Ap-pleby, C. A. 1976. A search for aleghaemoglobin-like compound in rootnodules of Trema cannabina Lour.Biochim. Biophys. Acta 420:105-11

46. Criswell, J. G., Havelka, U. D.,Quebedeaux, B., Hardy, R. W. F. 1976.Adaptation of nitrogen fixation by intactsoybean nodules to altered rhizospherepO2. Plant Physiol. 58:622-25

47. Daniel, R. M., Appleby, C. A. 1972.Anaerobic-nitrate, symbiotic, and aero-bic growth of Rhizobium japonicum:Effects on cytochrome P-ztS0, otherhaemoproteins, nitrate and nitrite reduc-

tases. Biochim. Biophys. Acre 275:347-54

48. Davenport, H. E. 1960. Haemoglobin inthe root nodules of Casuarina cunning-hamiana. Nature 186:653-54

49. de Hollander, J. A., Stouthamer, A. H.1980. The electron transport chain ofRhizobium trifolii. Eur. J. Biochem.111:473-78

50. Dickerson, R. E., Geis, I. 1983. Hemo-globin: Structure, Function, Evolution,and Pathology, p. 75, 82. Menlo Park,Calif:Benjamin/Cummings. 176 pp.

51. Dilworth, M. J. 1980. Leghemoglobins.Methods Enzymol. 69C:812-23

52. Dilworth, M. J., Appleby, C. A. 1979.Leghemoglobin and Rhizobiumhemoproteins. In A Treatise on Dinitro-gen Fixation, Sect. I & II, ed. R. W. F.Hardy, F. Bottomley, R. C. Bums, pp.691-764. New York:Wiley. 812 pp.

53. Dixon, R. O. D. 1972. Hydrogenase inlegume root nodule bacteroids: Occurr-ence and properties. Arch. Mikrobiol.85:193-201

54. Dreyfus, B. L., Elmerich, C., Dommer-gues, Y. R. 1983. Free-living Rhizobiumstrain able to grow on N2 as the solenitrogen source. Appl. Environ. Micro-blol. 45:711-13

55. Egorov, T. A., Kazakov, V. K., Shakh-paronov, M. I., Feigina, M. Y. 1980.Amino acid sequence of leghemoglobinsI and II from yellow lupine nodules.Bioorg. Chim. 6:666-83

56. Eisbrenner, G., Evans, H. J. 1982. Spec-tral evidence for a component involved inhydrogen metabolism of soybean nodulebacteroids. Plant Physiol. 70:1667-72

57. Eisbrenner, G., Evans, H. J. 1983.Aspects of hydrogen metabolism in nitro-gen-fixing legumes and other plant-microbe associations. Ann. Rev. PlantPhysiol. 34:105-36

58. Ellfolk, N. 1972. Leghaemoglobin, aplant haemoglobin. Endeavour 31:139-42

59. E1 Mokadem, M. T.. Keister, D. L.1982. Mutants of Rhizobium japonicumdeficient in cytochromes c and aa3. Stud-ies with aerobically grown cells. Proc.8th North Am. Rhizobium Conf., pp.200-21. Winnipeg: Univ. ManitobaPrint. Serv. 579 pp.

60. Emerich, D. W., Albrecht, S. L., Rus-sell, S. A., Ching, T. M., Evans, H. J.1980. Oxyhemoglobin-mediated hydro-gen oxidation by Rhizobium japonicumUSDA 122 DES bacteroids. Plant Phy-siol. 65:605-9

61. Erecinska, M., Oshino, N., Lob, P.,Brocklehurst, E. 1973. In vitro studies onyeast cytochrome c peroxidase and its


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


474 APPLEBY

possible function in the electron transferand energy coupling reactions. Biochim.Biophys. Acta 292:1-12

62. Evans, W. R., Crist, D. K. 1980. Com-parison of the cytochrome complementand cytochrome c oxidase activity offree-living rhizobia under N2-fixing andnon-fixing conditions. Plant Physiol.Suppl. 65:109 (Abstr.)

63. Fuchsman, W. H., Appleby, C. A. 1979.Separation and determination of the rela-tive concentrations of the homogeneouscomponents of soybean leghemoglobinby isoelectric focusing. Biochim. Bio-phys. Acta 579:314-24

64. Gebhardt, C., Turner, G. L., Gibson, A.H., Dreyfus, B. L., Bergersen, F. J.1984. Nitrogen fixing growth of a strainof Rhizobium sp. in continuous culture.J. Gen. Microbiol. 130: In press

65. Gibson, A. H., Newton, W. E., eds.1981. Current Perspectives in NitrogenFixation. Proc. 4th Int. Symp. NitrogenFixation, Canberra 1980. Canber-ra:Aust. Acad. Sci. 534 pp.

66. Gibson, A. H., Scowcroft, W. R.,Pagan, J. D. 1977. Nitrogen fixation inplants: An expanding horizon? See Ref.l13a, pp. 387-417

67. Gibson, J. F. 1978. Some recent studieson metalloproteins. J. Mol. Struct. 45:139-57

68. Glenn, A. R., Dilworth, M. J. 1978. Anexamination of Rhizobium leguminosar-um for the production of extracellular andperiplasmic proteins. J. Gen. Microbiol.112:405-9

69. Goodchild, D. J. 1977. The ultrastruc-rare of root nodules in relation to nitrogenfixation. Int. Rev. Cytol. Suppl. 6:235-88

70. Deleted in proof71. Hallenbeck, P. C. 1983. Nitrogenase re-

duction by electron carriers: Influence ofredox potential on activity and the ATP/2e ratio. Arch. Biochem. Biophys.220:657~0

72. Howard, K. S., Hales, B. J., Socolofsky,M. D. 1984. Structural interaction be-tween nitrogenase MoFe protein andmembrane in Azotobacter vinelandii andRhodospirillum rubrum. See Ref. 159, p.151 (Abstr.)

73. Hultquist, D. E. 1978. Methemoglobinreduction system of erythrocytes.Methods Enzymol. 52:463-73

74. Hunt, L. T., Hurst-Calderone, S., Day-hoff, M. O. 1978. Globins. In Atlas ofProtein Sequence and Structure, ed, M.O. Dayhoff, 5 (Suppl. 3): 229-49.Washington: Natl. Biomed. Res. Found.414 pp.

75. Hyldig-Nielsen, J. J., Jensen, E. ~.,Paludan, K., Wiborg, O., Garrett, R., etal. 1982. The primary structures of twoleghemoglobin genes from soybean.Nucleic Acids Res. 10:689-701

76. Inozemtseva, I. A., Melik-Sarkisyan, S.S., Kretovich, V. L. 1979. Can leghe-moglobin transfer oxygen to cytochromeP-450r~h? Dokl. Akad. Nauk. SSSR246:741-44

77. Jensen, E. ~., Paludan, K., Hyldig-Nielsen, J. J., J~rgensen, P., Marcker,K. A. 1981. The strneture of a chromo-somal leghemoglobin gene from soy-bean. Nature 291:677-79

78. Jing, Y., Paau, A. S., Brill, W. J. 1982.Leghemoglobins from alfalfa (Medicagosativa L. Vernal) root nodules. I. Puri-fication and in vitro synthesis of fiveleghemoglobin components. Plant Sci.Lett. 25:119-32

79. Jones, C. W. 1982. Bacterial Respira-tion and Photosynthesis. Wokingham:Van Nostrand. 106 pp.

79a. Jones, D. P., Mason, I-I. S. 1978. Gra-dients of 02 concentration in hepatocytes.J. Biol. Chem. 253:4874-80

80. Keister, D. L., Huber, T. A., Marsh, S.S., Agarwal, A. K. 1984. Phenotypiccharacteristics of ex planta nitrogenasepositive strains of Rhizobium japonicum.See Ref. 159, p, 264 (Abstr.)

81. Keister, D. L., Marsh, S. S., ElMokadem, M. T. 1983. Cytochromes ofRhizobium japonicum 61A76 bacteroidsfrom soybean nodules. Plant Physiol.71:194-96

82. Keister, D. L., Ranga Rao, V. 1977. Thephysiology of acetylene reduction in purecultures of rhizobia. See Ref. 113a, pp.419-30

83. Keithley, J. H., Nadler, K. D. 1983.Protoporphyrin formation in Rhizobiumjaponicum. J. Bacteriol. 154:838-45

84. Khurana, A. L., Vyas, S. R. 1977. Res-piratory and enzymatic activity of differ-ent dehydrogenases of bacteroids frompea as a function of plant age. Indian J.Microbiol. 17:4-8

85. Kijne, J. W. 1975. The fine structure ofpea root nodules. 2. Senescence and dis-integration of the bacteroid tissue. Phy-siol. Plant Pathol. 7:17-21

86. Klucas, R. V. 1974. Studies on soybeannodule senescence. Plant Physiol.54:612-16

87. Klucas, R. V., Arp, D. 1977. Physiolog-ical and biochemical studies on senescingtap root nodules of soybeans. Can. J.Microbiol. 23:1426-32

88. Klucas, R. V., Evans, H. J. 1968. Anelectron donor system for nitrogenase-


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


dependent acetylene reduction by ex-tracts of soybean nodules. Plant Physiol.43:1458-60

89. Kretovich, V. L., Matus, V. K., Melik-Sarkisyan, S. S. 1972. Cytochromes andrate of respiration in bacteroids of lupinenodules during the process of plantgrowth. Fiziol. Rast. 19:1060-65

90. Kretovich, W. L., Melik-Sarkisyan, S.S., Bashirova, N. F., Topunov, A. F.1982. Enzymatic reduction of leghemo-globin in lupin nodules. J. Appl.Biochem. 4:209-17

91. Kretovich, W. L., Melik-Sarkisyan, S.S., Raikchinstein, M. V., Archakov, A.I. 1974. The binding of microsomal hy-droxylation substrates to cytochrome P-450Rh and its effect on the nitrogen fixa-tion by lupin bacteroids. FEBS Lett. 44:305-8

92. Kretovich, W. L., Romanov, V. I.,Korolyov, A. V. 1973. Rhizobium legu-minosarum cytochromes (Vicia faba).Plant Soil 39:619-34

93. Kretovich, W. L., Romanov, V. I.,Yushkova, L. A., Shramko, V. I., Fedu-lova, N. G. 1977. Nitrogen fixation andpoly-[3-hydroxybutyric acid content inbacteroids of Rhizobium lupini and Rhi-zobium leguminosarum. Plant Soil 48:291-302

94. Laane, C., Haaker, H., Veeger, C. 1978.Involvement of the cytoplasmic mem-brane in nitrogen fixation by Rhizobiumleguminosarum, bacteroids. Eur. J.Biochem. 87:147-53

95. Laane, C., Haaker, H., Veeger, C. 1979.On the efficiency of oxidative phos-phorylation in membrane vesicles of Azo-tobacter vinelandii and of Rhizobiumleguminosarum bacteroids. Eur. J.Biochem. 97:36%77

96. Laane, C., Krone, W., Konings, W. N.,Haaker, H., Veeger, C. 1979. The in-volvement of the membrane potential innitrogen fixation by bacteroids of Rhizo-bium leguminosarum. FEBS Lett. 103:328-32

97. Laane, N. C. M. 1980. Energysupplyfordinitrogen fixation by Azotobacter vine-landii and by bacteroids of Rhizobiumlegurninosarum. PhD thesis. Dep.Biochem., Agricultural Univ., Wage-ningen. 93 pp.

98. Lambers, H., de Visser, R. 1984. Energymetabolism in nodulated roots. See Ref.159, pp. 453-60

99. Lee, J. S., Brown, G. G., Verma, D. P.S. 1983. Chromosomal arrangements ofleghemoglobin genes in soybean. Nucle-ic Acids Res. 11:5541-53

100. Lehtovaara, P., Ellfolk, N. 1975. Theamino-acid sequence of leghemoglobin

component a from Phaseolus vulgaris(kidney bean). Eur. J. Biochem. 54:577-84

101. Lehtovaara, P., Lappalainen, A.,Ellfolk, N. 1980. The amino acid se-quence of pea (Pisum sativum) leghe-moglobin. Biochirn. Biophys. Acta 623:98-106

102. Lehtovaara, P., Perttila, U. 1978. Bile-pigment formation from differentleghaemoglobins: Methine-bridge speci-ficity of coupled oxidation. Biochem. J.176:359-64

103. Livanova, G. I., Zhiznevskaya, G. Ya.,Andreeva, I. N. 1979. Intracellular local-ization of leghemoglobin in nodules ofLupinus luteus. Dokl. Akad. Nauk SSSR245:739-42

104. Maier, R. J., Brill, W. J. 1976. Ineffec-tive and non-nodulating mutant strains ofRhizobium japonicum. J. Bacteriol.127:763-69

105. Matus, V. K., Melik-Sarkisyan, S. S.,Kretovich, V. L. 1973. Cytochromes andrespiration rate of bacteroids fromnodules of lupine inoculated with effec-tive and ineffective strains of Rhizobiurnlupini. Mikrobiologiya 42:112-18

106. McRae, R. E., Hanus, J., Evans, H. J.1978. Properties of the hydrogenase sys-tem in Rhizobium japonicum bacteroids.Biochem. Biophys. Res. Commun. 80:384-90

107. Melik-Sarkisyan, S. S., Bashirova, N.F., Vladzievskaya, L. P., Karpilenko, G.P., Kretovich, V. L. 1975. Isoelectricfocusing in the study of lupine noduleleghemoglobin, lzv. Akad. Nauk. SSSRSer. Biol. 6:837-45

108. MelikoSarkisyan, S. S., Raikhinshtein,M. V., Vladzievskaya, L. P., Bashirova,N. F., Kretovich, V. L. 1976. Effect oflegoglobin on respiration and nitrogenfixing activity of lupine bacteroids in theprocess of plant development. Fiziol.Rast. 23:274-78

109. Melik-Sarkisyan, S. S., Tikhomirova,A. I., Kretovich, W. L. 1981. Proteins ofthe peribacteroidal space of Rhizobiumlupini bacteroids. DokL Akad. Nauk.SSSR. 259:1498-1501

110. Mohapatra, S. S. 1983. Genetical andphysiological analysis of the derepres-sion of nitrogenase activity in Paraspo-nia--Rhizobium strain ANU 289. PhDthesis. Australian Natl. Univ., Canberra.217 pp.

111. Mohapatra, S. S., Bender, G. L., Shine,J., Rolfe, B. G., Gresshoff, P. M. 1983.In vitro expression of nitrogenase activ-ity in Parasponia-Rhizobium strainANU 289. Arch. Microbiol. 134:12-16


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


476 APPLEBY

112. Mortenson, L. E. 1978. Regulation ofnitrogen fixation. Curr. Top. Cell. Reg-ul. 13:179-232

113. Nash, D. T., Schulman, I-I. M. 1976.The absence of oxidized leghemoglobinin soybean root nodules during noduledevelopment. Biochem. Biophys. Res.Commun. 68:781-85

l13a. Newton, W., Postgate, J. R., Rodri-guez-Ban-ucco, C., eds. 1977. RecentDevelopments in Nitrogen Fixation.Proc. 2nd Int. Symp. Nitrogen Fixation,Salamanca, 1976. London: Academic.622 pp.

114. Nichols, D. G. 1982. Bioenergetics: AnIntroduction to the Chemiosmotic The-ory. London: Academic. 190 pp.

115. Noel, K. D., Stacey, G., Tandon, S. R.,Silver, L. E., Brill, W. J. 1982. Rhizo-bium japonicum mutants defective insymbiotic nitrogen fixation. J. Bacteriol.152:485-94

116. O’Brian, M. R., Maier, R. J. 1983. In-volvement of cytochromes and a fla-voprotein in hydrogen oxidation in Rhi-zobium japonicum bacteroids. J. Bacter-iol. 155:481-87

117. Ollis, D. L., Appleby, C. A., Colman, P.M., Cutten, A. E. C., Guss, J. M., etal.1983. Crystal structure of soybean ferricleghaemoglobin a nicotinate at a resolu-tion of 3.3 ~. Aust. J. Chem. 36:451-68

118. Perttil~i, U. 1981. Effects of modifica-tions of leghemoglobin structure on itsability to oxidize indole-3-acetic acid.FEBS Lett. 128:100-2

119. Peterson, J. B., LaRue, T. A. 1981. Uti-lization of aldehydes and alcohols by soy-bean bacteroids. Plant Physiol. 61:489-93

120. Peterson, J. B., LaRue, T. A. 1982.Soluble aldehyde dehydrogenase andmetabolism of aldehydes by soybean bac-teroids. J. Bacteriol. 151:1473-84

121. Pleiffer, N. E., Malik, N. S. A., Wag-ner, F, W. 1983. Reversible dark-in-duced senescence of soybean rootnodules. Plant Physiol. 71:393-99

122. Planqu6, K., Kennedy, I. R., de Vries,G. E., Quispel, A., van Brussel, A. A.N. 1977. Location of nitrogenase andammonia-assimilatory enzymes in bac-teroids of Rhizobium leguminosarum andRhizobium lupini. J. Gen. Microbiol.102:95-104

123. Puppo, A., Dimitrijevic, L., Rigaud, J.1982. Possible involvement of nodule su-peroxide dismutase and catalase in leghe-moglobin protection. Planta 156:374-79

124. Puppo, A., Rigaud, J., Job, D. 1980.Leghemoglobin reduction by a nodule re-ductase. Plant Sei. Lett. 20:1-6

125. Quispel, A., ed. 1974. The Biology of

Nitrogen Fixation. New York: Elsevier/North Holland. 769 pp.

126. Raikhinshtein, M. V., Melik-Sarkisyan,S. S., Zaigraeva, G. G., Kretovich, V.L. 1976. Inhibition analysis of the re-spiration of bacteroids from yellowlupine nodules. Mikrobiologiya 45:210-16

127. Ratcliffe, H. D., Drozd, J. W., Bull, A.T., Daniel, R. M. 1980. Energy couplingin soybean bacteroids. FEMS Lett.8:111-15

128. Richardson, M., Dilworth, M. J.,Scawen, M. D. 1975. Amino acid se-quence of leghemoglobin I from rootnodules of broad bean (Viciafaba). FEBSLett. 51:33-37

128a. Robertson, J. G., Lyttleton, P., Tapper,B. A. 1984. The role of peribacteroidmembrane in legume root nodules. SeeRef. 159, pp. 475-81

129. Robertson, J. G., Warburton, M. P.,Lyttleton, P., Fordyce, A. M,, Bullivant,S. 1978. Membranes in lupin rootnodules. II.’Preparation and properties ofperibacteroid membranes and bacterialenvelope inner membranes from de-veloping lupin nodules. J. Cell Sci.30:151-74

130. Robson, R. L., Postgate, J. R. 1980.Oxygen and hydrogen in biological nitro-gen fixation. Ann. Rev. Microbiol.34:183-207

131. Ruiz-Argiieso, T., Emerich, D. W.,Evans, H. J. 1979, Characteristics of theH2 oxidizing system in soybean nodulebacteroids. Arch. Microbiol. 121:199-206

132. Saari, L. 1982. Leghemoglobin redu¢-tase from soybean nodules. PhD thesis.Univ. Nebraska, Lincoln. 132 pp.

133. Schwinghamer, E. A., Evans, H. J.,Dawson, M. D. 1970. Evaluation ofeffectiveness in mutant strains of Rhizo-bium by acetylene reduction relative toother criteria of N2 fixation. Plant Soil33:192-212

134. Shaw, B. D. 1983. Non-coordinate reg-ulation of Rhi~obium nitrogenase synthe-sis by oxygen: Studies with bacteroidsfrom nodulated Lupinus angustifolius. J.Gen. Microbiol. 129:849-57

135. Sidloi-Lumbroso, R., Kleiman, L.,Schulman, H. M. 1978. Biochemicalevidence that leghemoglobin genes arepresent in the soybean but not in Rhizo-bium genome. Nature 273:558-60Sievers, G., R6nnberg, M. 1978. Studyof the pseudoperoxidatic activity of soy-bean leghemoglobin and sperm whalemyoglobin. Biochim. Biophys. Acta533:293-301Smith, J. D. 1948. Haemoglobin and the

136.

137.


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.



oxygen uptake of leguminous root nod-ules. Biochem. J. 44:591-98

138. Soininen, R., Ellfolk, N. 1973. Pseudo-monas cytochrome c peroxidase. V.Absorption spectra of the enzyme and ofits compounds with ligands. Inhibition ofthe enzyme by cyanide and azide. ActaChem. Scand. 27:35-46

139. Stam, H., de Vries, W., van Verseveld,H. W., Stouthamer, A. H. 1984. Nitro-gen fixation and hydrogen oxidation inchemostat cultures of Rhizobium sesba-niae. See Ref. 159, p. 271 (Abstr.)

140. Stam, H., van Verseveld, H. W., Stou-thamer, A. H. 1983. Derepression of nit-rogenase in chemostat cultures of the fastgrowing Rhizobium leguminosarum.Arch. Microbiol. 135:19%204

141. Stokes, A. N. 1975. Facilitated diffu-sion: The elasticity of oxygen supply. J.Theor. Biol. 52:285-97

142. Stouthamer, A. H. 1984. Energy genera-tion and hydrogen metabolism in Rhizo-bium. See Ref. 159, pp. 189-97

143. Thulborn, K. R., Minasian, E., Leach,S. J. 1979. Leghemoglobin from Trifo-lium subterraneum. Purification andcharacterization. Biochim. Biophys. Acta578:476-83

144. Tikhomirova, A. I., Melik-Sarkisyan, S.S., Topunov, A. F., Lysenko, L. A.,Kretovich, V. L. 1982. Flavin ox-idoreductascs in lupine nodules. Prikl.Biokhim. Mikrobiol. 18:125-31

145. Tjepkema, J. D. 1983. Hemoglobins inthe nitrogen fixing root nodules of acti-norhizal plants. Can. J. Bot. 63:2924-29

146. Tjepkema, J. D. 1984. Oxygen, hemo-globins, and energy usage in actinorhizalnodules. See Ref. 159, pp. 467-73

147. Tjepkema, J. D., Cartica, R. J. 1982.Diffusion limitation of oxygen uptakeand nitrogenase activity in the rootnodules of Parasponia rigida. Merr. andPerry. Plant Physiol. 69:728-33

148. Tjepkema, J. D., Ormerod, W., Torrey,J. G. 1981. Factors affecting vesicleformation and acetylene reduction (nitro-genase activity) in Frankia sp. Cp.ll.Can. J. Bot. 27:815-23

149. Tjepkema, J. D., Yocum, C. S. 1970.Leghemoglobin facilitated oxygen diffu-sion in soybean nodule slices. Plant Phy-siol. Suppl. 45:44 (Abstr.)

150. Tjepkema, J. D., Yocum, C. S. 1974.Measurement of oxygen partial pressurewithin soybean nodules by oxygen mic-roelectrodes. Planta 119:351-60

151. Trinchant, J. C., Birot, A. M., Denis,M., Rigaud, J. 1983. C2H2 reduction,oxygen uptake and cytochrome c reduc-tion by bacteroids isolated from French-

bean nodules. Arch. Microbiol. 134:182 86

152. Trinchant, J. C., Birot, A..M., Rigaud,J. 1981. Oxygen supply and energy-yielding substrates for nitrogen fixation(acetylene reduction) by bacteroid prepa-rations. J. Gen. Microbiol. 125:159-65

153. Trinick, M. J. 1980. Effects of oxygen,temperature and other factors on the re-duction of acetylene by root nodulesformed by Rhizobium on Parasponiaandersonii Planch. New Phytol. 86:27-38

154. Trinick, M. J. 1981. The effectiveRhizo-bium symbiosis with the non-legume Pa-rasponia andersonii. See Ref. 65, p. 480(Abstr.)

155. Tuzimura, K., Watanabe, I. 1964. Elec-tron transport systems of Rhizobiumgrown in nodules and in laboratorymedium. Plant Cell Physiol. 5:157-70

156. Uheda, E., Sy6no, K. 1982. Physiologi-cal role of leghaemoglobin heterogeneityin pea root nodule development. PlantCell Physiol. 23:75-84

157. Uheda, E., Sy6no, K. 1982. Effects ofleghemoglobin components on nitrogenfixation and oxygen consumption. PlantCell Physiol. 23:85-90

158. Veeger, C., Haaker, H., Laane, C. 1981.Energy transduction and nitrogen fixa-tion. See Ref. 65, pp. 101-4

159. Vccgcr, C., Newton, W. E., eds. 1984.Advances in Nitrogen Fixation Research.Proc. 5th Int. Syrnp. Nitrogen FixationNoordwijkerhout, The Netherlands,1983. The Hague: Nijhoff. 760 pp.

160. Verma, D. P. S., Ball, S., Gu6rin, C.,Wanamaker, L. 1979. Leghemoglobinbiosynthesis in soybean root nodules.Characterization of the nascent and re-leased peptides and the relative rate ofsynthesis of the major leghemoglobins.Biochemistry 18:476-83

161. Verma, D. P. S., Kazazian, V., Zogbi,V., Bal, A. K. 1978. Isolation and char-acterization of the membrane envelopeenclosing the bacteriods in soybean rootnodules. J. Cell Biol. 78:919-36

162. Verma, D. P. S., Long, S. 1983. Themolecular biology of Rhizobium-legumesymbiosis. Int. Rev. Cytol. Suppl.14:211-45

163. Verma, D. P. S., Nash, D. T., Schul-man, H. M. i974. Isolation and in vitrotranslation of soybean leghaemoglobinmRNA. Nature 251:74-77

164. Wheeler, C. T., Gordon, J. C., Ching, T.M. 1979. Oxygen relations of the rootnodules of Alnus rubra Bong. New Phy-tol. 82:499-57

165. Whittaker, R. G., Lennox, S., Appleby,C. A. 1981. Relationship of the minor


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


478 APPLEBY

soybean lcghemoglobins dl, d2 and d~ tothe major leghemoglobins cl, c2, and c3.Biochem. Int. 3:117-24

166. Wibo’rg, O., Hyldig-Nielsen; J. J., Jen-sen, E. 0., Paludan, K., Marcker, K. A.1982. The nucleotide sequences of twoleghemoglobin genes from soybean.Nucleic Acids Res. 10:3487-94

167. Wiborg, O., Hyldig-Nielsen, J. J., Jen-sen, E. 0., Paludan, K., Marcker, K. A.1983. The structure of an unusual leghe-mogiobin gene from soybean. EMBO J.2:449-52

168. Williams, R. J. P. 1978. The history andthe hypotheses concerning ATP-forma-tion by energised protons. FEBS Lett.85:9-19

169. Wittenberg, J. B. 1980. Utilization ofleghemoglobin-bound oxygen by Rhizo-bium bacteroids. In Nitrogen Fixation II.Proc. 3rd Int. Syrup. Nitrogen Fixation,Madison, 1978, ed. W. E. Newton, W.H. Orme-Johnson, pp. 53-67. Balti-more: Univ. Park Press. 325 pp.

170. Wittenberg, J. B., Appleby, C. A., Wit-tenberg, B. A. 1972. The kinetics of thereactions of leghemoglobin with oxygenand carbon monoxide. J. Biol. Chem.247:527-31

171. Wittenberg, J. B., Bergersen, F. J., Ap-pleby, C. A., Turner, G. L. 1974. Facili-tated oxygen diffusion. The role of leghe-moglobin i n nitrogen fixation by bacter-oids isolated from soybean root nodules.J. Biol. Chem. 249:4057--66

172. Wong, P. P., Evans, H. J. 1971. Poly-13-hydroxybutyrate utilization by soybean(Glycine max Merr.) nodules and assess-ment of its role in maintenance of nitro-genase activity. Plant Physiol. 47:750-55

173. Yushkova, L. A., Romanov, V. I., Kre-tovich, V. L. 1975. 13-hydroxybutyricacid dehydrogenase of bacteroids of Rhi-zobium lupini. Dokl. Akad. Nauk 221:239-42


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.


Ann

u. R

ev. P

lant

. Phy

siol

. 198

4.35

:443

-478

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f V

erm

ont L

ibra

ries

on

09/0

7/05

. For

per

sona

l use

onl

y.

Leghemoglobin and Rhizobium Respiration - Semantic Scholar · It is misleading to describe the...

Documents

Transcript of Leghemoglobin and Rhizobium Respiration - Semantic Scholar · It is misleading to describe the...