Directory UMM :Data Elmu:jurnal:P:PlantScience:Plant Science_BioMedNet:141-160:

214

Biochemical dissection of photorespiration
Roland Douce* and Michel Neuburger
Progress has been made in the understanding of
photorespiration and related proteins (Rubisco, glycolate
oxidase and glycine decarboxylase) in the context of recent
structural information. Numerous shuttles exist to support
transamination, ammonia refixation and the supply or export of
reductants generated or consumed (via malate-oxaloacetate
shuttles) in the photorespiratory pathway. A porin-like channel
that is anion selective represents the major permeability
pathway of the peroxisomal membrane.
Addresses
DBMS, Laboratoire de Physiologie Cellulaire Végétale,
CEA Grenoble et Université Joseph Fourier, 17 rue des martyrs,
F 38054 Grenoble, Cedex 9, France
*e-mail: douce@dsvgre.cea.fr
Current Opinion in Plant Biology 1999, 2:214–222
http://biomednet.com/elecref/1369526600200214
© Elsevier Science Ltd ISSN 1369-5266

Abbreviation
RuBP
ribulose-1,5-bisphosphate

Introduction
The prime function of the C2 oxidative photosynthetic carbon cycle — inappropriately named ‘photorespiration’
[1•] — is to salvage glycolate-2-P produced continuously in
the light by the oxygenase activity of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco). In leaves under
ambient conditions the rate of oxygenation to carboxylation
has been estimated as high as 0.4. Low intercellular concentrations of CO2, as may occur, for example, under water
stress (e.g. whenever the stomata are closed), can result in
even higher ratios. Given the voluminous literature on photorespiration [2,3], in this short review we merely highlight
recent advances in this topic, laying emphasis on a few photorespiratory enzymes (Rubisco, glycolate oxidase and
glycine decarboxylase) and molecular traffic between peroxisomes, chloroplasts, and mitochondria.

each other. In the course of this pathway two molecules of
glycolate-2-P are metabolized to form one molecule each of
glycerate-3-P and CO2 and these carbon compounds are
used immediately for the regeneration of RuBP via the
Benson–Calvin cycle (C3 cycle) without the net synthesis

of triose phosphate. Once glycolate-2-P is formed, the photorespiratory cycle works forward to convert all the carbon
diverted out of the C3 cycle back to photosynthesis as
rapidly as possible [3]. Indeed, several reactions occuring in
chloroplasts and peroxisomes strongly favor product formation. Obviously, although very little is known about the
feed-back mechanisms that might operate in photorespiration [4], the most important control step is at the level of
competition between O2 and CO2 for binding to Rubisco.
In C3 plants the C2 cycle is operating in the photosynthetically active mesophyll cells. In C4 plants the C2 cycle
operates in the bundle sheath cells [5]. Using two genetically modified C4 plants, a mutant of Amaranthus edulis that
is deficient in PEP carboxylase and a transgenic plant
Flaveria bidentis which has reduced levels of Rubisco,
Marocco et al [6•] have demonstrated that when the C4
plant is ineffective in concentrating CO2 in the bundle
sheath cells there is a marked increase in photorespiration
and when the C4 plant exhibits low levels of Rubisco there
is a marked increase in bundle-sheath CO2 leakage. This
observation provides definitive evidence that photorespiration is insignifiant in C4 plants because they are capable
of concentrating CO2 in the bundle-sheath cells leading to
the suppression of the oxygenase reaction of Rubisco.

Functioning of key enzymes involved in

photorespiration
Few enzymes involved in this cycle have been studied
carefully. Only Rubisco, glycolate oxidase and a sophisticated set of proteins involved in glycine cleavage (glycine
decarboxylase system) have been studied in an exhaustive
manner. For this reason we have chosen to focus on a
restricted set of enzyme systems.

The photorespiratory pathway
The value of numerous mutant plants (Hordeum, pea, and
Arabidopsis thaliana) in the exquisite elucidation of the mechanism of photorespiration and its relationships with CO2
fixation and amino acid metabolism has been highlighted by
several groups (see [3] for a full list). These mutants were
unable to survive in air, but could thrive in atmospheres containing a high concentration of CO2 (or low [O2]).
The recycling of glycolate-2-P into glycerate-3-P via the
photorespiratory pathway and then further to ribulose-1,5bisphosphate (RuBP) is not only a very costly reaction, it
also requires a large machinery consisting of 16 enzymes
and more than six translocators, distributed over the chloroplast, peroxisome and mitochondrion in close proximity to

Mechanism of Rubisco: the triggering of photorespiration

Rubisco is present at a tremendous concentration in the
stroma of the chloroplasts (~0.2 g ml–1) stromal extract and
catalyses both the carboxylation (the enzyme exhibits a low
catalytic rate constant, 3.5 sec–1) and the oxygenation of
ribulose-1,5-bisphosphate [7–9,10••]. The two reactions
involve the competition of molecular CO2 with O2 for the
2,3-enediol(ate) form of RuBP which is first generated at the
active site of the enzyme. At any given [CO2][O2], the fractional partitioning of RuBP between the carboxylation and
oxygenation pathways is governed by the relative reactivity
of the enzyme-bound 2,3-enediol(ate) toward CO2 and
O2 [10••]. From biochemical analyses of Rubisco purified
from several species, including photosynthetic bacteria,

Biochemical dissection of photorespiration Douce and Neuburger

cyanobacteria, green algae, and higher plants, there are large
differences in specificity towards the substrates CO2 and O2:
evolutionary pressures seem to have directed Rubisco
towards more efficient utilization of CO2 [10••,11]. Rubisco
from cyanobacteria, green algae and higher plants is assembled from eight large (L) subunits and eight small (S)

subunits (four dimers of L subunits surrounded by two
tetramers of S subunits; (L2)4(S4)2) [12], whereas Rubisco
from Rhodospirillum rubrum (a nonsulfur purple bacteria)
consists of only two large subunits (L2). The large subunit
from spinach can be divided into two domains, an amino
terminal domain and a carboxy-terminal α/β-barrel domain.
Two active sites are located at the interface of the L-subunits in the L2 dimer (‘head to tail’ arrangement). The
catalytic center is mostly situated at the carboxy-terminal
end of the α/β-barrel. The enhancement of catalytic rate by
S subunits can only mediated through induced conformational changes in catalytic subunits because S subunits are
far removed from the active site [13].
A large number of crystal structures of Rubiscos from various sources including Rhodospirillum rubrum,
Synechococcus, and spinach have been reported along with
a variety of ligands (see [10••,], for a full list) and this, in
synergy with biochemical investigations, led to a careful
dissection of the carboxylation pathway. The carboxylation of RuBP involves multiple discrete steps and
associated transition states: removal of ligands such as 2carboxyarabinitol 1-phosphate from the inactive enzyme
form (this process occurs slowly by simple dissociation, or
rapidly when catalysed by the enzyme Rubisco activase);
carbamylation of the ε-amino group of Lys201 (spinach)

residue in the active site by an activator CO2 molecule;
stabilization of this protein-bound carbamate by monodentate coordination to Mg2+ (three water molecules,
Asp203 and Glu204 complete the octahedral coordination
sphere around this metal ion); binding of RuBP: it is oriented in the active site with the Si face of the C-2 (and
C-3) accesible to the bulk solution (for an explanation,
see [10••]); removal of the C-3 proton of RuBP to effect
enolisation (the deprotonating agent is still not identified); addition of CO2 to the Si face of C-2 and water to
the Si face of C-3 to yield the six-carbon hydrated intermediate
(2′-carboxy-3-keto-D-arabinitol
1,5
bisphosphate); and carbon–carbon cleavage between C-2
and C-3 to form two glycerate-3-P molecules] [10••,14••].
Higher resolution structures of both the Synecococchus [15]
and of the spinach enzyme [12,14••] demonstrated the
key role of the carbamate on K201 in the carboxylation
pathway. The role of Rubisco activase in limiting steady
state photosynthesis has been examined using transgenic
plants with reduced levels of activase [16,17]. It was concluded that Arabidopsis grown under high and low
irradiance does not contain Rubisco activase in great
excess of the amount required for optimal growth [16]. In

addition, a phase in the activation of Rubisco that represents the activation of the 2-carboxy arabinitol 1
phosphate inhibited form of Rubisco was discerned [17].

215

The oxygenation of RuBP yields one molecule each of
glycerate-3-P (formed from C-3, C-4 and C-5 of RuBP),
and glycolate-2-P. The oxygenation pathway has not been
dissected as deeply as the carboxylation counterpart [10••].
Very likely the oxygenation pathway is similar to the carboxylation pathway although the putative key labile
intermediate (2-peroxy-3-ketoarabinitol 1,5-bisphosphate)
[10••] postulated through the exquisite characterization of
two-different site directed mutants (E60→Q, K334→A)
[18,19], has never been characterized so far. This reaction
may be an inevitable consequence of Rubisco’s inability to
protect its ene-diolate reaction intermediate from O2.
Indeed, this notion is supported by the failure of numerous
efforts to eliminate selectively its oxygenase activity by
genetic manipulation. The partitioning of RuBP between
the carboxylation and oxygenation pathways is sensitive to

the active site microenvironment and does not involve
large movements within the structure [10••]. Given the
structural similarity of the two alternative substrates CO2
and O2 and the large difference in their concentration
within the chloroplasts, it is clear that Rubisco influences
the selectivity for CO2 in some way [12].
Glycolate oxidase

Glycolate oxidase (an octamer composed of identical subunits of approximately 40 kDa) is one of the very few
peroxisomal proteins for which a high resolution crystal
structure is available [20•]. The enzyme from spinach crystallizes in an octameric form and the subunit contains an
eight-fold β/α barrel motif corresponding to the flavin
mononucleotide (FMN) domain which is also found in
other FMN-dependent enzymes. The irreversible reaction
catalysed by the enzyme can be divided into two half-reactions. First glycolate is oxidised by the flavin which is
deeply burried in the barrel. In the second part FMN is
reoxidized by O2 to produce H2O2 which is, in turn,
decomposed by catalase (a heme-containing enzyme). The
active site is formed by the loops at the carboxy-terminal
end of the β-strands in the barrel. The amino acids

involved in the structure of the active site have been studied [21]. Thus, the replacement of Trp108 by Ser led to
dramatic effects on both the Km of substrate as well as on
the turnover number indicating that this amino acid is of
crucial importance in catalysis and in determining the substrate specificity of glycolate oxidase. Likewise Tyr24 is
involved in binding of the substrate by way of hydrogenbond formation between its hydroxyl group and the
carboxylate group of the substrate molecule.
The uptake of glycolate oxidase into peroxisomes has been
studied [22]. The signal for targeting glycolate oxidase into
the plant peroxisome is rather complex. Apparently the
amino-terminal 59 amino acids are dispensible for protein
import in an ATP-dependent and temperature-dependent
manner. This raises the question of the presence of a
carboxy-terminal hexapeptide (RAVARL) at the carboxyterminus of the protein which plays also a role in targeting
a protein to peroxisomes.

216

Physiology and metabolism

Figure 1

S
S
O
H N

COO
H 2C
NH 3

NADH

H ox

P

CO2

L

H met

H red

NH 3
O

NH

NAD

O

NH 3

NH

CH 2
S
SH

HS

T
H4FGlu5

HS

CH 2

H4FGlu5
COO

COO
HOH 2C

H 2C

H 2C
NH 3

SHMT

During the course of glycolate oxidation, proceeding in an
irreversible way, huge amounts of hydrogen peroxide are
released in the peroxisomes. Most of the hydrogen peroxide is degraded by catalase, but the high Km (millimolar
range) for the enzyme could result in low harmful residual
concentrations diffusing into contact with the inner surface
of the limiting peroxisomal membrane which contains an
ascorbate peroxidase [23]. Transgenic tobacco with 0.05 to
0.15 times the catalase activity of wild-type has been
reported [24], and it was shown that under high photorespiratory conditions necrotic lesions were produced in
leaves owing to dramatic accumulation of H2O2.

Reaction catalysed by the glycine
decarboxylase multienzyme complex
Rapid glycine oxidation, which requires the functioning of
two enzymatic complexes (glycine decarboxylase and serine hydroxymethyltransferase) working in concert, is a
key step of the C2 cycle because it results in the conversion of a two-carbon molecule into a three-carbon
molecule that thereafter, could be reintroduced in the C3
cycle [25]. The glycine decarboxylase multienzyme complex, present at tremendous concentration in the matrix of

Outline of the reactions involved in oxidative
decarboxylation and deamination of glycine in
plant mitochondria. Glycine decarboxylase
consists of four different component proteins:
P, T, H, and L. H-protein is a 14.1 kDa
monomer that plays a pivotal role in the
reaction mechanism, as it interacts sequentially
with each of the other three proteins through
its lipoic acid cofactor bound to a lysine
residue. The P-protein component (this
enzyme has a Mr of 210,000 and is a
homodimer of 105,000 Mr polypeptides)
catalyses the decarboxylation of glycine and
the reductive transfer of the resultant
methylamine moiety to the lipoyl-lysine
(lipoamide arm) of the H-protein. The lipoate
cofactor is located in the loop of a hairpin
configuration, but following methylamine
transfer, it is pivoted to bind into a cleft at the
surface of the H-protein. The lipoamidemethylamine arm is, therefore, not free to move
in the solvent. The lipoamide-methylamine arm
is then shuttled to the T-protein (a 45,000 Mr
monomer) where the methylene carbon is
transferred to tetrahydrofolate (H4FGlu5),
producing CH2-H4FGlu5 and releasing the
amino nitrogen as NH3. Finally, the reduced
lipoamide resulting from this transfer is
reoxidized by the FAD coenzyme bound to the
L-protein (a homodimer of 60,000 Mr
polypeptides), with the sequential reduction of
FAD and NAD+. SHMT, serine
hydroxymethyltransferase is involved in the
recycling of CH2-H4FGlu5 to H4FGlu5.

NH 3

Current Opinion in Plant Biology

plant mitochondria, has been purified and, like its mammalian counterpart, contains four different component
enzymes designated as the H-protein (a monomeric
lipoamide-containing protein, 14 kDa), P-protein (a
homodimer containing pyridoxal phosphate, 200 kDa), Tprotein (a monomer acting in concert with folate
[5,6,7,8-tetrahydropteroylpolyglutamate;
H4PteGlu],
45 kDa) and L-protein or lipoamide dehydrogenase (a
homodimer containing flavin adenine dinucleotide [FAD]
and a redox active cystine residue, 100 kDa) [25]. All the
protein components of the glycine decarboxylase system
dissociate very easily and behave as non-associated proteins following mitochondrial inner membrane rupture
after several cycles of freezing and thawing.
The H-protein acts as a mobile co-substrate that commutes between the other three proteins (Figure 1). Its
lipoyl moiety (attached by an amide linkage to the ε-amino
group of a lysine residue [Lys63 in the 131 amino acid pea
H-protein; 26] which is located in the loop of an hairpin
configuration [27]) undergoes a cycle of reductive methylamination, methylamine transfer and electron transfer.
The reaction commences with the amino group of glycine

Biochemical dissection of photorespiration Douce and Neuburger

forming a Schiff base with the pyridoxal phosphate of the
P-protein. The carboxyl group of glycine is lost as CO2 and
the remaining methylamine moiety is passed to the
lipoamide cofactor of the H-protein; when it is oxidized
the lipoyl moiety is free to move in the solvent and is
allowed to visit the active site of the P-protein. The rapid
methylamination of the H-protein is half-saturated at
micromolar concentrations of H-protein (Km H-protein
= 9 µM; Vmax = 5 µmol mg–1 protein min–1). During the
course of the reductive methylamination, the lipoamidemethylamine arm formed rotates to interact readily with
several specific amino acid residues located within a cleft
at the surface of the H-protein; the methylamine group
linked to the distal sulfur of the dithiolane ring is tightly
bound by ionic and hydrogen bonds to residues Glu14,
Ser12, and Asp67, whereas the carbon atoms of the
lipoamide arm interact through van der Waals contacts
with hydrophobic residues [27,28].
Such a situation locks the methylamine group into a very
stable conformation preventing the non-enzymatic release
of NH3 and formaldehyde which would otherwise take
place due to nucleophilic attack by OH– of the carbon
atom bearing NH 2 group until the reaction with
H4PteGlun and T-protein takes place. In the absence of
H4PteGlun in the incubation medium the T-protein causes
a change in the overall conformation of the H-protein,
leading to the release of the lipoamide-methylamine arm
from the cleft at the surface of the H-protein. These circumstances, therefore, favour, the nucleophilic attack by
OH– of the carbon atom bearing NH2 group; NH3 and
formaldehyde accumulate slowly in the incubation medium and the lipoamide arm becomes fully reduced
(Figure 2). On the other hand, in the presence of
H4PteGlun formaldehyde does not accumulate because
the methylamine group undergoes a preferential nucleophilic attack by the N-5 atom of the pterin ring of
H4PteGlun: NH3 and CH2H4PteGlun; accumulate rapidly
in the medium concomitantly with the reduction of the
lipoamide arm (Figure 2).
Plant mitochondria possess a powerful NAD-dependent formate dehydrogenase [29]. They also possess a formaldehyde
dehydrogenase. These enzymes are not believed to be
involved in the main route of carbon flow through the glycolate pathway. They could serve as rescue reactions,
neutralising the harmful effect of formaldehyde molecules
produced by the glycine cleavage system in a non-controlled
reaction. Finally, the L-protein (dihydrolipoamide dehydrogenase) catalyses the regeneration of the oxidised form of
lipoamide with the sequential reduction of FAD and NAD+.
This rapid oxidation is half-saturated at micromolar concentrations of H-protein (Km reduced H-protein = 20 µM). In
green leaf mitochondria, the pyruvate dehydrogenase and
glycine decarboxylase complexes share the same dihydrolipoamide dehydrogenase (E3 component of pyruvate
dehydrogenase, L-protein of glycine decarboxylase) [30] and
this raises some interesting questions about the regulation of

217

synthesis and control of the distribution of this unique
enzyme associated with different complexes. For example,
the distribution of L-protein among complexes may rely
upon various metabolic situations.
In leaf mitochondria, the major function of serine hydroxymethyltransferase (SHMT, a 220 kDa homotetramer) is
to recycle CH2H4PteGlun produced by the T-protein activity to H4PteGlun, to allow the continuous operation of the
glycine-oxidation reaction [31]. This reaction is permanently pushed out of equilibrium towards the production
of serine and CH2H4PteGlun [32] that is the forward
motion of the photorespiratory cycle. The T-protein and
SHMT do not associate and the reaction intermediates are
not directly transferred through a channeling mechanism
from the active site of T-protein to that of SHMT.

Photorespiratory nitrogen cycle
Quantitatively, the conversion of glycine to serine in the
C2 cycle is probably the most important metabolic process
that liberates ammonia within the mesophyll cells.
Nitrogen is inserted into the C2 cycle through a transamination step in the peroxisome catalysed by a
glutamate:glyoxylate aminotransferase [2]. Ammonia liberated in the matrix of mitochondria during the course of
glycine oxidation diffuses rapidly to the chloroplast where
it is used, with a very high affinity, by glutamine synthetase catalysing the ATP-dependent conversion of
glutamate to glutamine [33]. Indeed Mattson et al. [34]
demonstrated that in barley mutants with reduced glutamine synthetase the rate of ammonia emission correlated
with the concentration of ammonia in the leaves. In bacterial glutamine synthetase, the active site is located
between adjacent subunits and structural models for the
reaction mechanism based on five crystal structures of
enzyme–substrate complexes have shown that the reaction
occurs in two steps. First ATP binds to the active site followed by glutamate to yield γ-glutamyl phosphate and
ADP. Then NH4+ binds to the active site, which, after losing a proton, attacks the γ-glutamyl phosphate with the
liberation of glutamine and phosphate [35,36]. Whether a
similar mechanism also operates in eukaryotic octameric
glutamine synthetase is still a matter of debate. It is clear
now from the analysis of barley mutants deficient in glutamine synthetase that the chloroplastic isoform is directly
involved in the reassimilation of ammonia released during
the process of photorespiration [33]. On the other hand,
the cytoplasmic isoform is localized in the vascular system
and the phloem companion cells of the leaf [37,38], thus
precluding any role in photorespiration.
Ferredoxin-dependent glutamate synthase which is exclusively localised in the chloroplast of mesophyll cells
catalyses the reductant-dependent conversion of glutamine and 2-oxoglutarate to two molecules of glutamate.
This enzyme, therefore, functions coordinately with glutamine synthetase. One molecule of glutamate thus formed
is exported to the peroxisomes as an amino donor for

218

Physiology and metabolism

Figure 2

(a)

H

H

H

NH3

HS

HS

HCHO

H

H O CH2 S

NH 3 CH 2 S

HN

O

O

O

H

HN

HN

HN

O

HS

HS

S

HS

OH

(b)

H 4F Glun
H

5

N

H

10

H

NH R
H

H3N CH2 S
HS

NH

H

NH3

H

N
CH2 S

CH2 S

HS

HS

O

NH

H

N
NH R

NH R

H

O

H2C
H

NH

H

H2C

S

HS

HS

HS

O

n

H

N

N
NH R

CH 2H 4F Glu

H

NH

O

H

NH

NH R

O

H
Current Opinion in Plant Biology

Proposed model for the reaction catalysed by the T-protein. (a) In the
absence of H4FGlun in the incubation medium the T-protein causes a
change in the overall conformation of the H-protein, leading to the
release of the lipoamide-methylamine arm from the cleft at the surface
of the H-protein (see Figure 1). Such a situation favours, therefore, the
nucleophilic attack by OH– of the carbon atom bearing the NH2 group;

NH3 and formaldehyde accumulate slowly in the incubation medium
and the lipoamide arm becomes fully reduced. (b) On the other hand,
in the presence of H4FGlun, the methylamine group undergoes a
preferential nucleophilic attack by the N-5 atom of the pterin ring of
H4FGlun. CH2H4FGlun is, therefore, rapidly formed in place of
formaldehyde, concomitantly with the reduction of the lipoamide arm.

glutamate:glyoxylate amino transferase in exchange for
2-oxoglutarate. Arabidopsis contains two expressed genes
for this enzyme (Glu1 and Glu2) situated on different
chromosomes. Glu1 plays a major role in photorespiration
in Arabidopsis, as has been determined by the characterization of mutants deficient in this form [39]. Glu2 may play
a major role in primary nitrogen assimilation in roots. The
enzyme (monomeric with an Mr of ~160 kDa) contains one
FMN and one {3Fe-4S} cluster per molecule [40]. The
assay of this enzyme activity has been greatly facilitated by
the use of methyl viologen as a source of reductant [41]
which is recognized by the ferredoxin-binding site containing two critical lysine and arginine residues [41,42].

Molecular traffic

An interesting point recently raised by Migge et al. [43•]
was that key enzymes of the photorespiratory nitrogen
cycle were not affected either by growing plants in elevated CO2 partial pressure (short-term exposure) or by the
rate of photorespiratory ammonium production, thus
allowing C2-cycles and nitrogen-cycles to take place immediately following exposure to normal air.

During the course of photorespiration, massive traffic of
various molecules occurs between different cell organelles.
Unfortunately, the major characteristics of the transport
proteins (reconstitution of the transporter into liposomes,
kinetic parameters, multisubunit nature, high-resolution
structures, and multifaceted regulation) catalysing substrate travel through membranes to fulfil photorespiration
have been poorly studied. We must say that it is always a
real ‘tour de force’ to reconstitute a transporter into liposomes in an active form.
NH3 and CO2 movement

The NH4+ (and/or NH3) released during glycine oxidation
passes through the inner membrane of mitochondria and
chloroplasts. Whether this passage occurs by simple diffusion, or is brought about by specific ion channels or
translocators is still a matter of debate. In order to maintain
ammonia emission close to zero when carbon assimilation
is strongly limited by stomatal closure under drought conditions, we should expect a specific mechanism to divert

Biochemical dissection of photorespiration Douce and Neuburger

ammonia towards chloroplasts where it is assimilated. In
support of this suggestion a gene from Arabidopsis for a
high affinity ammonia transporter has been identified [44].
We can speculate, therefore, the presence of a specific
ammonia transporter on the inner membrane of the chloroplast envelope.
Likewise, one of the major unresolved aspects of the inner
membranes of mitochondria and chloroplasts in all eukaryotes concerns the CO2 permeability of the membranes. In
other words it is not known which carbon inorganic species
(CO2, HCO3-) is transported in cell organelles. In this connection Rolland et al. [45], using a mutant of Chlamydomonas
reinhardtii, have suggested the existence of a specific protein within the plastid envelope which promotes inorganic
carbon uptake into chloroplasts. Very likely, this protein is
the product of the chloroplast ycb10 gene which has been
localized in the inner membrane of the plastid envelope.
The disruption of this gene in Chlamydomonas using biolistic transformation was correlated with a decrease in
CO2-dependent photosynthesis and a reduced affinity of
the CO2 and HCO3-uptake system for their substrates.
Chloroplast transporters

Glycolate must move from the stroma to the peroxisome
across the inner envelope membrane and D-glycerate must
go in the opposite direction. Experiments with intact
chloroplasts have shown that a single carrier-type transporter is responsible for the movement of both glycolate
and D-glycerate across the chloroplast inner envelope
membrane. This transporter was solubilized by treatment
of the chloroplast inner membrane by sodium cholate and
reinserted into artificial vesicles [46] . The glycolate/glycerate transporter is interesting because it does not catalyse
a strictly coupled substrate exchange (however, glycolate
and D-glycerate stimulate one another’s transport from the
opposite side of the membrane); unidirectional influx or
efflux also occurs as a proton symport or hydroxyl antiport.
This flexibility allows the amount of glycerate returning to
the chloroplasts to be only half that of the glycolate
released from the chloroplasts.
During the course of photorespiration, 2-oxoglutarate is
massively imported into the chloroplasts, and glutamate,
deriving from the glutamine synthetase/glutamate synthase
cycle, is exported towards the peroxisome. Two different
dicarboxylate antiport systems with overlapping substrate
specificities are involved in this process. The 2-oxoglutarate/malate translocator imports 2-oxoglutarate in
exchange for stromal malate, whereas export of glutamate
from the chloroplast in exchange for malate is catalysed by
the glutamate/malate translocator. Malate is, therefore, the
counterion for both translocators, resulting in 2-oxoglutarate/glutamate exchange without net malate import [47].
A cDNA clone encoding the spinach chloroplast 2-oxoglutarate/malate translocator has been obtained by Weber et al.
[48]. The predicted protein with an apparent molecular
mass of 45 kDa contains a 12-helix motif and probably

219

functions as a monomer, in contrast to other known transporters of organellar origin, including mitochondria, that
have 5–7 transmembrane helices functioning as dimers.
The transit peptide of this translocator is extremely long
although its import characteristics closely resemble those of
other inner envelope membrane proteins. The 2-oxoglutarate/malate translocator could be functionally expressed
in the fission yeast Schizosaccharomyces pombe and subsequent reconstitution of the recombinant protein in
liposomes demonstrated definitively that this translocator
mediates the exchange of 2-oxoglutarate with malate.
Obviously the glutamate/malate carrier, which also plays a
critical role in the recycling of ammonia during the course
of photorespiration, requires an exhaustive study in order to
understand precisely the interplay of both carriers working
in concert.
Mitochondria transporters

The rate of glycine oxidation demands that green leaf mitochondria support a phenomenal rate of glycine transport
(0.8–1.6 µmol min–1 mg–1 protein). In the course of glycine
decarboxylation and deamination, one molecule of serine
leaves the mitochondrion and two molecules of glycine are
taken up. For the present, we have to admit that the details
of glycine and serine transport in green leaf mitochondria
remain a mystery and the question as to whether both
glycine and serine are transported by a single protein or by
two different ones cannot be answered at present.
The conversion of hydroxypyruvate to glycerate in the peroxisomal matrix requires NADH as reductant. Peroxisomes
are, therefore, dependent on the supply of reducing equivalents from the cytoplasmic compartment. On the other
hand, NADH produced during the course of glycine oxidation is reoxidized very rapidly by oxaloacetate owing to the
tremendous excess of NAD+-linked malate dehydrogenase
in the matrix space. The malate produced from this reaction is removed from the mitochondria in exchange for
cytosolic oxaloacetate by a specific oxaloacetate transporter.
Peroxisomes are supplied, therefore, with reducing equivalents not by direct uptake of NADH but by indirect
transfer via this malate–oxaloacetate shuttle [49]. A very
powerful phthalonate-sensitive oxaloacetate carrier has
been characterised in all the plant mitochondria isolated so
far [50]. This rapid phthalonate-sensitive uptake of oxaloacetate, which plays an important role in the C2 cycle, is
half-saturated at micromolar concentrations of oxaloacetate
(KmOAA = 5 µM; Vmax = 2 µmol min–1 per mg of protein).
The activity of this carrier appears to be high enough to
account for in vivo carbon fluxes through the inner mitochondrial membrane. The purification and functional
reconstitution, as well as the completion of detailed kinetic
analyses, of this specific transporter should be undertaken.
Porin of peroxisomal membrane

Apparently, peroxisomal membrane does not contain anion
exchangers able to sustain the high fluxes of organic anions
across the membrane of leaf peroxisomes. In fact, it has

220

Physiology and metabolism

been proposed that this membrane contains a slightly
anion-selective channel-forming component, in accordance
with its physiological function and distinct from other
known eukaryotic porins [51,52,53•]. For example, its single-channel conductance of about 300 pS (in 1 M KCl) is
one order of magnitude lower than that of the mitochondrial porin. The narrow diameter (0.6 nm) of this
pore-forming protein restricts the diffusion to anions (glycolate, glycerate, etc.). The characterization of a binding
site for dicarboxylate anions inside the peroxisomal channel, however, is puzzling. It is possible, in analogy with
inducible porins which have been characterized in some
gram-negative bacteria, that this binding site confers rather
selective properties to this peroxisomal channel, preventing
the diffusion of highly reactive intermediates of peroxisomal metabolism, such as glyoxylate and H2O2 [53•].

Conclusions
It has been claimed that Rubisco behaves as a
‘Schizophrenic’ enzyme because of its inability to protect
it’s ene-diolate reaction intermediate from O2 [13]. This
unfair statement should be reconsidered [1•], however,
because several groups have demonstrated that photorespiratory metabolism can prevent the formation of the excited
triplet state of chlorophyll and excess reactive O2 species
(superoxide radicals and singlet oxygen) which necessarily
occur in sunlight when CO2, the final electron acceptor, is
lacking [54]. In other words photorespiration, a very ‘wasteful’ process, in concert with other reactions including a
cycle utilizing monodehydroascorbate reductase, dehydroascorbate reductase and glutathione reductase
(Halliwell-Asada cycle) alleviates the damage that oxygen
radicals can cause in green leaves [55,56]. Wasteful and useful are not necessarily incompatibles and very likely
Rubisco is more ‘clever’ than we thought because when
stomata are closed (the CO2 concentration of the intercellular space of the leaves drops to the CO2 compensation
point) C3- and C2-cycles operate in perfect synchrony to
prevent excessive reduction, and, therefore, photoinactivation, of the chloroplast electron transport chain [55].
Our understanding of which structural features of Rubisco
control discrimination between the two gaseous substrates
is rather meagre, and identification of determinants which
influence CO2 and O2 substrate specificities is a prerequisite for redirecting and modifying fluxes of glycolate-2-P
and glycerate-3-P. Indeed, Rubisco is located at an ideal
strategic position for control of photorespiration [8,11]. It is
possible that the small subunit might influence both the
enzymatic turnover and the discrimination of the two
gaseous substrates [57,58•].
Despite a few impressive advances, it is fair to say that we
still do not have a clear idea as to how any of these
enzymes and transporters involved in photorespiratory
cycle function at the molecular level in establishing the coordinated function of the C3- C2- and nitrogen-cycles for
maximum efficiency. Likewise, an intriguing question is

how the co-ordinated control of a multitude of genes in a
precise spatial and temporal program, can lead to the
development of this exquisite photorespiratory cycle. It
appears certain that the introduction of a genetic approach
will complement the more classical methods used in the
study and regulation of photorespiration with regard to the
ultimate goal of engineering plants with superior growth
characteristics and devising new herbicides.

Acknowledgements
This article is dedicated to the memory of Professor NE Tolbert, tireless
champion of photorespiration.

References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:

• of special interest
•• of outstanding interest
1. Tolbert NE: The C2 oxidative photosynthetic carbon cycle. Annu
•
Rev Plant Physiol Plant Mol Biol 1997, 48:1-25.
This article provides an excellent overview of the C2 cycle.
2.

Husic DW, Husic HD, Tolbert NE: The oxidative photosynthetic
carbon cycle or C2 cycle. CRC Crit Rev Plant Sci 1987,
5:45-100.

3.

Leegood RC, Lea PJ, Adcok MD, Häusler RE: The regulation and
control of photorespiration. J Exp Bot 1995, 46:1397-1414.

4.

Leegood RC, Lea PJ, Haüsler RE: Use of Barley mutants to study
the control of photorespiratory metabolism. Biochem Soc Trans
1996, 24:757-761.

5.

Lea PJ, Ireland RJ: Nitrogen metabolism in higher plants. In The
Plant Amino Acids. Edited by Singh BK. New York: Marcel Dekker
Inc; 1998:1-47.

6.
•

Marocco JP, Ku MSB, Lea PJ, Dever LV, Leegood RC, Furbank RT,
Edwards GE: Oxygen requirement and inhibition of C4
photosynthesis. Plant Physiol 1998, 116:823-832.
This article clearly indicates that when the C4 cycle is ineffective in concentrating CO2 there is an increase in photorespiration.
7.

Von Caemmerer S, Evans JR, Hudson GS, Andrew TJ: The kinetics of
ribulose-1,5-bisphosphate carboxylase/oxygenase in vivo
inferred from measurements of photosynthesis in leaves of
transgenic tobacco. Planta 1994, 195:88-97.

8.

Bainbridge G, Madgwick P, Parmar S, Mitchell R, Paul M, Pitts J, Keys
AJ: Engineering Rubisco to change its catalytic properties. J Exp
Bot 1996, 46:1269-127.

9.

Gutteridge S, Gatenby AA: Rubisco synthesis, assembly,
mechanism and regulation. Plant Cell 1995, 7:809-819.

10. Cleland WW, Andrews TJ, Gutteridge S, Hartman FC, Lorimer, GH:
•• Mechanism of Rubisco: the carbamate as general base. Chem
Rev 1998, 2:549-561.
A striking example of enzyme functioning. The divalent metal ion coordinated lysyl carbamate plays the role of a cofactor and is central to our understanding of the carboxylation and oxygenation reactions.
11. Wildner GF, Schlitter J, Müller M: Rubisco, an old challenge with
new perspectives. Z Naturforsch 1996, 51c:263-276.
12. Andersson I: Large structures at high resolution: the 1.6A crystal
structure of spinach Ribulose-1,5-bisphosphate
carboxylase/oxygenase complexed with 2-carboxyarabinitol
bisphosphate. J Mol Biol 1996, 259:160-174.
13. Hartman FC, Harpel MR: Structure, function, regulation, and
assembly of D-ribulose 1,5-bisphosphate
carboxylase/oxygenase. Annu Rev Biochem 1994, 63:197-234.
14. Taylor TC, Andersson I: The structure of the complex between
•• Rubisco and its substrate ribulose 1,5-bisphosphate. J Mol Biol
1997, 265:432-444.
An impressive paper describing the mechanism of the enzyme in the light
of the high-resolution structure. This paper gives the main structural features of the active site which is situated at the carboxy-terminal end of the
α/β-barrel of the L subunit.

Biochemical dissection of photorespiration Douce and Neuburger

15. Newman J, Gutteridge S: The x-ray structure of Synechococcus
ribulose bisphosphate carboxylase/oxygenase activated
quaternary complex at 2.2 Å resolution. J Biol Chem 1993,
268:25876-25886.
16. Eckardt NA, Snyder GW, Portis AR, Ogren WL: Growth and
photosynthesis under high and low irradiance of Arabidopsis
thaliana antisense mutants with reduced ribulose-1,5bisphosphate carboxylase/oxygenase activase content. Plant
Physiol 1997, 113:575-586.
17.

Hammond ET, Andrews TJ, Woodrow IE: Regulation of ribulose-1,5bisphosphate carboxylase/oxygenase by carbamylation and 2carboxyarabinitol 1-phosphate in tobacco: insights from amounts
of Rubisco activase. Plant Physiol 1998, 118:1463-1471.

18. Chen YR, Hartman FC: Signature of the oxygenase intermediate in
catalysis by ribulose-bisphosphate carboxylase/oxygenase as
provided by a site-directed mutant. J Biol Chem 1995,
270:11741-11744.
19. Harpel, MR, Serpersu EH, Lamerdin JA, Huang ZH, Gage DA,
Hartman FC: Oxygenation mechanism of ribulose-bisphosphate
carboxylase/oxygenase. Structure and origin of 2-carboxytetritol
1,4-bisphosphate, a novel O2-dependent side product generated
by a site-directed mutant. Biochemistry 1995, 34:11299-11306.
20. Stenberg K, Lindqvist Y: Three-dimensional structures of glycolate
•
oxydase with bound active-site inhibitors. Protein Sci 1997,
6:1009-1015.
This paper gives the position of the isoalloxazine ring of FMN in the enzymeinhibitor complexes.
21. Stenberg K, Clausen T, Lindqvist Y, Macheroux P: Involvement of
Tyr24 and Trp 108 in substrate binding and substrate specificity of
glycolate oxidase. Eur J Biochem 1995, 228:408-416.
22. Horng JT, Behari R, Burke LECA, Baker A: Investigation of the
energy requirement and targeting signal for the import of
glycolate oxidase into glyoxysomes. Eur J Biochem 1995,
230:157-163
23. Bunkelmann J, Trelease RN: Ascorbate peroxidase: a prominent
membrane protein in oilseed glyoxysomes. Plant Physiol 1996,
110:589-598.
24. Chamnongpol S, Willekens H, Langebartels C, Van Montagu M,
Inzé D, Van Camp W: Transgenic tobacco with a reduced catalase
activity develops necrotic lesions and induces pathogenesisrelated expression under high light. Plant J 1996, 10:491-503.
25. Oliver DJ: The glycine decarboxylase complex from plant
mitochondria. Annu Rev Plant Physiol Plant Mol Biol 1994,
45:323-338.
26. Mérand V, Forest E, Gagnon J, Monnet C, Thibault P, Neuburger M,
Douce R: Characterization of the primary structure of H-protein
from Pisum sativum and location of a lipoic acid residue by
combined LC-MS and LC-MS-MS. Biol Mass Spectrometry 1993,
22:447-456.
27.

Cohen-Addad C, Pares S, Sieker L, Neuburger M, Douce R: The
lipoamide arm in the glycine decarboxylase complex is not freely
swinging. Nat Struct Biol 1995, 2:63-68.

28. Guilhaudis L, Simorre JP, Blackledge M, Neuburger M, Bourguignon J,
Douce R, Marion D, Gans P: Investigation of the solution structure
and dynamics of the H subunit of the mitochondrial glycine
decarboxylase using heteronuclear NMR spectroscopy.
Biochemistry 1999, in press.
29. Hourton-Cabassa C, Ambard-Bretteville F, Moreau F, Davy de Virville J,
Rémy R, Colas des Francs-Small C: Stress induction of
mitochondrial formate dehydrogenase in potato leaves. Plant
Physiol 1998, 116:627-635.

221

33. Ireland RJ, Lea P: The enzymes of glutamine, glutamate,
asparagine and aspartate metabolism. In Plant Amino Acids.
Edited by Singh BK: New York: Marcel Dekker; 1998:49-109.
34. Mattson M, Schjoerring JK, Häusler RE, Lea PJ, Leegood RC: Leafatmosphere ammonia exchange in barley mutants with reduced
activities of glutamine synthetase. Plant Physiol 1997,
114:1307-1312.
35. Liaw SH, Eisenberg D: Structural model for the reaction
mechanism of glutamine synthetase, based on five crystal
structures of enzyme-substrate complexes. Biochemistry 1994,
33:675-681.
36. Liaw SH, Kuo IC, Eisenberg D: Discovery of the ammonium
substrate site on glutamine synthetase, a third cation binding site.
Protein Sci 1995, 4:2358-2365.
37.

Dubois F, Brugiere N, Sangwan RS, Hirel B: Localization of tobacco
cytosolic glutamine synthetase enzymes and the corresponding
transcripts show organ-specific and cell-specific patterns of
protein synthesis and gene expression. Plant Mol Biol 1996,
31:803-817.

38. Sakurai N, Hayakawa T, Nakamura T, Yamaya T: Changes in the
cellular localization of cytosolic glutamine synthetase protein in
vascular bundles of rice leaves at various stages of development.
Planta 1996, 200:306-311.
39. Coschigano KT, Melo-Oliveira R, Lim J, Coruzzi GM: Arabidopsis gls
mutants and distinct Fd-GOGAT genes: implications for
photorespiration and primary nitrogen assimilation. Plant Cell
1998, 10:741-752.
40. Hirasawa M, Hurley JK, Salamon Z, Tollin G, Knaff DB: Oxidationreduction and transient kinetic studies of spinach ferredoxindependent glutamate synthase. Arch Biochem Biophys 1996,
330:209-215.
41. Hirasawa M, Knaff DB: The role of lysine and arginine residues at
the ferredoxin-binding site of spinach glutamate synthase.
Biochim Biophys Acta 1993, 1144:85-91.
42. Suzuki A, Vergnet C, Morot-Gaudry JF, Zehnacker C, Grosclaude J:
Immunological characterization of ferredoxin and methyl viologen
interacting domains of glutamate synthase using monoclonal
antibodies. Plant Physiol Biochem 1994, 32:619-626.
43. Migge A, Carrayol E, Kunz C, Hirel B, Foch H, Becker T: The
•
expression of the tobacco genes encoding plastidic glutamine
synthetase or ferredoxin-dependent glutamate synthase does
not depend on the rate of nitrate reduction, and is unaffected
by suppression of photorespiration. J Exp Bot 1997,
48:1175-1184.
The rate of level of photorespiratory ammonium production does not exert a
metabolic control on both glutamine synthetase and ferredoxin-dependent
glutamate synthase gene expression.
44. Ninnemann O, Jauniaux J-C, Frommer WB: Identification of a high
affinity NH3 transporter from plants. EMBO J 1994,
13:3464-3471.
45. Rolland N, Dorne AJ, Amoroso G, Sültemeyer DF, Joyard J,
Rochaix JD: Disruption of the plastid ycf10 open reading frame
affect uptake of inorganic carbon in the chloroplast of
Chlamydomonas. EMBO J 1997, 16:6713-6726.
46. Howitz KT, McCarty RE: Solubilization, partial purification and
reconstitution of the glycolate/glycerate transporter from
chloroplast inner envelope membranes. Plant Physiol 1991,
96:1060-1069.
47.

Flügge UI, Weber A, Fisher K, Häusler R, Kammerer B: Molecular
characterization of plastid transporters. CR Acad Sci 1996,
319:849-852.

30. Bourguignon J, Mérand V, Rawsthorne S, Forest E, Douce R: Glycine
decarboxylase and pyruvate dehydrogenase complexes share the
same dihydrolipoamide dehydrogenase in pea leaf mitochondria:
evidence from mass spectrometry and primary structure analysis.
Biochem J 1996, 313:229-234.

48. Weber A, Menzlaff E, Arbinger B, Gutensohn M, Eckerskorn C,
Flügge UI: The 2-oxoglutarate/malate translocator of chloroplast
envelope membranes: molecular cloning of a transporter
containing a 12-helix motif and expression of the functional
protein in yeast cells. Biochemistry 1995, 34:2621-2627.

31. Bourguignon J, Rébeillé F, Douce R: Serine and glycine metabolism
in higher plants. In Plant Amino Acids. Edited by Singh BK. New
York: Marcel Dekker; 1998:111-146.

49. Raghavendra AS, Reumann S, Heldt HW: Participation of
mitochondrial metabolism in photorespiration. Reconstituted
system of peroxisomes and mitochondria from spinach leaves.
Plant Physiol 1998, 116:1333-1337.

32. Rébeillé F, Neuburger M, Douce R. Interaction between glycine
decarboxylase, serine hydroxymethyltransferase and
tetrahydrofolate polyglutamates in pea leaf mitochondria.
Biochem J 1994, 302:223-228.

50. Oliver DJ, Neuburger M, Bourguignon J, Douce R: Glycine
metabolism by plant mitochondria. Physiologia Plantarum 1990,
80:487-491.

222

Physiology and metabolism

51. Reumann S, Maier E, Benz R, Heldt HW: The membrane of leaf
peroxisomes contains a porin-like channel. J Biol Chem 1995,
270:17559-17565.
52. Reumann S, Maier E, Benz R, Heldt HW: A specific porin is involved
in the malate shuttle of leaf peroxisomes. Biochem Soc Trans
1996, 24:754-757.
53. Reumann S, Maier E, Heldt HW, Benz R: Permeability properties of
•
the porin of spinach leaf peroxisomes. Eur J Biochem 1998,
251:359-366.
Characterization of a peroxisomal Porin-like channel in leaf peroxisomes that
is different from those of other cell organelles.
54. Osmond CB, Grace SC: Perspectives on photoinhibition and
photorespiration in the field: quintessential inefficiencies of the light
and dark reactions of photosynthesis. J Exp Bot 1995, 46:1351-1362.
55. Heber U, Bligny R, Streb P, Douce R: Photorespiration is essential
for the protection of the photosynthetic apparatus of C3 plants

against photoinactivation under sunlight. Bot Acta 1996,
109:307-315.
56. Kozaki A, Takeba G: Photorespiration protects C3 plants from
photooxidation. Nature 1996, 384:557-560.
57.

Horken KM, Tabita R: Closely related form I ribulose bisphosphate
carboxylase/oxygenase molecules that possess different CO2/O2
substrate specificities. Arch Biochem Biophys 1999,
361:183-1294.

58. Getzoff TP, Zhu G, Bohnert HJ, Jensen RG. Chimeric Arabidopsis
•
thaliana ribulose-1,5-bisphosphate carboxylase/oxygenase
containing a pea small subunit protein is compromised in
carbamylation. Plant Physiol 1998, 116:695-702.
This article indicates that heterologous expression and stable assembly in
the holoenzyme of a foreign small subunit in a transgenic plant is possible. These results also indicates that the presence of one foreign small
subunit in eight affected carbamylation and carboxylation of the
chimeric Rubisco.