BioSystems 107 (2012) 158166

Contents lists available at SciVerse ScienceDirect

BioSystems
journal homepage: www.elsevier.com/locate/biosystems

Feedforward non-MichaelisMenten mechanism for CO2 uptake by Rubisco:

Contribution of carbonic anhydrases and photorespiration to optimization of
photosynthetic carbon assimilation
Abir U. Igamberdiev a, , Marc R. Roussel b
a
b

Department of Biology, Memorial University of Newfoundland, St. Johns, NL A1B 3X9, Canada
Department of Chemistry and Biochemistry, University of Lethbridge, University Hall, Lethbridge AB T1K 3M4, Canada

a r t i c l e

i n f o

Article history:
Received 26 October 2011
Received in revised form
22 November 2011
Accepted 22 November 2011
Keywords:
Rubisco
Non-MichaelisMenten enzyme kinetics
Carbonic anhydrase
Photosynthesis
Redox balance

a b s t r a c t
Rubisco, the most abundant protein serving as the primary engine generating organic biomass on Earth,
is characterized by a low catalytic constant (in higher plants approx. 3 s1 ) and low specicity for CO2
leading to photorespiration. We analyze here why this enzyme evolved as the main carbon xation
engine. The high concentration of Rubisco exceeding the concentration of its substrate CO2 by 23 orders
of magnitude makes application of MichaelisMenten kinetics invalid and requires alternative kinetic
approaches to describe photosynthetic CO2 assimilation. Efcient operation of Rubisco is supported by
a strong ux of CO2 to the chloroplast stroma provided by fast equilibration of bicarbonate and CO2
and forwarding the latter to Rubisco reaction centers. The main part of this feedforward mechanism is
a thylakoidal carbonic anhydrase associated with photosystem II and pumping CO2 from the thylakoid
lumen in coordination with the rate of electron transport, water splitting and proton gradient across the
thylakoid membrane. This steady ux of CO2 limits photosynthesis at saturating CO2 concentrations. At
low ambient CO2 and correspondingly limited capacity of the bicarbonate pool in the stroma, its depletion
at the sites of Rubisco is relieved by utilizing O2 instead of CO2 , i.e. by photorespiration, a process which
supplies CO2 back to Rubisco and buffers the redox state and energy level in the chloroplast. Thus, the
regulation of Rubisco function aims to keep steady non-equilibrium levels of CO2 , NADPH/NADP and
ATP/ADP in the chloroplast stroma and to optimize the condition of homeostatic photosynthetic ux of
matter and energy.
2011 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Rubisco (ribulose-1,5-bisphosphate carboxylase oxygenase)
is the most abundant protein on Earth (Ellis, 1979), responsible
for practically all CO2 xation in the biosphere, and characterized
by a surprisingly low catalytic efciency with a low carboxylase
catalytic constant (kcat , on average 3 s1 in higher plants) and an
apparently wasteful side reaction with molecular oxygen. From
common sense it is difcult to rationalize why such an inefcient
pathway of carbon xation became prevailing on Earth. Before
the discovery of the CalvinBenson cycle, it was assumed that
a direct pathway of carbohydrate synthesis for CO2 reduction

Abbreviations:
CA, carbonic anhydrase; CA1P, 2-carboxyarabinitol-1phosphate; Enco, Rubisco-enediol complex; MM, MichaelisMenten; PGA,
3-phosphoglycerate; PS II, photosystem II; QSSA, quasi-steady state approximation;
RSA, reactant stationary approximation; RuBP, ribulose-1,5-bisphosphate.
Corresponding author. Tel.: +1 709 864 4567; fax: +1 709 864 3018.
E-mail addresses: igamberdiev@mun.ca (A.U. Igamberdiev), roussel@uleth.ca
(M.R. Roussel).
0303-2647/$ see front matter 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.biosystems.2011.11.008

via formate and formaldehyde could operate in photosynthetic

organisms. This was suggested by Baeyer (1870) based on the
initial discovery by Butlerow (1861) of carbohydrate synthesis
from formaldehyde. In fact, some rate of formate synthesis from
CO2 in photosynthetic tissues has been reported (reviewed in
Igamberdiev et al., 1999) but it never represented a signicant
ux. To date, ve natural metabolic pathways that perform carbon
xation in place of the classic reductive pentose phosphate cycle
have been identied. They include the reductive tricarboxylic
acid (ArnonBuchanan) cycle in green sulfur bacteria, the reductive acetyl-CoA pathway, the 3-hydroxypropionate cycle, the
3-hydroxypropionate/4-hydroxybutyrate cycle, and the recently
discovered dicarboxylate/4-hydroxybutyrate cycle (Bar-Even et al.,
2010). These pathways may specically x CO2 with a high rate
and, in most cases, without wasteful side processes.
The enzyme Rubisco experienced an early evolution before it
acquired the features common to modern forms of the enzyme.
All higher plants, cyanobacteria and most eukaryotic algae contain
the Form I enzyme, a hexadecamer of eight large and eight small
subunits which possesses high CO2 /O2 specicity (SC/O ) values and
low catalytic constants. In contrast, the Form II enzyme, which

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

lacks small subunits and usually presents itself as a dimer of large

subunits, has very low SC/O values but higher catalytic constants.
This form is characteristic for some photosynthetic proteobacteria, for chemoautotrophic bacteria, and for some dinoagellate
algae (Mueller-Cajar and Badger, 2007). The thermophylic Form
III Rubisco of Archaebacteria possesses a high catalytic constant,
includes both the highest and lowest reported SC/O values, and
exists as a dimer (Tabita et al., 2008). The Class III Rubiscos operate in the AMP regenerating pathway in the absence of a complete
CalvinBenson cycle. The Rubisco-like proteins, which make up
Form IV, do not catalyze CO2 xation but participate as enolases in
the methionine salvage pathway (Mueller-Cajar and Badger, 2007).
Of special interest is the coexistence of Form I and Form II in some
organisms. In this case, Form II is needed to balance the intracellular
redox potential when organic carbon is oxidized (Dubbs and Tabita,
2004), a function which, as we show below, is fullled exclusively
by the Form I enzyme in a low-CO2 atmosphere.
The reaction mechanism of Rubisco consists of several steps,
each regulated in its unique way. First, Rubisco activase catalyzes
the separation of inactive Rubisco from the stored ribulose-1,5bisphosphate (RuBP) (Woodrow et al., 1996) or other inhibitory
sugar phosphates (Robinson and Portis, 1988; Portis, 1995). This
allows the formation of the active enzyme by carbamylation and
addition of Mg2+ . The catalytic process then proceeds in two steps.
In step 1, the reaction of Rubisco with RuBP produces a Rubiscoenediol complex (the real carboxylase-oxygenase enzyme, Enco,
as emphasized by Farazdaghi, 2011) and in step 2, Enco captures
CO2 or O2 and forms intermediate products leading to release of
3-phosphoglycerate (PGA) or PGA and phosphoglycolate (in the
case of the oxygenase function) (Tcherkez et al., 2006). While
the activation of Rubisco is a very important mechanism of control, in particular, adapting the reaction to variable light intensity,
we concentrate here on the activated Rubisco bound with RuBP
(Rubisco-enediol complex, Enco) which is commonly saturating
in chloroplast stroma because the stromal RuBP concentration
(>5 mM, Woodrow and Berry, 1988) is comparable or higher than
that of Rubisco. Most important here is to analyze the reaction of
Enco with CO2 or O2 , whose concentrations (especially that of CO2 )
are always lower than the Enco concentration, which raises the
possibility of deviation from MichaelisMenten (MM) kinetics and
puts in question the validity of the widely used quasi-steady state
approximation (QSSA) for this enzyme.
2. Kinetics of the Rubisco-enediol Complex: High
Enzyme-Low Substrate
The decrease of CO2 concentration (or more correctly of the
CO2 /O2 ratio) in the atmosphere during evolution of the biosphere
(Igamberdiev and Lea, 2006) was compensated for by an increase in
the amount of Rubisco in chloroplasts, thus solving by a brute-force
method the problem of low substrate concentration. The consequent high-enzyme/low-substrate concentration regime in which
Rubisco operates, along with its other kinetic peculiarities, mean
that this enzyme can be expected to deviate substantially from MM
kinetics. This can be a problem for the interpretation of both in vitro
and in vivo experiments.
In vitro kinetic studies of the carboxylation reaction of Rubisco
are often carried out under conditions in which oxygen is excluded
and the enzyme has been preincubated with magnesium, carbon
dioxide and an excess of RuBP in order to maximize the concentration of Enco and the rate of its regeneration after a catalytic cycle.
Under these conditions, the mechanism of carboxylation is well
approximated by
k1

k2

fast

Enco + CO2  ECEAEnco + 2(PGA),

k1

(1)

159

where EC is the Enco-CO2 complex, EA is the enzyme complex

involving the intermediate 3-keto-2-carboxy-d-arabinitol-1,5bisphosphate, and PGA is 3-phosphoglycerate. We show the
conversion of the intermediate, the release of PGA, and the regeneration of Enco by binding to RuBP as a single fast process. It is
known that the conversion of the EA complex to an enzyme-(PGA)2
complex is fast (Pierce et al., 1986). Moreover, the rate constant
for the regeneration of Enco is one order of magnitude larger than
the rate constant for binding CO2 (McNevin et al., 2006), so that
under conditions of a large excess of RuBP, we would certainly
expect regeneration to be a fast process. It is however possible
that product release is either slower than or of a similar rate to
the conversion of the EC complex to EA. Given that the mechanism
is essentially irreversible from the EC to EA step to product release,
it is doubtful that this would have much effect on the overall kinetics of the reaction, so for simplicity we take the mechanism to be as
shown above. It may then be asked whether we should expect the
usual analysis of enzyme kinetics based on the MichaelisMenten
equation to be valid for Rubisco under typical in vitro conditions.
The MichaelisMenten (MM) equation (Michaelis and Menten,
1913) can be derived from either the equilibrium (Henri, 1902) or
steady-state approximation (Briggs and Haldane, 1925). However,
in common usage, this equation also involves the implicit use of the
reactant stationary approximation (RSA) (Cme, 1979). The RSA is
the assumption that the substrate concentration changes little during the transient approach to the quasi-steady state. If the RSA is
invalid but the QSSA holds, then it is necessary to measure both
the reaction rate and the concentration of the substrate simultaneously after the attainment of a steady state in order to apply the
QSSA correctly. Hanson and Schnell (2008) have analyzed these two
approximations for the MM mechanism. The QSSA is valid if

(e0 /KM ) + (c0 /KM )

s0
1
1+

KM
1 + s0 /KM

 1,

(2)

where = (k1 + k2 )/k2 , and e0 , c0 and s0 are, respectively, the initial

concentrations of free enzyme, of enzymesubstrate complex, and
of substrate. Note that e0 is not the total enzyme concentration in
this convention. The total enzyme concentration (eT ) is, rather,
eT = e0 + c0 .

(3)

We normally assume that c0 is zero, and of course this is so in

most in vitro assays. However, in an assay using crude extracts, it
is not impossible for c0 to be nonzero because of the presence of
some amount of endogenous substrate in the extract, the situation
contemplated by Hanson and Schnell (2008). This would be particularly likely to occur for enzymes catalyzing reversible reactions.
Using Eq. (3), Eq. (2) can thus be rewritten

eT /KM
s0
1
1+

KM
1 + s0 /KM

 1.

(4)

Hanson and Schnell (2008) also found that the RSA was valid
when



(e0 /KM ) + (1 1/)(c0 /KM )(KM /s0 )
1 + s0 /KM

 1.

(5)

If we dene fc as the fraction of the total enzyme in the form

of enzymesubstrate complex at time zero, then c0 = fc eT and
e0 = (1 fc )eT , such that

(eT /KM ) fc (1 1/)(KM /s0 ) (1 fc )

1 + s0 /KM

 1.

(6)

In typical in vitro kinetics experiments involving Rubisco, the

apoenzyme is preincubated with bicarbonate and magnesium ions
in order to ensure full carbamylation. Reaction is initiated by adding
RuBP. Thus, if we think of Enco as the enzyme and assume rapid

160

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

binding of RuBP, the enzyme itself is formed at time zero, so there is

certainly no enzymesubstrate complex at that time. Accordingly,
we are justied in setting fc = 0, which reduces Eq. (6) to
eT /KM
 1.
1 + s0 /KM

(7)

McNevin et al. (2006) have t kinetic time courses for the

carboxylation reaction of Rubisco in the absence of oxygen and
extracted a full set of kinetic parameters. Their best t parameters
give = 1.5. Their t is however not very sensitive to the value of
k1 . This rate constant can even be set to zero without affecting the
goodness of t to any appreciable extent. Unfortunately, they did
not directly test the sensitivity of their model to larger values of k1 .
The large uncertainty in this parameter (69% of the best-t value
of the rate constant) suggests that larger values are equally plausible. Moreover, Pierce et al. (1986) have pointed out that kinetic
isotope effect data suggest that the exchange of gaseous substrates
between the Enco-bound and solution phases is rapid. In what follows, we thus focus mainly on = 1.5, but also consider smaller and
larger values of this parameter as appropriate.
In order to study Eqs. (4) and (7) graphically, we need to decide
what we mean by much smaller than unity. Following Hanson
and Schnell (2008), we arbitrarily use 0.1 as the threshold separating large from small values. Fig. 1 shows the boundaries of the
regions where the QSSA and RSA are valid. As the enzyme concentration is increased or the substrate concentration is decreased,
the RSA becomes invalid rst, and eventually the QSSA becomes
invalid, too. In the region where the QSSA is valid but the RSA is not,
estimates of Vmax and KM are wildly inated (Hanson and Schnell,
2008). A typical in vitro assay for Rubisco might be as in Wang
et al. (2008). In their experiments, 520 g of enzyme was dissolved in a total volume of 1 mL, corresponding to 0.1 M of active
sites which, since KM (CO2 ) 10 M, gives eT /KM 102 . Referring

Fig. 1. Limits of validity of QSSA and RSA for the Rubisco reaction based on the
approach developed in Hanson and Schnell (2008) for = 1.5 and fc = 0. Area A: both
the RSA and QSSA are valid and the reaction obeys the MM equation. Area B: the QSSA
is valid but the RSA is not. Applying the MM equation in the usual way in this region
will lead to articially inated estimates of Vmax and KM . Area C: neither the QSSA nor
the RSA is valid, substrate is rapidly depleted by enzyme and its continuous inux
is necessary for steady reaction. The square marks typical in vitro assay conditions,
while the solid dot marks typical in vivo operating conditions.

to Fig. 1, we see that these assay conditions correspond to region

A. In vitro assays of Rubisco activity will therefore usually, because
of the very low enzyme concentrations used, fall within the range
where standard MM kinetics (and related approximations) apply.
Note that, for the case fc = 0, only enters into the condition for the
validity of the QSSA [Eq. (4)], and moreover that the curve dened
by Eq. (4) always lies above that dened by Eq. (7) for any given
smallness threshold. Accordingly, the conclusions presented for the
case = 1.5 in fact extend to any value of .
Rubisco operates in a different regime in vivo than is encountered in typical in vitro experiments, but some of the same
issues need to be considered. First of all, in the case of Rubisco,
which is preserved in an inactive form overnight by binding to
2-carboxyarabinitol-1-phosphate (CA1P) (Servaites, 1990), there
would be a brief induction period in the early morning which is
somewhat analogous to the in vitro situation. During this induction period, Rubisco activase starts to remove CA1P from the active
site such that a pool of free enzyme becomes available, leading to an
eventual buildup of the enzymesubstrate complex concentration
to daytime steady-state values. Of course, the kinetics of removal
of CA1P by Rubisco activase, of carbamylation and of Mg2+ binding
are also important in this phase. These processes are relatively slow
(Woodrow et al., 1996), the fastest responses in shade plants occurring on a time scale of about 1 min (Pons et al., 1992; Kursar and
Coley, 1993). One way to think about the activation of Rubisco in
the light is that it causes the total enzyme concentration, eT , to be a
slowly varying quantity rather than a xed parameter. We cannot
rigorously apply the theory of Hanson and Schnell (2008) to this
case, but qualitatively we can think of each incremental change in
eT as inducing a new initial value problem, i.e. a new relaxation
process corresponding to a new set of values of e0 , c0 and s0 or,
correspondingly, to new values of eT , fc and s0 . We can then ask
whether we expect to be able to use the MM equation under conditions prevailing in vivo. We can get some sense of a sensible range
of values of fc from the modeling results of Viil et al. (2001). The
concentration of their complex ERC, which can be roughly associated with the Enco-CO2 complex, is never more than 30% of the
total number of active sites, and can fall to negligibly low relative
values. For very small values of fc , Fig. 1 gives the domain of validity of the two approximations. Since the stromal concentration of
CO2 is about 10 M (Woodrow and Mott, 1993) and KM (CO2 ) is
of a similar size, we have s0 /KM 1. The in vivo concentration of
Rubisco active sites is much larger, of the order of 34 mM of active
sites, giving eT /KM 300400. These data put the operating conditions squarely in region C of Fig. 1, i.e. in the region where the
QSSA is invalid. If we take fc = 0.3, at the higher end of the realistic
range, we get Fig. 2. Typical operating conditions would again be
in region C where the QSSA is inapplicable. These conclusions are
insensitive to the value of , provided the latter parameter is not
too large. At very large values of (100), the curve expressing the
QSSA validity condition moves in such a way that the in vivo conditions may be inside region B, where the QSSA is valid but the RSA is
not. The validity of the QSSA alone might be sufcient for dynamic
modeling. However, a large implies the unrealistic (for Rubisco)
condition k1  k2 (McNevin et al., 2006). We conclude from these
simple calculations that the QSSA will not normally be applicable
to Rubisco kinetics in vivo. The total QSSA (tQSSA) has a broader
range of applicability (Borghans and De Boer, 1996; Tzafriri, 2003;
Tzafriri and Edelman, 2007; Pedersen and Bersani, 2010) and might
be used instead. However, an exploration of this idea is beyond the
scope of the present contribution.
Let us now leave the issue of approximate rate equations and
consider the expected dynamics of photosynthesis. Over the day, as
outside conditions vary and as the plants normal circadian program
unfolds, there will be a (typically) slow variation in various factors
inuencing the kinetics of Rubisco, including for instance stomatal

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

Fig. 2. Limits of validity of QSSA and RSA for the Rubisco reaction based on the
approach developed in Hanson and Schnell (2008) for = 1.5 and fc = 0.3. The regions
are labeled as in Fig. 1. The solid dot again marks a typical in vivo operating condition.

closure, which affects the resistance to CO2 transport from the

atmosphere to the chloroplasts. Neglecting for the moment the possibility of oscillations, we would expect the photosynthetic system,
including the Rubisco subsystem, to reach a slowly varying steady
state, in the following sense: if we write down the mass-action
rate equations for the evolution of the various species involved in
this system (cytosolic and chloroplastic CO2 and O2 , free enzyme,
enzyme bound to various effectors, substrates, intermediates,
and so on), then to a very good approximation, at any given time,
dx/dt 0, where x is the vector of concentrations. The time-varying
inuences would be treated as time-dependent parameters whose
slow variation is followed adiabatically by the components of the
concentration vector according to the steady-state condition.
Let us follow this idea with a very simple model for the function of Rubisco, namely the double MM mechanism in which one
enzyme catalyzes the transformation of two different substrates.
As above, Enco is the enzyme of interest, which we assume to be
rapidly regenerated, but we treat the total concentration of Enco as
a slowly varying parameter. To streamline the notation, we let S1
represent CO2 , and S2 represent O2 . We also consider transport of
the two substrates into the cell. Then we have
k0,i

Si,ext  Si ,
k0,i

k1,i

(8)

2,i
E + Si  Ci E
+ Pi ,

k1,i

where C1 and C2 represent, respectively, the Enco-CO2 complex and

the Enco-O2 complex, and the Pi are the appropriate products of the
two reactions. The total enzyme concentration is
eT = e + c1 + c2

(9)

where lower-case letters are used to represent the corresponding concentrations. The transport coefcients k0,i and k0,i are also
time-dependent quantities since they depend on stomatal opening. We could nd an adiabatic steady state of this system, i.e. one
that follows the slower changes in eT and in k0,i and k0,i , by writing
down the relevant rate equations, setting them equal to zero, and

161

solving for the free concentrations. Using Eq. (9) to eliminate e, this
procedure gives us a set of four equations in the four unknowns {s1 ,
c1 , s2 , c2 }. Since, in vivo, eT is much larger than the concentrations
of the substrates, none of the standard tricks for simplifying these
equations can be used, and we end up with a problem which is
equivalent to solving a cubic equation (details not shown). Accordingly, the in vivo rate will not reduce to the MM form, a point which
had been made previously by Farquhar (1979), albeit using a different model. Models of photosynthesis will therefore need to handle
Rubisco kinetics carefully. Ideally, any kinetically relevant steps
would be included explicitly, without any attempt to use MM-like
equations unless these had been carefully validated using in vivo
experimental data. The availability of rate constants for the individual steps of the reactions catalyzed by Rubisco will be critical to
the proper modeling of this central reaction of photosynthesis. At
this time, only rate constants for the carboxylase reaction are available (Viil and Prnik, 1995; Viil et al., 1999; McNevin et al., 2006).
We hope that someone will take up the challenge of estimating the
in vivo rate constants for the oxygenase reaction as well.
The issues highlighted above regarding the kinetics of Rubisco
are not unique to this enzyme. Indeed, several of the enzymes in the
chloroplast are present at high concentrations relative to the concentrations of the corresponding substrates (Harris and Kniger,
1997). In enzymatic reactions, this condition will generally result
in the depletion of substrate in the proximity of the active sites
of enzymes. In these cases, a great deal of the substrate will be
tied up in enzymesubstrate complexes. This will tend to favor
metabolite channeling, i.e. direct passing of the product of one
enzyme-catalyzed reaction to the next enzyme in a conversion
pathway, which ensures that once a substrate enters a pathway,
it is converted to the ultimate product of the pathway with high
probability (Easterby, 1989). Rubisco has been shown to exist in
a multi-enzyme complex in spinach (Rault et al., 1993), and this
complex has a higher activity than the isolated enzyme (Gontero
et al., 1993). It is also known that many Calvin cycle intermediates are primarily present as complexes with Rubisco, due to the
high concentration of the latter (Pettersson and Ryde-Pettersson,
1988). It is thus very likely that channeling is an important feature
of the Calvin cycle and related reactions occurring in chloroplasts.
In the case of Rubisco, there are the physico-chemical limits to
the CO2 concentration that can be reached in the chloroplast to
deal with. The extremely high concentration of this enzyme can
be effective in utilization of CO2 if there is a mechanism of homeostatic CO2 inux. We will show below that carbonic anhydrase
(CA) and the oxygenase reaction of Rubisco are essential parts of
this mechanism.
In simple terms, a huge buildup of enzyme concentration to
increase metabolic ux makes sense only in the case that there is
a mechanism of continuous pumping of the substrate to the active
site. In this case if the inux of substrate is equaled by the capacity
of its enzymatic conversion (dened by the carboxylase catalytic
constant and enzyme concentration), this corresponds to the condition of optimality of the enzymatic reaction. At a lower inux of
substrate, its concentration will be depleted near active sites which
can be avoided either by inactivation of a part of the pool of enzyme
molecules (decrease of the active fraction) or by use of an alternative substrate (like O2 in the oxygenase reaction of Rubisco). In
the case of high enzyme concentration we cannot expect the relationship between rate and substrate concentration to be similar to
MM but instead we get a curve where at low concentration of substrate the dependence will be more or less linear with substrate
concentration if an alternative substrate can be used, otherwise
low concentrations of substrate will give extremely low rates due
to substrate depletion. At high concentration of substrate when the
pump reaches its maximum capacity, the rate will stabilize. It may
even further decrease if the pump is inhibited by high substrate

162

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

concentration or if additional energy is consumed by the pump at

a higher abundance of substrate.
The general conclusion here is that at enzyme concentrations
for which the QSSA and RSA are not both valid, the enzyme operates efciently not because of steady substrate concentrations but
because of steady substrate uxes. In this case, for a continuous
and steady reaction, it is important that the substrate pumping
ux and the substrate turnover ux are equilibrated, which can
be achieved by feedback mechanisms. However, the main driving
force is a feedforward regulation of CO2 assimilation in which the
external pump greatly amplies the ux of CO2 available for carboxylation. This feedforward mechanism becomes a condition for
stable operation of the enzyme outside the range of validity of the
MM approximation.
It was shown before in the approach developed by Fridlyand
and Scheibe (1999) for analysis of metabolic cycles that limitations
in metabolism usually occur at reaction steps controlled by equilibrium enzymes. In the case of engine enzymes driving major
physiological processes and present at very high concentrations,
the preceding equilibrium reaction supplying substrate serves as
a control gate for the non-equilibrium ux through this enzyme.
This preceding reaction is catalyzed by a thermodynamic buffer
enzyme (Stucki, 1980). For the main energy-generating engine
of the cell, ATP synthase, the equilibrium reaction is catalyzed by
adenylate kinase (Igamberdiev and Kleczkowski, 2009) which provides a constant supply of ADP to ATP synthase from the total pool
of adenylates even at expense of a part of the ATP that the cell generates. By equilibrating adenylates, adenylate kinase also controls
free Mg2+ levels (Igamberdiev and Kleczkowski, 2003, 2006) and,
thus, indirectly, the activation state of Rubisco. For Rubisco, which
is the primary engine enzyme for converting CO2 to biomass, the
limiting step will be a supply of CO2 from the stromal bicarbonate
pool, which is catalyzed by CA. Rubisco kinetics will be ultimately
determined by the CO2 supply and the reaction will be saturated at
the point when the ux of CO2 approaches its maximum. Thermodynamic buffering that controls the homeostatic ux in the Calvin
cycle is supported also by equilibrium enzymes such as glyceraldehyde phosphate dehydrogenase, transaldolase and transketolase
acting to prevent depletion of RuBP, even at high CO2 , and to maintain conditions where the only control is exerted by the CO2 supply
(Igamberdiev and Kleczkowski, 2011).
The deviation of the CO2 assimilation curve from MM kinetics
has already been mentioned. The observed curve of CO2 assimilation deviates from MM kinetics markedly at high concentrations of
CO2 approaching the maximum rate earlier than predicted by MM
kinetics. Laisk (1977) and Farquhar et al. (1980) explained this by a
putative co-limitation of the Rubisco reaction by two processes, one
being limited by the enzyme (Rubisco-enediol) availability and the
other by the supply of the substrate RuBP. This model is widely
used and only rarely criticized (Farazdaghi and Edwards, 1988;
Farazdaghi, 2011). Farazdaghis criticism resulted in the development of the one-process two-step model (Farazdaghi, 2011) which
is still based on MM kinetics. Our arguments above show that this is
a problematic assumption. Moreover, it is possible that the deviation from MM kinetics following from the high concentration of the
enzyme can by itself explain the observed curve of CO2 assimilation.

3. The Role of the Carbonic Anhydrase Reaction: the

Thylakoidal CO2 Pump
CO2 is mostly stored in cells in the form of bicarbonate and this
pool represents an effective source of CO2 in chloroplasts. Bicarbonate buffering via carbonic anhydrase (CA) plays a signicant role in
many physiological processes from carbon xation in photosynthesis to respiration in animals (Igamberdiev and Kleczkowski, 2009).

Fig. 3. Roles of carbonic anhydrases (A) and of photorespiration (B) in CO2 supply
for photosynthetic assimilation. (A) The effect of the CA inhibitor ethoxyzolamide
(1 mM) on CO2 assimilation in pea protoplasts (Igamberdiev, unpublished results).
The isolation and incubation of protoplasts are described in Igamberdiev and
Gardestrm (2003), pH 8.0. (B) CO2 assimilation curves in wild type barley plant
and in the mutant of barley decient in glycine decarboxylase (GDC). Characteristics of the mutant and experimental conditions are given in Igamberdiev et al.
(2004). CO2 concentration in the stroma is calculated from the CO2 concentration in
the substomatal cavity (Ci) assuming a stromal pH of 8.0.

According to the HendersonHasselbach equation, at physiological

pH only 57% of CO2 is released from bicarbonate and carbonate. A
facilitation of the equilibrium reaction between CO2 and bicarbonate will promote a supply of CO2 to Rubisco, which is the role of
carbonic anhydrases. The data of Riazunnisa et al. (2006) show that
the inhibition of CA by its specic inhibitor ethoxyzolamide results
in an essential decrease of photosynthetic oxygen evolution. The
data on pea protoplasts (Igamberdiev, unpublished, Fig. 3a) show
that ethoxyzolamide affects CO2 assimilation in pea protoplasts by
a marked decrease of the CO2 afnity (S0.5 increased about by a
factor of two). This means that the supply of CO2 from bicarbonate
(limited by the CA activity) represents a limiting step in Rubisco
operation so the ux of CO2 to Rubisco (rather than the steady CO2
concentration) represents a mechanism limiting CO2 assimilation.
S0.5 here appears as a dynamic KM dependent on the single process
of CO2 ux to Rubisco consisting of the two steps described below,
one limited by photorespiration (Fig. 3b, see below) and other by
the capacity of the CA-associated CO2 pump (Fig. 3a). This mechanism (one process, two steps) has similarities to that described by
Farazdaghi (2011) but is based on different mechanisms of limitation of the CO2 assimilation.
There are different isoforms of CA present in photosynthetic
plant cells and distributed to different cell compartments. The
direct supply of CO2 to Rubisco is provided by the thylakoidal CA

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

Fig. 4. Coordination of supply of NADPH, ATP and CO2 to the Calvin cycle by photosystem II and its CA-associated activity. PS II supplies electrons to the chloroplast
electron transport chain which ultimately results in NADP reduction and generation
of proton gradient. A part of the proton gradient is used for bicarbonate transport in
the lumen where it forms CO2 with the help of the PS II-associated CA. CO2 is thus
supplied to the stroma and feeds Rubisco.

that utilizes bicarbonate, pumped inside the thylakoid by using

the proton gradient formed during operation of the photosynthetic electron transport chain. The thylakoidal CA converts this
bicarbonate to CO2 which escapes back to the stroma (Fig. 4).
This concentration mechanism was suggested rst by Fridlyand
(1995), see also (Fridlyand and Kaler, 1987, 1988). In this mechanism, the dehydration of bicarbonate occurs inside thylakoids in
the reaction with intrathylakoid H+ catalyzed by the membranebound CA (which was initially reported by Pronina and Semenenko,
1988, 1990). Because of a low diffusion resistance for CO2 between
the cytoplasm and chloroplast stroma, this mechanism, although
having some similarity with the CO2 concentration mechanism in
algae, cannot drastically decrease the CO2 compensation point. At
very high CO2 concentrations this mechanism will have a higher
energy expenditure (dissipation of proton gradient) which may
lead to a depletion of ATP needed for regeneration of the CO2 acceptor in the Calvin cycle and thus explain the observed decrease in
the rate of photosynthesis at oversaturating CO2 concentrations,
e.g. in wheat plants (Fridlyand and Tsyuryupa, 1992) and in pea
protoplasts (Riazunnisa et al., 2006).
The CA associated with photosystem II (Park et al., 1999) represents a central part of the mechanism supplying CO2 to Rubisco
(Fig. 4). This CA promotes proton removal from the Mn complex of
PS II by catalyzing the reaction with HCO3 transported from the
stroma (Shutova et al., 2008). There are probably four separate CA
activities associated with thylakoids (Rudenko et al., 2007). Hanson
et al. (2003) were rst to show explicitly that the Chlamydomonas
reinhardtii cia3 mutant lacking a thylakoid lumen-localized CA is
limited by the CO2 supply to Rubisco and not by photosystem II
function in vivo. Referring to Fig. 4, we note that decreasing the CO2
concentration in the stroma could generate a chemical potential
gradient that would favor transfer of bicarbonate into the thylakoid. This means that the facilitation of CO2 utilization by Rubisco
stimulates the bicarbonate inux into the thylakoid and regulates
efcient operation of water splitting in photosynthesis (Villarejo
et al., 2002). PS II in association with the lumenal CA operates to
provide an ample ux of CO2 for carboxylation (Park et al., 1999).
The HLA3 ABC type bicarbonate transporter may be involved in this
reaction (Duanmu et al., 2009).
The operation of chloroplast electron transport therefore not
only supplies NADPH and ATP to the Calvin cycle but also participates in pumping CO2 to Rubisco in the stroma by using
the CA mechanism which is linked to the CA activity of PS II
and correlates the intensity of chloroplast electron transport and

163

photophosphorylation with the intensity of the Calvin cycle. In

other words, the chloroplast electron transport is responsible
for a coordinated supply of ATP, NADPH and CO2 to the Calvin
cycle. In fact, coordination between the ux through Rubisco
and the chloroplast electron transport was demonstrated and it
was shown that the value of kcat of Rubisco strongly depends
on the electron ux through the b6 f complex (Eichelmann et al.,
2009). Bicarbonate inux to the thylakoid, the CO2 supply to
Rubisco, water splitting, and the build-up of proton gradient are
all coupled via the CA activity of PS II (Fig. 4). In line with this
coordination is the observation that light induces the stimulation of CA activity in pea thylakoids (Moskvin et al., 2000) and
that the most responsive protein to low CO2 in Chlamydomonas
is the mitochondrial CA (Wienkoop et al., 2010). This all means
that a PS II-driven electron transport is a feedforward regulator of CO2 assimilation and that PS II-CA greatly amplies the
ux of CO2 available for carboxylation (Park et al., 1999). The
stromal CA can also participate in supplying CO2 to Rubisco but
it apparently lacks the necessary efciency in CO2 concentration.

4. The Role of the Rubisco Oxygenase Reaction: the

Photorespiratory CO2 Pump
The mechanism of CO2 supply by CA that includes its pumping from the thylakoid lumen can efciently deliver CO2 to Rubisco
above its equilibrium concentration with the atmosphere. At lower
concentrations, we can still observe the fast depletion of CO2 at
Rubisco sites leading to a fall of the reaction rate. To avoid this, biological systems can exploit the homeostatic ux control mechanism
called thermodynamic buffering by Stucki (1980), which includes
a sequence of reactions rebuilding a substrate from the product of
an enzymatic reaction, contributing to feedforward build-up of the
substrate. In the Rubisco reaction, the oxygenase function can play
this role by taking an alternative substrate (O2 ) upon depletion of
CO2 and forming CO2 in the following sequence of reactions associated with photorespiration. This provides a stable non-equilibrium
condition for Rubisco operation and ultimately restricts variation of
CO2 and O2 concentrations within certain limits. This keeps the ux
through Rubisco constant even when the CA mechanism is incapable of supplying sufcient CO2 to saturate the Calvin cycle. At
low CO2 , the supply of this substrate to Rubisco becomes uncoupled from the activity of PS II in the sense that the generation of
NADPH and ATP exceeds the capacity of the Calvin cycle. The use
of oxygen keeps the ux through Rubisco steady and generates
a metabolic pathway that becomes a major sink of the reducing
power and ATP in photosynthetic plant cells. In fact, plants decient
in glycine decarboxylation exhibit unusually high ATP/ADP and
NADPH/NADP ratios in chloroplasts (Igamberdiev et al., 2001). They
also display lower CO2 afnity of photosynthesis (Igamberdiev
et al., 2004) resulting in decreased rates of CO2 xation at low
CO2 (Fig. 3b). The latter can be simply explained by the impairment of the process of CO2 recycling back to Rubisco. This means
that the oxygenation reaction of Rubisco plays both a regulating
role when the photochemical energy exceeds the carboxylation
capacity (Andr, 2011a,b) and a role in feedback delivery of CO2 to
Rubisco active sites (Roussel et al., 2007; Roussel and Igamberdiev,
2011).
The oxygenase and the carboxylase reactions of Rubisco are
reciprocally inhibited by the alternative substrates (Fig. 5). When
two substrates (CO2 and O2 in the case of Rubisco) are competing
to bind an enzyme active site, and when the ow of one substrate is controlled by a feedback device, e.g. photorespiratory
CO2 release, sustained oscillations can result (Ngo and Roussel,
1997). In the case of high enzyme and low substrate concentration,

164

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

Fig. 6. General scheme showing operation of Rubisco as an engine for generating

biomass from CO2 . The source of CO2 is a bicarbonate pool fed from the atmosphere
and buffered by carbonic anhydrase (CA) serving as a feedforward pump for Rubisco.
The latter is an engine producing biomass and at the same time generating a feedback
(photorespiration) to feed the bicarbonate pool in conditions of insufcient CO2
supply.

Fig. 5. The scheme showing CO2 supply to Rubisco by photorespiration. Rubisco participates both in reactions with CO2 (as carboxylase) and with O2 (as oxygenase),
the latter eventually leading to the appearance of glycine in the mitochondria and,
subsequently, to photorespiration. The CO2 produced by photorespiration is transported through the cytoplasm to the chloroplasts as bicarbonate after conversion by
the mitochondrial carbonic anhydrase. The equilibria between CO2 and bicarbonate established by carbonic anhydrases are shown in chloroplasts, thylakoid lumen,
mitochondria and cytosol.

oscillations can be established due to the depletion of the substrate

if a delayed feedback, which is particularly apt to generate oscillations (Stpn, 1989), exists, as it does in the case of photorespiration
(Ivlev, 1989; Roussel et al., 2007). The delay, clearly visible in the
phenomenon of post-illumination burst (Doehlert et al., 1979), is
introduced by the sequence of chemical transformations (Roussel,
1996) and transport processes intervening between oxygenation
and the appearance in the chloroplasts of the CO2 generated during the recycling of the oxygenation product 2-phosphoglycolate.
This oscillatory dynamics can improve overall process efciencies (Richter and Ross, 1981; Dolmetsch et al., 1998), manifested
through a variety of resonance effects in stochastic systems (Hahn
et al., 1974; Lindner et al., 2004). In the case of Rubisco, CO2 pulsing
might increase the efciency of CO2 delivery to the carboxylation
centers (Roussel and Igamberdiev, 2011).
For effective operation of the photorespiratory CO2 pump, it is
important that CO2 released in the glycine decarboxylase reaction
is effectively channeled back to chloroplast. In fact, it was shown
that the carbon released in photorespiration is mostly rexed
(Loreto et al., 1999). The photorespiratory CO2 loss thus should
be attributed to CO2 that is not xed due to competition from the
oxygenase reaction of Rubisco rather than to actual CO2 release.
Some estimates give a value of more than 80% CO2 rexed after
photorespiratory release in ambient air conditions (Loreto et al.,
1999). The rate of the photorespiratory CO2 rexation may be constrained by the mitochondrial CA activities, discovered rst as a
soluble matrix -form in Chlamydomonas and then as a membrane bound -form associated with complex I in higher plants
(Sunderhaus et al., 2006). Plants lacking the mitochondrial complex
I exhibited increased rates of photorespiration due to a decreased
mesophyll conductance to CO2 (Priault et al., 2006). This can be
linked in this mutant to the lack of CA activity associated with the
mitochondrial complex I. Both the complex I-linked CA as well as a
matrix-localized CA (Igamberdiev, unpublished) could be involved
in an intracellular CO2 transport system in higher plants.
For effective operation of photorespiratory CO2 channeling from
mitochondria, Raven (2001) postulated the presence of bicarbonate channels that facilitate CA-assisted HCO3 transport to the
mitochondrial intermembrane space and then to the cytosol and
chloroplast. There are several unresolved questions regarding this
transport but the observed values of a higher CO2 afnity of

photosynthesis in pea protoplasts as compared to chloroplasts

(Riazunnisa et al., 2006) indicate the existence of an effective
system of concentration of photorespiratory CO2 . Giordano et al.
(2003) showed that in Chlamydomonas reihardtii the mitochondrial
CA is involved in the inorganic carbon-concentrating mechanism
at low CO2 . If the photorespiratory carbon leaves the mitochondria
as HCO3 , this will result in limiting the potential for CO2 leakage
through the plasmalemma and increased CO2 supply to Rubisco
(Raven, 2001).
5. Conclusions
Rubisco is a striking example of an enzyme that operates at
much higher concentration than its substrate and CO2 pumping
to its active site is a major prerequisite of its stable operation.
The operation of Rubisco is highly correlated with chloroplast
electron transport and is strongly dependent on feedforward CO2
supply associated with the CA mechanism. CO2 is pumped from
the thylakoid lumen by a mechanism that involves PS II CA activity coordinated with water splitting, chloroplast electron transport,
and ATP synthesis. At low ambient CO2 , the concentrating mechanism of the thylakoidal CA does not itself produce enough CO2
to utilize all the reducing power generated in light reactions.
Under these conditions, the oxygenase reaction of Rubisco generates a sink for reducing power and energy, and photorespiration
supplies CO2 to keep the Calvin cycle operating and Rubisco
functioning. At all CO2 concentrations, there is a strong linkage
between NADPH production, ATP generation (adjusted also via
cyclic phosphorylation and non-coupled pathways), and Rubisco
activity associated with the C3 Calvin cycle at high CO2 and with
both C3 (Calvin) and C2 (photorespiratory) cycles at low CO2
(summarized in Fig. 6). This supports a stable non-equilibrium
state according to Bauer (1982), who wrote that the comparison of living system with waterfall becomes correct if we assume
that the difference in the levels of water, which is an indispensable condition for the fall, should be made and maintained
by the waterfall itself (p. 239). The ux through the activated
Rubisco-enediol is apparently steady and roughly independent of
different CO2 concentrations (considering that CO2 can be substituted by O2 ), due in large part to the pumping of the CO2
substrate via the thylakoid CA and/or through photorespiration.
At low CO2 the supply of this substrate is limited by the photorespiratory feedback while at high CO2 the supply is limited
by the capacity of the thylakoidal CO2 pump. Operating at lower
CO2 concentrations than its protein concentration, Rubisco can
stabilize the ux and coordinate it with NADPH and ATP produced in light reactions. By exploiting Rubisco, which appeared in
evolution as an inefcient enzyme with a low catalytic constant
and low specicity, having originated from a protein possessing
a different function, living systems had acquired the most optimal parameters of carbon xation, coordinated with building up

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

reducing power and ATP and oxygen utilization, to support a

stable non-equilibrium condition in such an energetically intensive
process as photosynthesis.
Acknowledgement
This work was supported by the Natural Sciences and Engineering Research Council of Canada.
References
Andr, M.J., 2011a. Modelling 18 O2 and 16 O2 unidirectional uxes in plants: I. Regulation of pre-industrial atmosphere. Biosystems 103, 239251.
Andr, M.J., 2011b. Modelling 18 O2 and 16 O2 unidirectional uxes in plants: II. Analysis of Rubisco evolution. Biosystems 103, 252264.
Bar-Even, A., Noor, E., Lewis, N.E., Milo, R., 2010. Design and analysis of synthetic
carbon xation pathways. Proc. Natl. Acad. Sci. U.S.A. 107, 88898894.
Baeyer, A.D., 1870. Ueber die Wasserentziehung und ihre Bedeutung fr das
Panzenleben und die Ghrung. Ber. Deutsch. Chem. Ges. 3, 6375.
Bauer, E.S., 1982. Theoretical Biology. Akadmiai Kiad, Budapest (Originally published 1935).
Borghans, J.A.M., De Boer, R.J., 1996. Extending the quasi-steady state approximation
by changing variables. Bull. Math. Biol. 58, 4363.
Briggs, G.E., Haldane, J.B.S., 1925. A note on the kinetics of enzyme action. Biochem.
J. 19, 338339.
Butlerow, A., 1861. Bildung einer zuckerartigen Substanz durch Synthese. Ann.
Chem. Pharm. 44, 295298.
Cme, G.M., 1979. Mechanistic modelling of homogeneous reactors: a numerical
method. Comput. Chem. Eng. 3, 603609.
Dolmetsch, R.E., Xu, K., Lewis, R.S., 1998. Calcium oscillations increase the efciency
and specicity of gene expression. Nature 392, 933936.
Doehlert, D.C., Ku, M.S.B., Edwards, G.E., 1979. Dependence of the post-illumination
burst of CO2 on temperature, light, CO2 , and O2 concentration in wheat (Triticum
aestivum). Physiol. Plant. 46, 299306.
Duanmu, D., Miller, A.R., Horken, K.M., Weeks, D.P., Spalding, M.H., 2009. Knockdown
of limiting-CO2 -induced gene HLA3 decreases HCO3 transport and photosynthetic Ci afnity in Chlamydomonas reinhardtii. Proc. Natl. Acad. Sci. U.S.A. 106,
59905995.
Dubbs, J.M., Tabita, F.R., 2004. Regulators of nonsulfur purple phototrophic bacteria and the interactive control of CO2 assimilation, nitrogen xation, hydrogen
metabolism and energy generation. FEMS Microbiol. Rev. 28, 353376.
Easterby, J.S., 1989. The analysis of metabolite channelling in multienzyme complexes and multifunctional proteins. Biochem. J. 264, 605607.
Eichelmann, H., Talts, E., Oja, V., Padu, E., Laisk, A., 2009. Rubisco in planta kcat is
regulated in balance with photosynthetic electron transport. J. Exp. Bot. 60,
40774088.
Ellis, R.J., 1979. The most abundant protein in the world. Trends Biochem. Sci. 4,
241244.
Farazdaghi, H., Edwards, G.E., 1988. A model for photosynthesis and photorespiration in C3 plants based on the biochemistry and stoichiometry of the pathways.
Plant Cell Environ. 11, 799809.
Farazdaghi, H., 2011. The single-process biochemical reaction of Rubisco: a unied
theory and model with the effects of irradiance, CO2 and rate-limiting step on
the kinetics of C3 and C4 photosynthesis from gas exchange. Biosystems 103,
265284.
Farquhar, G.D., 1979. Models describing the kinetics of ribulose biphosphate
carboxylase-oxygenase. Arch. Biochem. Biophys. 193, 456468.
Farquhar, G.D., von Caemmerer, S., Berry, J.A., 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 plants. Planta 149, 7890.
Fridlyand, L.E., Kaler, V.L., 1987. Possible CO2 concentration mechanism in chloroplasts of C3 plants. Role of carbonic anhydrase. Gen. Physiol. Biophys. 6, 617632.
Fridlyand, L.E., Kaler, V.L., 1988. Criteria indicating the existence of hypothetical
CO2 -concentrating mechanism in C3 plants and estimation of their efciency.
Sov. Plant Physiol. 35, 333339.
Fridlyand, L.E., 1995. On the possibility of existence of specic CO2 concentration
mechanis in chloroplasts of C3 plants. In: Mathis, P. (Ed.), Photosynthesis: From
Light to Biosphere, vol. 5. Kluwer, Netherlands, pp. 559562.
Fridlyand, L.E., Tsyuryupa, S.N., 1992. Inhibition of photosynthesis by supraoptimal
concentrations of CO2 and possible mechanisms of this phenomenon. Sov. Plant
Physiol. 39, 504507.
Fridlyand, L.E., Scheibe, R., 1999. Regulation of the Calvin cycle for CO2 xation as
an example for general control mechanisms in metabolic cycles. Biosystems 51,
7993.
Giordano, M., Norici, A., Forssen, M., Eriksson, M., Raven, J.A., 2003. An anaplerotic
role for mitochondrial carbonic anhydrase in Chlamydomonas reinhardtii. Plant
Physiol. 132, 21262134.
Gontero, B., Mulliert, G., Rault, M., Giudici-Orticoni, M.-T., Ricard, J., 1993. Structural
and functional properties of a multi-enzyme complex from spinach chloroplasts.
2. Modulation of the kinetic properties of enzymes in the aggregated state. Eur.
J. Biochem. 217, 10751082.
Hahn, H.-S., Nitzan, A., Ortoleva, P., Ross, J., 1974. Threshold excitations, relaxation
oscillations, and effect of noise in an enzyme reaction. Proc. Natl. Acad. Sci. U.S.A.
71, 40674071.

165

Hanson, D.T., Franklin, L.A., Samuelsson, G., Badger, M.R., 2003. The Chlamydomonas
reinhardtii cia3 mutant lacking a thylakoid lumen-localized carbonic anhydrase
is limited by CO2 supply to Rubisco and not photosystem II function in vivo.
Plant Physiol. 132, 22672275.
Hanson, S.M., Schnell, S., 2008. Reactant stationary approximation in enzyme kinetics. J. Phys. Chem. A 112, 86548658.
Harris, G.C., Kniger, M., 1997. The high concentrations of enzymes within the
chloroplast. Photosynth. Res. 54, 523.
Henri, V., 1902. Thorie gnrale de laction de quelques diastases. C. R. Acad. Sci.
135, 916919.
Igamberdiev, A.U., Bykova, N.V., Kleczkowski, L.A., 1999. Origins and metabolism of
formate in higher plants. Plant Physiol. Biochem. 37, 503513.
Igamberdiev, A.U., Bykova, N.V., Lea, P.J., Gardestrm, P., 2001. The role of photorespiration in redox and energy balance of photosynthetic plant cells: a study
with a barley mutant decient in glycine decarboxylase. Physiol. Plant. 111,
427438.
Igamberdiev, A.U., Gardestrm, P., 2003. Regulation of NAD- and NADP-dependent
isocitrate dehydrogenases by reduction levels of pyridine nucleotides in
mitochondria and cytosol of pea leaves. Biochim. Biophys. Acta 1606,
117125.
Igamberdiev, A.U., Kleczkowski, L.A., 2003. Membrane potential, adenylate levels
and Mg2+ are interconnected via adenylate kinase equilibrium in plant cells.
Biochim. Biophys. Acta 1607, 111119.
Igamberdiev, A.U., Mikkelsen, T.N., Ambus, P., Bauwe, H., Lea, P.J., Gardestrm, P.,
2004. Photorespiration contributes to stomatal regulation and carbon isotope
fractionation: a study with barley, potato and Arabidopsis plants decient in
glycine decarboxylase. Photosynth. Res. 81, 139152.
Igamberdiev, A.U., Kleczkowski, L.A., 2006. Equilibration of adenylates in the mitochondrial intermembrane space maintains respiration and regulates cytosolic
metabolism. J. Exp. Bot. 57, 21332141.
Igamberdiev, A.U., Lea, P.J., 2006. Land plants equilibrate O2 and CO2 concentrations
in the atmosphere. Photosynth. Res. 87, 177194.
Igamberdiev, A.U., Kleczkowski, L.A., 2009. Metabolic systems maintain stable nonequilibrium via thermodynamic buffering. Bioessays 31, 10911098.
Igamberdiev, A.U., Kleczkowski, L.A., 2011. Optimization of CO2 xation in photosynthetic cells via thermodynamic buffering. Biosystems 103, 224229.
Ivlev, A.A., 1989. On discreteness of CO2 assimilation by C3 plants in the light.
Biozika 34, 887891.
Kursar, T.A., Coley, P.D., 1993. Photosynthetic induction times in shade-tolerant
species with long and short-lived leaves. Oecologia 93, 165170.
Laisk, A., 1977. Kinetics of Photosynthesis and Photorespiration in C3-Plants. Nauka,
Moscow.
Lindner, B., Garca-Ojalvo, J., Neiman, A., Schimansky-Geier, L., 2004. Effects of noise
in excitable systems. Phys. Rep. 392, 321424.
Loreto, F., Delne, S., DiMarco, G., 1999. Estimation of photorespiratory carbon dioxide recycling during photosynthesis. Aust. J. Plant Physiol. 26, 733736.
McNevin, D., von Caemmerer, S., Farquhar, G., 2006. Determining RuBisCO activation
kinetics and other rate and equilibrium constants by simultaneous multiple nonlinear regression of a kinetic model. J. Exp. Bot. 57, 38833900.
Michaelis, L., Menten, M., 1913. Die Kinetik der Invertinwirkung. Biochem. Z. 49,
333369.
Moskvin, O.V., Ivanov, B.N., Ignatova, L.K., Kollmeier, M.A., 2000. Light-induced stimulation of carbonic anhydrase activity in pea thylakoids. FEBS Lett. 470, 375377.
Mueller-Cajar, O., Badger, M.R., 2007. New roads lead to Rubisco in Archaebacteria.
Bioessays 29, 722724.
Ngo, L.G., Roussel, M.R., 1997. A new class of biochemical oscillator models based
on competitive binding. Eur. J. Biochem. 245, 182190.
Park, Y.I., Karlsson, J., Rojdestvenski, I., Pronina, N., Klimov, V., Oquist, G., Samuelsson,
G., 1999. Role of a novel photosystem II-associated carbonic anhydrase in photosynthetic carbon assimilation in Chlamydomonas reinhardtii. FEBS Lett. 444,
102105.
Pedersen, M.G., Bersani, A.M., 2010. Introducing total substrates simplies theoretical analysis at non-negligible enzyme concentrations: pseudo rst-order
kinetics and the loss of zero-order ultrasensitivity. J. Math. Biol. 60, 267283.
Pettersson, G., Ryde-Pettersson, U., 1988. Effects of metabolite binding to ribulosebisphosphate carboxylase on the activity of the Calvin photosynthesis cycle. Eur.
J. Biochem. 177, 351355.
Pierce, J., Andrews, T.J., Lorimer, G.H., 1986. Reactions intermediate partitioning by
ribulose-bisphosphate carboxylases with differing substrate specicities. J. Biol.
Chem. 261, 1024810256.
Pons, T.L., Pearcy, R.W., Seemann, J.R., 1992. Photosynthesis in ashing light in soybean leaves grown in different conditions. I. Photosynthetic induction state and
regulation of ribulose-1,5-bisphosphate carboxylase activity. Plant Cell Environ.
15, 569576.
Portis Jr., A.R., 1995. The regulation of Rubisco by Rubisco activase. J. Exp. Bot. 46,
12851291.
Priault, P., Tcherkez, G., Cornic, G., DePaepe, R., Naik, R., Ghashghaie, J., Streb, P., 2006.
The lack of mitochondrial complex I in a CMSII mutant of Nicotiana sylvestris
increases photorespiration through an increased internal resistance to CO2 diffusion. J. Exp. Bot. 57, 31953207.
Pronina, N.A., Semenenko, V.E., 1988. Localization of bound carbonic anhydrase in
membranes of Chlorella cells. Sov. Plant Physiol. 35, 3846.
Pronina, N.A., Semenenko, V.E., 1990. Membrane-bound carbonic anhydrase takes
part in CO2 concentration in algae cells. In: Baltscheffsky, M. (Ed.), Current Research in Photosynthesis, vol. IV. Kluwer Academic Publishers, The
Netherlands, pp. 489492.

166

A.U. Igamberdiev, M.R. Roussel / BioSystems 107 (2012) 158166

Rault, M., Giudici-Orticoni, M.-T., Gontero, B., Ricard, J., 1993. Structural and functional properties of a multi-enzyme complex from spinach chloroplasts. 1.
Stoichiometry of the polypeptide chains. Eur. J. Biochem. 217, 10651073.
Raven, J.A., 2001. A role for mitochondrial carbonic anhydrase in limiting CO2 leakage
from low CO2 -grown cells of Chlamydomonas reinhardtii. Plant Cell Environ. 24,
261265.
Riazunnisa, K., Padmavathi, L., Bauwe, H., Raghavendra, A.S., 2006. Markedly low
requirement of added CO2 for photosynthesis by mesophyll protoplasts of pea
(Pisum sativum): possible roles of photorespiratory CO2 and carbonic anhydrase.
Physiol. Plant. 128, 763772.
Richter, P.H., Ross, J., 1981. Concentration oscillations and efciency: glycolysis.
Science 211, 715717.
Robinson, S.P., Portis Jr., A.R., 1988. Release of the nocturnal inhibitor,
carboxyarabinitol-1-phosphate, from ribulose bisphosphate carboxylase/oxygenase by Rubisco activase. FEBS Lett. 233, 413416.
Roussel, M.R., 1996. The use of delay differential equations in chemical kinetics. J.
Phys. Chem. 100, 83238330.
Roussel, M.R., Ivlev, A.A., Igamberdiev, A.U., 2007. Oscillations of the internal CO2
concentration in tobacco leaves transferred to low CO2 . J. Plant Physiol. 164,
11881196.
Roussel, M.R., Igamberdiev, A.U., 2011. Dynamics and mechanisms of oscillatory
photosynthesis. Biosystems 103, 230238.
Rudenko, N.N., Ignatova, L.K., Ivanov, B.N., 2007. Multiple sources of carbonic
anhydrase activity in pea thylakoids: soluble and membrane-bound forms. Photosynth. Res. 91, 8189.
Servaites, J.C., 1990. Inhibition of ribulose 1,5-bisphosphate carboxylase/oxygenase
by 2-carboxyarabinitol-1-phosphate. Plant Physiol. 92, 867870.
Shutova, T., Kenneweg, H., Buchta, J., Nikitina, J., Terentyev, V., Chernyshov, S., Andersson, B., Allakhverdiev, S.I., Klimov, V.V., Dau, H., Junge, W., Samuelsson, G.,
2008. The photosystem II-associated Cah3 in Chlamydomonas enhances the O2
evolution rate by proton removal. EMBO J. 27, 782791.
Stpn, G., 1989. Retarded Dynamical Systems: Stability and Characteristic Functions. Longman, NewYork.
Stucki, J.W., 1980. The thermodynamic-buffer enzymes. Eur. J. Biochem. 109,
257267.
Sunderhaus, Dudkina, S., Jansch, N.V., Klodmann, L., Heinemeyer, J., Perales, J.,
Zabaleta, M., Boekema, E., Braun, E.J.H.-P., 2006. Carbonic anhydrase subunits
form a matrix-exposed domain attached to the membrane arm of mitochondrial
complex I in plants. J. Biol. Chem. 281, 64826488.
Tabita, F.R., Hanson, T.E., Satagopan, S., Witte, B.H., Kreel, N.E., 2008. Phylogenetic
and evolutionary relationships of RubisCO and the RubisCO-like proteins and

the functional lessons provided by diverse molecular forms. Phil. Trans. R. Soc.
B 363, 26292640.
Tcherkez, G.G., Farquhar, G.D., Andrews, T.J., 2006. Despite slow catalysis
and confused substrate specicity, all ribulose bisphosphate carboxylases
may be nearly perfectly optimized. Proc. Natl. Acad. Sci. U.S.A. 103,
72467251.
Tzafriri, A.R., 2003. MichaelisMenten kinetics at high enzyme concentrations. Bull.
Math. Biol. 65, 11111129.
Tzafriri, A.R., Edelman, E.R., 2007. Quasi-steady-state kinetics at enzyme and substrate concentrations in excess of the MichaelisMenten constant. J. Theor. Biol.
245, 737748.
Viil, J., Ivanova, H., Prnik, T., 1999. Estimation of rate constants of the partial reactions of carboxylation of ribulose-1,5-bisphosphate in vivo. Photosynth. Res. 60,
247256.
Viil Yu, Ivanova, H., Prnik, T., Prsim, E., 2001. Measurement of the concentration of
the active centers of ribulose-1,5-bisphosphate carboxylase/oxygenase in vivo.
Russ. J. Plant Physiol. 48, 116121.
Viil, J., Prnik, T., 1995. The rate constant for the reaction of CO2 with enzyme-bound
ribulose 1,5-bisphosphate in vivo. J. Exp. Bot. 46, 13011307.
Villarejo, A., Shutova, T., Moskvin, O., Forssn, M., Klimov, V.V., Samuelsson, G., 2002. A photosystem II-associated carbonic anhydrase regulates the efciency of photosynthetic oxygen evolution. EMBO J. 21,
19301938.
Wang, D., Naidu, S.L., Portis, A.R., Moose Jr., S.P., Long, S.P., 2008. Can the cold tolerance of C4 photosynthesis in Miscanthus giganteus relative to Zea mays be
explained by differences in activities and thermal properties of Rubisco? J. Exp.
Bot. 59, 17791787.
Wienkoop, S., Weiss, J., May, P., Kempa, S., Irgang, S., Recuenco-Munoz, L., Pietzke,
M., Schwemmer, T., Rupprecht, J., Egelhofer, V., Weckwerth, W., 2010. Targeted
proteomics for Chlamydomonas reinhardtii combined with rapid subcellular protein fractionation, metabolomics and metabolic ux analyses. Mol. Biosyst. 6,
10181031.
Woodrow, I.E., Berry, J.A., 1988. Enzymatic regulation of photosynthetic CO2 xation
in C3 plants. Annu. Rev. Plant. Physiol. Plant. Mol. Biol. 39, 533594.
Woodrow, I.E., Kelly, M.E., Mott, K.A., 1996. Limitation of the rate of ribulosebisphosphate carboxylase activation by carbamylation and ribulosebisphosphate
carboxylase activase activity: development and tests of a mechanistic model.
Aust. J. Plant Physiol. 23, 141149.
Woodrow, I.E., Mott, K.A., 1993. Modelling C3 photosynthesis: a sensitivity analysis
of the photosynthetic carbon-reduction cycle. Planta 191, 421432.