Photorespiration: The Futile Cycle?

Abstract

Photorespiration, or C₂ photosynthesis, is generally considered a futile cycle that potentially decreases photosynthetic carbon fixation by more than 25%. Nonetheless, many essential processes, such as nitrogen assimilation, C₁ metabolism, and sulfur assimilation, depend on photorespiration. Most studies of photosynthetic and photorespiratory reactions are conducted with magnesium as the sole metal cofactor despite many of the enzymes involved in these reactions readily associating with manganese. Indeed, when manganese is present, the energy efficiency of these reactions may improve. This review summarizes some commonly used methods to quantify photorespiration, outlines the influence of metal cofactors on photorespiratory enzymes, and discusses why photorespiration may not be as wasteful as previously believed.

1. Introduction

Photorespiration involves the oxygenation of ribulose-1,5-bisphosphate (RuBP) to form 3-phosphoglycerate (3PGA) and 2-phosphoglycolate (2PG) and the subsequent carbon oxidation pathways that release CO₂ under light conditions [1,2,3,4,5]. Because it produces 2PG, a compound “toxic” to many enzymes in photosynthetic metabolism, and oxidizes organic carbon without seemingly generating ATP, photorespiration is generally considered a wasteful process. The following sections examines how the photorespiratory pathway converts 2PG into glycolate, the only carbon source for the photosynthetic carbon oxidation cycle [6], a cycle that together with nitrogen assimilation, C₁ metabolism, and sulfur assimilation generates essential amino acids and intermediate compounds [7]. Moreover, the three enzymes involved in the initial photorespiratory steps within chloroplasts—Rubisco, malic enzyme, and phosphoglycolate phosphatase—have metal binding sites that accommodate either Mg²⁺ or Mn²⁺, and balance between the binding of these enzymes to Mg²⁺ or Mn²⁺ may shift the relative rates and energy efficiencies of photosynthesis and photorespiration [8].

2. Photosynthesis vs. Photorespiration

2.1. Rubisco

Atmospheric CO₂ concentration has increased more than 20% during the past 35 years [9]. The major sink for this CO₂ is the approximately 258 billion tons per year that photosynthetic organisms convert into organic carbon compounds through carbon fixation via the Calvin–Benson pathway [10]. This pathway begins with Rubisco (Ribulose 1,5-bisphosphate carboxylase–oxygenase), the most abundant protein on the planet [11].

Rubisco comes in three forms [12]:

Form I, which is found in cyanobacteria, proteobacteria, chlorophyte algae, heterokont algae, and haptophyte algae, and higher plants, is the most common [13,14]. It is a hexadecamer containing eight identical large subunits (~55,000 M_r), each with a metal-binding site, and eight small subunits (~15,000 M_r). The large subunits are coded by a single plastomic gene, whereas the small subunits are coded by a nuclear multigene family that consists of 2 to 22 members, depending on the species [15]. Complex cellular machinery is required to assemble this form of Rubisco and to maintain its activity [16]. Form I Rubisco, until recently, had resisted all efforts to generate a functional holoenzyme in vitro or upon recombinant expression in E. coli [17].

Form II Rubisco, found in proteobacteria, archaea, and dinoflagellate algae, contains one or more isodimers with subunits that share about 30% identity to the large subunit of Form I Rubisco [8].

Form III Rubisco, found in archaea, has one or five isodimers composed of subunits homologous to the large subunit of Form I Rubisco [8].

Form II and Form III Rubisco show greater similarity in their primary sequence to one another than either do to the large subunit of Form I Rubisco [8].

All three forms of Rubisco catalyze not only the reaction in which the carboxylation of the five-carbon sugar RuBP generates two molecules of the three-carbon organic acid 3-phosphoglycerate (3PGA), but also an alternative reaction in which oxidation of RuBP generates one molecule of 3PGA and one of 2PG (Figure 1) [8]. The carboxylation pathway of photosynthesis expends 3 ATP and 2 NADPH per RuBP regenerated and produces a carbon in hexose [18], whereas the oxygenation pathway of photorespiration reportedly expends 3.5 ATP and 2 NADPH per RuBP regenerated but produces no additional organic carbon [19,20].

Rubisco must be activated before it can carboxylate or oxygenate RuBP. Activation of the three forms of Rubisco involves binding of Mn²⁺ or Mg²⁺ [21,22]. Binding of Mg²⁺ requires carbamylation of Rubisco by the addition of CO₂. One histidine at the active site of Rubisco rotates into an alternate conformation, resulting in a transient binding site where Mg²⁺ is partially neutralized by the conversion of two water molecules to hydroxide ions and coordinated indirectly by three histidine residues through the water molecules. Subsequently, the hydroxide ions cause a lysine residue at the active site to become deprotonated and rotate 120 degrees into the trans conformer, which brings its nitrogen into close proximity to the carbon of CO₂, allowing for the formation of a covalent bond that produces a carbamyl group. This carbamyl group causes the Mg²⁺ ion to transfer to a second binding site, after which the histidine that first rotated returns to its original conformation [23]. It is unclear whether binding Mn²⁺ follows a similar mechanism and whether it requires an activator CO₂ to be bound first [21,22]; hence, understanding the mechanism of Mn²⁺ binding to Rubisco is important to future research on Rubisco kinetics. During in vitro studies, Rubisco is often activated at pH 8.0 in the presence of CO₂ and either Mg²⁺ or Mn²⁺.

Rubisco can also bind to other metals. When bound to Fe²⁺, Ni²⁺, Cu²⁺, Ca²⁺, or Co²⁺, Rubisco may exhibit some carboxylase and oxygenase activity [24]. For example, one study found that Rubisco from R. rubrum, when bound to Co²⁺, was incapable of carboxylation but still capable of oxygenation [24]. Another study found that Rubisco from spinach performed both carboxylation and oxygenation when bound to Ni²⁺ or Co²⁺ [25]. When bound to some other metal ions, including Cd²⁺, Cr²⁺, and Ga²⁺, Rubisco cannot catalyze either carboxylation or oxygenation [24]. Although it is known that the metal ion plays a role in stabilizing the activator carbamate and determining the active site’s structure, its effect upon the reactions catalyzed by Rubisco is still not completely understood. One hypothesis is that Mg²⁺, because of its electron-withdrawing properties, polarizes the C2 carbonyl of RuBP, which favors the removal of the C3 proton and thereby contributes to enolization [21].

NADPH complexes strongly with Rubisco and acts as an effector molecule to maintain the Rubisco catalytic pocket in an open confirmation that more rapidly facilitates CO₂-Mg²⁺ activation when CO₂ and Mg²⁺ are present in suboptimal concentrations [26,27,28,29]. The crystal structure of Rubisco with both Mg²⁺ and NADPH as ligands indicates that NADPH binds to the catalytic site of Rubisco through metal-coordinated water molecules [26]. The activated enzyme catalyzes either carboxylation or oxygenation of the enediol form of the five-carbon sugar ribulose-1,5-bisphosphate (RuBP) [14,21,22,30,31].

2.2. Balance between Carboxylation and Oxygenation and Metal Cofactors

Several factors alter the balance between Rubisco carboxylation and oxygenation and, thereby, alter the relative rates of photosynthesis and photorespiration. These include the concentrations of CO₂ and O₂ at the active site of Rubisco, the specificity of the enzyme for each gas, and whether the enzyme is associated with Mg²⁺ or Mn²⁺ [32]. These divalent cations share the same binding site in Rubisco [14,22,33], and in tobacco, Rubisco associates with both metals and rapidly exchanges one metal for the other [32]. Nonetheless, nearly all recent studies on the photosynthetic and photorespiratory pathways have been conducted in the presence of Mg²⁺ and absence of Mn²⁺ [8]. Rubisco binding of Mg²⁺ accelerates carboxylation, whereas binding of Mn²⁺ slows carboxylation [25,34,35,36,37,38]. Chloroplastic Mg²⁺ and Mn²⁺ activities seem to be regulated at the cellular level because in isolated tobacco chloroplasts, activities of the metals were directly proportional to their concentrations in the medium [32]. The thermodynamics of binding Mg²⁺ to Rubisco were similar for enzymes isolated from a Form I and a Form II species [32]. By contrast, the thermodynamics of binding differed greatly between the two Rubisco forms when the enzymes were associated with Mn²⁺ [32].

Mg²⁺ and Mn²⁺ have nearly identical ionic radii but highly disparate electron configurations: Mg²⁺ (1s²2s²2p⁶ or [Ne]) has a very stable outer shell [8], whereas Mn²⁺ has five unpaired d electrons (1s²2s²2p⁶3s²3p⁶3d⁵ or [Ar]3d⁵) that are susceptible to many redox reactions. An aerated solution of activated Mn²⁺-Rubisco exhibits a long-lived chemiluminescence when RuBP is added [39,40]. This chemiluminescence was attributed to a spin-flip within the Mn²⁺ 3d manifold, leading to an excited quartet (S = ³/₂) d⁵ electronic configuration that decays over the course of 1 to 5 min back to the sextet (S = ⁵/₂) ground state electronic configuration [39]. Excited states are intrinsically better oxidants and reductants (larger reduction/oxidation potentials) than their corresponding ground states [41,42,43]; thus, the observed chemiluminescence opens the possibility that the RuBP-O₂-Mn²⁺—Rubisco excited state may be quenched via electron transfer. Consequently, the liberated reducing equivalent could participate in the reduction of NADP⁺ to NADPH (Figure 2, blue pathway). In this way, oxidation of RuBP via O₂ may proceed in a spin-allowed manner, while the Mn²⁺ remains “innocent” in the generation of the oxygenated RuBP precursor. Mn²⁺-centered redox may still proceed, with oxidation of excited Mn²⁺ to Mn³⁺ occurring in a manner independent of, but parallel to, substrate oxidation.

In wheat leaves, the ratio of Mn²⁺ to Mg²⁺ contents increased as the CO₂ levels increased and when the nitrogen source was nitrate rather than ammonium [32]. Nitrate assimilation into amino acids in shoots is heavily dependent on photorespiration, whereas ammonium assimilation is much less so. This indicates that plants shifted to Rubisco Mn²⁺ binding in order to compensate for the slower photorespiration rates and slower amino acid production that would otherwise occur under elevated CO₂ and nitrate nutrition.

2.3. The Photorespiratory Pathway

The 3-phosphoglycerate produced during photorespiration, like that produced during photosynthesis, is converted to triose phosphate and used to regenerate RuBP. On the other hand, 2-phosphoglycolate is converted to glycolate by phosphoglycolate phosphatase. In the peroxisome and mitochondrion, a series of reactions converts glycolate to glycerate, which is ultimately returned to the chloroplast to regenerate RuBP (Figure 2) [8]. In addition to Rubisco, several other chloroplast enzymes in the photorespiratory pathway, including malic enzyme and phosphoglycolate phosphatase, bind either Mg²⁺ or Mn²⁺ [8]. The plastid isoform of malic enzyme in Arabidopsis and tobacco catalyzes the reverse pyruvate synthesis reaction (pyruvate + CO₂ + NADPH → malate + NADP) [44,45]. Phosphoglycolate phosphatase, which is responsible for the hydrolysis of 2-phosphoglycolate to glycolate, binds to and is activated by either metal [46]. Hypothesized is an alternative photorespiratory pathway that increases photorespiration energy efficiency by generating malate (RuBP + O₂ + CO₂ + H₂O → glycolate + malate + 2Pi) when Mn²⁺ binds to these enzymes (Figure 2) [8].

3. Estimating Rates of Photorespiration

Many different methods have been employed for estimating rates of photorespiration. The following sections outline the general approach of each method and highlights the assumptions and potential errors in each. The hope is that certain methods might be better suited for assessing the influence of Mn²⁺ vs. Mg²⁺ on the relative rates of oxygenation and carboxylation in situ.

3.1. Traditional Methods for Estimating Photorespiration

3.1.1. Post Illumination CO₂ Burst

This method measures the evolution of CO₂ from a leaf for 1 to 4 min after turning off the light because glycine metabolism continues longer in the dark than CO₂ assimilation [47]. The rate of CO₂ generation is measured by a transient CO₂ analyzer [48] when the light has just been turned off or at the maximum rate of CO₂ evolution observed. CO₂ assimilation, however, does not stop immediately after the light is off. Separating CO₂ assimilation from the CO₂ burst effects during this time is difficult, and hence this method underestimates photorespiratory rates [49,50]. This method also fails to consider variations in mitochondrial respiration, leading to overestimates of photorespiratory rates [51].

3.1.2. O₂ Inhibition of Net CO₂ Assimilation

This method aims to assess the photorespiration rate from the increase in the CO₂ assimilation rate after switching from normal to low O₂ concentrations. Yet, changes in CO₂ assimilation with O₂ concentration may derive from components of the photosynthetic pathway other than photorespiration [4]. For example, when starch and sucrose synthesis limit photosynthesis, increasing or decreasing the photorespiration does not affect net CO₂ assimilation [52].

3.1.3. Photorespiration CO₂ Efflux into CO₂-Free Air

This method estimates photorespiration from the CO₂ efflux rate in CO₂-free air. A high-O₂ and low-CO₂ environment, however, promotes photorespiration [4]. Additionally, a CO₂-free atmosphere inhibits both the activity of Rubisco [53] and the regeneration of its substrate RuBP [54], leading to underestimates of photorespiration.

3.1.4. Ratio of ¹⁴CO₂ to ¹²CO₂ Uptake

In this method, ¹⁴CO₂ and ¹²CO₂ fluxes are measured after feeding a leaf with ¹⁴CO₂ for a short period of time. Gross photosynthesis is estimated from ¹⁴CO₂ uptake measured using an ionization chamber attached to an electrometer, while net photosynthesis is estimated from ¹²CO₂ measured using an infrared gas analyzer. Photorespiration is estimated as the difference between gross and net photosynthesis [55] (Figure 3).

There are several uncertainties associated with this method. The recycling effect on the specific activity of CO₂ inside the leaf can cause about a 20% error. One must consider the specific activity of CO₂ inside the leaf to obtain an accurate estimate of the gross photosynthesis rate because CO₂ efflux through photorespiration dilutes the ¹⁴C label in the intercellular spaces, decreasing the specific activity of CO₂. The activity might be even lower at the actual carboxylation site than in the interleaf spaces because of photorespiratory CO₂ loss [56]. Moreover, Rubisco carboxylation discriminates about 2.9% against ¹³C [57,58] and about 5.5% against ¹⁴C [4], resulting in errors in estimations of photorespiration rates that exceed 25% [4,57].

3.2. Recent Methods for Estimating Photorespiration

3.2.1. Calculation from Kinetics Models

Rubisco reaction kinetics can provide an estimate of the photorespiration rate [4,59]. This method can provide accurate estimates of photorespiration rates if the CO₂ compensation point in the absence of mitochondrial respiration (Γ*) being known for a given plant species. The rate of oxygenation, which is assumed to be twice the rate of photorespiration, is given by:

$$v_o=\frac{A+R_d}{\frac{1}{\mathsf{\Phi }}-0.5}$$

where A is the rate of photosynthetic CO₂ assimilation, R_d is the rate of respiration other than photorespiration, and:

$$\mathsf{\Phi }=\frac{v_o}{v_c}=\frac{2{\mathsf{\Gamma }}^{*}}{C}$$

where C is the CO₂ concentration.

The principal drawbacks of this method are that it does not directly measure photorespiration and depends on estimates of C and Γ*. There are several techniques for estimating C at the site of Rubisco activity, but estimating Γ* is more difficult. Values for Γ* are known for only a few species, and depend on estimates of kinetic parameters, which themselves rely on estimates of photorespiration [59].

3.2.2. CO₂ Efflux into ¹³CO₂-Air

The gas exchange method is based on the FvCB (Farquhar, van Caemmerer, and Berry) model [4,59]. First, ambient air is rapidly replaced with air containing ¹³CO₂ and no ¹²CO₂. The levels of released ¹²CO₂ can be measured either using an infrared gas analyzer or a membrane inlet mass spectrometer. Because the rate of ¹²CO₂ release includes both photorespiration and mitochondrial respiration, additional effort is needed to separate these effects. For example, in one approach, the rate of ¹²CO₂ release is calculated as:

$$R_{{12}_C}=\frac{F\left({}^{12}{C}_R-{}^{12}{C}_S\frac{1-W_R}{1-W_S}\right)}{a}$$

where F is the gas flow rate, ¹²C_R and ¹²C_S are the mole fractions of ¹²CO₂ in the chamber without and with a leaf, W_R and W_S are the corresponding water mole fractions, and a is the illuminated leaf area in the chamber [60]. To divide this quantity into photorespiration and mitochondrial respiration, the air is replaced with air containing 10,000 ppm ¹³CO₂ and the concentration of ¹³CO₂ over 2 min is fitted to an exponential curve. The mitochondrial respiration R_d is taken to be the rate of ¹²CO₂ release after 2 min.

This method can provide estimates of both carboxylation and oxygenation if one assumes that the rate of mitochondrial respiration (R_d) is not affected by the sudden high CO₂ concentration, that 0.5 CO₂ is generated per oxygenation reaction when the CO₂ released per oxygenation varies widely with temperature and light level and among species [7], and that leaves do not naturally contain any ¹³C [59]. If intracellular reassimilation is significant and it often is [60], substantial errors in the estimate can result. These errors can be accounted for by monitoring the release of ¹²CO₂ after switching from ambient air to air with a high concentration of ¹³CO₂; however, high CO₂ concentrations could affect mitochondrial respiration and thus produce error in the estimate of photorespiration. The presence of naturally occurring ¹³C also generates additional errors [59].

3.2.3. Labelling of Photosynthates with ¹⁴C

Leaves at a photosynthetic steady state are exposed to ¹⁴CO₂ for different lengths to label primary and stored photosynthates. Exposing the leaf to an ambient concentration of ¹⁴CO₂ for 10 to 15 min will label primary photosynthates, such as the metabolites from the Calvin cycle, glycolate cycle, and intermediates of starch and sucrose synthesis and of glycolysis [44]. Longer exposures (2 to 3 h) will label stored photosynthates, such as starch, sucrose, fructans, and vacuolar acids. ¹⁴CO₂ efflux into different backgrounds containing various combinations of O₂ and CO₂ concentrations provides an estimate of photorespiration [61,62]. Four different backgrounds are used: first, 21% O₂ and ambient CO₂ to measure the steady-state release of CO₂ from both photosynthesis and photorespiration; second, 1.5% O₂ and ambient CO₂ to measure the rate of photorespiration only; third, 21% O₂ and 30,000 μmol/mol CO₂ to limit CO₂ reassimilation; and fourth, 21% O₂ with no CO₂ to measure the specific radioactivity of CO₂ efflux [63].

The assumptions for this method are that all photosynthates must be labeled during the labeling time frames and that R_d is not affected by the percentage of O₂ in the air. A recent report indicated that R_d was actually lower at a lower O₂ concentration (2%) than at an ambient concentration (21%) [64]. One also has to assume that the mitochondrial respiration (R_d) value does not change upon transient exposure to high CO₂ levels.

3.2.4. Measuring Photorespiratory Ammonia

Photorespiration generates NH₃ in addition to CO₂ during the conversion from glycine to serine in mitochondria [65]. Adding glutamine synthetase (GS) inhibitors methionine sulphoximine [35] or phosphinothricin [66] prevents ammonia reassimilation in chloroplasts, and NH₃ subsequently accumulates in the leaf. The advantages of this method also include the prevention of CO₂ refixation and uncertainties in R_d values under the experimental conditions [35,66]. This approach, however, depends on several assumptions: (1) The GS inhibitors do not inhibit photorespiration, and (2) they can prevent NH₃ refixation completely.

Other factors might limit the diffusion of NH₃ out of the leaves, leading to an underestimation of photorespiration [59]. GS inhibitors will disrupt the C₂ cycle under photorespiratory conditions, and glycolate will rapidly accumulate, which in turn will inhibit photosynthesis. Feeding the plant an amino acid donor, such as glutamine, together with GS inhibitors will help minimize this inhibition effect [66,67].

Quantification of ammonia poses some challenges. The commonly used ion chromatography method to quantify NH₄⁺ may overestimate the amount of NH₄⁺ because methylamine, ethylamine, ethanolamine, and some non-protein amino acids co-elute with NH₄⁺. Degradation of labile nitrogen metabolites in leaf extract, xylem sap, and apoplastic fluid to NH₄⁺ during extraction will cause further overestimation of NH₄⁺ levels [68].

3.2.5. Measuring ¹⁸O₂ Consumption and Labeled Metabolites

Replacing ambient air in a chamber containing a leaf with air containing ¹⁸O₂ provides another estimate of the photorespiration rate. A mass spectrometer measures levels of ¹⁶O₂ and ¹⁸O₂. The rate of oxygenation is estimated as:

$$v_o=\frac{2}{3}\left({}^{18}{O}_2 uptake in light-{ }^{18}{O}_2 uptake in dark\right)$$

and carboxylation as:

$$v_c={}^{16}{O}_{2}evolution-v_o$$

Ref. [69,70].

Unfortunately, this method cannot separate photorespiration from other light-dependent O₂-consuming processes, such as light-dependent differences in the rate of mitochondrial respiration [46,48]. To diminish these errors, the mass spectrometer can quantify ¹⁸O-labeled metabolites, such as glycolate, glycine, and serine; with several assumptions about the photorespiratory pathway, such as the pool sizes of the labeled metabolites [49], one can then use the amounts of labeled metabolites to calculate the photorespiration rate [71,72].

3.2.6. NMR Measurements on ¹³C-Labeled Metabolites

This method requires that plants receive fertilizer labeled with ¹⁵N and that leaves subsequently be exposed to ¹³CO₂. Rotational-echo double resonance (REDOR) detects ¹³C within two covalent bonds of ¹⁵N and thus assesses the formation of organic nitrogen metabolites labeled with ¹³C [59,73]. The ratio of ¹³C-labeled to unlabeled phosphorylated Calvin–Benson cycle metabolites between 2 and 4 min after exposure to ¹³CO₂ indicates the ratio of photosynthesis to photorespiration [50]. This assumes that metabolites produced from photosynthesis are fully labeled in less than 2 min after being exposed to ¹³CO₂ and that those produced from photorespiration do not become labeled until after 4 min. These assumptions may lead to errors because photosynthesis may re-assimilate some of the ¹²CO₂ generated by photorespiration and because photorespiration may produce intermediates labeled with ¹³C in less than 2 min [60]. Furthermore, this method is based on the premise that photorespiration releases one CO₂ for every two oxygenations, when the CO₂ released per oxygenation varies widely with temperature and light level and among species [7].

3.2.7. Quantification of 2-Phosphoglycolate (2PG) and Photorespiratory Metabolites by Mass Spectrometry

This method uses LC-MS/MS to measure directly the first intermediate, 2PG, of photorespiration when Rubisco oxygenates RuBP, and GC-MS to measure other photorespiratory metabolites. In the LC-MS/MS portion, 2PG is separated from other molecules in three steps: First, liquid chromatography separates 2PG based on its physiochemical properties; second, mass spectrometry separates 2PG based on its m/z ratio; and third, mass spectrometry separates 2PG based on its m/z ratio after being fragmented [74]. Readings from the LC-MS/MS samples are compared with 2PG standard solutions [75]. Additionally, GC-MS is used to quantify additional photorespiratory metabolites, such as glycolate, glyoxylate, glycine, serine, hydroxypyruvate, and glycerate [74,76].

This approach has estimated photorespiratory rates in plant mutants deficient in expression of genes coding for photorespiratory enzymes. The gaseous environment of the aerial part of the plant, but not the root, was altered before experimentally determining the changes in the metabolite (2PG) content [77,78,79,80,81].

This method has several problems [74,82,83,84,85]. First, non-volatile salts and metabolites were deposited at the inlet of MS/MS after eluting from the LC step, which is very common when using anion-exchange chromatography [84]. Second, numerous metabolites eluted from the LC step had overlapping and asymmetrical peaks resulting from the matrix effect (interference in the ionization between compounds with similar elution times) [82,85], which significantly affects the sensitivity and accuracy of the measurements on a specific metabolite, such as 2PG. Third, post-harvest changes in metabolite concentrations can severely affect the quantification of 2PG [74,83]. Fourth, the GC-MS step is not targeted and therefore is potentially prone to error if other compounds with a similar molar mass as the photorespiratory metabolites are present [76].

3.2.8. CO₂ Labeling and MS Analysis

Isotopically nonstationary metabolic flux analysis (INST-MFA) can trace ¹³C-labeled photorespiratory metabolites in plants exposed to ¹³CO₂ to assess the photorespiration rate [86,87,88,89,90]. Monitoring the isotope incorporation in downstream metabolites over time assesses the relative contributions of different pathways after administration of the tracer. The turnover rates of each enzyme determine the labeling dynamics (Figure 4). Mathematical metabolic models specific for each pathway are often used to enumerate mass and isotopomer balances and ensure atoms’ conservation within the system. The models’ proposed metabolic fluxes are compared with those measured experimentally, and differences are minimized with each subsequent iteration.

The INST-MFA approach presents several challenges. A minimum of three sample time points is needed for precise measurements of metabolic fluxes [91]. This makes experimental design more complex and time-consuming. To ensure accurate and precise measurements, the pool size for each component of a metabolic pathway has to be very specific. Absolute quantification of intracellular pool sizes, however, is not yet possible even with pool size measurements made with optimized mathematical modeling [91]. A second challenge of this approach is isotopic transients. Some intracellular metabolites can exhibit short isotopic transients that last only for a few minutes or seconds. Rapid sampling and quenching have to be achieved to obtain precise and meaningful INST-MFA measurements [92].

3.2.9. Micro-Optode Measurement of O₂ Consumption

We have been conducting direct oxygenation rate measurements using a needle-type O₂ micro-optode to examine the effects of metal cofactors on Rubisco photorespiration reactions. In this instrument, a polymer optical fiber transmits the excitation wavelength to the tip of the sensor and at the same time transmits the fluorescence response of an oxygen-sensitive dye that is immobilized in a polymer matrix at the tip. The rate of oxygenation can be calculated easily by comparing the amount of quenching of the excitation light by dissolved O₂. The micro-optode has a 50–70-µm tip diameter, which makes it possible for a micro-scale setup, such as in a micro-cuvette or plate. The most important advantages for this type of sensor are that the micro-optode does not consume O₂ in contrast to the other commonly used O₂ sensors, such as a Clark electrode [93,94]; it has no stirring sensitivity; and it is resistant to most corrosive environments. The micro-optode also works in both gas (%O₂) and liquid phases (DO), which makes it possible to measure O₂ exchanges accurately up to 250% air O₂ saturation in intact plant leaves, bioreactors, cell cultivation, microtiter plates, and many general oxygen measurements in liquids [95,96,97,98,99,100].

4. Photorespiration and Other Metabolic Pathways

4.1. NO₃⁻ Assimilation

Multiple lines of evidence link shoot NO₃⁻ assimilation to photorespiration:

(a)

Elevated CO₂ or low O₂ levels inhibited shoot NO₃⁻ reduction [101].

(b)

In independent ¹⁴N and ¹⁵N labeling experiments, assimilation of either ¹⁴N–NO₃⁻ or ¹⁵N–NO₃⁻ decreased under CO₂ enrichment [102].

(c)

Under elevated CO₂ conditions, NO₃^– nutrient absorption and organic N accumulation levels in various plant species declined when plants received NO₃⁻ as a sole N source [102,103,104,105,106].

(d)

C₃ plants receiving NO₃⁻ as their sole N source experienced slower growth under CO₂ enrichment than those receiving NH₄⁺ [9,107,108].

In wheat and Arabidopsis plants grown under CO₂ enrichment and receiving NO₃⁻ containing ¹⁵N at natural abundance levels, shoot tissues became less enriched with ¹⁵N organic compounds [102,109]: elevated CO₂ inhibited shoot NO₃⁻ reduction so it was less limited by nitrate availability, and NO₃⁻ reductase discriminated more strongly against ¹⁵N–NO₃⁻ [110].

The assimilatory quotient (AQ) is the ratio of net CO₂ consumption to net O₂ evolution in plant shoots [111]. During shoot NO₃⁻ assimilation, ferredoxin generated from the photosynthetic electron chain reduces NO₂⁻ to NH₄⁺ rather than producing NADPH, and so net O₂ evolution increases without a change in net CO₂ consumption. Therefore, the change in assimilatory quotient (∆AQ) when a plant receives NH₄⁺ instead of NO₃⁻ as a sole N source provides an estimate of shoot NO₃⁻ assimilation [106]. ∆AQ decreased as the shoot internal CO₂ concentration increased in C₃ plants (Figure 5) [9,104,112,113].

Shoot CO₂ and O₂ fluxes at ambient and elevated CO₂ were contrasted between stages of plant development or genotypes that have significantly different NO₃⁻ reductase activities in situ (i.e., 36- vs. 48-day-old wild-type Arabidopsis, Arabidopsis NO₃⁻ reductase knockout mutants vs. transgenic Arabidopsis overexpressing NO₃⁻ reductase, and NO₃^– reductase-deficient barley mutants vs. wild-type barley) [104,112]. ∆AQ, a measure of shoot NO₃^– assimilation, differed between these stages of development and genotypes under ambient CO₂ but not under elevated CO₂. This indicates that none of the stages of development or genotypes were assimilating NO₃⁻ under elevated CO₂ [104,112].

Maximum NO₃⁻ reductase activity in vitro generally declined under CO₂ enrichment [105,114]. Nonetheless, shoot NO₃⁻ reductase activity seldom limits NO₃⁻ assimilation in planta [115,116]. Accordingly, NO₃⁻ assimilation significantly declined only in genotypes with mutations that nearly eliminated enzyme activities [104,117,118], and genotypes with 50% higher NO₃⁻ reductase activities did not assimilate more NO₃⁻ [119]. Studies that have confused rates of enzyme activities with those of NO₃⁻ assimilation as a whole have drawn false conclusions [120,121].

One physiological mechanism that may be responsible for the interdependency of photorespiration and shoot NO₃^– assimilation involves the reduction of the Mn²⁺-RuBP complex during oxidation of RuBP. This increases the redox potential of the chloroplast [101], thereby stimulating the production of malate [122,123] and promoting its export from chloroplasts to the cytoplasm. Malate dehydrogenase in the cytoplasm converts malate to oxaloacetate, generating NADH [124,125,126] to empower the initial step of NO₃⁻ assimilation [127]. Consequently, mutations that alter malate transport or metabolism influence both photorespiration and NO₃⁻ assimilation [122,128,129].

4.2. C₁ Metabolism

The photorespiratory pathway within mitochondria involve reactions with glycine. In one reaction, serine hydroxymethyltransferase 1 (SHMT1) converts glycine to serine and converts CH₂-THF (5,10-methylene-tetrahydrofolate) to THF (Figure 2). In the other reaction, the glycine decarboxylase complex reduces NAD⁺ to NADH and catabolizes glycine to CO₂, NH₃, and CH₂-THF (Figure 2). These C₁ units, in the form of CH₂-THF, serve as precursors in the synthesis of tetrahydrofuran (THF) derivatives [130,131,132,133]. One derivative of CH₂-THF, 5-CH₃-THF, is used to produce methionine, an essential amino acid. Methionine is a powerful antioxidant and is involved in protein synthesis and methylation of DNA, RNA, proteins, phospholipids, and other substrates [132]. In addition, about 5% of the total assimilated carbon in many secondary metabolites, such as glycine betaine, nicotine, and lignin, derive from C₁ metabolism [131].

4.3. Sulfur Assimilation

Photorespiration stimulates sulfur assimilation, although the effects are relatively small. By tracing ³³S in reactions involved in sulfur assimilation (such as sulfate reduction and synthesis of cysteine), and ¹³C in glycine and serine, a positive linear relationship was derived between relative photorespiration and sulfur assimilation. Sulfur assimilation decreases as photorespiration declines and photosynthesis increases [134].

Cysteine, the major product from sulfur assimilation, uses the sulfur element converted from serine generated from photorespiratory pathways [134,135]. H₂S, produced from sulfite reduced by sulfite reductase, is incorporated into O-acetylserine (OAS) via a protein complex consisting of serine acetyl transferase and OAS thiol-lyase to form cysteine [135,136]. Cysteine is essential in methionine synthesis, glutathione metabolism, sulfur-rich protein synthesis, glucosinolate biosynthesis, and the synthesis of phytoalexins (Figure 6) [137]. Cysteine is the precursor of methionine through o-phosphohomo-serine and homocysteine. Using methyl tetrahydrofolate as a cofactor, homocysteine is methylated by methionine synthase to yield methionine. Cysteine and methionine are the major sulfur contributors found in downstream metabolites, the most important of which is S-adenosyl methionine (SAM), which is a donor in methyl group transfers, transsulfuration, and aminopropylation [135,138].

5. Conclusions

Is photorespiration simply a futile cycle? The answer is “no”. Multiple lines of evidence show its crucial role in many plant processes. Despite heroic efforts to suppress photorespiration, disrupting any photorespiratory reaction usually proves detrimental to plants [139,140]. The reassimilation of CO₂ from photorespiration [60] and the important role played by photorespiration in the acclimation of plants to conditions, such as salinity [141] and elevated CO₂ [142], are topics that are beyond the scope of this review but nevertheless provide important evidence showing that photorespiration is not a wasteful process. There are many promising directions for further studies on photorespiration; for example, examining Mn²⁺ interactions with Rubisco, further exploring the reassimilation of photorespired CO₂, and exploring how the biochemical processes related to photorespiration contribute to its role in adaptation to various conditions will probably reveal that plant carbon fixation and respiration is more energy efficient than what has been previously assumed.