Photorespiration: The Futile Cycle?

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.


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 CO 2 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, C 1 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 Mg 2+ or Mn 2+ , and balance between the binding of these enzymes to Mg 2+ or Mn 2+ may shift the relative rates and energy efficiencies of photosynthesis and photorespiration [8].
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].
Plants 2021, 10, x FOR PEER REVIEW 2 of 17 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 Mn 2+ or Mg 2+ [21,22]. Binding of Mg 2+ 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 Mg 2+ 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 Rubisco must be activated before it can carboxylate or oxygenate RuBP. Activation of the three forms of Rubisco involves binding of Mn 2+ or Mg 2+ [21,22]. Binding of Mg 2+ requires carbamylation of Rubisco by the addition of CO 2 . One histidine at the active site of Rubisco rotates into an alternate conformation, resulting in a transient binding site where Mg 2+ 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 CO 2 , allowing for the formation of a covalent bond that produces a carbamyl group. This carbamyl group causes the Mg 2+ 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 Mn 2+ follows a similar mechanism and whether it requires an activator CO 2 to be bound first [21,22]; hence, understanding the mechanism of Mn 2+ 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 CO 2 and either Mg 2+ or Mn 2+ .
Rubisco can also bind to other metals. When bound to Fe 2+ , Ni 2+ , Cu 2+ , Ca 2+ , or Co 2+ , Rubisco may exhibit some carboxylase and oxygenase activity [24]. For example, one study found that Rubisco from R. rubrum, when bound to Co 2+ , 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 Ni 2+ or Co 2+ [25]. When bound to some other metal ions, including Cd 2+ , Cr 2+ , and Ga 2+ , 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 Mg 2+ , 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 CO 2 -Mg 2+ activation when CO 2 and Mg 2+ are present in suboptimal concentrations [26][27][28][29]. The crystal structure of Rubisco with both Mg 2+ 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].

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 CO 2 and O 2 at the active site of Rubisco, the specificity of the enzyme for each gas, and whether the enzyme is associated with Mg 2+ or Mn 2+ [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 Mg 2+ and absence of Mn 2+ [8]. Rubisco binding of Mg 2+ accelerates carboxylation, whereas binding of Mn 2+ slows carboxylation [25,[34][35][36][37][38]. Chloroplastic Mg 2+ and Mn 2+ 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 Mg 2+ 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 Mn 2+ [32].
In wheat leaves, the ratio of Mn 2+ to Mg 2+ 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 Mn 2+ 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.

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) The solid red lines represent reactions of the photorespiratory pathway, the solid blue lines represent reactions of the proposed alternative photorespiratory pathway, the solid purple lines represent reactions of amino acid synthesis, and the dotted lines represent associated transport processes. Numbered reactions are catalyzed by the following enzymes: 1. Rubisco, 2. Malic enzyme, 3. Phosphoglycolate phosphatase, 4. Glycerate kinase, 5. Glycolate oxidase, 6. Glutamate:glyoxylate aminotransferase, 7. Glycine decarboxylase complex, 8. Serine hydroxymethyltransferase-1, 9. Serine:glyoxylate aminotransferase, 10. Hydroxypyruvate reductase-1, 11. Malate dehydrogenase, 12, Nitrate reductase, 13 Nitrite reductase, and 14. Glutamine synthetase. PETC designates photosynthetic electron transport chain and RETC, respiratory electron transport chain. Adapted from ref. [8]. Copyright 2018 Springer Nature Ltd.
In wheat leaves, the ratio of Mn 2+ to Mg 2+ contents increased as the CO 2 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 Mn 2+ binding in order to compensate for the slower photorespiration rates and slower amino acid production that would otherwise occur under elevated CO 2 and nitrate nutrition.

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 Mg 2+ or Mn 2+ [8]. The plastid isoform of malic enzyme in Arabidopsis and tobacco catalyzes the reverse pyruvate synthesis reaction (pyruvate + CO 2 + 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 + O 2 + CO 2 + H 2 O → glycolate + malate + 2Pi) when Mn 2+ binds to these enzymes ( Figure 2) [8].

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 Mn 2+ vs. Mg 2+ on the relative rates of oxygenation and carboxylation in situ.

Post Illumination CO 2 Burst
This method measures the evolution of CO 2 from a leaf for 1 to 4 min after turning off the light because glycine metabolism continues longer in the dark than CO 2 assimilation [47]. The rate of CO 2 generation is measured by a transient CO 2 analyzer [48] when the light has just been turned off or at the maximum rate of CO 2 evolution observed. CO 2 assimilation, however, does not stop immediately after the light is off. Separating CO 2 assimilation from the CO 2 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].

O 2 Inhibition of Net CO 2 Assimilation
This method aims to assess the photorespiration rate from the increase in the CO 2 assimilation rate after switching from normal to low O 2 concentrations. Yet, changes in CO 2 assimilation with O 2 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 CO 2 assimilation [52].

Photorespiration CO 2 Efflux into CO 2 -Free Air
This method estimates photorespiration from the CO 2 efflux rate in CO 2 -free air. A high-O 2 and low-CO 2 environment, however, promotes photorespiration [4]. Additionally, a CO 2 -free atmosphere inhibits both the activity of Rubisco [53] and the regeneration of its substrate RuBP [54], leading to underestimates of photorespiration.

Ratio of 14 CO 2 to 12 CO 2 Uptake
In this method, 14 CO 2 and 12 CO 2 fluxes are measured after feeding a leaf with 14 CO 2 for a short period of time. Gross photosynthesis is estimated from 14 CO 2 uptake measured using an ionization chamber attached to an electrometer, while net photosynthesis is estimated from 12 CO 2 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 14 C 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 13 C [57,58] and about 5.5% against 14 C [4], resulting in errors in estimations of photorespiration rates that exceed 25% [4,57].

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: where A is the rate of photosynthetic CO2 assimilation, Rd is the rate of respiration other than photorespiration, and: 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]. There are several uncertainties associated with this method. The recycling effect on the specific activity of CO 2 inside the leaf can cause about a 20% error. One must consider the specific activity of CO 2 inside the leaf to obtain an accurate estimate of the gross photosynthesis rate because CO 2 efflux through photorespiration dilutes the 14 C label in the intercellular spaces, decreasing the specific activity of CO 2 . The activity might be even lower at the actual carboxylation site than in the interleaf spaces because of photorespiratory CO 2 loss [56]. Moreover, Rubisco carboxylation discriminates about 2.9% against 13 C [57,58] and about 5.5% against 14 C [4], resulting in errors in estimations of photorespiration rates that exceed 25% [4,57].

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 CO 2 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: is the rate of photosynthetic CO 2 assimilation, R d is the rate of respiration other than photorespiration, and: where C is the CO 2 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].

CO 2 Efflux into 13 CO 2 -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 13 CO 2 and no 12 CO 2 . The levels of released 12 CO 2 can be measured either using an infrared gas analyzer or a membrane inlet mass spectrometer. Because the rate of 12 CO 2 release includes both photorespiration and mitochondrial respiration, additional effort is needed to separate these effects. For example, in one approach, the rate of 12 CO 2 release is calculated as: where F is the gas flow rate, 12 C R and 12 C S are the mole fractions of 12 CO 2 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 13 CO 2 and the concentration of 13 CO 2 over 2 min is fitted to an exponential curve. The mitochondrial respiration R d is taken to be the rate of 12 CO 2 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 CO 2 concentration, that 0.5 CO 2 is generated per oxygenation reaction when the CO 2 released per oxygenation varies widely with temperature and light level and among species [7], and that leaves do not naturally contain any 13 C [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 12 CO 2 after switching from ambient air to air with a high concentration of 13 CO 2 ; however, high CO 2 concentrations could affect mitochondrial respiration and thus produce error in the estimate of photorespiration. The presence of naturally occurring 13 C also generates additional errors [59].

Labelling of Photosynthates with 14 C
Leaves at a photosynthetic steady state are exposed to 14 CO 2 for different lengths to label primary and stored photosynthates. Exposing the leaf to an ambient concentration of 14 CO 2 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. 14 CO 2 efflux into different backgrounds containing various combinations of O 2 and CO 2 concentrations provides an estimate of photorespiration [61,62]. Four different backgrounds are used: first, 21% O 2 and ambient CO 2 to measure the steady-state release of CO 2 from both photosynthesis and photorespiration; second, 1.5% O 2 and ambient CO 2 to measure the rate of photorespiration only; third, 21% O 2 and 30,000 µmol/mol CO 2 to limit CO 2 reassimilation; and fourth, 21% O 2 with no CO 2 to measure the specific radioactivity of CO 2 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 O 2 in the air. A recent report indicated that R d was actually lower at a lower O 2 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 CO 2 levels.

Measuring Photorespiratory Ammonia
Photorespiration generates NH 3 in addition to CO 2 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 NH 3 subsequently accumulates in the leaf. The advantages of this method also include the prevention of CO 2 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 NH 3 refixation completely. Other factors might limit the diffusion of NH 3 out of the leaves, leading to an underestimation of photorespiration [59]. GS inhibitors will disrupt the C 2 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 NH 4 + may overestimate the amount of NH 4 + because methylamine, ethylamine, ethanolamine, and some non-protein amino acids co-elute with NH 4 + . Degradation of labile nitrogen metabolites in leaf extract, xylem sap, and apoplastic fluid to NH 4 + during extraction will cause further overestimation of NH 4 + levels [68].

Measuring 18 O 2 Consumption and Labeled Metabolites
Replacing ambient air in a chamber containing a leaf with air containing 18 O 2 provides another estimate of the photorespiration rate. A mass spectrometer measures levels of 16  Ref. [69,70]. Unfortunately, this method cannot separate photorespiration from other light-dependent O 2 -consuming processes, such as light-dependent differences in the rate of mitochondrial respiration [46,48]. To diminish these errors, the mass spectrometer can quantify 18 O-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].

NMR Measurements on 13 C-Labeled Metabolites
This method requires that plants receive fertilizer labeled with 15 N and that leaves subsequently be exposed to 13 CO 2 . Rotational-echo double resonance (REDOR) detects 13 C within two covalent bonds of 15 N and thus assesses the formation of organic nitrogen metabolites labeled with 13 C [59,73]. The ratio of 13 C-labeled to unlabeled phosphorylated Calvin-Benson cycle metabolites between 2 and 4 min after exposure to 13 CO 2 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 13 CO 2 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 12 CO 2 generated by photorespiration and because photorespiration may produce intermediates labeled with 13 C in less than 2 min [60]. Furthermore, this method is based on the premise that photorespiration releases one CO 2 for every two oxygenations, when the CO 2 released per oxygenation varies widely with temperature and light level and among species [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]. Addition-ally, 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].

CO 2 Labeling and MS Analysis
Isotopically nonstationary metabolic flux analysis (INST-MFA) can trace 13 C-labeled photorespiratory metabolites in plants exposed to 13 CO 2 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].
Plants 2021, 10, x FOR PEER REVIEW 9 of 17 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].

CO2 Labeling and MS Analysis
Isotopically nonstationary metabolic flux analysis (INST-MFA) can trace 13 C-labeled photorespiratory metabolites in plants exposed to 13 CO2 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].

Micro-Optode Measurement of O 2 Consumption
We have been conducting direct oxygenation rate measurements using a needle-type O 2 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 oxygensensitive 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 O 2 . 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 O 2 in contrast to the other commonly used O 2 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 (%O 2 ) and liquid phases (DO), which makes it possible to measure O 2 exchanges accurately up to 250% air O 2 saturation in intact plant leaves, bioreactors, cell cultivation, microtiter plates, and many general oxygen measurements in liquids [95][96][97][98][99][100].

NO 3 − Assimilation
Multiple lines of evidence link shoot NO 3 − assimilation to photorespiration: (a) Elevated CO 2 or low O 2 levels inhibited shoot NO 3 − reduction [101].
(b) In independent 14 [102,109]: elevated CO 2 inhibited shoot NO 3 − reduction so it was less limited by nitrate availability, and NO 3 − reductase discriminated more strongly against 15 N-NO 3 − [110]. The assimilatory quotient (AQ) is the ratio of net CO 2 consumption to net O 2 evolution in plant shoots [111]. During shoot NO 3 − assimilation, ferredoxin generated from the photosynthetic electron chain reduces NO 2 − to NH 4 + rather than producing NADPH, and so net O 2 evolution increases without a change in net CO 2 consumption. Therefore, the change in assimilatory quotient (∆AQ) when a plant receives NH 4 + instead of NO 3 − as a sole N source provides an estimate of shoot NO 3 − assimilation [106]. ∆AQ decreased as the shoot internal CO 2 concentration increased in C 3 plants ( Figure 5) [9,104,112,113].
Shoot CO 2 and O 2 fluxes at ambient and elevated CO 2 were contrasted between stages of plant development or genotypes that have significantly different NO 3 − reductase activities in situ (i.e., 36-vs. 48-day-old wild-type Arabidopsis, Arabidopsis NO 3 − reductase knockout mutants vs. transgenic Arabidopsis overexpressing NO 3 − reductase, and NO 3 reductase-deficient barley mutants vs. wild-type barley) [104,112]. ∆AQ, a measure of shoot NO 3 assimilation, differed between these stages of development and genotypes under ambient CO 2 but not under elevated CO 2 . This indicates that none of the stages of development or genotypes were assimilating NO 3 − under elevated CO 2 [104,112].
planta [115,116]. Accordingly, NO3 − assimilation significantly declined on with mutations that nearly eliminated enzyme activities [104,117,118], and 50% higher NO3 − reductase activities did not assimilate more NO3 − [119]. St confused rates of enzyme activities with those of NO3 − assimilation as a wh false conclusions [120,121]. One physiological mechanism that may be responsible for the inter photorespiration and shoot NO3assimilation involves the reduction of complex during oxidation of RuBP. This increases the redox potential of [101], thereby stimulating the production of malate [122,123] and promo from chloroplasts to the cytoplasm. Malate dehydrogenase in the cytoplasm ate to oxaloacetate, generating NADH [124][125][126] to empower the initial s similation [127]. Consequently, mutations that alter malate transport or me ence both photorespiration and NO3 − assimilation [122,128,129].

C1 Metabolism
The photorespiratory pathway within mitochondria involve reaction In one reaction, serine hydroxymethyltransferase 1 (SHMT1) converts g and converts CH2-THF (5,10-methylene-tetrahydrofolate) to THF (Figure reaction, the glycine decarboxylase complex reduces NAD + to NADH and cine to CO2, NH3, and CH2-THF ( Figure 2). These C1 units, in the form of C as precursors in the synthesis of tetrahydrofuran (THF) derivatives [130-1 ative of CH2-THF, 5-CH3-THF, is used to produce methionine, an essent Methionine is a powerful antioxidant and is involved in protein synthesi tion of DNA, RNA, proteins, phospholipids, and other substrates [132]. In 5% of the total assimilated carbon in many secondary metabolites, such as nicotine, and lignin, derive from C1 metabolism [131].

Sulfur Assimilation
Photorespiration stimulates sulfur assimilation, although the effect small. By tracing 33 S in reactions involved in sulfur assimilation (such as su Maximum NO 3 − reductase activity in vitro generally declined under CO 2 enrichment [105,114]. Nonetheless, shoot NO 3 − reductase activity seldom limits NO 3 − assimilation in planta [115,116]. Accordingly, NO 3 − assimilation significantly declined only in genotypes with mutations that nearly eliminated enzyme activities [104,117,118], and genotypes with 50% higher NO 3 − reductase activities did not assimilate more NO 3 − [119]. Studies that have confused rates of enzyme activities with those of NO 3 − assimilation as a whole have drawn false conclusions [120,121].
One physiological mechanism that may be responsible for the interdependency of photorespiration and shoot NO 3 assimilation involves the reduction of the Mn 2+ -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 NO 3 − assimilation [127]. Consequently, mutations that alter malate transport or metabolism influence both photorespiration and NO 3 − assimilation [122,128,129].

C 1 Metabolism
The photorespiratory pathway within mitochondria involve reactions with glycine. In one reaction, serine hydroxymethyltransferase 1 (SHMT1) converts glycine to serine and converts CH 2 -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 CO 2 , NH 3 , and CH 2 -THF ( Figure 2). These C 1 units, in the form of CH 2 -THF, serve as precursors in the synthesis of tetrahydrofuran (THF) derivatives [130][131][132][133]. One derivative of CH 2 -THF, 5-CH 3 -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 C 1 metabolism [131].

Sulfur Assimilation
Photorespiration stimulates sulfur assimilation, although the effects are relatively small. By tracing 33 S in reactions involved in sulfur assimilation (such as sulfate reduction and synthesis of cysteine), and 13 C 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]. H 2 S, 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 Sadenosyl methionine (SAM), which is a donor in methyl group transfers, transsulfuration, and aminopropylation [135,138].
Plants 2021, 10, x FOR PEER REVIEW 12 of 17 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 Sadenosyl methionine (SAM), which is a donor in methyl group transfers, transsulfuration, and aminopropylation [135,138].

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 Figure 6. An outline of sulfur assimilation and its role in producing sulfur-containing defense compounds. Adapted with permission from ref. [137]. Copyright 2005 Elsevier Ltd.

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 CO 2 from photorespiration [60] and the important role played by photorespiration in the acclimation of plants to conditions, such as salin-ity [141] and elevated CO 2 [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 Mn 2+ interactions with Rubisco, further exploring the reassimilation of photorespired CO 2 , 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.