Abstract
C4 photosynthesis concentrates CO2 around Rubisco in the bundle sheath, favouring carboxylation over oxygenation and decreasing photorespiration. This complex trait evolved independently in >60 angiosperm lineages. Its evolution can be investigated in genera such as Flaveria (Asteraceae) that contain species representing intermediate stages between C3 and C4 photosynthesis. Previous studies have indicated that the first major change in metabolism probably involved relocation of glycine decarboxylase and photorespiratory CO2 release to the bundle sheath and establishment of intercellular shuttles to maintain nitrogen stoichiometry. This was followed by selection for a CO2-concentrating cycle between phosphoenolpyruvate carboxylase in the mesophyll and decarboxylases in the bundle sheath, and relocation of Rubisco to the latter. We have profiled 52 metabolites in nine Flaveria species and analysed 13CO2 labelling patterns for four species. Our results point to operation of multiple shuttles, including movement of aspartate in C3–C4 intermediates and a switch towards a malate/pyruvate shuttle in C4-like species. The malate/pyruvate shuttle increases from C4-like to complete C4 species, accompanied by a rise in ancillary organic acid pools. Our findings support current models and uncover further modifications of metabolism along the evolutionary path to C4 photosynthesis in the genus Flaveria.
Keywords: C4 photosynthesis, Calvin–Benson cycle, evolution, Flaveria, metabolites, photorespiration
Metabolite profiling and 13CO2labelling studies of Flaveriaspecies on the spectrum between C3and C4photosynthesis reveal progressive re-wiring of photorespiration, nitrogen metabolism, carbon concentration shuttles, and the Calvin–Benson cycle.
Introduction
In C3 photosynthesis, CO2 is assimilated by Rubisco in mesophyll cells. In a side reaction with O2, Rubisco catalysis produces 2-phosphoglycolate (2PG) that must be recycled at the expense of energy, CO2, and NH3 (Lorimer and Andrews, 1973; Bauwe et al., 2010). C4 photosynthesis appeared 25–30 million years ago as an adaptation to a global climate change, which included a sharp decrease in atmospheric CO2 (Christin et al., 2008; Zachos et al., 2008), and has since then evolved independently in >60 angiosperm lineages (Sage et al., 2011; Sage, 2016). In C4 photosynthesis, carbonic anhydrase (CA) converts CO2 into HCO3– that is combined with phosphoenolpyruvate (PEP) by PEP carboxylase (PEPC) to form oxaloacetate in the mesophyll. PEPC has a high HCO3– affinity and no side reaction with O2. Oxaloacetate is transformed into more stable four-carbon (C4) metabolites, malate and aspartate (Asp), that diffuse from the mesophyll to the bundle sheath (BS) and are decarboxylated to generate a high CO2 concentration around Rubisco, which is located only in the BS. This high CO2 concentration suppresses the wasteful side reaction with O2. Three-C metabolites, such as pyruvate (Pyr) and alanine (Ala), move back to the mesophyll to regenerate the initial HCO3− acceptor, PEP.
Under current atmospheric CO2 concentrations, C4 photosynthesis has several advantages over C3 photosynthesis (Sage et al., 2018); energetically wasteful photorespiration is decreased (Osmond and Harris, 1971; Sage et al., 2012), water use efficiency is increased (Ghannoum, 2009), and nitrogen (N) use efficiency is improved (Ghannoum et al., 2011). C4 species contain less Rubisco than C3 species because the high CO2 internal environment allowed evolution of lower specificity forms of Rubisco with a higher turnover number (kcat) (Brown, 1978; Schmitt and Edwards, 1981; Kubien et al., 2008; Kapralov et al., 2011). These advantages have attracted great interest in engineering C4 photosynthesis into C3 crops (Matsuoka et al., 2001; Häusler et al., 2002; Miyao et al., 2011; Ermakova et al., 2019). Such crop improvement strategies may profit from understanding how C4 photosynthesis evolved.
C4 photosynthesis is a complex trait involving anatomical as well as biochemical differences from C3 photosynthesis. Almost all C4 species have a ‘Kranz’ anatomy with a ring of mesophyll cells surrounding photosynthetically active BS cells. Furthermore, compared with C3 species, C4 plants have a higher vein density, a large increase in PEPC expression in the mesophyll cells, loss of Rubisco and most of the rest of the Calvin–Benson cycle (CBC) components from mesophyll cells, and increased numbers of plasmodesmata between the BS and mesophyll cells (Muhaidat et al., 2007; Lundgren et al., 2014; Danila et al., 2016, 2018). C4 photosynthesis is thought to have evolved in a stepwise manner, with pre-conditioning steps such as closer vein spacing occurring in a C3 background, followed by the acquisition of genetic regulatory elements, and the establishment of intermediary biochemical states (Langdale, 2011; Sage et al., 2012; Mallmann et al., 2014). Investigation of the evolution of C4 photosynthesis has been aided by genera including Flaveria, Heliotropium, Salsola, Steinchisma, and Neurachne, which contain species with intermediate modes of photosynthesis that are thought to represent steps on the evolutionary path from the ancestral C3 condition to a complete C4 syndrome (Ku et al., 1983; Muhaidat et al., 2011; Khoshravesh et al., 2016, 2019;Schüssler et al., 2017). Bayesian modelling indicated that different genera may have acquired the various anatomical, ultrastructural, and metabolic traits in different time sequences (Williams et al., 2013). This flexibility may have facilitated convergent evolution of the complex C4 trait across so many lineages. Modelling also indicated that metabolic subtraits evolved as successive modules, with each module being associated with an increase in fitness defined as the amount of CO2 fixed by a given amount of Rubisco (Heckmann et al., 2013; Heckmann, 2016).
Relocation of glycine (Gly) decarboxylation is thought to have been an initial metabolic step that increased the chance of C4 photosynthesis evolving. The glycine decarboxylase complex (GDC) is involved in photorespiration and converts two molecules of Gly into one molecule of serine (Ser) plus CO2 and NH3. In C3 plants, GDC is mainly located in mesophyll cells. Rawsthorne and colleagues noticed that Flaveria and Moricandia species with conspicuous Kranz anatomy had little or no GDC activity in their mesophyll cells and high GDC activity in their BS cells (Hylton et al., 1988; Rawsthorne et al., 1988a; Rawsthorne and Hylton, 1991), and predicted that they operate a photorespiration-driven carbon-concentrating mechanism (CCM). In this photorespiratory-based CCM (termed the C2 cycle, as Gly has two C atoms), Gly diffuses from the mesophyll to the BS cells, where it is decarboxylated, leading to a higher CO2 concentration within BS cells (Keerberg et al., 2014), and Ser diffuses to the mesophyll. In agreement, expression of genes encoding proteins in photorespiration and the glutamine–oxoglutarate aminotransferase (GOGAT) pathway for NH3 assimilation in C2 species is as high as in C3 species and only falls in species that are considered to represent later steps in the evolution of C4 photosynthesis (Mallmann et al., 2014). The C2 cycle improves recapture of photorespired CO2, lowers the photosynthetic CO2 compensation point, and would have been of selective advantage in a low CO2 world (Keerberg et al., 2014; Lundgren, 2020).
Operation of the C2 cycle is thought to have created a context in which the next step towards C4 photosynthesis could occur (Monson et al., 1986; Bauwe and Kolukisaoglu, 2003; Sage, 2004; Bauwe et al., 2010; Sage et al., 2012; Busch et al., 2013; Heckmann et al., 2013; Schulze et al., 2013; Williams et al., 2013); consequently, species with a C2 cycle represent an evolutionarily significant subset among C3–C4 intermediates. In addition to modifications in the exchange of C between mesophyll and BS cells, the progressive evolution of a complete C4 syndrome involved changes to N stoichiometry that required large fluxes of organic acids and amino acids between mesophyll and BS cells (Monson and Rawsthorne, 2000). In a widely accepted version of the C2 cycle, two N atoms move (in two Gly molecules) into the BS but only one (in one Ser) moves back to the mesophyll, whilst the other is released as NH3 in the BS cells. Mallmann et al. (2014) employed flux balance analysis modelling to investigate the consequences of refixing NH3 in the BS cells. They predicted that N stoichiometry could be maintained by large-scale movement of 2-oxoglutarate (2OG) to the BS cells and glutamate (Glu) to the mesophyll or, if their movement is constrained, exchange of Pyr and Ala, or malate and Asp. N fluxes could also be balanced by allowing a C4-like cycle, with malate being synthesized by PEPC in the mesophyll and decarboxylated in the BS cells with Ala returning to the mesophyll cells. This analysis pointed to a possible causal relationship between the C2 cycle and the emergence of C4 photosynthesis.
C3 plants contain orthologues of genes encoding C4-related enzymes, transporters, and regulatory proteins (Aubry et al., 2011) that were co-opted during the transition from C3–C4 to C4 photosynthesis (Sage, 2004; Mallmann et al., 2014; Schlüter and Weber, 2016, 2020). Early C4 species are thought to have possessed a functional C4 cycle, but with non-optimal distribution of C4 enzymes and Rubisco between cell types. This is exemplified by some C4-like members of the genus Flaveria, including F. palmeri, F. vaginata, and F. brownii (Cheng et al., 1988; Moore et al., 1989), and Sesuvium sesuvioides (Bohley et al., 2019).
The final step towards complete C4 status is thought to have included fine tuning of compartmentation and the kinetic and regulatory properties of enzymes (Sage, 2004; Heckmann et al., 2013; Williams et al., 2013). For example, compared with C3 species, Rubisco from C4 species generally has a lower CO2 affinity and a higher Vmax (see above), and PEPC from C4 species is less sensitive to feedback inhibition by malate (Westhoff and Gowik, 2004; Gowik and Westhoff, 2011).
Current thinking about the evolution of C4 photosynthesis is based on studies of photosynthetic traits (e.g. CO2 fixation rates, CO2 compensation point, and water use efficiency), anatomy, gene expression, enzyme activities, and genome structure (reviewed in Schlüter and Weber, 2020). There have been studies of a small number of metabolites in a few species (Leegood and von Caemmerer, 1994; Gowik et al., 2011; Arrivault et al., 2019). However, we lack a comprehensive study of metabolite levels in multiple species on the evolutionary path between C3 and C4 photosynthesis. This study investigated the absolute contents of >50 primary metabolites in nine Flaveria species, representing all modes of photosynthesis found in this genus, and uses this information to test current ideas and provide new insights into the evolution of C4 photosynthesis.
Materials and methods
Materials
Solvents and chemicals were from Merck (https://www.merckmillipore.com) and Roche Applied Science (https://lifescience.roche.com/en_gb.html). N2, O2, and unlabelled CO2 were from Air Liquide (https://www.airliquide.com/), and 13CO2 (isotopic purity 99%) was from Campro Scientific GmbH (www.campro.eu). A commercial soil mix containing white peat, clay, coconut husk fibre, and Osmocote Start slow release fertilizer (1 g l−1) was obtained from Stender AG (Schermbeck, Germany).
Plant material and harvest for metabolite measurements
Seeds of Flaveria robusta (C3 proto-Kranz), F. anomala (C3–C4 intermediate Type I), F. ramosissima (C3–C4 intermediate Type II), F. palmeri (C4-like), F. brownii (C4-like), F. trinervia (C4), and F. bidentis (C4), and cuttings of F. vaginata (C4-like) and F. cronquistii (C3) were kindly provided by Professor Peter Westhof (Heinrich Heine University, Düsseldorf, Germany). Cuttings of F. vaginata were directly rooted in soil in 30 cm plastic pots, and plants were grown through to seed production in a naturally illuminated polytunnel with the temperature maintained above a minimum of 10 °C. Seeds were germinated on moist soil in a controlled-environment chamber with a 16 h day (20 °C)/8 h night (18 °C) cycle, 150 µmol m−2 s−1 irradiance, and relative humidity (RH) of 70%. After 2 weeks, individual seedlings were transferred to 6 cm diameter plastic pots containing soil, repotted into 18 cm diameter plastic pots 1 week later, and into 30 cm diameter plastic pots 2 weeks before harvest. All growth was under a 16 h day (26 °C)/8 h night (22 °C) cycle, average irradiance 350 µmol m−2 s−1, and 70% RH. Due to seed sterility, F. cronquistii was propagated from sterile cuttings [sterilization medium: 0.5% (w/v) Murashige and Skoog medium containing 5% (w/v) Plant Preservation Medium (PPM®; Plant Cell Technologies Inc.)] grown initially in tissue culture [growth medium: 1% (w/v) Murashige and Skoog medium, 6.8% (w/v) agarose, 1% (w/v) PPM®] for 10 d before transferring well-rooted cuttings to 10 cm diameter plastic pots containing soil. After 2 weeks under a 16 h day (20 °C)/8 h night (18 °C) cycle, 150 µmol m−2 s−1, and 70% RH, plantlets were repotted into 18 cm diameter plastic pots, moved to a 16 h day (26 °C)/8 h night (22 °C) cycle, 350 µmol m−2 s−1 irradiance, and 70% RH, and repotted in 30 cm diameter plastic pots 2 weeks before harvest.
Leaves of most species were harvested 55–56 d after sowing, the slower growing F. ramosissima and F. trinervia 79–80 d after sowing, and F. cronquistii 55 d after transfer to soil. First fully expanded leaves were removed under an irradiance of 350 µmol m−2 s−1 and quickly (<2 s) plunged in a bath of liquid N2 without shading or turning. Multiple leaves from four different plants (four biological replicates), were harvested (Supplementary Table S1).
Plant material and harvest for 13CO2 labelling
Four species from Clade A (Supplementary Fig. S1; McKown et al., 2005; Lyu et al., 2015) were chosen to represent different photosynthesis modes: F. robusta (C3), F. ramosissima (C3–C4), F. palmeri (C4-like), and F. bidentis (C4). Seeds were germinated on moist soil (as above) under a 16 h day (20 °C)/8 h night (18 °C) cycle, 150 µmol m−2 s−1, and 70% RH. After 3 weeks, individual seedlings were transferred into 10 cm diameter plastic pots and grown under a 16 h day (26 °C)/8 h night (22 °C) cycle, average irradiance 500 µmol m−2 s−1, and 70% RH, until labelling 49/50 d after sowing. Labelling was conducted by placing a single leaf per plant (i.e. a biological replicate) in a custom-made transparent gas-tight acrylic chamber connected to a gas flowmeter, infrared CO2 analyser (LI-800 Gashound; LI-COR Biosciences), humidifier bottle, and gas bottles (Ermakova et al., 2020). Irradiance at leaf level within the chamber was 500 µmol m−2 s−1. After 2 min acclimation with 420 µl l−1 CO2, 21% O2, 78% N2, the gas flow was switched to 420 ppm 13CO2, 21% O2, 78% N2 for 40 min or 60 min before harvest. Metabolism was quenched by rapidly flooding the chamber with liquid N2 without opening the chamber and avoiding any shading of the leaf (Ermakova et al., 2020; Supplementary Table S1). Unlabelled (time zero) control samples were quenched with liquid N2 at the end of the acclimation phase without introducing 13CO2 into the labelling chamber.
Metabolite analyses
Samples were ground to a fine powder at liquid N2 temperature using a ball mill (Retsch GmbH) at 25 Hz speed and stored at –80 °C until analysis. Pyruvate, PEP, 3-phosphoglycerate (3PGA), and ATP were measured enzymatically in trichloracetic acid extracts from 50 mg FW aliquots of tissue (Merlo et al., 1993). Other organic acids, sugar phosphates, and cofactors were measured, after methanol/chloroform extraction from 15 mg FW aliquots, using reverse phase HPLC coupled with a tandem MS (LC-MS/MS) platform (Arrivault et al., 2009). Stable isotopically labelled internal standards were added to correct for matrix effects for a subset of metabolites (Arrivault et al., 2015). Amino acids were measured by HPLC on 20 mg FW aliquots after ethanolic extraction and o-phthalaldehyde derivatization (Carillo et al., 2005). Supplementary Table S2 provides a list of metabolites, abbreviations, and analytical methods.
13C enrichment analyses
13CO2-labelled samples were ground as above and extracted for GC-time-of-flight (TOF)-MS as in Lisec et al. (2006) and for LC-MS/MS as in Arrivault et al. (2009). Peaks obtained from GC-TOF-MS analyses were assigned to metabolites by comparing mass spectra and GC retention times with database entries and those of authentic standards available in a reference library from the Golm Metabolome Database (Kopka et al., 2005). Chromatograms obtained from LC-MS/MS were analysed as in Arrivault et al. (2017). Natural abundance was corrected using the CORRECTOR software (www.mpimp-golm.mpg.de/719693/Bioinformatik-Tools; Huege et al., 2014).
Statistical analyses
Clustering and principal component (PC) analyses were performed in R Studio version 1.2.5033 (www.rstudio.com) integrated with R version 3.6.1 (www.r-project.org/). One-way ANOVAs with Holm–Sidak post-hoc tests were computed using SigmaPlot vers. 14.0 (Systat Software, Inc.).
Results
Global analysis of metabolite levels in nine Flaveria species
The Flaveria genus includes two basal C3 species and two phylogenetically separated clades, with Clade A containing C3–C4 intermediate, C4-like, and C4 species, and Clade B containing C3–C4 intermediate and a C4-like species (Supplementary Fig. S1; McKown et al., 2005; Lyu et al., 2015). For our analyses, nine species were selected, comprising the two basal C3 species (F. robusta and F. cronquistii), two C3–C4 intermediates (F. ramosissima and F. anomala; from Clades A and B, respectively), three C4-like species (F. palmeri, F. vaginata, and F. brownii; Clades A, A, and B, respectively), and two complete C4 species (F. bidentis and F. trinervia; both Clade A). After establishment, all species were grown in the same controlled conditions. The first fully expanded leaves were harvested by flash-freezing under growth irradiance. We measured 52 metabolites, encompassing sugar phosphates, organic acids, and amino acids, and including most of the metabolites in the CBC, C4 photosynthesis (from here onward termed ‘CCM-related’ metabolites), photorespiration, the tricarboxylic acid (TCA) cycle, and sucrose and starch synthesis (Supplementary Table S2). Absolute metabolite contents are provided in Supplementary Dataset S1. The data for CBC metabolites and 2PG levels in the two complete C4 species F. bidentis and F. trinervia were previously published in Arrivault et al. (2019).
To provide a first overview, metabolite contents (per unit FW) were Z-score normalized and used in two-way clustering (Fig. 1; Supplementary Dataset S1) and PC analysis (Fig. 2A, B; Supplementary Dataset S1). PC analysis was also performed on data that were transformed to express each metabolite as the amount of C in that metabolite normalized to total C in all measured metabolites, termed a ‘dimensionless’ dataset (Fig. 2C, D; Supplementary Dataset S1; Arrivault et al., 2019; Borghi et al., 2019). This normalization removes possible bias due to differences in leaf protein and/or water content. These global analyses separated species based on photosynthetic mode and their phylogeny.
Fig. 1.
Heatmap and dendrograms of absolute metabolite levels in nine Flaveria species. The heatmap was generated by two-way clustering: each column represents a Flaveria species, and each row represents a metabolite. The Flaveria species is indicated at the bottom of each column, as well as its photosynthesis mode and taxonomical grouping according to current knowledge. Species (top) and metabolite (left-hand side) dendrograms were calculated using Z-scored data. The data were Z-scored by expressing the average level of a metabolite for (n=3–4; each replicate obtained by harvesting several newly fully expanded leaves from one Flaveria plant) in a given species as a fraction of the average across all species. The heatmap cell colour intensity represents the average Z-score for the metabolite in that species. The colour key and data distribution plot for the heatmap are shown in the insert in the upper-left corner. The right-hand subpanel indicates the sector of metabolism in which a given metabolite is involved; metabolic sectors are indicated at the top of the subpanel, and the involvement of a metabolite by a grey box. In some cases, a metabolite is assigned to more than one sector. This aspect of the data is explored further in Fig. 4. For metabolite abbreviations, refer to Supplementary Table S2, and for the original data see Supplementary Dataset S1.
Fig. 2.
Principal component (PC) analysis of metabolite profiles in nine Flaveria species. (A, B) PC analysis of the FW-normalized dataset. (C, D) PC analysis using a dimensionless dataset. (A, C) Distribution of samples along the two first two PCs, with each sample being represented by a coloured label indicating the species and biological replicate number. (B, D) Metabolite eigenvectors driving sample separation are shown in yellow, while individual samples appear as coloured dots. In all panels, the colour code represents the different photosynthetic modes, as indicated by the key below the figure. In (A) and (C), a box denotes Clade B species, as indicated by the key below the figure. Species abbreviations are, alphabetically: ano, F. anomala (Clade B, C3–C4); bid, F. bidentis (Clade A, C4); bro, F. brownii (Clade B, C4-like); cro, F. cronquistii (basal, C3); pal, F. palmeri (Clade A, C4-like); ram, F. ramosissima (Clade A, C3–C4); rob, F. robusta (basal, C3); tri, F. trinervia (Clade A, C4); vag, F. vaginata (Clade A, C4-like). For metabolite abbreviations, refer to Supplementary Table S2, and for the original data see Supplementary Dataset S1.
The clustering analysis is shown as a heatmap in Fig. 1. The basal C3 species F. cronquistii and F. robusta grouped with the C3–C4F. ramosissima of Clade A in a polytomy. The C4-like species F. palmeri and F. vaginata from Clade A formed a sister group to that composed of the two Clade A C4 species, F. bidentis and F. trinervia, used in the study. This group of C4-like and complete C4 species are sister to the basal C3F. ramosissima cluster, while the Clade B C4-like F. brownii and C3–C4F. anomola formed a distinct cluster.
In the PC analysis, PC1 and PC2 captured 22.45% and 16.66%, respectively, of the total variance in the FW-normalized dataset (Fig. 2A, B; Supplementary Table S3), and 41.68% and 14.1%, respectively, of the total variance in the dimensionless dataset (Fig. 2C, D; Supplementary Table S3). Both datasets separated species based on mode of photosynthesis and on phylogeny. For Clade A species, the two complete C4 species (F. bidentis and F. trinervia) were clearly separated from the two C4-like species (F. vaginata and F. palmeri) and the C3–C4 species (F. ramosissima), as well as from the basal C3 species F. cronquistii and F. robusta. This separation was found in PC1 and PC2 (Fig. 2A, C). Compared with the basal C3 species, the C3–C4F. ramosissima was slightly displaced in a negative direction in PC2, while the two C4-like species (F. vaginata and F. palmeri) were strongly displaced in a positive direction. This suggests that the evolutionary path from C3 through C3–C4 and C4-like to complete C4 may not involve a progressive change in metabolite levels, but instead the establishment of discrete states at each stage in the evolutionary process, especially the C3–C4 stage. For Clade B species, C3–C4F. anomala was largely separated from the basal C3 species, again mainly in PC2, but, unlike F. ramosissima, in a positive direction, lying quite close to the Clade B C4-like species F. brownii (Fig. 2A, C). In the analysis with FW-normalized data (Fig. 2A, B), PC1 and PC2 captured 39.11% of the total variance. We extended the analysis to include PC3 (12.52%) and PC4 (11.34%) (Supplementary Fig. S2), which, together with PC1 and PC2, explained >60% of the total variance (Supplementary Table S3). In PC3 there was no clear pattern. PC4 separated C3 species from C3–C4F. anomola and F. ramosissima, but not entirely from C4-like and C4 species (Supplementary Fig. S2).
We next examined which metabolites drove the separations in the clustering and PC analyses. The separation of the C3 species from other species, except Clade A C3–C4F. ramosissima (see below), was driven by 2PG, Gly, Ser, and glycerate, several CBC intermediates (e.g. RuBP, Ru5P+Xu5P, R5P, and S7P), some amino acids (Pro, Thr, His, Gln, Arg, and Glu), and the aromatic amino acid precursor shikimate (Figs 1, 2B, D). Flaveria ramosissima, which is an advanced C3–C4 species in Clade A, lays close to the basal C3 species, resembling the hierarchical clustering shown in Fig. 1. Its separation from the other species was due to comparatively high levels of some CBC (S7P, F6P, and R5P) and photorespiratory (2PG, Gly, and glycerate) metabolites (Figs 1, 2B, D). Separation of the two Clade A C4-like species from the C3–C4 intermediates and C3 species occurred mainly in PC1 and was driven by higher levels of CCM-related metabolites (malate and Pyr) and some amino acids (Ala and Asn). High levels of TCA and other respiratory intermediates such as 2OG, citrate, isocitrate. and aconitate and, to a lesser extent, Pyr and malate (only clear in the analysis with FW-normalized data), were largely responsible for the separation of complete C4 species from C4-like species. The separation of F. anomala, an early branching C3–C4 species in Clade B, from basal C3 species was driven by high Ser, SBP, and FBP, and a distinctive amino acid profile (high Ala, Asn, Met, Trp, Tyr, and Gln), a pattern that was also largely responsible for the placement of the Clade B C4-like F. brownii in the PC analyses.
In summary, species in Clade B differ from those in Clade A in having higher levels of some amino acids and sugar phosphates. Compared with basal C3 species, the Clade A C3–C4F. ramosissima tends to have higher CBC and photorespiratory metabolites; C4-like species, regardless of clade, have higher CCM-related intermediates; and the change from C4-like to a complete C4 syndrome is associated with higher levels of other organic acids (Figs 1, 2B, D). Overall, changes between C3 and C3–C4 species are captured mainly in PC2 and are smaller and partly orthogonal to the changes presumably associated with subsequent evolutionary steps to C4-like and complete C4 species that are captured mainly in PC1. This is especially so for Clade A species.
Changes of individual metabolites between species
To investigate responses that might be masked in the above global analyses, the responses of individual metabolites were examined (Fig. 3; Supplementary Fig. S3)
Fig. 3.
Absolute amounts of metabolites in nine Flaveria species. (A) CCM-related metabolites, (B) selected CBC metabolites, and (C) photorespiratory metabolites. The colour of each bar represents the different photosynthetic modes, as indicated by the key below the figure. In the sets of C3–C4 and C4-like species, the Clade B species (F. anomala and F. brownii, respectively) are placed to the left of the Clade A species; clade is also indicated as ‘A’ or ‘B’ below the species name. The amounts are plotted as average (nmol g–1 FW) ±SD (n=3–4). Letters above each bar represent post-hoc pairwise comparison grouping (Holm–Sidak method, P<0.05). For metabolite abbreviations. refer to Supplementary Table S2, and for the original data see Supplementary Dataset S1. The data in (B) for CBC metabolite levels in F. bidentis and F. trinervia were included in a previous publication (Arrivault et al., 2019).
Figure 3A shows amounts of malate, Asp, Pyr, Ala, and PEP which are directly involved in the C4 CCM, and levels of Glu, 2OG, and Gln, which potentially facilitate aminotransferase reactions in C4 photosynthesis and/or could be involved in shuttles that return amino groups from the BS to the mesophyll in C3–C4 photosynthesis. These metabolites were relatively low in both C3 species, except for Asp and Glu, whose levels resembled those in most other Flaveria species. The only major difference between the two C3 species was for Gln, which was significantly higher in F. cronquistii than in F. robusta. There were some noticeable differences between the two C3–C4 species, with Gln being 4-fold being higher, and Ala 5-fold lower in F. anomala than in F. ramosissima. Comparison of all four species indicated that the transition from C3 to C3–C4 was not accompanied by consistent and significant increases in the content of any CCM-related metabolites. The transition from C3–C4 to a C4-like state was accompanied by an increase (significant in some pairwise comparisons) in the contents of malate, Pyr, and Ala. That said, the three C4-like species were quite diverse: for example, they had differing levels of malate, Ala, PEP, and Gln. These differences might be due to differing evolutionary histories, reflected in F. brownii (high Ala, low PEP, low Glu, and high Gln) belonging to Clade B, while the two other C4-like species are in Clade A. In the transition from C4-like to a complete C4 syndrome, there was a trend towards even higher malate, Pyr, and PEP levels, no consistent change in the amounts of Ala and Glu, a decrease in Asp and Gln, and an increase in 2OG and other TCA cycle intermediates such as aconitate, citrate, and isocitrate (Supplementary Fig. S3).
CBC metabolite levels were less dependent on photosynthesis mode (Fig. 3B). Comparing the two C3 species, RuBP, DHAP, and F6P were significantly higher in F. cronquistii than in F. robusta, whereas no significant differences were found for 3PGA, FBP, S7P, SBP, and R5P. It might be noted that some of these metabolites (e.g. DHAP and F6P) are also involved in sucrose synthesis. Relative to the two C3 species, the C3–C4 species (F. anomala and F. ramosissima) showed few significant or consistent changes in CBC metabolite levels. Both C3–C4 species had significantly higher RuBP, DHAP, F6P, and S7P compared with F. robusta, but not with F. cronquistii. Compared with C3–C4 species, C4-like species showed a trend towards lower RuBP (significant in the comparison with F. anomala but not with F. ramosissima). The Clade B species F. brownii was the most divergent of the three C4-like species, with significantly higher FBP, F6P, and SBP than the Clade A C4-like species. Complete C4 species showed a trend (significant for some metabolites and pairwise comparisons) towards lower RuBP, FBP, and SBP than in Clade A C4-like species. Moving from C3–C4 to C4-like to complete C4 species, there was a consistent trend towards lower amounts of RuBP, FBP, and SBP, but not of other CBC intermediates.
Figure 3C shows contents of photorespiratory intermediates. No significant differences were detected between the two C3 species (F. cronquistii and F. robusta). Both C3–C4 intermediates showed significant increases in Ser compared with the C3 species. No consistent differences were found for 2PG and Gly; neither of these metabolites increased in F. anomala relative to the C3 species, while F. ramosissima had significantly higher amounts of 2PG compared with F. cronquistii and F. robusta, and significantly higher amounts of Gly compared with F. robusta. It is noteworthy that Ser levels are far higher than Gly levels (15- to 50-fold) in the C3–C4 species. Compared with C3–C4 intermediates, C4-like Flaveria species showed lower levels of Gly (significant in some pairwise comparisons), Ser (significant in all pairwise comparisons), and glycerate (significant in some pairwise comparisons). Compared with C4-like species, complete C4 species had similar levels of 2PG, consistently but non-significantly higher Gly and lower Ser, and higher glycerate (significant in some pairwise comparisons). Maybe unexpectedly, the complete C4 species had the highest glycerate levels of all nine species.
Changes of classes of metabolites between species
The above analyses of the levels of individual metabolites revealed some significant differences, but also indicated that others did not show consistent differences across the nine Flaveria species (Fig. 3; Supplementary Fig. S3). The results shown in Figs 1 and 3 indicated that changes in metabolite levels may be focused in particular sectors of metabolism. This possibility was explored further using the ‘dimensionless’ dataset (Fig. 4; Supplementary Dataset S1). Metabolites were grouped based on pathways or sectors of metabolism: CBC, photorespiratory, CCM-related, TCA cycle intermediates, and amino acid. The class ‘CCM-related’ contained not only metabolites that participate in intercellular shuttles in C4 photosynthesis, but also metabolites potentially involved in shuttles in C3–C4 species. Some metabolites were assigned to more than one class (e.g. malate and 2OG were assigned to ‘CCM-related’ and ‘TCA cycle’). The analysis shown in Fig. 4 revealed that the evolution of C4 photosynthesis is accompanied by major changes in C allocation between sectors of metabolism.
Fig. 4.
Carbon allocation to different sectors of metabolism in nine Flaveria species. C allocation was calculated using a dimensionless dataset. The percentage of total C in a given metabolite is depicted in the stacked bar diagrams as the average (n=3–4); metabolites are ordered according to their sequence in the metabolic pathway or sector. A metabolite can appear in multiple panels. The colour assigned to each metabolite is indicated in the insert of each panel: (A) CBC intermediates, (B) photorespiratory intermediates, (C) CCM-related intermediates, (D) TCA cycle intermediates, and (E) amino acids (only amino acids containing >5% of the total C pool). The colour of each species’ name represents the corresponding photosynthetic mode as indicated by the key at the bottom right. For metabolite abbreviations, refer to Supplementary Table S2, and for the original data see Supplementary Dataset S1.
Metabolites in the CBC accounted for only a small fraction of the C in central metabolism (Fig. 4A), reflecting their small pool sizes and rapid turnover (Szecowka et al., 2013; Ma et al., 2014; Arrivault et al., 2017). They accounted for 2.4–3% of total C in C3 species and 3.0–3.8% in C3–C4 species, falling to 1.1–2.0% in C4-like species and to 0.4–0.8% in complete C4 species. Photorespiratory metabolites accounted for 5–10% of total C in central metabolism in Flaveria species using C3 photosynthesis (Fig. 4B). This reflects the large pool sizes of Gly, Ser, and glycerate in C3 species (Szecowka et al., 2013; Ma et al., 2014). C allocation to photorespiratory metabolites rose to 13–20% in C3–C4 species, and fell to <2.5% in C4-like and complete C4 species. This was mainly due to changes in Ser. Metabolites assigned to the CCM-related category accounted for 44–47% of all C in central metabolism in C3 and C3–C4 species, slightly more (53–58%) in C4-like species, and slightly less (38–44%) in complete C4 species (Fig. 4C). It should be noted that in any given species, some of these metabolites may not be involved in intercellular shuttles and, when they are, only part of the total pool may be involved (see Introduction and below). C allocation to 2OG was highest in complete C4 species. Ala showed a unique pattern, being relatively high in both Clade B species (5–15%) and the C4-like Clade A species, F. palmeri (just below 5%). TCA cycle intermediates accounted for 20–45% of total C in central metabolism in C3 species and C3–C4 species, 45–70% in C4-like species, and >75% in complete C4 species (Fig. 4D). In addition to malate and 2OG, the TCA cycle metabolite showing the greatest difference between species was citrate, which accounted for >30% of total C in C4-like F. vaginata and the complete C4 species. Amino acids that contributed >5% of total C in at least one species were also examined (Fig. 4E). Amino acids showed an opposite trend to TCA cycle metabolites, accounting for 40–50% of total C in C3 species, 45–60% in C3–C4 species, 20–40% in C4-like species, but only 5–10% in complete C4 species.
13CO2 labelling experiments in four selected Flaveria species
As many metabolites are compartmented, with large pools in vacuoles or non-photosynthetic cells, total content may not provide accurate information about the size of the pool that is directly involved in photosynthesis (Szecowka et al., 2013; Arrivault et al., 2017). One way to investigate compartmentation is by 13CO2 labelling. Experiments in C3 tobacco (Hasunuma et al., 2010), C3 Arabidopsis (Szecowka et al., 2013; Ma et al., 2014), C3 cassava (Arrivault et al., 2019), and the C4 plant maize (Weissmann et al., 2016; Arrivault et al., 2017) have shown that CBC metabolites label to high enrichment within minutes. Enrichment in photorespiratory metabolites rises more slowly, but usually plateaus by 40 min. In maize, labelling of metabolites involved in the CCM plateaus by 15–20 min. For organic acids and amino acids, in particular, enrichment can plateau at quite a low value. The labelled portion is thought usually to reflect the pool that is actively involved in photosynthesis, namely the pool that is directly downstream of C assimilation, although in some cases incomplete labelling can result from entry of unlabelled C from downstream metabolites (e.g. starch or photorespiratory intermediates) into a single pool that is directly involved in photosynthesis (Sharkey et al., 2020).
Four Flaveria species were chosen for 13CO2 labelling experiments; a basal C3 species (F. robusta), and three species from Clade A, a C3–C4 (F. ramosissima), a C4-like (F. vaginata), and a complete C4 species (F. bidentis). We focused on Clade A to cover the spectrum of photosynthetic types, including complete C4, in species with a more recent shared evolutionary history, thereby minimizing any differences that reflect phylogeny rather than photosynthetic type. Pulse durations of 40 min or 60 min were chosen, with the expectation that enrichment would be at or close to a plateau (see above). Enrichment was determined for two metabolites in the CBC and 10 metabolites in CCM or photorespiratory pathways (highlighted in red in Supplementary Table S2; data are provided in Supplementary Dataset S2).
Supplementary Fig. S4 shows the differences in 13C enrichment in the four Flaveria species. Although the estimates of enrichment are approximate, the values for the C3 and C4Flaveria species generally resemble those reported in extensive studies of other C3 and C4 species (Hasunuma et al., 2010; Szecowka et al., 2013, Ma et al., 2014; Weissmann et al., 2016; Arrivault et al., 2017, 2019). Enrichment in the CBC intermediates 3PGA and DHAP was high (>80%) in all four species (Supplementary Fig. S4A, B). Enrichment was also high in PEP, which is important as it means that low enrichment in any of the downstream CCM-related metabolites provides evidence for slow synthesis and/or the presence of compartmented pools. Enrichment in malate was weak (<10%) in the C3 species F. robusta, only marginally higher in the complete C4 species F. bidentis and the C3–C4F. ramosissima, and slightly higher (~25%) in the C4-like F. palmeri. This resembles previous studies of malate in C3 and C4 species (Hasunuma et al., 2010; Szecowka et al., 2013; Ma et al., 2014; Weissmann et al., 2016; Arrivault et al., 2017, 2019). Enrichment in Asp was low (~20%) in the C3Flaveria species, and higher in the C3–C4, C4-like, and C4 species (~60, 85, and 75%, respectively), which all lay in the range previously reported for maize (Weissmann et al., 2016; Arrivault et al., 2017). Enrichment in Pyr was quite low (~30%) in the C3Flaveria species (similar to Arabidopsis, see Szecowka et al., 2013) and C3–C4Flaveria species, and higher (>50%) in the C4-like and complete C4Flaveria species (similar to maize; Weissmann et al., 2016; Arrivault et al., 2017). Enrichment in Ala was ~50% in the C3Flaveria species, 60% in the C3–C4 species, and 70–80% in the C4-like and complete C4 species. Overall, these analyses of enrichment point to increased C flow from the CBC to Asp in C3–C4, C4-like, and C4 species, and to Pyr and Ala in C4-like and C4 species. Glu and 2OG were weakly and incompletely labelled in all four Flaveria species, resembling previous studies of other C3 and C4 species (Szecowka et al., 2013; Ma et al., 2014; Ishihara et al., 2015; Arrivault et al., 2017, 2019). For photorespiratory metabolites, enrichment was high in Gly (75–90%) and Ser (70–80%) in all four Flaveria species, as found previously for other C3 and C4 species (Szecowka et al., 2013; Ma et al., 2014; Ishihara et al., 2015; Arrivault et al., 2017, 2019). Enrichment in glycerate was comparatively low (~20%) in the C3Flaveria species, and higher in the C3–C4 (>60%), C4-like, and complete C4 species (~50%).
Estimation of active pools in Flaveria species, and comparison with maize
To estimate the active pool of each metabolite in each of the four Flaveria species, total metabolite content was multiplied by 13C enrichment (Fig. 5; Supplementary Table S4). For comparison, Fig. 5 also shows active pools for maize, estimated from published data (Arrivault et al., 2017). In cases where a clear labelling plateau was not reached, the estimated active pool is a minimum value. The estimates highlight differences in active pool size between the four Flaveria species, revealing at what stage in the C3 to C4 evolutionary process major changes in photosynthetic metabolism occurred and indicating the extent to which the complete C4Flaveria species resembles maize.
Fig. 5.
Estimated active pool of metabolites in four Flaveria species and maize. Amounts (nmol g–1 FW) are shown for metabolites involved in the CCM, the photorespiratory pathway, and two metabolites from the CBC. Details for calculations are presented in Supplementary Table S4. Estimates for maize are from Arrivault et al. (2017) and additional 3PGA quantification. For metabolite abbreviations, refer to Supplementary Table S2.
The malate active pool was small in the C3 species (85 nmol g–1 FW), ~70% higher in the C3–C4 species, and >10-fold higher in C4-like and complete C4 species, reaching values almost as high as in maize. The Asp active pool was relatively high in the C3 species (180 nmol g–1 FW), at least 3-fold higher in the C3–C4 and C4-like species, and ~3-fold lower in the C4 species, but still higher than in maize. The Pyr active pool was very low in the C3 species (4 nmol g–1 FW), ~80% higher in the C3–C4 species, and was dramatically higher (20- to 40-fold) in the C4-like and C4 species, where the pool resembled that in maize. In comparison with other CCM-related metabolites, the active PEP pool was small, but increased progressively from C3 to C3–C4, C4-like, and C4 species, with the F. bidentis pool resembling that in maize. The Ala active pool in the C3 species (80 nmol g–1 FW) was about half that of Asp, remained low in the C3–C4 species, and was >15-fold higher in the C4-like and complete C4 species, again like the active pool in maize.
Active pools of 2OG and Glu were low in the C3 species, ~2-fold higher in the C3–C4 and C4-like species, and Glu increased by another 2-fold in the complete C4 species. The estimated active pool sizes of 2OG and Glu in the C4Flaveria species resembled those reported for maize.
Among photorespiratory metabolites, the Gly active pool was low in the C3 species (~50 nmol g–1 FW), ~2-fold higher in the C3–C4 species, and low in C4-like and complete C4 species, which was about half that previously reported for maize. The Ser active pool was large in the C3 species (710 nmol g–1 FW), ~70% higher in the C3–C4 species, but considerably lower in the C4-like and C4Flaveria species, which resembled maize. It is noteworthy that the active pool of Ser in the C3–C4 species is 12-fold higher than that for Gly, and was also 5- to 10-fold higher than that of any other analysed metabolite in this species, except for Asp. The glycerate active pool was ~120 nmol g–1 FW in the C3 species and >3-fold higher in the C3–C4 intermediate species, which is consistent with rapid photorespiration, as well as the possibility that glycerate may contribute to shuttling of amino groups. It was very low (~60 nmol g–1 FW) in the C4-like species, as expected if photorespiration is decreased. Unexpectedly, the complete C4 species contained a large glycerate active pool (420 nmol g–1 FW). The active pool in the C4Flaveria species was ~4-fold larger than in maize, which contained an active glycerate pool similar to that in the C3Flaveria species. Finally, the active pools of 3PGA and DHAP increased progressively from C3 to C3–C4, C4-like, and C4Flaveria species, but were consistently much lower than in maize (for more analysis see the next section and the Discussion).
Cross-species comparison of CBC intermediate profiles
Arrivault et al. (2019) recently presented a cross-species comparison using CBC metabolites and 2PG amounts in five C3 plants and four NADP-malic enzyme (NADP-ME) subtype C4 plants, including the two C4Flaveria species used in the current study. In PC analyses with this subset of metabolites, C3 species separated from C4 species, and there was divergence between different C4 species, and between different C3 species. These analyses pointed to changes in CBC operation between C3 and C4 plants, probably as a result of the CBC adapting to a high CO2 environment. We were interested to learn how Flaveria C3, C3–C4 intermediate, and C4-like species map onto this landscape.
We performed PC analysis on a combined dataset containing CBC metabolites and 2PG in the nine species from Arrivault et al. (2019) and the seven additional C3, C3–C4, and C4-like Flaveria species from the current study (Fig. 6). The analysis was performed on a dimensionless dataset (Supplementary Dataset S3). We chose this normalization method because some of the species used in Arrivault et al. (2019) had a high protein or chlorophyll content, leading to them being systematically displaced when PC analysis was performed with metabolites expressed on a FW basis.
Fig. 6.
Principal component (PC) analysis of CBC metabolites and 2PG amounts, combining data from the current study of nine Flaveria species with data for five further C3 species and two further NADP-ME subtype C4 species. PC analysis was performed on dimensionless datasets, combining data from this study and from Arrivault et al. (2019). The data for CBC metabolite levels in F. bidentis and F. trinervia were already included in the analysis of Arrivault et al. (2019). (A) Distribution of samples along the two first two PCs, with each sample being represented by a coloured label indicating the species and biological replicate number. (B) Metabolite eigenvectors driving sample separation are shown in yellow, while individual samples appear as coloured dots. In both panels, the colour code represents the different photosynthetic modes, as indicated by the key below the figure. Species abbreviations (and photosynthesis mode) are, alphabetically: At, Arabidopsis thaliana (C3); Fan, F. anomala (C3–C4); Fbi, F. bidentis (C4); Fbr, F. brownii (C4-like); Fcr, F. cronquistii (C3); Fpa, F. palmeri (C4-like); Fra, F. ramosissima (C3–C4); Fro, F. robusta (C3); Ftr, F. trinervia (C4); Fva, F. vaginata (C4-like); Me, Manihot esculenta (C3); Nt, Nicotiana tabacum (C3); Os, Oryza sativa (C3); Sv, Setaria viridis (C4); Ta, Triticum aestivum (C3); Zm, Zea mays (C4). For metabolite abbreviations, refer to Supplementary Table S2, and for the original data see Supplementary Dataset S3.
PC1 and PC2 captured 30.4% and 24.5% of total variance in this large dataset (Fig. 6; see Supplementary Table S5 for the first 10 PC contributions). The distribution of the nine species in the study of Arrivault et al. (2019) was largely retained in this extended Flaveria analysis (compare Fig. 6A with fig. 4D in Arrivault et al., 2019). Of the seven added Flaveria species, the two C3 species lay close to other C3 species (especially Triticum aestivum and Oryza sativa). The two Flaveria C3–C4 species and the Clade B C4-like species F. brownii fell into a diagonal between the C3 and complete C4 species (Fig. 6A). The Clade A C4-like species F. palmeri and F. vaginata lay close to the Clade A C4 species F. bidentis and F. trinervia. The Clade B C3–C4F. anomala showed some scatter, but most samples were closer to the Clade A C3–C4F. ramosissima and the two Clade A C4-like species than the Clade B C4-like species F. brownii. Overall, C3, C3–C4, C4-like, and complete C4 species were separated mainly along PC1, driven by high 3PGA and low levels of 2PG and several CBC intermediates including S7P, F6P, SBP, FBP, and, especially, RuBP (Fig. 6B). The Clade B C4-like F. brownii is somewhat of an outlier, separating from all other Flaveria species in PC1 and especially PC2, driven by high FBP and SBP (Fig. 6B).
Discussion
Using a phylogenetically informed approach, we profiled >50 metabolites in nine Flaveria species that use different photosynthetic modes to provide new insights into the evolution of C4 photosynthesis in this genus from a metabolic perspective. Our study reveals substantial metabolic diversity within the genus. Importantly, hierarchical clustering (Fig. 1) and PC analysis (Fig. 2; Supplementary Fig. S2) revealed that metabolite profiles suffice to classify these nine species according to photosynthesis mode and phylogenetic relationship, providing intrinsic support for our approach.
In some cases, total metabolite content overestimates the pool that is actively involved in photosynthetic metabolism (Szecowka et al., 2013; Ma et al., 2014; Arrivault et al., 2017; Allen and Young, 2020). We therefore performed 40–60 min 13CO2 pulse labelling experiments in one basal C3 species and three species from Clade A, one C3–C4 intermediate, one C4-like, and one complete C4 species (Supplementary Fig. S1; McKown et al., 2005; Lyu et al., 2015). This allowed us to estimate the active pool sizes for a subset of the metabolites. The active pools were noticeably smaller than the total content for several metabolites, for example malate, 2OG, and Glu, as previously reported for other species (Hasunuma et al., 2010; Szecowka et al., 2013; Weissmann et al., 2016; Arrivault et al., 2017). However, total content and active pool sizes usually showed similar trends between Flaveria species.
Our results provide new insights into the current model of NADP-ME-type C4 evolution in the genus Flaveria, based on early studies by Ku et al. (1983) and Monson et al. (1986) and general models of C4 evolution (Sage et al., 2012, 2018; Sage, 2016). By revealing which metabolite levels differ in species representing the stages of C4 evolution, clues are obtained about how nature built a C4 pathway (see Fig. 7). In particular, two related questions can be addressed. (i) Are the observed changes in metabolite levels those expected as new reactions are co-opted into photosynthetic C fixation? (ii) Do metabolites that are involved in predicted intercellular shuttles show an increase in content, as might be expected if intercellular movement occurs by diffusion and is driven by intercellular concentration gradients?
Fig. 7.
Schematic representation of stepwise changes in primary metabolism along the proposed path from C3 to C4 photosynthesis, based on current phylogeny of the genus Flaveria. The display builds on the work of Ku et al. (1983) and Monson et al. (1986) and resulting concepts developed in Sage et al. (2012, 2018). It is proposed that the photorespiration-dependent C2 CCM established in C3–C4 intermediate (indicated by red arrows) species may have been rather flexible, with different metabolites moving from the MC to the BSC. This would affect the extent to which glycine and possibly glycolate may move from the MC to the BSC, which in turn would affect the magnitude and direction of intercellular exchanges that are required to maintain N stoichiometry. There may also be flexibility in which organic acid/amino acid exchanges maintain N stoichiometry. At this step, PEPC activity in the MC would be needed to build up pools of C4 shuttle metabolites but would not be required to generate a net flux during steady-state photosynthesis. Similarly, decarboxylation would serve to reduce the levels of C4 metabolites or increase the levels of C3 metabolites, but net flux would not be needed in steady-state photosynthesis. A small draw-down of CO2 in the MC by PEPC would lead to increased RuBP oxygenation and strengthen the C2 shuttle, but a large decrease would result in a futile cycle and decreased net C fixation. This dichotomy might be broken by increased relocation of the CBC to the BSC accompanied, crucially, by increased C4 acid decarboxylation rates in the BSC. This shift from a C2 to C4 cycle represents a very large switch in metabolism, and would require increased PEPC activity in the MC, relocation of at least part of CBC capacity to the BSC, increased decarboxylation in the BSC, and increased levels of organic acids and other metabolites required for the C4 cycle. The final optimization of C4 photosynthesis would have included not only optimization of the expression patterns and regulation of enzymes involved in the C4 cycle, but also changes in the balance between different shuttles and optimization of their operation by establishing reserve pools of organic acids and amino acids. In parallel, CBC operation would have progressively adjusted to a gradual increase in the CO2 concentration around Rubisco. It should be noted that the scheme considers the proposed path to NADP-ME-type C4 photosynthesis. Shuttling of amino acids remains a crucial aspect of C4 cycles in NAD-ME- and PCK-type C4 species where much of the C4 flux is carried by Asp. Abbreviations: Ala, alanine; Asp, aspartate; BSC, bundle sheath cell; CBC, Calvin–Benson cycle; CCM, carbon-concentrating mechanism; GDC, glycine decarboxylase; MC, mesophyll cell; PEPC, phosphoenolpyruvate carboxylase; Pyr, pyruvate; Mal, malate.
From C3 to C3–C4 intermediacy
The two C3Flaveria in our study differ in terms of anatomy: F. cronquistii has classical C3 anatomy with abundant palisade mesophyll and inconspicuous BS cells, while F. robusta has a proto-Kranz anatomy, with slightly larger BS cells containing more organelles than those of F. cronquistii (Sage et al., 2013). These C3 species showed differences in metabolic profiles (Figs 1, 2), including higher levels of several CBC metabolites, Gln, as well as metabolites that are unrelated to photosynthesis such as higher Lys, Thr, Tyr, Pro, and shikimate in F. cronquistii and higher Arg content in F. robusta (Fig. 3; Supplementary Fig. S3). It is unlikely that these differences are related to the evolution of C3–C4 photosynthesis and subsequent steps towards C4 photosynthesis.
In Flaveria, the transition from C3 to C3–C4 photosynthesis has been proposed to involve a shift of GDC activity from the mesophyll to the BS cells, leading to the establishment of a photorespiratory CO2 pump in which Gly moves from the mesophyll to the BS, where it is decarboxylated, generating an elevated CO2 concentration, and Ser moves back to the mesophyll (see the Introduction; Fig. 7). The NH3 released by GDC is assimilated by the GOGAT pathway in the BS, and an intercellular shuttle is required to return an amino acid to the mesophyll and recycle an organic acid to the BS (Rawsthorne et al., 1988b; Rawsthorne and Hylton, 1991; Gowik et al., 2011; Mallmann et al., 2014).
We had expected to find an increase in Gly and Ser levels in C3–C4 intermediate Flaveria species relative to C3 species that would support a photorespiratory CO2 pump. However, total Gly content (Fig. 3C) and the active Gly pool (95 nmol g–1 FW; Fig. 5) were low in C3 species and were not consistently higher in C3–C4 species. Gly was previously reported to be low in the Clade B C3–C4 intermediate species F. floridiana, whereas it was high in the C3–C4 intermediate species, Moricandia arvensis (Leegood and von Caemmerer, 1994). Ser was present at >10-fold higher levels than Gly in the C3 species (Fig. 3C), and the absolute (both) and active (F. ramosissima) pool of Ser was noticeably even higher in the C3–C4 intermediates (Figs 3C, 5). These high Ser levels are consistent with a C2 cycle in which Ser diffuses from the BS to the mesophyll; but questions remain with regard to movement of Gly from the mesophyll to the BS (see below for further discussion). That said, Ser was not noticeably higher than Gly in an earlier study of the C3–C4 intermediate species F. floridiana and M. arvensis (Leegood and von Caemmerer, 1994).
There were no consistent or marked increases between the C3 and C3–C4 species in the levels of organic acids and amino acids that might be involved in a shuttle that maintains N stoichiometry between the BS and mesophyll. However, F. anomala and F. ramosissima are in different clades of the Flaveria phylogeny (Supplementary Fig. S1; McKown et al., 2005; Lyu et al., 2015), indicating that they represent independent origins of C3–C4 photosynthesis. Also, although both species were classified as type II intermediates by Mallmann et al. (2014), they may represent slightly different stages in the establishment of C3–C4 photosynthesis. Flaveria anomola has sometimes been considered an early C3–C4 intermediate (Moore et al., 1987), whereas F. ramosissima is an advanced intermediate with relatively high PEPC and Asp aminotransferase (AspAT) activities (Ku et al., 1983, 1991; Bauwe, 1984), and has been reported to assimilate up to 50% of its C through PEPC (Moore et al., 1987; Chastain and Chollet, 1989). We therefore inspected the metabolite profiles in F. ramosissima and F. anomala separately.
The active pool of Ser in F. ramosissima (~1200 nmol g–1 FW; Fig. 5) might be taken as a benchmark for the pool size required to support formation of a concentration gradient large enough to drive intercellular movement at a rate that balances N between the BS and mesophyll. Among organic acids and amino acids that might be involved in intercellular shuttles in F. ramosissima, the largest active pool was found for Asp (Fig. 5). Whilst the total pool of Asp did not differ between C3 and C3–C4 species (Fig. 3A), due to much higher enrichment, the active pool was ~4-fold larger in F. ramosissima than in F. robusta (Fig. 5), reaching a level (~750 nmol g–1 FW) that was about two-thirds of the F. ramosissima Ser active pool. The active pools of malate and Ala were several times smaller, and that of Pyr was very small compared with the active Ser pool in F. ramosissima (~8, 10, and 200 times smaller, respectively; Fig. 5). The total malate content in F. ramosissima was similar to or slightly lower than in C3 species (Fig. 3A), and the active pool (~150 nmol g–1 FW; Fig. 5) showed at the most only a slight increase compared with the C3 species F. robusta, and was ~10-fold smaller than the malate active pool in C4-like and complete C4Flaveria species (Fig. 5). The overall (Fig. 3) and active (Fig. 5) pools of Ala did not differ markedly between C3 species and F. ramosissima, and the active pool (~120 nmol g–1 FW; Fig 5) was >10 times smaller than in C4-like and complete C4 species (Figs 3A, 5). The total (Fig. 3) and active (Fig. 5) pools of Pyr in F. ramosissima resembled those in C3 species, and were ~3- and >10-fold smaller than the corresponding Pyr pools in C4-like species, and even smaller compared with those in complete C4 species. The 2OG content and active pool were very low in C3 species and, although the active pool was twice as large in F. ramosissima, it was also low in C4-like and complete C4 species (130 nmol g–1 FW; Figs 3, 5). Taken together, our results are consistent with the idea that high PEPC and AspAT activities in F. ramosissima (Ku et al., 1983, 1991; Bauwe, 1984) lead to formation of Asp, which acts as a major N equilibrator between the BS and mesophyll cells. They fit with a submodel of the C2 cycle that has a constraint on intercellular movement of Glu, 2OG, Ala, and Pyr (Mallmann et al., 2014). However, our analyses do not identify a clear single candidate for the C skeleton that moves from the mesophyll to the BS. Further, they do not provide strong evidence for rapid malate/Ala exchange, although slow exchange might be supported without increasing the Ala pool much above that in C3 plants.
Relative to the Clade A C3–C4 intermediate F. ramosissima, the Clade B C3–C4 intermediate F. anomala had much higher overall levels of Gln and Asn (~600 nmol g–1 FW and 900 nmol g–1 FW, respectively) and relatively high Asp and Glu, but a trend to even lower malate, Pyr, 2OG, and Ala amounts (Fig. 3A). An earlier study of another Clade B C3–C4 intermediate, F. floridiana, revealed slightly higher Asp, as well as Pyr, relative to the C3F. pringlei (Leegood and von Caemmerer, 1994). The most plausible candidates for intercellular shuttles to transfer amino groups in C3–C4 intermediate species are a Gln/Glu or an Asn/Asp exchange.
Altogether, our data for both studied C3–C4 species are consistent with the operation of a C2 cycle in which Ser diffuses from the BS to the mesophyll (Fig. 7). However, Gly levels are low and there is no consistent increase of Gly in the C3–C4 species compared with the C3 species. This raises the question of whether earlier metabolites in the photorespiratory pathway, such as glycolate, also move from the mesophyll to the BS cells, which in turn would decrease the need to recycle amino groups to the mesophyll. An earlier study of M. arvensis and F. floridana (Leegood and von Caemmerer, 1994) points to C3–C4 species in separate lineages differing in which C2 metabolite moves from the mesophyll to the BS cells. Our data also point to different sets of metabolites being involved in maintaining N balance in the two C3–C4 species studied here, with Clade A F. ramosissima using mainly Asp and Clade B F. anomala using Gln or Asn to return amino groups to the mesophyll. Collectively, our data and those of Leegood and von Caemmerer (1994) indicate that multiple metabolites may have been co-opted as carriers during the C3 to C3–C4 intermediate transition (Fig. 7), in which case the overall level of single metabolite pools would not need to increase markedly. This fits with the general similarity of the metabolite profiles of C3 and C3–C4 species, as seen in our clustering and PC analyses.
From C3–C4 intermediacy to C4-like
Advanced C3–C4 species such as F. ramosissima possess elements of a C4-like CCM, with increased PEPC activity relative to C3 and Type I C3–C4 intermediate species. The transition from a C3–C4 pathway to a C4-like state is thought to have involved a gradual shift of Rubisco from the mesophyll to BS cells and increased enzymatic activity in the later part of the C4 cycle, especially pyruvate,orthophosphate dikinase (PPDK) (Ku et al., 1983) to boost regeneration of PEP. It would also require increased decarboxylation of C4 metabolites in the BS.
Co-opting multiple metabolites to the intercellular shuttles in C3–C4 photosynthesis (see previous section) might have provided important metabolic flexibility that increased the chance of establishing a primitive C4 cycle. For example, movement of glycolate in parallel with Gly would decrease the need to shuttle amino groups back from the BS cells to the mesophyll; indeed, if more glycolate were to move than glycine, amino groups would need to be shuttled from the BS to the mesophyll. This would create a set of alternative scenarios where a primitive C4 cycle could operate simultaneously with a C2 cycle. Hypothetical options would include not only an Asp/Pyr exchange that transfers amino groups to the BS, but also amino group-neutral Asp/Ala or malate/Pyr exchanges, or even a malate/Ala cycle with net transfer of amino groups from the BS to the mesophyll.
Our metabolite data reveal a substantial change in the primary metabolome of C4-like species compared with C3–C4 intermediates. This is clearly displayed in the clustering and PC analyses (Figs 1, 2). C4-like species differ from C3–C4 intermediate species by having higher contents of C4-related metabolites, such as malate, Pyr, and Ala, and generally lower contents of photorespiratory intermediates, such as Ser and glycerate (Fig. 3). Our labelling analyses on the C4-like species F. palmeri suggest that this evolutionary step involved a large increase in the active pools of malate, Pyr, and Ala (Fig. 5), providing strong support for the operation of an expanded NADP-ME-type C4 cycle with integration of aminotransferases (Fig. 7; Furbank, 2011; Schlüter et al., 2018). Our analyses also point to increased exchange of C between the CBC and the C1–C3 positions of the carrier metabolites, presumably as a result of increased PEPC activity. Whilst Asp levels are slightly lower in F. palmeri than in F. ramosissima (Figs 3, 5), they may be sufficient to support some intercellular movement of Asp. Engelmann et al. (2003) suggested that a switch from Asp to malate as a carrier happened soon after the transition from C3–C4 to C4-like photosynthesis; malate has the advantage that its decarboxylation by NADP-ME provides the BS cells with reducing equivalents, which can power reactions in the CBC.
There were noticeably higher levels of Asn in C4-like species than in C3 and C3–C4 intermediates, and these high levels are in part seen in complete C4 species (Supplementary Fig. S3). Asn represents a potential N-rich compound for shuttling (Mallmann et al., 2014), possibly in combination with Asp or malate. 13CO2 labelling data for Asn would be required to provide direct evidence for its involvement, including assessing if the rate of labelling resembles that of Asp and malate, and determining the size of the active pool.
The overall levels of photorespiratory metabolites such as Gly, Ser, and glycerate (Fig. 3C) tended to be lower in C4-like species than in C3 and C3–C4 species. The active pools of Gly and Ser were also consistently smaller in F. palmeri than in F. ramosissima (Fig. 5), indicating that C4-like species do not employ a C2 cycle and instead benefit from lower rates of photorespiration (Ku et al., 1991).
In summary, our results point to a major reorganization of metabolism between C3–C4 and C4-like photosynthesis, including large increases in the pools of malate, Pyr, and Ala, retention of a large pool of Asp, and an increase in the pool of Asn. This is consistent with the establishment of a mixed C4-like CCM involving both organic acids and amino acids (Fig. 7). The decrease in levels of photorespiratory metabolites is consistent with a substantial decrease in photorespiration.
From C4-like to complete C4
The last step in the evolution of C4 photosynthesis is proposed to have included optimization of the compartmentation and the properties of key CCM-related enzymes such as Rubisco and PEPC (Kubien et al., 2008; Gowik and Westhoff, 2011; Kapralov et al., 2011). Our metabolite analyses capture further changes that probably occurred in parallel. One appears to be a refinement of the NADP-ME-type C4 cycle toward one that is less reliant on aminotransferase reactions (Fig. 7). Compared with C4-like species, complete C4 species show a trend to higher malate and Pyr content, a 2- to 3-fold lower amount of Asp, and a trend to lower levels of Ala (Fig. 3A). Labelling studies revealed that, compared with C4-like F. palmeri, the complete C4 species F. bidentis contained a similar or marginally higher active malate pool, an almost 2-fold higher active Pyr pool, and smaller active pools of Asp and Ala (Fig. 5).
Another striking shift between C4-like species and those with a complete C4 syndrome was an increase in the levels of several TCA cycle intermediates (citrate, isocitrate, aconitate, and 2OG) (Figs 1, 2, 3A; Supplementary Figs S2, S3). Although the total Glu pool did not increase, more 13C moved into Glu in the C4 plant F. bidentis than in C4-like F. palmeri (Fig. 5; Supplementary Fig. S4), indicating increased flux through the TCA cycle to this amino acid. It is notable that levels of 2OG and Glu in F. bidentis resembled those in the model NADP-ME-type C4 species maize (Fig. 5). One possible role for 2OG and Glu in C4 photosynthesis might be to support flux through AspAT or Ala aminotransferase in the mesophyll or the BS cells. This might have contributed to the decline in Ala and especially Asp in the transition from C4-like to C4 photosynthesis in Flaveria (Fig. 7). It might be noted that for participation of Glu and 2OG in aminotransferase reactions, the overall rather than the active pool (as defined by incorporation of 13C) may be more relevant. At least in steady-state conditions, this function can be supported by existing pools, as the C skeleton of Glu and 2OG is recycled during the complete amino transfer cycle and does not need to be synthesized de novo from newly fixed C.
A further interesting change in the transition from C4-like to C4 in Flaveria concerns glycerate, whose level is higher in complete C4 species than in C4-like species (Fig. 3C). Glycerate presumably has different roles in the C4 and C3 pathways, as photorespiration is greatly reduced in C4 plants. One possibility is that glycerate acts as a C reserve that can be used to replenish the levels of CBC- and CCM-related metabolites (Usuda and Edwards, 1980; Weber and von Caemmerer, 2010), for example in transitions between different C availability states (Levey et al., 2019).
Overall, our results indicate that, in addition to changes in the location and properties of enzymes, optimization of C4 photosynthesis in Flaveria also involved (i) a shift towards increased use of a malate/Pyr compared with an Asp/Ala exchange, implying that NADP-ME makes an increasing contribution to decarboxylation compared with other decarboxylation pathways; (ii) an increase in the overall levels of 2OG that might potentially aid aminotransferase reactions in one or both cell types; and (iii) an increase in the levels of TCA cycle acids and glycerate that might serve as C reservoirs to buffer against fluctuations in the environment.
Clade-dependent metabolite differences and considerations
The absence of complete C4 species in Flaveria Clade B raises questions about the evolution of C4 photosynthesis in the genus. Does it reflect a difference between the two clades in the timing of the evolutionary process where Clade B species are younger and none has yet had time to complete the passage to C4 photosynthesis, or are Clade B species not found in habitats that favour C4 photosynthesis, or was C3-C4 or C4-like photosynthesis established in Clade B using a combination of biochemical pathways that was unfavourable for evolution of a complete C4 syndrome, such as the lack of genetic enablers (Christin and Osborne, 2014)? The finding that the Clade B C3–C4F. anomola and C4-like F. brownii formed a distinct cluster from all Clade A species (Fig. 1) is consistent with them having a different balance of biochemical pathways. Curiously, in the PC analysis of Fig. 2, the Clade A C3–C4 intermediate F. ramosissima grouped quite close to the basal C3 species, and was clearly separated from the Clade B C3–C4F. anomola that grouped closer to the C4-like species, including those in Clade A. This indicates that if there are clade-dependent differences in metabolism that hinder the move towards full C4 photosynthesis, they may be rather specific.
As already mentioned, the Clade B C3–C4 intermediate F. anomala may use Gln/Glu or Asn/Asp shuttles to return an amino acid to the mesophyll and recycle an organic acid to the BS. This might be less conducive to evolution of a PEPC-based CCM than a shuttle based on Asp and one or a mix of organic acids. Separation of the Clade B C4-like species F. brownii from the Clade A C4-like species was also driven by high amino acid levels, including Ala and Asn (Fig. 2). However, Ala, Asn, and in one case (C4-like F. vaginata) Gln levels were higher in C4-like than in C3–C4 species in Clade A, indicating that high levels of these amino acids do not per se preclude evolution of full C4 photosynthesis. Other clade-dependent differences included a trend to higher RuBP, FBP, and SBP in Clade B species. It is not obvious how this would affect the propensity to evolve a complete C4 pathway.
While our analyses reveal clade-dependent differences in metabolism, too few species were studied to conclusively identify specific features that might hinder evolution of complete C4 photosynthesis in Clade B. It is also noteworthy that F. brownii dampens its C4 cycle at low irradiance (Cheng et al., 1989). This observation is consistent with the idea that the gains from C4-like photosynthesis in this species are not large, especially in low irradiance when the cost of operating its CCM may be high compared with available light energy.
Changes in CBC operation along the evolutionary path from C3 to C4 photosynthesis
It is well established that CBC operation differs between C3 and C4 species (see the Introduction). Recently, profiling of CBC metabolites revealed widespread differences in levels between C4 and C3 species, between different C4 species, and between different C3 species (Arrivault et al., 2019; Borghi et al., 2019). This indicates that in addition to adjustment of the CBC during the evolution of C4 photosynthesis, there is also interspecies diversity in how the CBC operates in different C3 species and in different C4 species. This diversity may be relevant to current attempts to engineer C4 photosynthesis into C3 crops, and to exploit photosynthetic diversity for improving crops (Lawson et al., 2012; Driever et al., 2014; Simkin et al., 2019).
To investigate how CBC operation was modified during the evolutionary progression from C3 to C4 photosynthesis, we used PC analysis to compare CBC metabolite profiles in the nine Flaveria species and in several C3 and C4 species from outside the genus (Fig. 6). These analyses showed that it is possible to separate plants with differing photosynthetic strategies on the basis of their CBC metabolite profile. The C3Flaveria species have a CBC profile close to that of the other C3 species examined, especially rice and wheat. Intermediate Flaveria species lie on the path between C3 and C4 species, with the C3–C4 intermediate species being closer to C3 species and C4-like species closer to complete C4 species. Further, diversity in CBC function may exist between Flaveria clades, as the metabolic profiles of the Clade B C4-like F. brownii and C3–C4F. anomala differ from those of Clade A C4-like and C3–C4 species. More generally, these results indicate that CBC operation had already started to adapt to enhanced internal CO2 status in the C3–C4Flaveria species and that this adaptation progressed furthest in C4-like and C4 species in Clade A.
The evolutionary transition from C3 to C4 photosynthesis was mainly captured in PC1, and was driven by increasing 3PGA and decreasing levels of 2PG and several CBC intermediates including S7P, F6P, SBP, FBP, and, especially, RuBP. RuBP is highest in C3 and C3–C4 species, and decreases through C4-like to C4 species (Fig. 3B). Evolution of C4 photosynthesis is accompanied by a decrease in Rubisco abundance, made possible by relaxed selection for catalytic fidelity for CO2 and an associated increase in kcat (Sage, 2002; Ghannoum et al., 2005; Kapralov et al., 2011). As most RuBP is bound in the active site of Rubisco (Salvucci, 1989), the lower RuBP may in part reflect the lower Rubisco abundance. Lower 2PG presumably reflects the decrease in photorespiration.
The trend to higher 3PGA in C4-like and C4 species may be linked to operation of a 3PGA/DHAP shuttle between mesophyll and BS cells (Leegood, 1985; Stitt and Heldt, 1985; Arrivault et al., 2017). High levels of both 3PGA and DHAP are found in Setaria viridis and maize (see Fig. 5). This shuttle allows BS chloroplasts to outsource part of the reductive phase of the CBC to the mesophyll chloroplasts where ATP and NADPH are readily available, and is required in NADP-ME-type C4 species such as maize that have dimorphic chloroplasts and little or no PSII activity in the BS cells (Woo et al., 1970; Andersen et al., 1972; Munekage, 2016). It is still unproven if S. viridis makes use of this shuttle and has the same chloroplastic properties as maize, but our metabolite data suggest this is the case. The C4Flaveria species have PSII activity in their BS chloroplasts to a varying extent, depending on conditions (Laetsch and Price, 1969; Höfer et al., 1992; Meister et al., 1996; Nakamura et al., 2013). There is only a weak trend to higher 3PGA in Flaveria C4-like and C4 species (Fig. 6; see also Fig. 3B). This might reflect partial operation of an intercellular 3PGA/DHAP shuttle (Leegood and von Caemmerer, 1994).
General conclusions
Our comparative analysis of metabolite levels in nine Flaveria species representing steps along the evolutionary path from C3 to complete C4 photosynthesis supports and extends current ideas about how C4 photosynthesis evolved through several semi-stable stages (Fig. 7). At the level of metabolism, the first step was preferential location of GDC activity to the BS cells and the associated appearance of intercellular shuttles of photorespiratory metabolites. High levels of Ser in C3–C4Flaveria species support the idea that this metabolite moves from the BS to the mesophyll cells following GDC activity. However, whilst Asp levels were high, we did not find substantial increases in the levels of Gly or of other metabolites that are required to maintain N stoichiometry between the mesophyll and BS, indicating that intercellular fluxes may be partitioned between two or more metabolites. The transition from C3–C4 to a C4-like state was associated with a major change in metabolism, including clear increases of malate and three-C metabolites that are required to shuttle CO2 into the BS cells. The transition from C4-like to a complete C4 syndrome was associated with changes in metabolites indicative of a shift in the balance between malate/Pyr and Asp/Ala shuttles, improved operation of aminotransferase reactions, and the formation of C reservoirs to buffer C4 metabolism against fluctuations in the environment. In addition, focused analysis of CBC metabolites revealed progressive adaptation of the CBC to an increasingly effective CCM.
Supplementary data
The following supplementary data are available at JXB online.
Fig. S1. Phylogenetic tree of 16 Flaveria species.
Fig. S2. Principal component (PC) analyses of metabolite profiles in nine Flaveria species: PC1 in combination with PC3 or PC4.
Fig. S3. Absolute amounts of additional metabolites in nine Flaveria species.
Fig. S4. 13C enrichment in key metabolites in four Clade A Flaveria species with different modes of photosynthesis.
Table S1. Germination, growth, and harvest conditions.
Table S2. List of measured metabolites, their abbreviations, and analytical methods used.
Table S3. Summary of principal components for analyses on the entire Flaveria dataset.
Table S4. Use of metabolite amounts and 13C enrichments to estimate active and inactive pools in four Flaveria species and maize.
Table S5. Summary of first 10 PCs for PC analyses on the dimensionless multispecies dataset for CBC metabolites and 2PG.
Dataset S1. Metabolite amounts in nine Flaveria species.
Dataset S2. 13C enrichment (%) in a selection of metabolites in four Flaveria species.
Dataset S3. Metabolite amounts of CBC intermediates plus 2PG in nine Flaveria species combined with data for another five C3 species and two C4 species from Arrivault et al. (2019).
Acknowledgements
We wish to thank Christin Abel and Sandra Stegemann (MPI-MP) for the Flaveria cultivation and propagation help and advice. Many thanks go to Hirofumi Ishihara (MPI-MP) for helping during the 13CO2 labelling system setup, and to lots of labelling helping hands from AG Stitt at the MPI-MP: Jana Verbančič, Thiago Moraes, Rubén Vicente, Haim Trebes, Armin Schlereth, Virginie Mengin, Sandra Kerbler, and Chiara Baccolini.
Glossary
Abbreviations
- BS
bundle sheath
- CA
carbonic anhydrase
- CBC
Calvin–Benson cycle
- CCM
CO2-concentrating mechanism
- GC-TOF-MS
GC coupled with time-of-flight MS
- GDC
glycine decarboxylase
- GOGAT
glutamine–oxoglutarate aminotransferase
- NADP-ME
NADP-dependent malic enzyme
- PC
principal component
- PCK
phosphoenolpyruvate carboxykinase
- PEPC
phosphoenolpyruvate carboxylase
- RH
relative humidity
- PPM®
Plant Preservation Medium
- TCA
tricarboxylic acid
Author contributions
GLB, SA, ML, JEL, and MS: conceptualization and planning the experiments; GLB: growing plants and harvesting samples; GLB, SA, MG, and DBM: extracting the samples for downstream analyses: GLB, SA, MG, DBM, ED, GMF, and PC: processing and analysis; GLB: statistical analyses; GLB, SA, ML, and MS: writing; ML, SA, ARF, JEL, and MS: supervision.
Conflict of interest
The authors declare that they have no conflicts of interest.
Funding
This work was financially supported by the Max Planck Society (GLB, SA, MG, DBM, ARF, JEL, and MS), by C4 Rice Project grants from the Bill & Melinda Gates Foundation to the University of Oxford [2015–2019, OPP1129902; 2019–2024, INV-002870 (to GLB and SA)], the German Federal Ministry of Education and Research (BMBF grant 031B0205C to ARF, MS, SA, and DBM), the Valere grant from the Università degli Studi della Campania Luigi Vanvitelli (EDA, GMF, and PC), the German Academic Exchange Service (DAAD grant 57396919 to PC, JEL, and GLB), and the Australian Research Council Discovery Project Grant DP150101037 (ML, JEL, and MS).
Data availability
All data supporting the findings of this study are available within the paper and within its supplementary data published online.
References
- Allen DK, Young JD.. 2020. Tracing metabolic flux through time and space with isotope labeling experiments. Current Opinion in Biotechnology 64, 92–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Andersen KS, Bain JM, Bishop DG, Smillie RM.. 1972. Photosystem II activity in agranal bundle sheath chloroplasts from Zea mays. Plant Physiology 49, 461–466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Arrivault S, Alexandre Moraes T, Obata T, et al. 2019. Metabolite profiles reveal interspecific variation in operation of the Calvin–Benson cycle in both C4 and C3 plants. Journal of Experimental Botany 70, 1843–1858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Arrivault S, Guenther M, Fry SC, Fuenfgeld MM, Veyel D, Mettler-Altmann T, Stitt M, Lunn JE.. 2015. Synthesis and use of stable-isotope-labeled internal standards for quantification of phosphorylated metabolites by LC-MS/MS. Analytical Chemistry 87, 6896–6904. [DOI] [PubMed] [Google Scholar]
- Arrivault S, Guenther M, Ivakov A, Feil R, Vosloh D, van Dongen JT, Sulpice R, Stitt M.. 2009. Use of reverse-phase liquid chromatography, linked to tandem mass spectrometry, to profile the Calvin cycle and other metabolic intermediates in Arabidopsis rosettes at different carbon dioxide concentrations. The Plant Journal 59, 826–839. [DOI] [PubMed] [Google Scholar]
- Arrivault S, Obata T, Szecówka M, Mengin V, Guenther M, Hoehne M, Fernie AR, Stitt M.. 2017. Metabolite pools and carbon flow during C4 photosynthesis in maize: 13CO2 labeling kinetics and cell type fractionation. Journal of Experimental Botany 68, 283–298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Aubry S, Brown NJ, Hibberd JM.. 2011. The role of proteins in C3 plants prior to their recruitment into the C4 pathway. Journal of Experimental Botany 62, 3049–3059. [DOI] [PubMed] [Google Scholar]
- Bauwe H. 1984. Photosynthetic enzyme activities and immunofluorescence studies on the localization of ribulose-1,5-bisphosphate carboxylase/oxygenase in leaves of C3, C4, and C3–C4 intermediate species of Flaveria (Asteraceae). Biochemie und Physiologie der Pflanzen 179, 253–268. [Google Scholar]
- Bauwe H, Hagemann M, Fernie AR.. 2010. Photorespiration: players, partners and origin. Trends in Plant Science 15, 330–336. [DOI] [PubMed] [Google Scholar]
- Bauwe H, Kolukisaoglu U.. 2003. Genetic manipulation of glycine decarboxylation. Journal of Experimental Botany 54, 1523–1535. [DOI] [PubMed] [Google Scholar]
- Bohley K, Schröder T, Kesselmeier J, Ludwig M, Kadereit G.. 2019. C4-like photosynthesis and the effects of leaf senescence on C4-like physiology in Sesuvium sesuvioides (Aizoaceae). Journal of Experimental Botany 70, 1553–1565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Borghi GL, Moraes TA, Günther M, Feil R, Mengin V, Lunn JE, Stitt M, Arrivault S.. 2019. Relationship between irradiance and levels of Calvin–Benson cycle and other intermediates in the model eudicot Arabidopsis and the model monocot rice. Journal of Experimental Botany 70, 5809–5825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brown RH. 1978. A difference in N use efficiency in C3 and C4 plants and its implications in adaptation and evolution. Crop Science 18, 93–98. [Google Scholar]
- Busch FA, Sage TL, Cousins AB, Sage RF.. 2013. C3 plants enhance rates of photosynthesis by reassimilating photorespired and respired CO2. Plant, Cell & Environment 36, 200–212. [DOI] [PubMed] [Google Scholar]
- Carillo P, Mastrolonardo G, Nacca F, Fuggi A.. 2005. Nitrate reductase in durum wheat seedlings as affected by nitrate nutrition and salinity. Functional Plant Biology 32, 209–219. [DOI] [PubMed] [Google Scholar]
- Chastain CJ, Chollet R.. 1989. Interspecific variation in assimilation of 14CO2 into C4 acids by leaves of C3, C4 and C3–C4 intermediate Flaveria species near the CO2 compensation concentration. Planta 179, 81–88. [DOI] [PubMed] [Google Scholar]
- Cheng SH, Moore BD, Edwards GE, Ku MS.. 1988. Photosynthesis in Flaveria brownii, a C4-like species: leaf anatomy, characteristics of CO2 exchange, compartmentation of photosynthetic enzymes, and metabolism of CO2. Plant Physiology 87, 867–873. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cheng SH, Moore BD, Wu J, Edwards GE, Ku MS.. 1989. Photosynthetic plasticity in Flaveria brownii: growth irradiance and the expression of C4 photosynthesis. Plant Physiology 89, 1129–1135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Christin PA, Besnard G, Samaritani E, Duvall MR, Hodkinson TR, Savolainen V, Salamin N.. 2008. Oligocene CO2 decline promoted C4 photosynthesis in grasses. Current Biology 18, 37–43. [DOI] [PubMed] [Google Scholar]
- Christin PA, Osborne CP.. 2014. The evolutionary ecology of C4 plants. New Phytologist 204, 765–781. [DOI] [PubMed] [Google Scholar]
- Danila FR, Quick WP, White RG, Furbank RT, von Caemmerer S.. 2016. The metabolite pathway between bundle sheath and mesophyll: quantification of plasmodesmata in leaves of C3 and C4 monocots. The Plant Cell 28, 1461–1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Danila FR, Quick WP, White RG, Kelly S, von Caemmerer S, Furbank RT.. 2018. Multiple mechanisms for enhanced plasmodesmata density in disparate subtypes of C4 grasses. Journal of Experimental Botany 69, 1135–1145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Driever SM, Lawson T, Andralojc PJ, Raines CA, Parry MA.. 2014. Natural variation in photosynthetic capacity, growth, and yield in 64 field-grown wheat genotypes. Journal of Experimental Botany 65, 4959–4973. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Engelmann S, Bläsing OE, Gowik U, Svensson P, Westhoff P.. 2003. Molecular evolution of C4 phosphoenolpyruvate carboxylase in the genus Flaveria—a gradual increase from C3 to C4 characteristics. Planta 217, 717–725. [DOI] [PubMed] [Google Scholar]
- Ermakova M, Arrivault S, Giuliani R, et al. 2020. Installation of C4 photosynthetic pathway enzymes in rice using a single construct. Plant Biotechnology Journal 19, 575–588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ermakova M, Danila FR, Furbank RT, von Caemmerer S.. 2019. On the road to C4 rice: advances and perspectives. The Plant Journal 101, 940–950. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Furbank RT. 2011. Evolution of the C4 photosynthetic mechanism: are there really three C4 acid decarboxylation types? Journal of Experimental Botany 62, 3103–3108. [DOI] [PubMed] [Google Scholar]
- Ghannoum O. 2009. C4 photosynthesis and water stress. Annals of Botany 103, 635–644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ghannoum O, Evans JR, Chow WS, Andrews TJ, Conroy JP, von Caemmerer S.. 2005. Faster Rubisco is the key to superior nitrogen-use efficiency in NADP-malic enzyme relative to NAD-malic enzyme C4 grasses. Plant Physiology 137, 638–650. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ghannoum O, Evans JR, von Caemmerer S.. 2011. Nitrogen and water use efficiency of C4 plants. In: Raghavendra AS, Sage RF, eds. C4 photosynthesis and related CO2 concentrating mechanisms. Dordrecht: Springer, 129–146. [Google Scholar]
- Gowik U, Bräutigam A, Weber KL, Weber AP, Westhoff P.. 2011. Evolution of C4 photosynthesis in the genus Flaveria: how many and which genes does it take to make C4? The Plant Cell 23, 2087–2105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gowik U, Westhoff P.. 2011b. C4 phosphoenolpyruvate carboxylase. In: Raghavendra AS, Sage RF, eds. C4 photosynthesis and related CO2 concentrating mechanisms. Dordrecht: Springer, 257–275. [Google Scholar]
- Hasunuma T, Harada K, Miyazawa S, Kondo A, Fukusaki E, Miyake C.. 2010. Metabolic turnover analysis by a combination of in vivo 13C-labelling from 13CO2 and metabolic profiling with CE-MS/MS reveals rate-limiting steps of the C3 photosynthetic pathway in Nicotiana tabacum leaves. Journal of Experimental Botany 61, 1041–1051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Häusler RE, Hirsch HJ, Kreuzaler F, Peterhänsel C.. 2002. Overexpression of C4-cycle enzymes in transgenic C3 plants: a biotechnological approach to improve C3-photosynthesis. Journal of Experimental Botany 53, 591–607. [DOI] [PubMed] [Google Scholar]
- Heckmann D. 2016. C4 photosynthesis evolution: the conditional Mt. Fuji. Current Opinion in Plant Biology 31, 149–154. [DOI] [PubMed] [Google Scholar]
- Heckmann D, Schulze S, Denton A, Gowik U, Westhoff P, Weber AP, Lercher MJ.. 2013. Predicting C4 photosynthesis evolution: modular, individually adaptive steps on a Mount Fuji fitness landscape. Cell 153, 1579–1588. [DOI] [PubMed] [Google Scholar]
- Höfer MU, Santore UJ, Westhoff P.. 1992. Differential accumulation of the 10-, 16- and 23-kDa peripheral components of the water-splitting complex of photosystem II in mesophyll and bundle-sheath chloroplasts of the dicotyledonous C4 plant Flaveria trinervia (Spreng.) C. Mohr. Planta 186, 304–312. [DOI] [PubMed] [Google Scholar]
- Huege J, Goetze J, Dethloff F, Junker B, Kopka J.. 2014. Quantification of stable isotope label in metabolites via mass spectrometry. Methods in Molecular Biology 1056, 213–223. [DOI] [PubMed] [Google Scholar]
- Hylton CM, Rawsthorne S, Smith AM, Jones DA, Woolhouse HW.. 1988. Glycine decarboxylase is confined to the bundle-sheath cells of leaves of C3–C4 intermediate species. Planta 175, 452–459. [DOI] [PubMed] [Google Scholar]
- Ishihara H, Obata T, Sulpice R, Fernie AR, Stitt M.. 2015. Quantifying protein synthesis and degradation in Arabidopsis by dynamic 13CO2 labeling and analysis of enrichment in individual amino acids in their free pools and in protein. Plant Physiology 168, 74–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kapralov MV, Kubien DS, Andersson I, Filatov DA.. 2011. Changes in Rubisco kinetics during the evolution of C4 photosynthesis in Flaveria (Asteraceae) are associated with positive selection on genes encoding the enzyme. Molecular Biology and Evolution 28, 1491–1503. [DOI] [PubMed] [Google Scholar]
- Keerberg O, Pärnik T, Ivanova H, Bassüner B, Bauwe H.. 2014. C2 photosynthesis generates about 3-fold elevated leaf CO2 levels in the C3–C4 intermediate species Flaveria pubescens. Journal of Experimental Botany 65, 3649–3656. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Khoshravesh R, Stata M, Busch FA, et al. 2019. The evolutionary origin of C4 photosynthesis in the grass subtribe Neurachninae. Plant Physiology 182, 566–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Khoshravesh R, Stinson CR, Stata M, Busch FA, Sage RF, Ludwig M, Sage TL.. 2016. C3–C4 intermediacy in grasses: organelle enrichment and distribution, glycine decarboxylase expression, and the rise of C2 photosynthesis. Journal of Experimental Botany 67, 3065–3078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kopka J, Schauer N, Krueger S, et al. 2005. GMD@CSB.DB: the Golm Metabolome Database. Bioinformatics 21, 1635–1638. [DOI] [PubMed] [Google Scholar]
- Ku MS, Wu J, Dai Z, Scott RA, Chu C, Edwards GE.. 1991. Photosynthetic and photorespiratory characteristics of Flaveria species. Plant Physiology 96, 518–528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ku MSB, Monson RK, Littlejohn RO, Nakamoto H, Fisher DB, Edwards GE.. 1983. Photosynthetic characteristics of C3–C4 intermediate Flaveria species - I. Leaf anatomy, photosynthetic responses to O2 and CO2 and activities of key enzymes in the C3 and C4 pathways. Plant Physiology 75, 993–996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kubien DS, Whitney SM, Moore PV, Jesson LK.. 2008. The biochemistry of Rubisco in Flaveria. Journal of Experimental Botany 59, 1767–1777. [DOI] [PubMed] [Google Scholar]
- Laetsch WM, Price I.. 1969. Development of the dimorphic chloroplasts of sugar cane. American Journal of Botany 56, 77–87. [Google Scholar]
- Langdale JA. 2011. C4 cycles: past, present, and future research on C4 photosynthesis. The Plant Cell 23, 3879–3892. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lawson T, Kramer DM, Raines CA.. 2012. Improving yield by exploiting mechanisms underlying natural variation of photosynthesis. Current Opinion in Biotechnology 23, 215–220. [DOI] [PubMed] [Google Scholar]
- Leegood RC. 1985. The intercellular compartmentation of metabolites in leaves of Zea mays L. Planta 164, 163–171. [DOI] [PubMed] [Google Scholar]
- Leegood RC, von Caemmerer S.. 1994. Regulation of photosynthetic carbon assimilation in leaves of C3–C4 intermediate species of Moricandia and Flaveria. Planta 192, 232–238. [Google Scholar]
- Levey M, Timm S, Mettler-Altmann T, Borghi GL, Koczor M, Arrivault S, Weber AP, Bauwe H, Gowik U, Westhoff P.. 2019. Efficient 2-phosphoglycolate degradation is required to maintain carbon assimilation and allocation in the C4 plant Flaveria bidentis. Journal of Experimental Botany 70, 575–587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lisec J, Schauer N, Kopka J, Willmitzer L, Fernie AR.. 2006. Gas chromatography mass spectrometry-based metabolite profiling in plants. Nature Protocols 1, 387–396. [DOI] [PubMed] [Google Scholar]
- Lorimer GH, Andrews TJ.. 1973. Plant photorespiration—an inevitable consequence of the existence of atmospheric oxygen. Nature 243, 359–360. [Google Scholar]
- Lundgren MR. 2020. C2 photosynthesis: a promising route towards crop improvement? New Phytologist 228, 1734–1740. [DOI] [PubMed] [Google Scholar]
- Lundgren MR, Osborne CP, Christin PA.. 2014. Deconstructing Kranz anatomy to understand C4 evolution. Journal of Experimental Botany 65, 3357–3369. [DOI] [PubMed] [Google Scholar]
- Lyu MJA, Gowik U, Kelly S, et al. 2015. RNA-Seq based phylogeny recapitulates previous phylogeny of the genus Flaveria (Asteraceae) with some modifications. BMC Ecology and Evolution 15, 116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ma F, Jazmin LJ, Young JD, Allen DK.. 2014. Isotopically nonstationary 13C flux analysis of changes in Arabidopsis thaliana leaf metabolism due to high light acclimation. Proceedings of the National Academy of Sciences, USA 111, 16967–16972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mallmann J, Heckmann D, Bräutigam A, Lercher MJ, Weber AP, Westhoff P, Gowik U.. 2014. The role of photorespiration during the evolution of C4 photosynthesis in the genus Flaveria. eLife 3, e02478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Matsuoka M, Furbank RT, Fukayama H, Miyao M.. 2001. Molecular engineering of C4 photosynthesis. Annual Review of Plant Physiology and Plant Molecular Biology 52, 297–314. [DOI] [PubMed] [Google Scholar]
- McKown AD, Moncalvo JM, Dengler NG.. 2005. Phylogeny of Flaveria (Asteraceae) and inference of C4 photosynthesis evolution. American Journal of Botany 92, 1911–1928. [DOI] [PubMed] [Google Scholar]
- Meister M, Agostino A, Hatch MD.. 1996. The roles of malate and aspartate in C4 photosynthetic metabolism of Flaveria bidentis (L). Planta 199, 262–269. [Google Scholar]
- Merlo L, Geigenberger P, Hajirezaei M, Stitt M.. 1993. Changes of carbohydrates, metabolites and enzyme activities in potato tubers during development, and within a single tuber along a stolon–apex gradient. Journal of Plant Physiology 142, 392–402. [Google Scholar]
- Miyao M, Masumoto C, Miyazawa S, Fukayama H.. 2011. Lessons from engineering a single-cell C4 photosynthetic pathway into rice. Journal of Experimental Botany 62, 3021–3029. [DOI] [PubMed] [Google Scholar]
- Monson RK, Moore BD, Ku MS, Edwards GE.. 1986. Co-function of C3- and C4-photosynthetic pathways in C3, C4 and C3–C4 intermediate Flaveria species. Planta 168, 493–502. [DOI] [PubMed] [Google Scholar]
- Monson RK, Rawsthorne S.. 2000. CO2 assimilation in C3–C4 intermediate plants. In: Leegood RC, Sharkey TD, von Caemmerer S, eds. Photosynthesis. Advances in photosynthesis and respiration. Dordrecht: Springer; 9, 533–550. [Google Scholar]
- Moore BD, Ku MSB, Edwards GE.. 1987. C4 photosynthesis and light-dependent accumulation of inorganic carbon in leaves of C3–C4 and C4Flaveria species. Australian Journal of Plant Physiology 14, 657–68. [Google Scholar]
- Moore B, Ku MSB, Edwards GE.. 1989. Expression of C4‐like photosynthesis in several species of Flaveria. Plant, Cell & Environment 12, 541–549. [Google Scholar]
- Muhaidat R, Sage RF, Dengler NG.. 2007. Diversity of Kranz anatomy and biochemistry in C4 eudicots. American Journal of Botany 94, 362–381. [DOI] [PubMed] [Google Scholar]
- Muhaidat R, Sage TL, Frohlich MW, Dengler NG, Sage RF.. 2011. Characterization of C₃–C₄ intermediate species in the genus Heliotropium L. (Boraginaceae): anatomy, ultrastructure and enzyme activity. Plant, Cell & Environment 34, 1723–1736. [DOI] [PubMed] [Google Scholar]
- Munekage YN. 2016. Light harvesting and chloroplast electron transport in NADP-malic enzyme type C4 plants. Current Opinion in Plant Biology 31, 9–15. [DOI] [PubMed] [Google Scholar]
- Nakamura N, Iwano M, Havaux M, Yokota A, Munekage YN.. 2013. Promotion of cyclic electron transport around photosystem I during the evolution of NADP-malic enzyme-type C4 photosynthesis in the genus Flaveria. New Phytologist 199, 832–842. [DOI] [PubMed] [Google Scholar]
- Osmond CB, Harris B.. 1971. Photorespiration during C4 photosynthesis. Biochimica et Biophysica Acta 234, 270–282. [DOI] [PubMed] [Google Scholar]
- Rawsthorne S, Hylton CM.. 1991. The relationship between the post-illumination CO2 burst and glycine metabolism in leaves of C3 and C3–C4 intermediate species of Moricandia. Planta 186, 122–126. [DOI] [PubMed] [Google Scholar]
- Rawsthorne S, Hylton CM, Smith AM, Woolhouse HW.. 1988a. Photorespiratory metabolism and immunogold localization of photorespiratory enzymes in leaves of C3 and C3–C4 intermediate species of Moricandia. Planta 173, 298–308. [DOI] [PubMed] [Google Scholar]
- Rawsthorne S, Hylton CM, Smith AM, Woolhouse HW.. 1988b. Distribution of photorespiratory enzymes between bundle-sheath and mesophyll cells in leaves of the C3–C4 intermediate species Moricandia arvensis (L.) DC. Planta 176, 527–532. [DOI] [PubMed] [Google Scholar]
- Sage RF. 2002. Variation in the kcat of Rubisco in C3 and C4 plants and some implications for photosynthetic performance at high and low temperature. Journal of Experimental Botany 53, 609–620. [DOI] [PubMed] [Google Scholar]
- Sage RF. 2004. The evolution of C4 photosynthesis. New Phytologist 161, 341–370. [DOI] [PubMed] [Google Scholar]
- Sage RF. 2016. A portrait of the C4 photosynthetic family on the 50th anniversary of its discovery: species number, evolutionary lineages, and Hall of Fame. Journal of Experimental Botany 67, 4039–4056. [DOI] [PubMed] [Google Scholar]
- Sage RF, Christin PA, Edwards EJ.. 2011. The C4 plant lineages of planet Earth. Journal of Experimental Botany 62, 3155–3169. [DOI] [PubMed] [Google Scholar]
- Sage RF, Monson RK, Ehleringer JR, Adachi S, Pearcy RW.. 2018. Some like it hot: the physiological ecology of C4 plant evolution. Oecologia 187, 941–966. [DOI] [PubMed] [Google Scholar]
- Sage RF, Sage TL, Kocacinar F.. 2012. Photorespiration and the evolution of C4 photosynthesis. Annual Review of Plant Biology 63, 19–47. [DOI] [PubMed] [Google Scholar]
- Sage TL, Busch FA, Johnson DC, Friesen PC, Stinson CR, Stata M, Sultmanis S, Rahman BA, Rawsthorne S, Sage RF.. 2013. Initial events during the evolution of C4 photosynthesis in C3 species of Flaveria. Plant Physiology 163, 1266–1276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Salvucci ME. 1989. Regulation of Rubisco activity in vivo. Plant Physiology 77, 164–171. [Google Scholar]
- Schlüter U, Bräutigam A, Droz JM, Schwender J, Weber APM.. 2018. The role of alanine and aspartate aminotransferases in C4 photosynthesis. Plant Biology 21, 64–76. [DOI] [PubMed] [Google Scholar]
- Schlüter U, Weber AP.. 2016. The road to C4 photosynthesis: evolution of a complex trait via intermediary states. Plant & Cell Physiology 57, 881–889. [DOI] [PubMed] [Google Scholar]
- Schlüter U, Weber APM.. 2020. Regulation and evolution of C4 photosynthesis. Annual Review of Plant Biology 71, 183–215. [DOI] [PubMed] [Google Scholar]
- Schmitt MR, Edwards GE.. 1981. Photosynthetic capacity and nitrogen use efficiency of maize, wheat, and rice: a comparison between C3 and C4 photosynthesis. Journal of Experimental Botany 32, 459–466. [Google Scholar]
- Schulze S, Mallmann J, Burscheidt J, Koczor M, Streubel M, Bauwe H, Gowik U, Westhoff P.. 2013. Evolution of C4 photosynthesis in the genus Flaveria: establishment of a photorespiratory CO2 pump. The Plant Cell 25, 2522–2535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schüssler C, Freitag H, Koteyeva N, Schmidt D, Edwards G, Voznesenskaya E, Kadereit G.. 2017. Molecular physlogeny and forms of photosynthesis in tribe Salsoleae (Chenopodiaceae). Journal of Experimental Botany 68, 207–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sharkey TD, Preiser AL, Weraduwage SM, Gog L.. 2020. Source of 12C in Calvin Benson cycle intermediates and isoprene emitted from plant leaves fed with 13CO2. The Biochemical Journal 477, 3237–3252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Simkin AJ, López-Calcagno PE, Raines CA.. 2019. Feeding the world: improving photosynthetic efficiency for sustainable crop production. Journal of Experimental Botany 70, 1119–1140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stitt M, Heldt HW.. 1985. Generation and maintenance of concentration gradients between the mesophyll and bundle sheath in maize leaves. Biochimica et Biophysica Acta 80, 400–414. [Google Scholar]
- Szecowka M, Heise R, Tohge T, et al. 2013. Metabolic fluxes in an illuminated Arabidopsis rosette. The Plant Cell 25, 694–714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Usuda H, Edwards GE.. 1980. Localization of glycerate kinase and some enzymes for sucrose synthesis in C3 and C4 plants. Plant Physiology 65, 1017–1022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weber AP, von Caemmerer S.. 2010. Plastid transport and metabolism of C3 and C4 plants—comparative analysis and possible biotechnological exploitation. Current Opinion in Plant Biology 13, 257–265. [DOI] [PubMed] [Google Scholar]
- Weissmann S, Ma F, Furuyama K, Gierse J, Berg H, Shao Y, Taniguchi M, Allen DK, Brutnell TP.. 2016. Interactions of C4 subtype metabolic activities and transport in maize are revealed through the characterization of DCT2 mutants. The Plant Cell 28, 466–484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Westhoff P, Gowik U.. 2004. Evolution of C4 phosphoenolpyruvate carboxylase. Genes and proteins: a case study with the genus Flaveria. Annals of Botany 93, 13–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Williams BP, Johnston IG, Covshoff S, Hibberd JM.. 2013. Phenotypic landscape inference reveals multiple evolutionary paths to C4 photosynthesis. eLife 2, e00961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Woo KC, Anderson JM, Boardman NK, Downton WJS, Osmond CB, Thorne SW.. 1970. Deficient photosystem II in agranal bundle sheath chloroplasts of C4 plants. Proceedings of the National Academy of Sciences, USA 67, 18–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zachos JC, Dickens GR, Zeebe RE.. 2008. An early Cenozoic perspective on greenhouse warming and carbon-cycle dynamics. Nature 451, 279–283. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data supporting the findings of this study are available within the paper and within its supplementary data published online.