TY - JOUR
AU1 - Majeran, Wojciech
AU2 - Friso, Giulia
AU3 - Ponnala, Lalit
AU4 - Connolly, Brian
AU5 - Huang, Mingshu
AU6 - Reidel, Edwin
AU7 - Zhang, Cankui
AU8 - Asakura, Yukari
AU9 - Bhuiyan, Nazmul H.
AU1 - Sun, Qi
AU1 - Turgeon, Robert
AU1 - van Wijk, Klaas J.
AB - Abstract C4 grasses, such as maize (Zea mays), have high photosynthetic efficiency through combined biochemical and structural adaptations. C4 photosynthesis is established along the developmental axis of the leaf blade, leading from an undifferentiated leaf base just above the ligule into highly specialized mesophyll cells (MCs) and bundle sheath cells (BSCs) at the tip. To resolve the kinetics of maize leaf development and C4 differentiation and to obtain a systems-level understanding of maize leaf formation, the accumulation profiles of proteomes of the leaf and the isolated BSCs with their vascular bundle along the developmental gradient were determined using large-scale mass spectrometry. This was complemented by extensive qualitative and quantitative microscopy analysis of structural features (e.g., Kranz anatomy, plasmodesmata, cell wall, and organelles). More than 4300 proteins were identified and functionally annotated. Developmental protein accumulation profiles and hierarchical cluster analysis then determined the kinetics of organelle biogenesis, formation of cellular structures, metabolism, and coexpression patterns. Two main expression clusters were observed, each divided in subclusters, suggesting that a limited number of developmental regulatory networks organize concerted protein accumulation along the leaf gradient. The coexpression with BSC and MC markers provided strong candidates for further analysis of C4 specialization, in particular transporters and biogenesis factors. Based on the integrated information, we describe five developmental transitions that provide a conceptual and practical template for further analysis. An online protein expression viewer is provided through the Plant Proteome Database. INTRODUCTION Plant species can be classified as C3 or C4 based on the primary product of carbon fixation in photosynthesis. In most C4 plants, the photosynthetic apparatus is partitioned over two cell types, the mesophyll cells (MCs) and bundle sheath cells (BSCs) surrounding the vascular bundle and organized in the Kranz anatomy. Active carbon transport (in the form of organic acids) between the MCs to BSCs and specific expression of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) in the BSCs allow Rubisco, the carboxylating enzyme in the Calvin-Benson cycle, to operate in a high CO2 cellular concentration in the BSCs. This higher intracellular CO2 concentration suppresses the oxygenation reaction by Rubisco (photorespiration), resulting in increased photosynthetic yield, more efficient use of water and nitrogen, and increased biomass production. Therefore, C4 species such as maize (Zea mays), switchgrass (Panicum virgatum), and sugarcane (Saccharum officinarum) are preferred for biofuels (Carpita and McCann, 2008), whereas genetic engineering of C4 features into C3 plants such as rice (Oryza sativa) has the potential to increase crop productivity (Hibberd et al., 2008; Taniguchi et al., 2008; Weber and von Caemmerer, 2010). C4 differentiation in maize and other grasses occurs along a developmental gradient with proplastids at the base of the leaf blade and differentiated C4 MC and BSC chloroplasts at the tip (Nelson and Langdale, 1992; Evert et al., 1996; Sheen, 1999; Edwards et al., 2001b; Majeran and van Wijk, 2009). Several features are required for a functioning C4 maize leaf: (1) formation of BSCs with thick cell walls that have high resistance against CO2 leakage, (2) BSCs containing large photosynthetic chloroplasts specialized in carbon fixation and lacking photosystem II (PSII) activity, (3) a high density of plasmodesmata (PD) as well as (C4-specific) metabolite transporters to mediate high metabolic fluxes, (4) short interveinal distances to facilitate efficient metabolic exchange between BSCs and MCs, and (5) differential BSC and MC expression of carbon fixation, the C4 malate-pyruvate shuttle, and other metabolic activities. Maize is a NADP+ malic enzyme-type C4 grass and has been used as a model system for C4 differentiation (Nelson and Langdale, 1992; Sheen, 1999; Edwards et al., 2001b; Majeran and van Wijk, 2009). The first draft of the maize genome sequence was recently published (Schnable et al., 2009) and now allows extensive and detailed molecular analysis of maize with attention for neofunctionalization of members of gene families (see Discussion). The use of new genomics and proteomics tools has resulted in several insights into the fully differentiated state of the maize leaf (Majeran et al., 2005, 2008; Bräutigam et al., 2008; Covshoff et al., 2008; Friso et al., 2010), most of them related to chloroplasts at the leaf tip. However, the networks that regulate leaf development, C4 structures, metabolic functions, and cell type–specific differentiation, as well as molecular aspects of sink-source relationships in C4 leaves, are still poorly understood. Furthermore, knowledge of the development of proplastids into specialized BSC and MC chloroplasts is scattered and lacks a systems-level understanding that incorporates the diverse metabolic functions of plastids. A systematic linkage between the formation of anatomical and ultrastructural features and detailed molecular functions of developing maize leaves is also lacking. To answer these long-standing questions and obtain a better systems-level understanding of the formation of the C4 maize leaf, we identified and quantified the dynamics of the leaf proteome as well as the proteome of the BSCs and vascular bundle along the developmental gradient using large-scale proteome analyses. This was complemented by detailed microscopy image analysis of anatomical and structural features. Proteins were annotated for function and subcellular localization to allow for interpretation of the underlying biology. Developmental protein accumulation profiles and hierarchical cluster analysis then allowed determination of the kinetics of organelle biogenesis, formation of cellular structures, and coexpression of metabolic pathways. Two major expression clusters were observed, each divided into a number of subclusters, suggesting that a limited number of developmental regulatory networks organize a concerted accumulation of the quantified proteins along the leaf gradient. The complementary light and electron microscopy analyses of cross sections of similar leaves facilitated determination of the formation of key structural features, including the Kranz anatomy, PD density and size, and formation and differentiation of subcellular structures and organelles. Particularly interesting was the observation that BSC- and MC-specific protein expression patterns were observed relatively late (several centimeters from the base of the leaf blade) around the sink-source transition point. The microscopy analysis also showed differential kinetics of thylakoid formation and chloroplast expansion between BSCs and MCs. Coexpression of proteins with BSC and MC markers provided strong candidates for further analysis of C4 specialization, in particular for chloroplast and mitochondrial transporters, as well as biogenesis factors. This study thus provides a systems-level analysis of C4 leaf development and differentiation and a molecular template and resource for further C4 leaf studies. Protein data with annotations, as well as an interactive interface to visualize protein expression profiles, are available via the Plant Proteome Database (PPDB; http://ppdb.tc.cornell.edu/). RESULTS Selection of the Optimal Leaf The 3rd leaf of 9-d-old maize seedlings (see Supplemental Figure 1A online) grown under controlled conditions possessed all necessary features for our objectives and included (1) a complete developmental gradient as determined by microscopy and pilot proteome experiments (see below), (2) a clear sink-source transition zone (at ~4 cm from the ligule) as determined by 14CO2 labeling (see Supplemental Figure 1B online), and (3) consistent leaf length (~15 cm), providing sufficient material and spatial resolution for proteomics. We also note that the sheaths of the selected 3rd leaves (below the ligule) were only a few millimeters long. The sheath lengthened in older leaves. Our analysis involves only the leaf blade starting just above the ligule. In the remainder of the article, we will use the term leaf base as the first centimeter (leaf section) or first 1.5 cm (vascular bundle with BSCs) just above the ligule. Qualitative and Quantitative Light and Transmission Electron Microscopy Analysis of the Developing Leaf To determine the developmental state and degree of (sub)cellular (C4) differentiation, extensive light microscopy (LM) and transmission electron microscopy (TEM) analyses of cross sections perpendicular to the leaf axis were performed (Figure 1). Sampling was done using 1-cm sections along the leaf gradient starting right above the ligule; throughout the article, we will refer to these sections always counting from the base (0 to 1 cm) to the tip of the leaf blade (15 cm). Qualitative and quantitative observations are summarized in Figures 1A to 1P and Supplemental Figures 1A, 1B, and 2A to 2F online. Figures 1A to 1L show LM and TEM images of the four most informative zones/sections (0 to 1 cm, 4 to 5 cm, 8 to 9 cm, and 13 to 14 cm); four additional leaf zones (2 to 3 cm, 5 to 6 cm, 6 to 7 cm, and 7 to 8 cm) were analyzed (see Supplemental Figure 2 online) and were included in the quantitative analysis of cell elongation and cell wall thickness (Figure 1M), plastid size and shape (Figure 1N), and thylakoid membranes and granal stacks (see Supplemental Figure 2F online). Figure 1. Open in new tabDownload slide Structural Analysis of the Maize Leaf along the Developmental Gradient. Transverse sections of maize leaves viewed by LM ([A] to [D]) and TEM ([E] to [L], [O], and [P]) at different positions along the developmental gradient. At the leaf base ([A], [E], and [I]), the small veins are immature, each consisting of a cluster of undifferentiated vascular cells (arrows in [A]) surrounded by a ring of undifferentiated BSCs (marked as b) and MCs (marked as m). In (I), many plasmodesmata are seen in transverse section along the cell wall (arrow), and small starch grains are visible in some proplastids (marked as p; asterisk). At 4 to 5 cm above the leaf base ([B], [F], and [J]), leaf tissue has expanded greatly, with prominent vacuoles and airspaces between cells. Arrows in (B) identify veins. In (F), peroxisomes or plastids (arrow) and mitochondria (asterisks) are visible in vascular parenchyma and companion cells, respectively; sieve elements are marked with an s. (J) also shows that thylakoids in BSCs are singular, but those in the MCs have formed small granal stacks (arrow). At 8 to 9 cm above the leaf base ([C], [G], and [K]), the chloroplasts of both MCs and BSCs have enlarged (arrows in [C]). In (G), sieve elements and xylem tracheary elements are indicated with s and x, respectively. Plasmodesmata (indicated by arrows) crossing the cell wall (cw) are visible in (K). Near the leaf tip ([D], [H], and [L]), the MC chloroplasts have continued to enlarge (arrow in [D]). Chloroplasts in the BSCs are distinctly narrowed and curved ([H] and [L]). At the end of the day ([O] and [P]), starch grains (asterisks) are visible in MC chloroplasts, 2 to 3 cm from the leaf base (O) and in BSC chloroplasts near the leaf tip (P). (M) shows morphometric analyses of cell wall thickness (open squares) and plasmodesmata length (closed circles) of the interface between BSC and MC, whereas epidermal cell length is shown by filled triangles. The vertical dotted line in (M) and (N) indicates the 4-cm point, as measured from the ligule. Chloroplast length (squares) and width (triangle) for BSC and MC are plotted in (N). Bars = 25 μm in (A) to (D), 10 μm in (E), 5 μm in (F) to (H), 1.0 μm in (I) to (L), 3 μm in (O), and 6 μm in (P). Figure 1. Open in new tabDownload slide Structural Analysis of the Maize Leaf along the Developmental Gradient. Transverse sections of maize leaves viewed by LM ([A] to [D]) and TEM ([E] to [L], [O], and [P]) at different positions along the developmental gradient. At the leaf base ([A], [E], and [I]), the small veins are immature, each consisting of a cluster of undifferentiated vascular cells (arrows in [A]) surrounded by a ring of undifferentiated BSCs (marked as b) and MCs (marked as m). In (I), many plasmodesmata are seen in transverse section along the cell wall (arrow), and small starch grains are visible in some proplastids (marked as p; asterisk). At 4 to 5 cm above the leaf base ([B], [F], and [J]), leaf tissue has expanded greatly, with prominent vacuoles and airspaces between cells. Arrows in (B) identify veins. In (F), peroxisomes or plastids (arrow) and mitochondria (asterisks) are visible in vascular parenchyma and companion cells, respectively; sieve elements are marked with an s. (J) also shows that thylakoids in BSCs are singular, but those in the MCs have formed small granal stacks (arrow). At 8 to 9 cm above the leaf base ([C], [G], and [K]), the chloroplasts of both MCs and BSCs have enlarged (arrows in [C]). In (G), sieve elements and xylem tracheary elements are indicated with s and x, respectively. Plasmodesmata (indicated by arrows) crossing the cell wall (cw) are visible in (K). Near the leaf tip ([D], [H], and [L]), the MC chloroplasts have continued to enlarge (arrow in [D]). Chloroplasts in the BSCs are distinctly narrowed and curved ([H] and [L]). At the end of the day ([O] and [P]), starch grains (asterisks) are visible in MC chloroplasts, 2 to 3 cm from the leaf base (O) and in BSC chloroplasts near the leaf tip (P). (M) shows morphometric analyses of cell wall thickness (open squares) and plasmodesmata length (closed circles) of the interface between BSC and MC, whereas epidermal cell length is shown by filled triangles. The vertical dotted line in (M) and (N) indicates the 4-cm point, as measured from the ligule. Chloroplast length (squares) and width (triangle) for BSC and MC are plotted in (N). Bars = 25 μm in (A) to (D), 10 μm in (E), 5 μm in (F) to (H), 1.0 μm in (I) to (L), 3 μm in (O), and 6 μm in (P). At the leaf base, the vascular tissue was immature (with the exception of the large veins,~9 per leaf; see Supplemental Figure 1B online), but the cellular architectures of the intermediate and small veins could be recognized, with clusters of undifferentiated vascular cells surrounded by future BSCs (Figures 1A and 1E). Vascular differentiation occurred rapidly, such that formation of metaxylem and metaphloem was complete within 3 cm, including enucleate sieve elements (Figures 1B and 1F). Little evidence for cell division was observed, but cell elongation was dramatic in the first 4 cm and was essentially complete at 6 cm (Figure 1M). PDs were already established at the base, and PD numbers (across the BSC and MC interface) did not increase toward the tip. However, PD length and cell wall thickness at the BSC-MC interface increased from ~90 to 200 nm in the first 8 cm, without further change toward the tip (Figure 1M); this indicates that cell wall deposition was complete at 8 cm, following completion of cell elongation. C4 plastid development and differentiation was determined by quantification of plastid dimensions, thylakoid formation, granal cross section, starch, and plastoglobules. At the base, only proplastids (~2.5-μm diameter) with one to two small prothylakoid membranes were observed, without visible differences between cell types (Figures 1E and 1I). No etioplasts were observed, as evidenced by the lack of prolamellar bodies. At 2 to 3 cm, the prothylakoid system had expanded and the first grana membrane stacks were visible, but still without differences between MCs and BSCs. At 4 to 5 cm, BSC and MC plastids now showed distinct differences in thylakoid membrane organization, with up to six stacks in MC plastids and no further stacking in BSCs (Figures 1I and 1J; see Supplemental Figure 2A online); thus, in this section, a signal transduction pathway(s) must be operating to drive differential BSC and MC plastid biogenesis. Since grana stacks are known to be formed by accumulation of PSII complexes (Chow et al., 2005), this suggests that PSII or light-harvesting complex II (LHCII) complexes began to accumulate in MC thylakoids around 3 cm. In both plastid types, vesicle-like structures were visible along the inner envelope membranes; these structures have previously been assigned as peripheral reticulum (Wise, 2006) (see Supplemental Figure 2D online). The function of these structures is not clear; they could indicate fatty acid or lipid movement or serve to increase the surface area of the inner envelope so as to accommodate a higher rate of metabolic exchange (for discussion, see Leegood, 2008). BSC chloroplasts showed continuous increase in cross section between 2 and 8 cm, mostly due to lengthening of the chloroplasts, after which they did not change much further (Figure 1N). Consistent with this, the number of thylakoids per BSC chloroplast increased 3-fold in the first 8 cm from the base, after which it was constant (see Supplemental Figure 2F online). By contrast, MC chloroplasts showed a fairly constant expansion in width and length up to the tip of the leaf (Figure 1N). Surprisingly, the grana stack cross section per MC chloroplast was constant up to 6 cm from the base, followed by a rapid 4-fold increase up to the leaf tip (see Supplemental Figure 2F online). Finally, BSC chloroplasts clearly had grana stacks, but they remained short and were seldom more than three membranes thick (Figures 1K and 1L) suggesting low accumulation levels of PSII and LHCII. By contrast, MC chloroplast grana stacks near the tip were longer and between ~10 and 20 stacks thick, reflecting high levels of PSII (Figure 1L). Thylakoid-associated lipoprotein particles, named plastoglobules, involved in quinone and tocopherol metabolism as well as several less defined functions (Bréhélin et al., 2007), appeared in low numbers starting from ~2 cm. In BSC chloroplasts, clusters of five to seven plastoglobules were often observed at the poles of the plastids (see Supplemental Figure 2E online); the functional significance of this remains to be determined. At the leaf tip, the ratio of the cross-sectional areas of MC:BSC chloroplasts was 1.61 (based on 27 BSC chloroplast measurements with average cross section of 265 μm2, sd = 7.3; and 74 MC chloroplast measurements with average cross section of 165 μm2, sd = 1.5) as determined from the LM images. The presence and distribution of starch particles depended on the time of harvest of the leaves. Leaves harvested within 2 h of the onset of the light period showed low but constant levels of starch particles in BSC chloroplasts between 6 cm and the tip, but starch particles were essentially absent closer to the base. By contrast, leaves harvested 2 h before the end of the light period showed basal proplastids with high amounts of starch particles in both future BSCs and future MCs. Moreover, between 2 and 5 cm, both BSC and MC chloroplasts contained clearly observable starch particles, with highest levels in MC chloroplasts (Figure 1O), whereas at the leaf tip, starch accumulation was massive in BSC chloroplasts but low in MC chloroplasts (Figure 1P). Thus, the distribution of transient starch across MC and BSC chloroplasts and plastids was determined by the developmental and differentiation state of the chloroplasts and the time of day. Mitochondria (on average 0.3-μm diameter) accumulated at high levels in vascular parenchyma cells (Figure 1F; see Supplemental Figure 2B online), suggesting a large need for respiratory or other mitochondrial activity within the vasculature. Calculation of the ratio of the mitochondrial cross section in the vascular cells to the cross section in BSCs showed that this ratio narrowly peaked between 4 and 5 cm and was otherwise constant. The ratio between the cross sections of plastids and mitochondria in BSCs and in MCs changed dramatically along the leaf gradient with a ratio of ~7.5 for both BSCs and MCs at the base, increasing to 35 and 120 in the leaf tip for BSCs and MCs, respectively. This suggested that the relative importance of mitochondria compared with chloroplasts decreased 5-fold (BSCs) to 16-fold (MCs) from base to tip, also indicating the particular importance of mitochondria in BSCs. Nonphotosynthetic plastids (S and P type) have previously been shown to be present within the vascular bundles of maize leaves (Williams, 1974; Walsh and Evert, 1975), but nothing is known about their function or protein content. These plastids are small (comparable in size to mitochondria) and have no internal membranes. However, some plastids contain crystalline structures of unknown composition (Williams, 1974), appearing quite similar to catalase crystals in peroxisomes, whereas other plastids appear to contain proteinaceous inclusions or plastoglobuli-like structures (Williams, 1974; Walsh and Evert, 1975). We also observed organelles with crystalline structures in parenchyma cells starting several centimeters from the base (Figure 1F; see Supplemental Figure 2B online); they could be such nonphotosynthetic plastids as described above, or they could be peroxisomes containing catalase crystals, indicative of the need for detoxification of H2O2. Later in the article, we discuss the various metabolic functions of the organelles based on the comparative proteome analysis. Dynamics of the Leaf and Vascular Proteome along the Maize Leaf Developmental Gradient For analysis of the cellular proteome along the leaf developmental gradient, 1-cm sections were sampled from a single leaf for each biological replicate (Figure 2A). The leaf sampling was focused on the six most dynamic zones (0 to 1, 2 to 3, 3 to 4, 4 to 5, 8 to 9, and 13 to 14 cm). In addition, we sampled the leaf vascular system, including the BSCs, from here on named BS strands, collecting four sections as follows: 0 to 1.5, 2.5 to 4, 4 to 5, and 12 to 13 cm from the ligule. In a pilot study, we also analyzed the soluble proteome of BS strands from the base, the greening zone, and tip of the leaf (see Supplemental Figure 3A). These data were used only to support the evaluation of gene models and add to the coverage of the maize proteome (see Supplemental Data Sets 1A and 1B online; and see PDDB). A list of full names and abbreviations of proteins specifically mentioned in the text or figures can be found in Supplemental Data Set 1C online. Figure 2. Open in new tabDownload slide Proteome Analysis of Leaf and BS Strands along the Leaf Developmental Gradient. (A) Cellular protein accumulation patterns along the leaf (left gel, with a representative leaf showing positions of the samples above) and BS strands (right gel) as evidenced by one-dimensional gels and staining with Coomassie blue. A few marker enzymes are indicated. RBCL, Rubisco large subunit; RBCS, Rubisco small subunit; PAL, phenylalanine ammonia lyase; TKL, transketolase. These markers were identified by MS analysis. (B) Protein investment in various cellular functions per leaf section, calculated from NadjSPC per function, as percentage of total NadjSPC. The percentage of protein mass from mitochondrial (mito. %) and plastid proteins (plastid %) in each section are indicated (explanation for the color scheme, see [C]). (C) Dendrogram of the hierachical cluster analysis of protein expression (calculated from NadjSPC) along the developmental gradient of the leaf and BS strand. The two main clusters (I and II) and their assigned subclusters are indicated. The average profile for the accessions in each cluster is indicated, with closed and open circles for leaf and BS strands, respectively. Protein investments in various cellular functions for all accessions in each cluster are indicated, and the percentage of plastid proteins among the identified proteins (p/t %) in each cluster is indicated. N-S-AA refers to nitrogen and sulfur assimilation and amino acid metabolism (Mapman bins 12, 13, and 14). Construction refers to bins 10, 11, 16, 19, 23, and 25 (see also [D]). Catabolism refers to glycolysis, gluconeogenesis/glyoxylate cycle, OPPP, TCA cycle, mitochondrial electron transport, and oxidative phosphorylation (bins 4 to 9), and DNA and RNA (bins 27 and 28). Regulation and signaling includes hormone metabolism, stress response, redox regulation, signaling, and development (bins 17, 20, 21, 30, 31, and 33). Carbon metabolism includes dark reactions of photosynthesis and major and minor carbohydrate metabolism (bins 1 to 3). (D) Distribution across the clusters of functional groups of proteins involved in construction. These functional groups are metabolism of lipid and fatty acid (bin 11), tetrapyrroles (bin 19), nucleotides (bin 23), as well as C1 metabolism (bin 25), secondary metabolism (bin 16), and cell wall biogenesis (bin 10). Figure 2. Open in new tabDownload slide Proteome Analysis of Leaf and BS Strands along the Leaf Developmental Gradient. (A) Cellular protein accumulation patterns along the leaf (left gel, with a representative leaf showing positions of the samples above) and BS strands (right gel) as evidenced by one-dimensional gels and staining with Coomassie blue. A few marker enzymes are indicated. RBCL, Rubisco large subunit; RBCS, Rubisco small subunit; PAL, phenylalanine ammonia lyase; TKL, transketolase. These markers were identified by MS analysis. (B) Protein investment in various cellular functions per leaf section, calculated from NadjSPC per function, as percentage of total NadjSPC. The percentage of protein mass from mitochondrial (mito. %) and plastid proteins (plastid %) in each section are indicated (explanation for the color scheme, see [C]). (C) Dendrogram of the hierachical cluster analysis of protein expression (calculated from NadjSPC) along the developmental gradient of the leaf and BS strand. The two main clusters (I and II) and their assigned subclusters are indicated. The average profile for the accessions in each cluster is indicated, with closed and open circles for leaf and BS strands, respectively. Protein investments in various cellular functions for all accessions in each cluster are indicated, and the percentage of plastid proteins among the identified proteins (p/t %) in each cluster is indicated. N-S-AA refers to nitrogen and sulfur assimilation and amino acid metabolism (Mapman bins 12, 13, and 14). Construction refers to bins 10, 11, 16, 19, 23, and 25 (see also [D]). Catabolism refers to glycolysis, gluconeogenesis/glyoxylate cycle, OPPP, TCA cycle, mitochondrial electron transport, and oxidative phosphorylation (bins 4 to 9), and DNA and RNA (bins 27 and 28). Regulation and signaling includes hormone metabolism, stress response, redox regulation, signaling, and development (bins 17, 20, 21, 30, 31, and 33). Carbon metabolism includes dark reactions of photosynthesis and major and minor carbohydrate metabolism (bins 1 to 3). (D) Distribution across the clusters of functional groups of proteins involved in construction. These functional groups are metabolism of lipid and fatty acid (bin 11), tetrapyrroles (bin 19), nucleotides (bin 23), as well as C1 metabolism (bin 25), secondary metabolism (bin 16), and cell wall biogenesis (bin 10). Total cellular proteins were quantitatively extracted, separated by SDS-PAGE, and visualized by Coomassie blue staining (Figure 2A). In particular for the BS strands, several abundant proteins could be recognized in the gel after mass spectrometry analysis and included members of the C4 shuttle (e.g., NADPH-malic enzymes [NADP-MEs]), methionine synthase (MetS) in the S-adenosylmethionine cycle, sucrose synthase (SuSy) in sucrose degradation, and Rubisco (Figure 2A). They display the general pattern of sink-source transition, induction of the Calvin-Benson cycle and of the C4 carbon shuttle. For qualitative and quantitative proteome analysis, the gel lanes of the selected samples were processed (by tryptic digestions) for data-dependent tandem mass spectrometry (MS/MS) analysis using a nanoLC-ESI-LTQ-Orbitrap mass spectrometer operating at its highest resolution for precursor analysis (100,000). About 5.2 million MS/MS spectra were acquired, in addition to 3.4 million spectra from the soluble BS strand proteome. Spectral data were searched against the maize genome sequence (Schnable et al., 2009), supplemented with organellar genomes. The search results were further processed to (1) reduce false positive identification, (2) avoid overidentification of members of protein families, (3) select the best gene model for each gene, and (4) group proteins for quantification that shared more than ~80% of their matched MS/MS spectra, using the workflow developed by Friso et al. (2010) (Figure 3). We calculated the relative amount (mass) of each identified protein within each biological replicate based on the normalized number of adjusted matched MS/MS spectra (NadjSPC). The adjustment refers to a proportional distribution of shared SPC between closely related homologs (Friso et al., 2010). This workflow resulted in quantification of 2637 proteins and 667 protein groups (Figure 3). Proteins were annotated for subcellular location and function using the MapMan classification system (Thimm et al., 2004), similar to our previous maize studies (Majeran et al., 2005, 2008; Sun et al., 2009; Friso et al., 2010). We annotated 874 proteins (and protein groups) to plastids, 166 to mitochondria, 26 to peroxisomes, 24 to vacuoles, and 35 to the plasma membrane. The BS strands were enriched for vascular proteins as evidenced by well-known vascular markers, such as the sucrose transporter SUT1 (Carpaneto et al., 2005; Slewinski et al., 2009), homologs of vascular nicotianamine synthase (likely in companion cells) involved in iron homeostasis (Inoue et al., 2003), a xylem-specific Ser protease, and a vascular lectin protein (see Supplemental Figure 3B online). The 4430 identified proteins with their annotations are provided in Supplemental Data Set 1A online, and the quantified 2637 proteins and 667 protein groups are provided in Supplemental Data Set 1B online. For interactive information and an interactive protein expression viewer, see the PPDB at http://ppdb.tc.cornell.edu. Figure 3. Open in new tabDownload slide Experimental and Bioinformatics Workflow of the Proteome Analysis. Figure 3. Open in new tabDownload slide Experimental and Bioinformatics Workflow of the Proteome Analysis. Reproducibility and Variation in Proteomics Data Whereas there has now been ample demonstration that label-free spectral counting is a viable method for protein quantification (Liu et al., 2004; Old et al., 2005; Zybailov et al., 2005; Lu et al., 2007; Majeran et al., 2008; Friso et al., 2010), it was important to test the variation between biological replicates for both the leaf gradient set and the BS strand set. Such variation could result from the selection and sectioning of the leaves, the procedure for isolation of BS strands, and the complete proteomics workflow from SDS-PAGE, in-gel digests, and MS/MS itself. We performed two tests to determine variation prior to the grouping by similarity matrix of closely related proteins (Figure 3). The grouping will decrease variation; therefore, our test measured the worst-case scenario. In the first test, we calculated for each protein the Pearson's linear correlation for the NadjSPC values across replicates; a few individual examples for both soluble and integral membrane proteins are shown in Figures 4A and 4B. Figure 4A shows the accumulation profiles along the leaf gradient for the two individual biological replicates (based on NadjSPC) for MetS and SuSy2-2 (both decrease from base to tip) and for phosphoenolpyruvatecarboxylase (PEPC) and the hydrophobic D2 integral membrane protein of the PSII complex (both increase from base to tip). Figure 4B shows the same data but now as cross-correlation plot and correlation coefficients for each protein. The significance of each correlation coefficient was assessed by calculating its P value using a Student's t distribution (see Supplemental Data Set 1A online). Based on our previous experience with spectral counting (using similar instrumentation and workflow) (Majeran et al., 2008; Kim et al., 2009; Zybailov et al., 2009b; Friso et al., 2010) and evaluation of our current data sets, as well as published literature (Liu et al., 2004; Old et al., 2005; Zybailov et al., 2005; Lu et al., 2007), it was established that quantification is generally more reliable for proteins with a higher number of matched spectra. Therefore, we calculated the median correlation coefficient for intervals of proteins across the abundance spectrum (as measured by the total adjSPC across both biological replicates) (Table 1). We used those proteins (2029 and 2166 proteins for leaf and BS strands, respectively) that had at least two non-zero expression values among all the sections within each biological replicate. The reproducibility between the two biological replicates of accumulation profiles of proteins with the highest number of adjSPC showed a very high correlation (0.951 for leaf and 0.975 for BS strand). The reproducibility decreased to still significant levels with decreasing protein abundance (Table 1). Figure 4. Open in new tabDownload slide Reproducibility between Biological Replicates. Examples of reproducibility for both soluble and integral membrane proteins between biological replicates of the comparative proteome analysis. (A) Accumulation profiles along the leaf gradient for the two individual biological replicates (based on NadjSPC) for MetS (GRMZM2G149751_P01), SuSy (SuSy2-2; GRMZM2G152908_P01), PEPC (GRMZM2G083841_P01), and the D2 integral membrane protein of the PSII complex (NP_043009). Open and closed symbols are used for replicates 1 and 2, respectively. (B) Cross-correlation plot for the four proteins shown in (A). The inset shows the total number of AdjSPC in the leaf gradient samples and the correlation coefficients for each protein. Figure 4. Open in new tabDownload slide Reproducibility between Biological Replicates. Examples of reproducibility for both soluble and integral membrane proteins between biological replicates of the comparative proteome analysis. (A) Accumulation profiles along the leaf gradient for the two individual biological replicates (based on NadjSPC) for MetS (GRMZM2G149751_P01), SuSy (SuSy2-2; GRMZM2G152908_P01), PEPC (GRMZM2G083841_P01), and the D2 integral membrane protein of the PSII complex (NP_043009). Open and closed symbols are used for replicates 1 and 2, respectively. (B) Cross-correlation plot for the four proteins shown in (A). The inset shows the total number of AdjSPC in the leaf gradient samples and the correlation coefficients for each protein. Table 1. Pearson's Linear Correlation for the NadjSPC Values across the Biological Replicates Leaf BS Strand Sum (adjSPC) n a Median Correlation Coefficient P Value n a Median Correlation Coefficient P Value >1000 20 0.951 0.0035 22 0.975 0.0247 500 to 1000 58 0.898 0.0150 45 0.913 0.0867 100 to 500 467 0.727 0.1019 436 0.874 0.1256 50 to 100 343 0.599 0.2094 351 0.815 0.1854 20 to 50 513 0.451 0.3690 592 0.766 0.2335 10 to 20 308 0.415 0.4139 364 0.632 0.3682 5 to 10 210 0.492 0.3215 214 0.479 0.5211 <5 110 0.521 0.2887 142 0.497 0.5026 Leaf BS Strand Sum (adjSPC) n a Median Correlation Coefficient P Value n a Median Correlation Coefficient P Value >1000 20 0.951 0.0035 22 0.975 0.0247 500 to 1000 58 0.898 0.0150 45 0.913 0.0867 100 to 500 467 0.727 0.1019 436 0.874 0.1256 50 to 100 343 0.599 0.2094 351 0.815 0.1854 20 to 50 513 0.451 0.3690 592 0.766 0.2335 10 to 20 308 0.415 0.4139 364 0.632 0.3682 5 to 10 210 0.492 0.3215 214 0.479 0.5211 <5 110 0.521 0.2887 142 0.497 0.5026 The significance of each correlation coefficient was assessed by calculating its P value using a Student's t distribution. a The number of proteins in this abundance interval. Open in new tab Table 1. Pearson's Linear Correlation for the NadjSPC Values across the Biological Replicates Leaf BS Strand Sum (adjSPC) n a Median Correlation Coefficient P Value n a Median Correlation Coefficient P Value >1000 20 0.951 0.0035 22 0.975 0.0247 500 to 1000 58 0.898 0.0150 45 0.913 0.0867 100 to 500 467 0.727 0.1019 436 0.874 0.1256 50 to 100 343 0.599 0.2094 351 0.815 0.1854 20 to 50 513 0.451 0.3690 592 0.766 0.2335 10 to 20 308 0.415 0.4139 364 0.632 0.3682 5 to 10 210 0.492 0.3215 214 0.479 0.5211 <5 110 0.521 0.2887 142 0.497 0.5026 Leaf BS Strand Sum (adjSPC) n a Median Correlation Coefficient P Value n a Median Correlation Coefficient P Value >1000 20 0.951 0.0035 22 0.975 0.0247 500 to 1000 58 0.898 0.0150 45 0.913 0.0867 100 to 500 467 0.727 0.1019 436 0.874 0.1256 50 to 100 343 0.599 0.2094 351 0.815 0.1854 20 to 50 513 0.451 0.3690 592 0.766 0.2335 10 to 20 308 0.415 0.4139 364 0.632 0.3682 5 to 10 210 0.492 0.3215 214 0.479 0.5211 <5 110 0.521 0.2887 142 0.497 0.5026 The significance of each correlation coefficient was assessed by calculating its P value using a Student's t distribution. a The number of proteins in this abundance interval. Open in new tab In the second test, we calculated correlations between the two biological replicates across the proteins identified per individual leaf section or BS strand section (Table 2). We found high correlations (0.797 to 0.941 in leaf; 0.886 to 0.972 in BS strands) across replicates in each section (Table 2), providing further support for the reproducibility of our experimental observations. We note that, for both sample types, the highest correlation was found at the tip and the lowest in the most dynamic developmental zone, between 2 and 4 cm from base. This provides additional support for the robustness of our analysis. Table 2. Pearson's Linear Correlation Coefficient between the Biological Replicates of NadjSPC per Tissue Section Leaf BS Strand Section n Correlation Coefficient Section n Correlation Coefficient 0 to 1cm 1714 0.878 0 to 1.5 cm 2065 0.921 2 to 3 cm 1751 0.797 2.5 to 4 cm 2094 0.886 3 to 4 cm 1703 0.803 4 to 5 cm 2046 0.955 4 to 5 cm 1749 0.877 12 to 13 cm 1448 0.972 8 to 9 cm 1418 0.941 13 to 14 cm 1163 0.867 Leaf BS Strand Section n Correlation Coefficient Section n Correlation Coefficient 0 to 1cm 1714 0.878 0 to 1.5 cm 2065 0.921 2 to 3 cm 1751 0.797 2.5 to 4 cm 2094 0.886 3 to 4 cm 1703 0.803 4 to 5 cm 2046 0.955 4 to 5 cm 1749 0.877 12 to 13 cm 1448 0.972 8 to 9 cm 1418 0.941 13 to 14 cm 1163 0.867 The P value for every correlation is 0, indicating strong significance. n indicates the number of proteins that have at least one non-zero value of NadjSPC among the two replicates for the relevant section. Open in new tab Table 2. Pearson's Linear Correlation Coefficient between the Biological Replicates of NadjSPC per Tissue Section Leaf BS Strand Section n Correlation Coefficient Section n Correlation Coefficient 0 to 1cm 1714 0.878 0 to 1.5 cm 2065 0.921 2 to 3 cm 1751 0.797 2.5 to 4 cm 2094 0.886 3 to 4 cm 1703 0.803 4 to 5 cm 2046 0.955 4 to 5 cm 1749 0.877 12 to 13 cm 1448 0.972 8 to 9 cm 1418 0.941 13 to 14 cm 1163 0.867 Leaf BS Strand Section n Correlation Coefficient Section n Correlation Coefficient 0 to 1cm 1714 0.878 0 to 1.5 cm 2065 0.921 2 to 3 cm 1751 0.797 2.5 to 4 cm 2094 0.886 3 to 4 cm 1703 0.803 4 to 5 cm 2046 0.955 4 to 5 cm 1749 0.877 12 to 13 cm 1448 0.972 8 to 9 cm 1418 0.941 13 to 14 cm 1163 0.867 The P value for every correlation is 0, indicating strong significance. n indicates the number of proteins that have at least one non-zero value of NadjSPC among the two replicates for the relevant section. Open in new tab In both tests, we found that the BS strand analysis showed consistently higher correlation coefficients than did the leaf analysis, which likely related to reduced complexity and increased specialization within the BS strand compared with total leaf section. The consistent high correlation coefficient between the replicate sections showed that we were able to reproducibly select and process the different developmental sections. Protein Investment along the Leaf and BS Strand Gradient To discover patterns of leaf development and BS strand differentiation, we first determined the protein mass investment per function along the leaf gradient. Proteins were pooled into 11 functions based on physiological relevance (Figure 2B). The most dramatic transitions occurred for (1) extraplastidic protein synthesis and homeostasis, ranging from >30% at the leaf base to <5% at the leaf tip, (2) regulation/signaling, ranging from 15% in the first 3 cm and decreasing to 6% at the tip, (3) the thylakoid electron transport chain, ranging from <2% at the base and increasing to >30% at the leaf tip, and (4) carbon metabolism, ranging from <4% at the base and >20% at the tip. These strong and dominant transitions indicate the massive investment in protein synthesizing machinery in the first 4.5 cm, followed by the pronounced accumulation of the photosynthetic machinery in the chloroplast particularly beyond the first 4.5 cm. Consistent with this, proteins involved in DNA and RNA metabolism continuously decreased from 9% at the base to ~1% at the tip, whereas metabolic pathways (lipids/fatty acids, cell wall components, and secondary metabolites) responsible for synthesis of the major leaf structures (cell wall, membranes, isoprenoids, etc.) showed a broad peak between 2 and 5 cm (Figure 2B). The protein mass investment in mitochondria and plastids changed dramatically from base to tip, with mitochondrial protein mass decreasing from ~6% at the base to 1.6% at the tip and plastid protein mass increasing from 15% at the base to 78% at the tip (Figure 2B). Mitochondrial proteins were consistently overrepresented in the BS strand compared with leaf (Figure 2B), in agreement with the image analysis. Protein Expression Profiles of Functional Pathways along the Leaf and BS Strand Developmental Gradient Cluster analysis of large-scale quantitative transcript or protein data allows identification of groups of genes/proteins that share similar spatial or temporal expression profiles (i.e., they coexpress). Genes or proteins involved in related biological pathways or complexes often accumulate simultaneously, and information on their coexpression is key to understanding biological systems, such as C4 leaf development and cellular differentiation. Conversely, coexpression in many cases implies the presence of functional linkages between gene or proteins, allowing for identification of new components of processes or protein complexes. Cluster analysis has been used extensively for transcripts (Eisen et al., 1998; Belacel et al., 2006; Long et al., 2008) and more recently for proteomics (Dong et al., 2008; Huang et al., 2009; Pontén et al., 2009; Quintana et al., 2009; Olinares et al., 2010). Cluster analysis is based on the notion of unsupervised learning in which data objects within the same cluster are similar to one another and dissimilar to the objects in other clusters. Whereas many clustering algorithms have been developed, hierarchical clustering is the most appropriate for analysis of the proteomics data sets such as in this study, because no prior assumptions about the number of clusters have to be made (Belacel et al., 2006). The hierarchical clustering algorithms also provide a natural way for graphical representation of data, in the form of a dendogram in which each branch forms a group of genes or proteins that share similar behavior. To understand better leaf development and cellular differentiation and to discover spatial and temporal protein accumulation patterns and novel components of the C4 system, we performed a hierachical cluster analysis based on standardized NadjSPC per leaf and BS section, resulting in a dendrogram (Figure 2C). We selected 1043 proteins (out of 3304) that were above the minimal threshold of an average 4 and 2 adjSPC/section for each protein for the leaf and BS strand samples, respectively. This minimal threshold for both sample types ensured meaningful quantifications and clustering, and testing different thresholds (e.g., 5 and 10 adjSPC/section) suggested that the selected combined threshold for leaf and BS strand was optimal. Two main clusters, clusters I and II, were detected, containing 640 and 403 proteins, respectively (Figure 2C). The levels of proteins in cluster I were high in the base but low at the tip, with their highest expression levels either just above the ligule (clusters I-2a and I-2b) or more broadly along the first ~2.5 cm (clusters I-1 and I-3). Proteins in cluster II increased in abundance with progressive leaf development, either peaking between 3.5 and 8.5 cm (clusters II-1 and II-2) or increasing gradually and peaking near the leaf tip (cluster II-3) (Figure 2C). Proteins in clusters I-1 and II-2 were on average underexpressed in BS strands but overexpressed in clusters I-2b and II-1. The dendrogram shows that clusters II-1 and II-3 each had a subset of proteins under- and overrepresented in BS strands. Cluster II-1 was therefore split in two distinct subclusters (II-1a,b), whereas cluster II-3 was split in five subclusters, II-3a-e (see further below). Importantly, proteins in clusters II3c and II3e were enriched in known MC and BSC markers, respectively, facilitating discovery of proteins involved in C4-specifc functions (see further below). Cluster I functionally stands out by containing proteins involved in extraplastidic protein synthesis and homeostasis (36%), regulation/signaling (15%), DNA/RNA-related processes (8%), and catabolic reactions (glycolysis, oxidative pentose phosphate pathway [OPPP], and respiration; 9%), whereas cluster II stands out by the high portion of photosynthetic thylakoid proteins (20%) and proteins involved in plastid protein synthesis and homeostasis (20%) and carbon metabolism (Calvin-Benson cycle, C4 cycle, and starch metabolism; 15%) (Figure 2C). This illustrates that the basal region was focused on the buildup of nonplastid structures and functions, whereas the region beyond 4 cm was primarily focused on formation of chloroplasts and photosynthetic capacity. This clearly showed that developmental programs organizing the buildup of the cell preceded chloroplast development until the sink/source transition (see below). The functional group termed construction was dominated by cell wall biogenesis, lipid metabolism, tetrapyrrole synthesis, and nucleotide metabolism (Figure 2D). Cell wall metabolism fell mostly in clusters I-1 and I-3, with a smaller portion in cluster II-1, whereas tetrapyrrole synthesis was exclusively found in clusters II-1 and II-2, similar to the chloroplast biogenesis machinery (Figure 2D). The other secondary metabolic pathways were more broadly distributed. Finally, both clusters I and II had a sizable set of unassigned proteins and proteins with miscellaneous functions and provide a treasure trove for future discovery. Of the clustered proteins, 406 (39%) were chloroplast localized, most of which belonged to cluster II (72%) and were strongly overrepresented in clusters II-1b (76%), II-2 (83%), and II-3 (89%) but underrepresented in clusters I-1, I-3, and I-2a (16 and 6%, respectively) (Figure 2D). Fifty-eight of the clustered proteins were located in mitochondria, and all but six belonged to cluster I. This illustrated the switch from mitochondrial to plastid dominance from base to tip, reflecting the switch from nonphotosynthetic tissue (dependent on mitochondrial respiration and import of carbohydrates) to photosynthetic tissue. In the remainder of the Results section, we will use the quantitative proteome information and clustering to resolve the kinetics of BSC and MC organellar biogenesis, induction of photosynthesis and the C4 pathway, and transport and secondary metabolism, as well as to discover new C4-associated proteins. Expression of Photosynthesis, the C4 Malate-Pyruvate Shuttle, and Envelope Transporters Figure 5A compares the induction of proteins involved in thylakoid electron transport and photophosphorylation (85 proteins), the C4 malate-pyruvate shuttle (six proteins), and enzymes specific for the Calvin-Benson cycle (eight proteins) or shared with plastid glycolysis or OPPP (eight proteins). Except for three light stress proteins (Lil1.2, Lil3.2, and PsbS; see section below) and one of the ferredoxins (GRMZM2G359127_P01; see section below), all these proteins (above the minimum threshold for hierarchical clustering analysis) belonged to cluster II-3, indicating general coexpression along the leaf gradient. However, normalizing protein accumulation of these functions to their maximal levels showed that the thylakoid-bound photosynthetic apparatus was more rapidly induced than the Calvin-Benson cycle and C4 shuttle and reached maximum relative biomass at 8 to 9 cm from the base of the leaf. By contrast, the Calvin-Benson cycle and C4 shuttle further increased substantially between 8 and 9 cm and the tip (Figure 5A, inset). Importantly, the induction kinetics of the C4 shuttle enzymes and Calvin-Benson cycle were very similar, suggesting tightly regulated accumulation (Figure 5A, inset). Careful evaluation of the proteomics data (see Supplemental Data Set 1B online) showed that enzymes of the C4 shuttle (pyruvate phosphate dikinase [PPDK], PPDK regulator proteins [PPDL-RP], PEPC, and NADP-ME) and those unique to the Calvin-Benson cycle (RBCL, RBCS, Rubisco activase, fructose-1,6-biphosphatase, sedoheptulase-1,7-biphosphatase, and phosphoribulokinase), as well as plastid carbonic β-anhydrase (CA2) were undetected in the first centimeter of the leaf, although they were among the most abundant proteins at the developed 2leaf tip. This shows that biochemically the plastids at the base are indeed completely nonphotosynthetic heterotrophic plastids (and not etioplasts) and that we captured a very early stage of leaf development, as intended. Furthermore, this showed that the C4 malate-pyruvate shuttle was not yet induced at the base of the leaf blade and that C4 differentiation had not yet initiated. Figure 5. Open in new tabDownload slide Quantitative Protein Expression Analysis of the Light and Dark Reactions of Photosynthesis and the C4 Shuttle. (A) Expression of proteins (based on NadjSPC) involved in the light (squares) and dark (triangles) reactions of photosynthesis and the C4 shuttle (asterisks). The inset shows a comparison of the three pathways with NadjSPC values normalized to the maximum value for each pathway to better compare their accumulation kinetics. (B) Accumulation of thylakoid complexes in the developing BS strand distributed over the subclusters. (C) Accumulation pattern of the five thylakoid complexes in the developing BS strand. (D) Accumulation of NDH, PSII, PSI, and cytb6f normalized to proteins of the thylakoid ATP-synthase complex (CF). The inset compares the protein ratio of NDH/PSII along the developmental gradient in the BS strands (open squares) and leaf (closed squares). Figure 5. Open in new tabDownload slide Quantitative Protein Expression Analysis of the Light and Dark Reactions of Photosynthesis and the C4 Shuttle. (A) Expression of proteins (based on NadjSPC) involved in the light (squares) and dark (triangles) reactions of photosynthesis and the C4 shuttle (asterisks). The inset shows a comparison of the three pathways with NadjSPC values normalized to the maximum value for each pathway to better compare their accumulation kinetics. (B) Accumulation of thylakoid complexes in the developing BS strand distributed over the subclusters. (C) Accumulation pattern of the five thylakoid complexes in the developing BS strand. (D) Accumulation of NDH, PSII, PSI, and cytb6f normalized to proteins of the thylakoid ATP-synthase complex (CF). The inset compares the protein ratio of NDH/PSII along the developmental gradient in the BS strands (open squares) and leaf (closed squares). Taking advantage of the cluster analysis (Figure 2C), we assessed more detailed coexpression patterns (Figures 6A and 6B). Most of the Calvin-Benson cycle enzymes (10 out of 16), in particular those not part of the reductive phase, were in cluster II-3e, together with C4-NADP-Malic enzyme (C4-ME) and phosphoenol pyruvate carboxykinase (PEPCK), as well as the envelope triosephosphate translocator TPT-1 and envelope transporters MEP1, 2, and 4, and a plasma membrane proton-ATPase (see Supplemental Data Set 1B online). Proteins in this cluster show preferential accumulation in BSCs near the leaf tip, as indicated by expression profiles for C4-ME, PRK-2, sedoheptulose-bisphosphatase (S17BPase) (Figure 6B), and TPT-1 and MEP1,2,4 (Figure 7). The Calvin-Benson cycle enzymes of the reductive phase were found together with PPDK and PEPC in cluster II-3d, whereas C4-MDH was found together with plastidic β-carbonic anhydrase 2 (CA2) in cluster II-3c, strongly enriched in MCs. Expression profiles for C4-MDH and CA2 are shown in Figure 6B. This distribution at the leaf tip was consistent with known BSC and MC localization of the Calvin-Benson cycle and C4 shuttle enzymes (Friso et al., 2010). The coexpression of the envelope transporters with the BS-localized Calvin-Benson cycle and C4 shuttle enzymes was particularly exciting since the chloroplast transporters involved in metabolic exchange between BSC and MC chloroplasts are still not identified (Majeran and van Wijk, 2009; Bräutigam et al., 2010; Weber and von Caemmerer, 2010). MEP4 was by far the most abundant and should receive highest priority for functional analysis (Figure 7) (see further below). Figure 6. Open in new tabDownload slide Division of Cluster II into Subclusters and Expression pProfile of MC and BS Strand Marker Proteins. (A) Part of the dendrogram (cluster II) shown in Figure 2C indicating the subclusters for cluster II. The average profile for the accessions in each subcluster is indicated, with closed and open circles for leaf and BS strands, respectively. Protein investments in various cellular functions for all accessions in each cluster are indicated, and the percentage of plastid proteins among the identified proteins in each cluster is indicated. The color coding for the different function is the same as described in Figure 2. The 4-cm point, calculated from the ligule of the leaf base, is marked. (B) Expression profiles of MC and BSC marker proteins in clusters II-3c and II-3e, respectively. Protein names and accession numbers are indicated. Closed symbols and solid lines indicate leaf samples, whereas open symbols and dashed lines indicate BS strand samples. The 4-cm point, calculated from the ligule of the leaf base, is marked. Figure 6. Open in new tabDownload slide Division of Cluster II into Subclusters and Expression pProfile of MC and BS Strand Marker Proteins. (A) Part of the dendrogram (cluster II) shown in Figure 2C indicating the subclusters for cluster II. The average profile for the accessions in each subcluster is indicated, with closed and open circles for leaf and BS strands, respectively. Protein investments in various cellular functions for all accessions in each cluster are indicated, and the percentage of plastid proteins among the identified proteins in each cluster is indicated. The color coding for the different function is the same as described in Figure 2. The 4-cm point, calculated from the ligule of the leaf base, is marked. (B) Expression profiles of MC and BSC marker proteins in clusters II-3c and II-3e, respectively. Protein names and accession numbers are indicated. Closed symbols and solid lines indicate leaf samples, whereas open symbols and dashed lines indicate BS strand samples. The 4-cm point, calculated from the ligule of the leaf base, is marked. Figure 7. Open in new tabDownload slide Plastid Envelope Transporters Involved in Carbohydrate Balance That Passed the Minimum Threshold for Clustering. (A) Relative molar abundance of plastid envelope transporters in the leaf calculated from the normalized spectral abundance factor, NSAF*1000. Proteins in gray are in cluster I and proteins in black in cluster II. PPT1-1, phosphate/phosphoenolpyruvate translocator1-1 (GRMZM2G174107_P02; PPT1-2 quantification includes closely related GRMZM2G103047_P01 and GRMZM2G047404_P01); MEP1,2,3,4, inner envelope transporter proteins (GRMZM2G071423_P01, GRMZM2G077222_P01, GRMZM2G305851_P01, and GRMZM2G138258_P01, respectively); TPT1, phosphate/triose-phosphate translocator-1 (GRMZM2G070605_P01); GLT1-1, glucose transporter 1-1 (also named GlcT1-1) (GRMZM2G153704_P02); MEX1, maltose exporter (formerly root cap 1 [RCP1]; GRMZM2G156356_P01); PHT4, Pi transporter (also named ANTR1; GRMZM2G088196_P01). (B) Expression of transporter proteins along the developmental gradient in leaves (closed squares and solid lines) and BS strands (open squares and dashed lines). The cluster number is indicated. The 4-cm point, calculated from the ligule of the leaf base, is marked. Figure 7. Open in new tabDownload slide Plastid Envelope Transporters Involved in Carbohydrate Balance That Passed the Minimum Threshold for Clustering. (A) Relative molar abundance of plastid envelope transporters in the leaf calculated from the normalized spectral abundance factor, NSAF*1000. Proteins in gray are in cluster I and proteins in black in cluster II. PPT1-1, phosphate/phosphoenolpyruvate translocator1-1 (GRMZM2G174107_P02; PPT1-2 quantification includes closely related GRMZM2G103047_P01 and GRMZM2G047404_P01); MEP1,2,3,4, inner envelope transporter proteins (GRMZM2G071423_P01, GRMZM2G077222_P01, GRMZM2G305851_P01, and GRMZM2G138258_P01, respectively); TPT1, phosphate/triose-phosphate translocator-1 (GRMZM2G070605_P01); GLT1-1, glucose transporter 1-1 (also named GlcT1-1) (GRMZM2G153704_P02); MEX1, maltose exporter (formerly root cap 1 [RCP1]; GRMZM2G156356_P01); PHT4, Pi transporter (also named ANTR1; GRMZM2G088196_P01). (B) Expression of transporter proteins along the developmental gradient in leaves (closed squares and solid lines) and BS strands (open squares and dashed lines). The cluster number is indicated. The 4-cm point, calculated from the ligule of the leaf base, is marked. The relatively high accumulation of cytosolic PEPCK (GRMZM2G001696_P01; 474 adjSPC in leaf) was surprising to us and warranted more attention. PEPCK was strongly induced from base to tip and preferentially expressed in the BS strands, only visible above 4 cm from the base, hence, its location in cluster II-3e. This expression pattern is consistent with an earlier report showing that maize leaves contain appreciable amounts of PEPCK in the BS strands (Wingler et al., 1999). PEPCK in animals and non-C4 plants is believed to function primarily in gluconeogenesis (decarboxylating OAA), and in PEPCK-type C4 plants, it is part of the C4 cycle (decarboxylating OAA in the cytosol), whereas in maize and other NADP-ME-type C4 plants, its role is less clear (for a discussion see, Leegood and Walker, 2003). The expression pattern that we observed in maize is incompatible with a role in gluconeogenesis (as there is no lipid pool in the tip). It was shown for maize BS strands that decarboxylation of Asp (into OAA in the mitochondria) is indirectly dependent on the activity of PEPCK (whereas decarboxylation of malate depends on NADP-ME); thus, PEPCK could help increase the CO2 concentration in the BSCs. However, since our analysis was done on BS strands, we don’t know if PEPCK was within the vascular bundle or in the BSCs. Interestingly, PEPCK in the C3 plant cucumber (Cucumis sativus) was mainly present in the companion cells of the phloem of minor and major veins, where its function may be involved in amino acid metabolism/degradation (Chen et al., 2004). Several other chloroplast proteins with putative functions in plastid protein homeostasis, plastid gene expression, or plastid metabolism clearly coexpressed with MC and BSC markers in clusters II-3c and II3e, respectively. Examples for MCs are two lumenal PPR proteins (GRMZM2G436710_P01 and GRMZM2G010929_P01), and examples for BSCs are an integral thylakoid DnaJ protein (GRMZM2G068316_P01) and a stromal protein phosphatase 2C (PP2C; GRMZM2G071087_P01), both with unknown function (see Supplemental Data Set 1B online). Cell Type–Specific Differences in Photosynthetic Linear and Cyclic Electron Flow In both C3 and C4 plants, an NDH-dependent cyclic electron transport system operates in the thylakoid to increase the ATP/NADPH ratio generated by the light reactions (Takabayashi et al., 2005; Rumeau et al., 2007; Livingston et al., 2010); this is especially important in C4 BSC chloroplasts (Majeran and van Wijk, 2009). The NDH and PSII complexes operate strictly in cyclic and linear electron flow, respectively, whereas photosystem I (PSI) and cytb6f complexes participate in both. The cluster analysis allowed comparison of the kinetics of cell-specific accumulation of the five complexes of the photosynthetic electron transport chain. The 80 identified proteins (both nuclear and plastid encoded) of these five complexes were distributed across subclusters II-3c, II-3d, and II-3e, again indicating tight regulation of protein accumulation (Figure 5B). Figure 5C shows the differential accumulation levels of the five complexes in the BS strand along the leaf gradient; after 4.5 cm, PSII did not further increase much, whereas PSI, cytb6f, and CF doubled in abundance and NDH increased another 3.5-fold. Consequently, between 4.5 cm and the tip, the PSI/CF and cytb6f/CF ratios remained constant in the BSC gradient, whereas the NDH/CF ratio increased by 70%, but the PSII/CF decreased >200% (Figure 5D), further illustrating active upregulation of NDH and active downregulation of PSII in BSCs once BSC-MC chloroplast differentiation had initiated. Consistently, the NDH/PSII protein ratios were indistinguishable between BS strands and leaf samples up to 4 cm from the base (increasing from 0 to ~0.1) and then increased 4-fold in the BSCs toward the tip but remained constant for the leaf (Figure 5D, inset). The NDH/CF ratio was 85% higher in BS strands than in the total leaf, whereas the PSII/CF and cytb6f/CF ratios were 200 and 30% lower, respectively, but the PSI/CF ratio was the same. This clearly showed that ATP production in BSC chloroplasts required increased NDH levels per CF complex, but there was no change in PSI per CF; this strongly supports the notion of increased cyclic electron flow in the BSC chloroplasts. Testing the importance of the NDH complex in maize is urgently needed. Sink-Source Relationships in Primary Carbon Metabolism During leaf expansion and maturation, leaves undergo a sink-source transition, which is defined as the time at which the export of metabolites from photosynthesis exceeds import of carbohydrates (Turgeon, 1989). The sink-source transition must be well integrated with other developmental processes (e.g., chloroplast biogenesis and maturation of the vascular system); indeed, carbon metabolism and sugar signaling are known to impact gene expression strongly at the transcriptional and translational level (Rolland et al., 2006; Hummel et al., 2009; Smeekens et al., 2010). However, the relationship between the sink-source transition and (maize) C4 leaf development and differentiation is poorly understood; defining a molecular template for these metabolic transitions will provide a framework for identification of regulatory networks and metabolic signals that may contribute to C4 leaf development and differentiation. Therefore, we assembled all enzymes and transporters in glycolysis, OPPP, sucrose, and starch metabolism present in our data set; leaf and BS strand accumulation patterns for these pathways as well as for the Calvin-Benson cycle were compared (Figures 8A to 8D) and integrated in a sink-source diagram (Figure 9). We identified and quantified many carbon transporters in the chloroplast envelope as well as the well-studied plasma membrane sucrose transporter SUT1 (also named SUC2) (Lalonde et al., 2004; Slewinski et al., 2009) and the tonoplast sucrose transporter TMT2 (Wormit et al., 2006); these transporters were also integrated in the sink-source diagram. Enlarged expression plots for relevant envelope transporters that passed the clustering threshold are shown in Figure 7, and they are discussed in more detail in the next section. The assignments of the functions of these enzymes were based on information of these enzymes or their homologs in the literature (Buchanan et al., 2000; Bowsher et al., 2008). Figure 8. Open in new tabDownload slide Expression Profiles of Glycolysis, Starch, and Sucrose Metabolism. (A) Cumulative expression of marker proteins involved in sugar synthesis (squares) and degradation (diamonds) along the developmental gradient in leaves (closed symbols and solid lines) and BS strands (open symbols and dashed lines). The signal for sucrose synthesis was multiplied by a factor 30 for better visibility. Marker proteins for sucrose synthesis were sucrose phosphatase (SP1; GRMZM2G055489_P01), sucrose phosphate synthase 2 (SPS-2; GRMZM2G013166_P03 and GRMZM2G140107_P01), sucrose phosphate synthase 3 (SPS-3; GRMZM2G008507_P01), and d-fructose-1,6-bisphosphate 1-phosphohydrolase (F16BPase; GRMZM2G322953_P01). Markers for sucrose degradation were sucrose synthase 1 (SUS1; GRMZM2G089713_P01), sucrose synthase 2-2 (SUS2-2; GRMZM2G152908_P01), sucrose synthase 2-1 (SUS2-1; GRMZM2G060659_P01), fructokinase-1 (FK1; GRMZM2G086845_P01), and fructokinase-2 (FK2; GRMZM2G051677_P01). (B) Comparative analysis of the kinetics of the Calvin-Benson cycle (open squares), starch synthesis (filled squares), starch degradation (filled circles), sucrose synthesis (open diamond), and sucrose gradation (filled diamond) in the BS strands along the developmental gradient. The point of 50% capacity of the various pathways is indicated. NadjSPC for each function are normalized to the maximum value for each function. (C) Cumulative expression of proteins involved in starch synthesis (squares), starch degradation (circles) and β-amylase 5 (BAM5; GRMZM2G058310_P01) (filled triangles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). Included in the calculation for starch synthesis were ADP-glucose pyrophosphorylase large subunit 1,2 (GRMZM2G391936_P02 and GRMZM2G027955_P01), ADP-glucose pyrophosphorylase small subunit 1 (APS1; GRMZM2G163437_P01), granule-associated starch synthase (GBSS; GRMZM2G008263_P01), starch synthase I (SSI; GRMZM2G129451_P01), starch synthase IIa (SSIIa; GRMZM2G105791_P01), starch branching enzyme class IIb-2 (BEIIb; GRMZM2G073054_P01), starch synthase IIIb (SSIIIb; GRMZM2G121612_P01), and starch (amylose) binding protein (GRMZM2G042245_P01). Included in starch degradation are α-glucan phosphorylase-2-1,2 (PHS2-1,2; GRMZM2G147770_P01 and GRMZM2G085577_P01), glucan water dikinase (GWD also named Sex1 or R1 protein; GRMZM2G412611_P01), glucan-phosphorylase 1 (PHS1; GRMZM2G074158_P01), phosphoglucan water dikinase (PWD; GRMZM2G040968_P04), α-amylase 3 (AMY3; AC207628.4_FGP006), and dual-specificity protein phosphatase 4 (DSP4 or SEX4; GRMZM2G052546_P03). (D) Cumulative expression of proteins involved in glycolysis (squares) and the irreversible steps of OPPP (triangles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). To calculate the profiles for glycolysis and irreversible steps in OPPP, 29 and eight protein accessions were used, respectively. The signal for OPPP was multiplied by a factor 10 for better visibility. Figure 8. Open in new tabDownload slide Expression Profiles of Glycolysis, Starch, and Sucrose Metabolism. (A) Cumulative expression of marker proteins involved in sugar synthesis (squares) and degradation (diamonds) along the developmental gradient in leaves (closed symbols and solid lines) and BS strands (open symbols and dashed lines). The signal for sucrose synthesis was multiplied by a factor 30 for better visibility. Marker proteins for sucrose synthesis were sucrose phosphatase (SP1; GRMZM2G055489_P01), sucrose phosphate synthase 2 (SPS-2; GRMZM2G013166_P03 and GRMZM2G140107_P01), sucrose phosphate synthase 3 (SPS-3; GRMZM2G008507_P01), and d-fructose-1,6-bisphosphate 1-phosphohydrolase (F16BPase; GRMZM2G322953_P01). Markers for sucrose degradation were sucrose synthase 1 (SUS1; GRMZM2G089713_P01), sucrose synthase 2-2 (SUS2-2; GRMZM2G152908_P01), sucrose synthase 2-1 (SUS2-1; GRMZM2G060659_P01), fructokinase-1 (FK1; GRMZM2G086845_P01), and fructokinase-2 (FK2; GRMZM2G051677_P01). (B) Comparative analysis of the kinetics of the Calvin-Benson cycle (open squares), starch synthesis (filled squares), starch degradation (filled circles), sucrose synthesis (open diamond), and sucrose gradation (filled diamond) in the BS strands along the developmental gradient. The point of 50% capacity of the various pathways is indicated. NadjSPC for each function are normalized to the maximum value for each function. (C) Cumulative expression of proteins involved in starch synthesis (squares), starch degradation (circles) and β-amylase 5 (BAM5; GRMZM2G058310_P01) (filled triangles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). Included in the calculation for starch synthesis were ADP-glucose pyrophosphorylase large subunit 1,2 (GRMZM2G391936_P02 and GRMZM2G027955_P01), ADP-glucose pyrophosphorylase small subunit 1 (APS1; GRMZM2G163437_P01), granule-associated starch synthase (GBSS; GRMZM2G008263_P01), starch synthase I (SSI; GRMZM2G129451_P01), starch synthase IIa (SSIIa; GRMZM2G105791_P01), starch branching enzyme class IIb-2 (BEIIb; GRMZM2G073054_P01), starch synthase IIIb (SSIIIb; GRMZM2G121612_P01), and starch (amylose) binding protein (GRMZM2G042245_P01). Included in starch degradation are α-glucan phosphorylase-2-1,2 (PHS2-1,2; GRMZM2G147770_P01 and GRMZM2G085577_P01), glucan water dikinase (GWD also named Sex1 or R1 protein; GRMZM2G412611_P01), glucan-phosphorylase 1 (PHS1; GRMZM2G074158_P01), phosphoglucan water dikinase (PWD; GRMZM2G040968_P04), α-amylase 3 (AMY3; AC207628.4_FGP006), and dual-specificity protein phosphatase 4 (DSP4 or SEX4; GRMZM2G052546_P03). (D) Cumulative expression of proteins involved in glycolysis (squares) and the irreversible steps of OPPP (triangles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). To calculate the profiles for glycolysis and irreversible steps in OPPP, 29 and eight protein accessions were used, respectively. The signal for OPPP was multiplied by a factor 10 for better visibility. Figure 9. Open in new tabDownload slide Sink-Source Relationships in Primary Carbon Metabolism. This figure describes the pathways and identified transporters involved in carbon metabolism that are critical for the sink-source transition along the developing leaf. The expression profiles along the developmental gradient in leaves (closed symbols) and BS strands (open symbols) for key enzymes and transporters are shown in line plots. The figure summarizes carbon metabolism in photosynthetic chloroplasts in the source tissue (toward the tip) and carbon metabolism on nonphotosynthetic, heterotrophic plastids at the base of the leaf, the sink region. For simplicity, we did not show the metabolic exchange and specialization of source BSC and MC chloroplasts; instead, they are collectively summarized under the term “source plastid.” Reduced carbohydrates that are exported from the source plastids are converted into sucrose and then transported via the phloem to the sink region. In the sink region, the sucrose can be transiently stored (e.g., in the vacuoles), and sucrose is then degraded into hexose phosphates (G6P, G1P, and F6P) through two parallel pathways either involving invertase or SuSy. The hexosphosphates are then used in the cytosol or imported into the heterotrophic plastids for glycolysis or the OPPP. Blue arrows indicate the generation of reduced carbohydrates and sucrose in sink tissue. DHAP, dehydroxyacetone phosphate; G1P, glucose-1-phosphate; GAP, glyceraldehyde-3-phosphate; F16BP, fructose-1,6-biphosphate; F26BP, fructose-2,6-biphosphate; Pi, inorganic phosphate; PEP, phosphoenolpyruvate; R5P, ribulose-5-phosphate. The inset (right-hand side) shows a close-up of the expression line plot for MEX1; all other plots have a similar x and y axis, and the vertical dotted line indicates the 4-cm point. Figure 9. Open in new tabDownload slide Sink-Source Relationships in Primary Carbon Metabolism. This figure describes the pathways and identified transporters involved in carbon metabolism that are critical for the sink-source transition along the developing leaf. The expression profiles along the developmental gradient in leaves (closed symbols) and BS strands (open symbols) for key enzymes and transporters are shown in line plots. The figure summarizes carbon metabolism in photosynthetic chloroplasts in the source tissue (toward the tip) and carbon metabolism on nonphotosynthetic, heterotrophic plastids at the base of the leaf, the sink region. For simplicity, we did not show the metabolic exchange and specialization of source BSC and MC chloroplasts; instead, they are collectively summarized under the term “source plastid.” Reduced carbohydrates that are exported from the source plastids are converted into sucrose and then transported via the phloem to the sink region. In the sink region, the sucrose can be transiently stored (e.g., in the vacuoles), and sucrose is then degraded into hexose phosphates (G6P, G1P, and F6P) through two parallel pathways either involving invertase or SuSy. The hexosphosphates are then used in the cytosol or imported into the heterotrophic plastids for glycolysis or the OPPP. Blue arrows indicate the generation of reduced carbohydrates and sucrose in sink tissue. DHAP, dehydroxyacetone phosphate; G1P, glucose-1-phosphate; GAP, glyceraldehyde-3-phosphate; F16BP, fructose-1,6-biphosphate; F26BP, fructose-2,6-biphosphate; Pi, inorganic phosphate; PEP, phosphoenolpyruvate; R5P, ribulose-5-phosphate. The inset (right-hand side) shows a close-up of the expression line plot for MEX1; all other plots have a similar x and y axis, and the vertical dotted line indicates the 4-cm point. We observed 12 enzymes involved in sucrose degradation, five involved in sucrose synthesis, multiple cytosolic enzymes involved in conversion reactions of the central hexose-monophosphates (glucose-1-phosphate, glucose-6-phosphate [G6P], and fructose-6-phosphate [F6P]), as well as F16BP-aldolases and trioseP-isomerases required for condensation of glyceraldehyde 3-phosphate and the triosephosphate dehydroxyacetonephosphate (DHAP) into F1,6BP (Figure 9). Only a subset of these enzymes are true markers for sucrose synthesis or degradation because others overlap with glycolysis (e.g., cytosolic pyrophosphate-F6P-6-phosphate 1-phosphotransferase-α,β) and because a number of enzymes are bidirectional in sucrose metabolism, such as UDP-glucose pyrophosphorylase-1,2, cytosolic phosphoglucomutase-1, and hexokinase-1 (Figure 9). The cytosolic enzymes sucrose phosphase-1, sucrose phosphate synthase-1,2,3, and D-F1,6BP-1-phosphohydrolase (F1,6BPase) are markers for sucrose synthesis and clearly displayed an increased accumulation from base to tip (cluster II-3d) (Figures 8A and 9). Markers for sucrose degradation were sucrose synthase-1,2 (SuSy1,2) and fructokinase-1,2 (FK1,2), and they decreased from base to tip (Figures 8A and 9) and belonged to cluster I-1 or I2a. Notably, investments in sucrose synthesis and degradation enzymes were very similar between BS strands and the rest of the leaf, based on the distribution of SPS and SPase. This is perhaps surprising since based on the distribution of SPS and labeling, it was previously concluded that most of the sucrose (in mature leaves) is synthesized in MCs (Usuda and Edwards, 1980; Lunn and Furbank, 1997). However, a later immunohistological study came to a completely different conclusion and showed that SPS is localized in both BSCs and MCs (Cheng et al., 1996). In fact, in young leaves, SPS protein was predominantly in BSCs, whereas mature leaves showed nearly equal levels of signal in both BSCs and MCs (Cheng et al., 1996). Activity measurements suggested that SPS may function in both MCs and BSCs for sucrose synthesis in the light, particularly at high light intensity, while in the dark, the major function may be in the BSCs during starch degradation (Ohsugi and Huber, 1987). We point out that SPS is considered the controlling enzyme in sucrose synthesis (Winter and Huber, 2000). The explanation for the apparent discrepancy between these various studies likely relates to several aspects: (1) in the light period, sucrose is synthesized from triosephosphate DHAP and exported from the chloroplasts into the cytosol. Since the reductive phase of the Calvin cycle (ending with the generation of DHAP) occurs predominantly in the MC chloroplasts, it would be most effective to have sucrose synthesized from excess DHAP in this cell type. DHAP is also transported back into the BSC chloroplasts where it is used to complete the Calvin-Benson cycle. Therefore, it is critical that sucrose synthesis in MCs be controlled such that there is sufficient DAHP for the BSC chloroplasts to maintain high rates of CO2 fixation through the Calvin cycle. By contrast, during the night, sucrose is synthesized from transient starch (broken down though the hydrolytic pathway into Maltose; Figure 9) that is nearly exclusively stored in the BSC chloroplasts in the source region (Figure 1P). Therefore, the SPS observed in the MC and BSC chloroplasts is likely mostly active in the light and night period, respectively. (2) SPS is under allosteric regulation by G6P (stimulation) and Pi (inhibition) and is also regulated by phosphorylation and redox state (reviewed in Winter and Huber, 2000); (3) SPS is represented by four expressed homologs (GRMZM2G055489_P01, GRMZM2G013166_P03, GRMZM2G140107_P01, and GRMZM2G008507_P01). Induction of sucrose synthesis capacity closely tracked accumulation of the Calvin-Benson cycle, except in the first 3 cm where sucrose synthesis enzymes were not detected, whereas accumulation of Calvin-Benson cycle enzymes was already induced, albeit at low levels (Figure 8B). Sucrose degradation capacity decreased to 50% at 5.5 cm, whereas Calvin-Benson cycle and sucrose synthesis enzymes reached 50% of their capacity at around 7 cm from the base (Figure 8B). These kinetics are consistent with our 14CO2 labeling experiments that indicated that the sink-source transition point was around 4 cm from the base. Starch metabolism was represented by 24 proteins, with nine identified only in the BSC soluble fractions (Supplemental Data Sets 1A and 1B online); these nine are not further discussed here. The identified starch synthesis enzymes (10) were all plastid localized, whereas eight out of 14 starch degradation enzymes were plastidic. Figure 8C shows the total protein mass invested in starch synthesis and starch degradation. While starch synthesis was most prominent in the BSC chloroplasts above 4 cm, there was clearly also some accumulation of both the small and large subunit of ADP-glucose pyrophosphorylase at the base (~12% of the level at the tip), likely using imported sucrose as substrate. This observation was consistent with the presence of starch granules (seen by TEM) in the base (Figure 1O). Except for β-amylase 5 (BAM5), the starch degradation enzymes were all strongly enriched in the BS strands, with very strong expression in the leaf tip (II-3e) (Figure 8C). A big surprise was BAM5; it was the most abundant putative starch degradation enzyme in our data set, peaking at 3 to 4 cm from the base (II-2) and was restricted to the MCs (Figure 8C). Neither maize BAM5 nor its best homologs in Arabidopsis thaliana and rice have a predicted cTP, and Arabidopsis BAM5 was reported to be vacuolar and represented the bulk of all β-amylase activity in the Arabidopsis leaf (Laby et al., 2001). Our observations in maize suggest that either (1) significant extraplastidic starch degradation occurs around the transition zone, (2) BAM5 has a function unrelated to starch metabolism or degrades shorter glucose polymers than starch, or (3) despite the absence of a predicted transit peptide, the protein is imported into chloroplasts. Functional analysis of BAM5 in maize may reveal an unexpected feature of the sink-source transition. We identified 35 enzymes involved in glycolysis (F1,5BP was used as the starting point), nearly all of which were cytosolic; 15 fell in cluster I-2a,b and seven in cluster II (mostly II-2). Cytosolic glycolytic enzymes were highest in the base and remained high all along the leaf gradient (Figure 8D). Levels of plastid-localized glycolytic enzymes were insignificant compared with those in the cytosol. We also quantified eight enzymes of the irreversible OPPP reactions as well as several OPPP enzymes involved in the reversible reactions. The OPPP enzymes involved in the irreversible reactions were relatively highly expressed in the first 2 cm, followed by a rapid decrease; however, near the tip, OPPP, although low, was enriched in the BS strands, possibly functioning in the nonphotosynthetic plastids in the vasculature or in the BSCs in the night (Figure 8D). Among the OPPP enzymes involved in reversible reactions, plastidic transaldolase-1 (GRMZM2G134256_P01) was severalfold more abundant than the other PPP enzymes and was equally expressed across BSCs and MCs (I-2b). The high level of OPPP enzymes in the sink region is consistent with its role in providing reducing equivalents and shikimate precursors in nonphotosynthetic tissues. The base of the leaf was devoid of Calvin-Benson cycle enzymes, and components of the thylakoid electron transport are either absent or present at insignificant levels, consistent with the lack of thylakoid membranes as shown by the TEM analysis (Figure 1). Therefore, the base of the leaf completely depends on the tip of the leaf or older leaves for reduced carbohydrates. The base of the leaf imports in particular sucrose, which is broken down in the glycolytic pathway, used in mitochondrial respiration, or used in the OPPP, thus providing both the energy [ATP and NAD(P)H] and carbon backbones for growth and development. Sucrose arriving at the base of the leaf is either degraded by SuSy into fructose and UDP-glucose or imported into vacuoles and degraded by invertase, resulting into glucose and fructose (Winter and Huber, 2000). FK1 is an excellent marker for the combined sucrose degradation pathways and showed a gradual decrease from base to tip. The protein expression pattern for SuSY and FK1,2 was comparable and very distinct from the quantified vacuolar invertase (for discussion on maize invertases, see Kim et al., 2000). We also noted that this invertase peaked very sharply between 2 and 3 cm from the base (but not in the base itself) and that the invertase was not observed in the BS strands, suggesting a specific role in MCs slightly before the sink/source metabolic reorganization. Sucrose degradation products F6P and G6P are either consumed by glycolysis or imported as G6P into the sink plastids via GPT2 (Figure 9). Transporters in Carbohydrate Metabolism and Plastid Transporters for Other Types of Substrates The distribution of CO2 assimilation over two distinct cell types requires a high flux of metabolites between MCs and BSCs (Majeran and van Wijk, 2009; Weber and von Caemmerer, 2010). Thus, the identification and expression patterns of these chloroplast transporters are of great importance to understand C4 leaf development and differentiation. Moreover, the nonphotosynthetic plastids in the sink region must import carbohydrates and require a specific set of envelope transporters. Sucrose is transported from source to sink and must be loaded and unloaded into the vascular bundle and possible transiently stored in vacuoles; these transport steps require also specific transporters. Finally, mitochondria must adapt their transporters to accommodate cell-type and developmental differences (see next section). Twenty-three transporter proteins that participate in carbohydrate metabolism in sink and/or source regions were identified. These include the plasma membrane sucrose translocator involved in phloem loading (SUT1) (Lalonde et al., 2004; Carpaneto et al., 2005; Slewinski et al., 2009) and the tonoplast sucrose transporter (TMT2) (Wormit et al., 2006), and the rest were plastid envelope translocators for glucose (GLT1-1,2), glucose-6P (GPT2), maltose (MEX1), triosephosphates/Pi (TPT), phosphoenol pyruvate/Pi (PPT1,2), ATP/ADP (AATP1 or NTT1), Pi (PHT4), as well as the MEP family with undefined substrates. Many of these were integrated in the sink-source scheme (Figure 9). Ten of these accessions passed the threshold for clustering and provided the most reliable information with respect to sink-source distribution and cell-specific differentiation; their relative molar abundance in the leaf (calculated from the normalized spectral abundance factor [NSAF]) is shown in Figure 7A. PPT1-1, MEP4, and TPT1 were severalfold more abundant than all others, and MEP4 and TPT1 showed tight coexpression with BSC Calvin-Benson cycle and C4 shuttle proteins (cluster II-3e). This makes MEP4 and TPT1 the most likely candidates for export of pyruvate or import of DHAP (MEP4) and export of 3PGA (TPT1) (Figure 7B). PPT1-1 was part of cluster II-1b and showed maximal expression around the transition zone rather than at the leaf tip and was consistently higher in the BS strand. PPT1-1 seems to be the homolog of Arabidopsis PPT, which is a PEP/Pi antiporter, exporting PEP from the plastid. If maize PPT1-1 is the true functional homolog of Arabidopsis PPT, then one would predict MC chloroplast enrichment in particular close to the leaf tip. It is possible that the PPT1-1 expression pattern is due to a contribution of nonphotosynthetic plastids in the vascular bundle, where it could export PEP for glycolysis (see further in Discussion). The PPT1-1 and PPT1-2 proteins were both enriched in BS strands but otherwise behaved oppositely in their expression along the developmental gradient (Figures 7B and 9), as evidenced by the cluster analysis (in clusters II-1b and I-2b). MEP1,2 and 3 showed equal relative abundance (Figure 7A), with MEP1,2 both behaving as BSC-specific transporters highly expressed in the tip. These MEPs are good candidates for export of glycolate (for removal in the photorespiratory pathway). MEP3 accumulation decreased from base to tip (cluster I-1), and its substrate remains to be determined. The transporter GLT1-1 (represented by two close homologs GRMZM2G153296_P02 and GRMZM2G153704_P02) (also named GLC1-1; I-2b) involved in export of plastidic glucose was clearly a sink-localized transporter and was consistently enriched in the BS strands. We observed some starch in the plastids of the leaf base as well as low levels of starch metabolic enzymes. Therefore, we postulate that despite being heterotrophic, the nongreen plastids in the BSCs in the sink region of the leaf export some glucose to the rest of the cell and thus require low levels of GLT in the inner envelope. However, it should be noted that a molecular characterization of the GLT (GlcT) family is lacking, even in Arabidopsis (Linka and Weber, 2010). The homolog of GLT1-1, GLT1-2 (GRMZM2G098011_P01), was below the threshold for cluster analysis, but its expression was observed only in photosynthetic chloroplasts at the leaf tip, and its function is export of glucose during the day (Figure 9). We identified another 12 envelope proteins involved in transport of copper (PAA1), iron (PIC/TIC21), S-adenosylmethionine (SAMT1), or unknown substrates (see Supplemental Data Sets 1A and 1B online); five (TIM17/22, an anion ATPase, OEP24-II, PAA1, and SAMT1) passed the threshold for clustering, and their relative molar abundance and expression profiles are shown in Supplemental Figures 4A and 4B online. These five transporters were also identified in our previous study on BSC and MC chloroplasts isolated from the fully developed maize leaf tip (Friso et al., 2010). Clearly, the role of transport in leaf development and cellular differentiation is poorly understood and represents a challenge for future research (for discussion, see Weber and von Caemmerer, 2010). The Photorespiration Cycle Involving Peroxisomes and Mitochondria Repression of photorespiration is one of the hallmarks and benefits of C4 photosynthesis (Edwards et al., 2001a, 2001b). Based on the accumulation of two specific photorespiratory markers, the chloroplast 2-phosphoglycolate phosphatase and peroxisomal glycolate oxidase (GOX), photorespiration occurred in the BS strand starting at ~4.5 cm from the base and strongly (>5-fold) increased toward the leaf tip, concomitant with induction of the Calvin-Benson cycle (Figure 10A). Indeed, GOX was part of cluster II-3e, together with BSC markers. Mitochondrial glycine decarboxylase-P (GDC-P) was also part of cluster II-3e, in agreement with an earlier observation of BSC enrichment in a C3-C4 intermediate species (Rylott et al., 1998) (GDC-H homologs were below the threshold for clustering). The plastid envelope transporter of glycolate is unknown but should be part of II-3e (if sufficiently highly expressed). The BS strand-enriched MEP1 and MEP2 proteins in cluster II-3e, with ~10-fold lower abundance than MEP4 (Figure 7A), have unknown functions, making them good candidates for the glycolate exporter. Recent analysis of a maize line with a mutation in GOX showed that the photorespiratory pathway is indeed required to prevent accumulation of toxic glycolate and seedling death (Zelitch et al., 2009). Figure 10. Open in new tabDownload slide Induction of Photorespiration and Expression of the Vacuolar V-Type ATPase. (A) Expression patterns of the specific photorespiratory enzymes phosphoglycolate-phosphatase-2 (PGP-2; GRMZM2G018441_P01; squares) and glycolate oxidase 2 (GOX-2; GRMZM2G129246_P01; circles), in plastids and peroxisomes, respectively, along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (B) Cumulative expression pattern of the (near identical) ammonia transporter homologs in the tonoplast (TIP2; GRMZM2G027098_P01, GRMZM2G121275_P01, and GRMZM2G056908_P01) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (C) Cumulative expression pattern of the 11 subunits of the V-type ATPase (20 accession numbers; see Supplemental Data Set 1B online) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). Figure 10. Open in new tabDownload slide Induction of Photorespiration and Expression of the Vacuolar V-Type ATPase. (A) Expression patterns of the specific photorespiratory enzymes phosphoglycolate-phosphatase-2 (PGP-2; GRMZM2G018441_P01; squares) and glycolate oxidase 2 (GOX-2; GRMZM2G129246_P01; circles), in plastids and peroxisomes, respectively, along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (B) Cumulative expression pattern of the (near identical) ammonia transporter homologs in the tonoplast (TIP2; GRMZM2G027098_P01, GRMZM2G121275_P01, and GRMZM2G056908_P01) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (C) Cumulative expression pattern of the 11 subunits of the V-type ATPase (20 accession numbers; see Supplemental Data Set 1B online) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). We also quantified several other peroxisomal and mitochondrial enzymes of the respiratory cycle (e.g., mitochondrial Gly cleavage enzyme H and peroxisomal Ala-glyoxylate aminotransferase), as well as chloroplast Glu-ammonia ligase and Fd- and NADH-Glu synthase. These enzymes are involved in both primary N-assimilation and photorespiration, and they are therefore less diagnostic of strictly photorespiratory activity (see further below). Nevertheless, the mitochondrial and peroxisomal proteins clustered in II-3d or 3e. The tonoplast ammonia transporter (TIP2) was exclusively expressed in the BS strand with highest expression near the leaf tip, where it likely functions to store excess ammonia produced by photorespiration (Figure 10B). Respiration, Coexpression Analysis of Mitochondrial Transporters, and Vacuolar Transport We identified 89 proteins involved in the mitochondrial tricarboxylic acid (TCA) cycle and oxidative phosphorylation. Fifty-two enzymes passed the threshold for cluster analysis, with all except four belonging to cluster I (mostly in I-1 and I-2a), indicative of high basal expression (Figure 2C). Indeed, protein mass investments in these pathways were high and constant in the first 2.5 cm and then decreased relatively to 8 cm, followed by stabilization. Investments in the electron transport chain were ~2-fold higher than the TCA cycle (Figure 11A). The TCA investment was enriched in the BS strand, whereas relative investments in electron transport were similar between BS strands and MCs (Figure 11A). This suggests a higher demand for TCA cycle intermediates in BS strands, for instance, for synthesis of amino acids in the Glu family. Figure 11. Open in new tabDownload slide Quantitative Protein Expression of Mitochondrial Respiration and Transporters, Protein Synthesis and Organelle Biogenesis, and Isoprenoid and Tetrapyrrole Metabolism. (A) Cumulative expression of proteins involved in the TCA cycle (circles) and the mitochondrial electron chain and oxidative phosphorylation (squares) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). The inset shows a comparison of the two pathways with NadjSPC values normalized to the maximum value for each pathway. (B) Cumulative expression of mitochondrial outer membrane porins (circles) and inner membrane transporters (squares) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). The inset shows a comparison of these groups with NadjSPC values normalized to the maximum value for each group. (C) Cumulative expression of proteins involved in cytosolic translation (squares) and plastid translation (circles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). The inset shows a comparison of the two groups with NadjSPC values normalized to the maximum value for each group. (D) Cumulative expression of proteins involved in biogenesis in mitochondria (squares) and plastids (circles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). Values for mitochondria are multiplied by 10. (E) Cumulative expression of proteins involved in the plastidic deoxyxylulose phosphate (MEP) pathway (squares), the cytosolic mevalonate (MVA) pathway (triangles), and four plastid enzymes operating immediately downstream of the MEP pathway (post-MEP; circles) along the developmental gradient in leaves (closed symbols and solid lines) and BS strands (open symbols). (F) Cumulative expression of proteins involved in the tetrapyrrole pathway (squares) and of protochlorophyllide reductase A (triangles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). Figure 11. Open in new tabDownload slide Quantitative Protein Expression of Mitochondrial Respiration and Transporters, Protein Synthesis and Organelle Biogenesis, and Isoprenoid and Tetrapyrrole Metabolism. (A) Cumulative expression of proteins involved in the TCA cycle (circles) and the mitochondrial electron chain and oxidative phosphorylation (squares) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). The inset shows a comparison of the two pathways with NadjSPC values normalized to the maximum value for each pathway. (B) Cumulative expression of mitochondrial outer membrane porins (circles) and inner membrane transporters (squares) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). The inset shows a comparison of these groups with NadjSPC values normalized to the maximum value for each group. (C) Cumulative expression of proteins involved in cytosolic translation (squares) and plastid translation (circles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). The inset shows a comparison of the two groups with NadjSPC values normalized to the maximum value for each group. (D) Cumulative expression of proteins involved in biogenesis in mitochondria (squares) and plastids (circles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). Values for mitochondria are multiplied by 10. (E) Cumulative expression of proteins involved in the plastidic deoxyxylulose phosphate (MEP) pathway (squares), the cytosolic mevalonate (MVA) pathway (triangles), and four plastid enzymes operating immediately downstream of the MEP pathway (post-MEP; circles) along the developmental gradient in leaves (closed symbols and solid lines) and BS strands (open symbols). (F) Cumulative expression of proteins involved in the tetrapyrrole pathway (squares) and of protochlorophyllide reductase A (triangles) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). Eight outer membrane voltage-dependent anion channels (VDACs or porins) and seven inner membrane mitochondrial metabolite transporters were observed. The porins facilitate metabolite exchange between the cytosol and inter-membrane space (Kusano et al., 2009). Whereas the porins and inner membrane carriers show strong differences in absolute expression levels, they had comparable expression profiles; they were high in the first few centimeters, followed by a dramatic decrease in the MCs, while they stayed relative constant in the BS strands (Figure 11B). Three porins, the very abundant ATP/ADP carrier (export of ATP to cytosol), a phosphate transporter (PHT3-1), and two isoforms of the dicaboxylate/tricarboxylate carriers (DTC1,2), passed the threshold for cluster analysis, and their relative molar abundance (from NSAF) and expression profiles are shown in Supplemental Figures 5A and 5B online. All transporters fell in cluster I-1, except for one porin, which belonged to cluster I-2b (see Supplemental Figure 5B online). The adenine nucleotide carrier ACC1 (exchanging of mitochondrial ATP with cytosolic AMP; also named ADNT1) was expressed mostly right at the base of the leaf and decreased rapidly along the leaf axis, suggesting that mitochondria supply the cells at the base of the leaf with ATP (see Supplemental Figure 5B online). More than 20 vacuolar proteins were identified, including the monosaccharide transporter TM2, the ammonia transporter (TIP2;3), a putative aquaporin, Na+-Ca2+ exchanger, 11 subunits of the V-type ATPase, and an aspartyl and Cys protease. Nine proteins passed the threshold for cluster analysis, most of them V-type ATPase subunits, and seven proteins belonged to cluster I-1. The V-ATPase and V-PPase provide the driving force for active transport across the tonoplast. The expression profile of the sum of the V-ATPase subunits along the leaf gradient is shown in Figure 10C. Protein Synthesis and Homeostasis and Biogenesis of Chloroplasts and Mitochondria Based on identification of 168 cytosolic and 61 plastid proteins involved in translation, we observed dramatic shifts in investments in the cytosolic and plastid protein translation machinery (Figure 11C); the ratio of protein mass in cytosolic translation to plastid translation changed from 21:1 in base to 1.5:1 at the tip. Chloroplast translation was highest between 4 and 9 cm (cluster II), whereas cytosolic translation was highest immediately at the base (cluster I), with no general under- or overaccumulation in the BS strands (Figure 11C). Ten importins for protein sorting in/out of the nucleus also followed expression of the cytosolic translational machinery. Plants rely on proteolysis to regulate the abundance of proteins and to adjust the cellular proteome to respond to a changing environmental or to developmental needs. The ubiquitin–26S proteasome system (UPS) is the central proteolytic system outside the mitochondria and plastids (Vierstra, 2009). We identified at least 145 maize components of the UPS, and its total abundance closely followed cytosolic translation, with all but one component belonging to cluster I. In addition to the UPS, we identified 130 proteases, placed into 105 protein groups for quantification purposes. These proteases included all major protease and peptidase families, including metallo-, Ser, Cys, and Asp proteases, subtilases, and amino peptidases. This collection of proteases provides a great resource to investigate systematically the role of proteolysis in leaf development and C4 differentiation. As an example, we will discuss the mitochondrial and plastid proteases below. Proteins involved in plastid gene expression and biogenesis were well represented with 243 proteins; these included proteins associated with plastid DNA, proteins involved in plastid RNA synthesis, processing, or stabilization, translational components, as well as chaperones and assembly factors, proteases, and proteins with unknown functions. Total investments in this function were low in the base and peaked at 5 cm (Figure 11D); consistently, the majority (77%) of the clustered proteins belonged to cluster II, in particular II-1b and II-2 (Figures 2C and 6A). An exception was the plastid division proteins FtsZ1,2, which both clustered in I-2A, indicative of highest plastid division rates at the leaf base. Some of the clustered proteins show a pronounced under- or overexpression in BS strands, making them particularly interesting candidates for regulatory components of C4 expression; we already mentioned the lumenal PPR proteins and protein isomerases, as well the BSC-enriched integral thylakoid DnaJ protein and a stromal PP2C. Here, we highlight the plastid proteases since they may be involved in the regulation of BSC and MC chloroplast differentiation, such as BSC repression of PSII accumulation. As the heterotrophic proplastid is remodeled into BSC or MC specialized chloroplasts during leaf development, a subset of proteins must be removed; examples are protochlorophyllide reductase A and the carbohydrate transporters (GLT1-1 and PPT1-2). We evaluated the identified maize plastid proteases according to their subchloroplast location and their protease type, with functional annotation mostly based on Arabidopsis homologs (Sakamoto, 2006; Kato and Sakamoto, 2010). Figure 12A shows the nine plastid protease systems that have the highest relative concentration (based on NSAF). The most prominent system was the soluble ATP-dependent Ser-type Clp protease system. In Arabidopsis, this is the most abundant stromal protease system, and it is essential for plant growth and development (Sjögren et al., 2006; Kim et al., 2009). In Arabidopsis and other Brassiceae, the stromal Clp system has been shown to consist of a tetradecameric Clp proteolytic core with ClpP and ClpR subunits and two peripherally attached ClpT proteins (Peltier et al., 2004). Based on information from the bacterial Clp system, substrates are delivered to the Clp protease core by the ClpC/D chaperone system. We identified maize homologs for Arabidopsis ClpP1,4,5,6, ClpR1,2,3,4, ClpPT1,2, two ClpP2-like proteins, as well as ClpC chaperones. (We also identified low levels of maize homologs of the mitochondrial ClpP2.) The expression profile shows that the ClpPRT core is already present in the undifferentiated proplastids and remained quite constant with peak expression in the leaf just before the 4-cm point (Figure 12B). For comparison, we showed the accumulation of the total plastid and thylakoid proteomes (Figure 12C). When calculating the amount of Clp protein on a total plastid proteome basis, it is clear that the Clp system is particularly important in the early stages of plastid development (Figure 12D). This is consistent with the observation that the ClpPR core was very prominent in etioplasts of pea (Pisum sativum; Kanervo et al., 2008) and also consistent with the particular importance of the ClpPR core in early stages of leaf development in Arabidopsis (Rudella et al., 2006; Sjögren et al., 2006; Kim et al., 2009). The ClpC chaperones behaved similarly to the ClpPRT core proteins, except that the Clp chaperones increased their accumulation in the BS strands, suggesting a particular important role in establishing the BS chloroplast proteome. Figure 12. Open in new tabDownload slide Expression of Plastid Proteolytic Systems along the Leaf Developmental Gradient. (A) Relative molar abundance of nine different plastid protease systems along the leaf gradient calculated from the NSAF. Color coding is explained in the figure. 5x, The NSAF values are multiplied by 5 to make the bars better visible; AP, aminopeptidase; EP, endopeptidase (B) and (E) to (G) Expression profile of ClpC homologs and ClpPRT subunits (A), thylakoid FtsH proteases (E), stromal PreP1 (F), and stromal DegP2 (G) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (C) Accumulation pattern of the total plastid and thylakoid proteomes along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (D) Expression profile of the ClpC chaperones, ClpPRT subunits, and thylakoid FtsH subunits normalized to the total plastid proteome (based on NadjSPC) along the developmental gradient in leaves. Figure 12. Open in new tabDownload slide Expression of Plastid Proteolytic Systems along the Leaf Developmental Gradient. (A) Relative molar abundance of nine different plastid protease systems along the leaf gradient calculated from the NSAF. Color coding is explained in the figure. 5x, The NSAF values are multiplied by 5 to make the bars better visible; AP, aminopeptidase; EP, endopeptidase (B) and (E) to (G) Expression profile of ClpC homologs and ClpPRT subunits (A), thylakoid FtsH proteases (E), stromal PreP1 (F), and stromal DegP2 (G) along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (C) Accumulation pattern of the total plastid and thylakoid proteomes along the developmental gradient in leaves (closed symbols) and BS strands (open symbols). (D) Expression profile of the ClpC chaperones, ClpPRT subunits, and thylakoid FtsH subunits normalized to the total plastid proteome (based on NadjSPC) along the developmental gradient in leaves. Concomitant with the development of the chloroplast and its internal thylakoid, the thylakoid-bound FtsH protease system increased strongly (20×) in abundance and became the system with the highest relative concentration (Figure 12A). The thylakoid members (FtsH1,2,5,8) are well studied in Arabidopsis, and their inactivation leads to a variegated phenotype; they are believed to have a general quality control function but have also been reported to be involved in D1 degradation (Kato et al., 2009; Kato and Sakamoto, 2009; Liu et al., 2010). The accumulation patterns of the thylakoid FtsH proteins showed a strong positive correlation with the accumulation of the thylakoid-bound photosynthetic apparatus (Figure 12E). Normalization of thylakoid FtsH on a plastid proteome basis (Figure 12D) showed a relative increase in FtsH in the first 4 cm and then remained constant, consistent with the household protease role, but not supportive of a specialized D1 degradation function, since the proteases did not show preferential MC accumulation (Figure 12E). We also identified three soluble stromal aminopeptidases, namely, eucyl aminopeptidase (LAP1), glycyl aminopeptidase, and glutamyl endopeptidase (cGEP) (Figure 12A). Little is known about the first two peptidases, whereas there is some in vitro evidence that cGEP can degrade the N-terminal part of LHCII proteins (Kato and Sakamoto, 2010). These peptidases generally showed highest accumulation in the 3 to 5 cm from the base, at the peak of chloroplast biogenesis. A better characterized soluble stromal protease was the Prep1 metalloprotease; its homolog was reported to be dual targeted to mitochondria and chloroplasts (Glaser et al., 2006), but our previous proteomics data from maize and Arabidopsis strongly indicate that most Prep1 is localized to the (more abundant) chloroplast. It has been suggested that its role in Arabidopsis is to degrade short unstructured peptides, such as cleaved chloroplast transit peptides (Bäckman et al., 2009; Nilsson Cederholm et al., 2009). Prep1 clearly peaked in the transition zone and was strongly enriched in the BS strands (Figure 12F), suggesting an important role in early stages of BS chloroplast protein homeostasis. The pronounced enrichment in BS strands makes it an excellent candidate for involvement in C4 differentiation; targeted analysis should determine if and how it plays a role in C4 differentiation. We identified maize homologs of all four DegP proteases known to be located in Arabidopsis chloroplasts, namely, lumenal DegP1,2,5 and stromal DegP2. DegP2 was the most abundant member, followed by DegP1, and we could determine the expression profile along the leaf gradient (Figures 12A and 12G). The thylakoid lumenal DegP1 protease increased from base to tip, concomitant with accumulation of the thylakoid system (Figure 12A). DegP1 was shown to be involved in the repair cycle of the PSII reaction protein D1 in the thylakoid (Kapri-Pardes et al., 2007; Sun et al., 2010); the observed maize leaf expression profile is consistent with such a role. However, the expression profile of DegP2 suggests a primary role during chloroplast biogenesis, rather than a role in D1 repair (Figure 12G). We quantified investments in mitochondrial biogenesis and homeostasis based on 25 proteins, including peptidases and proteases, protein import components of the TIM/TOM complex, HSP70 and CPN60 chaperones, elongation factors, and a few ribosomal proteins (Figure 11D). Those passing the minimum threshold were all members of cluster I (mostly I-2a), with not much difference between BS strands and total leaf. Maximum relative protein investments (based on NadjSPC) were ~10-fold lower in mitochondria than in plastids (Figure 11D). Thus, investments in mitochondrial biogenesis were maximal at the base (cluster I), consistent with the high ratio between mitochondrial and plastid cross sections at the base determined by image analysis. Coordination of Synthesis of Isoprenoids and Chlorophylls Terpenoids have key functions in photosynthesis and electron transport (e.g., carotenoids, chlorophylls, and quinones) and growth regulation (brassinosteroids and gibberellin). The first stage of terpenoid synthesis occurs in the cytosolic mevalonate (MVA) pathway and in the plastidic methyl-erythritol-4-phosphate (MEP) pathway and produces isopentenyl pyrophosphate and dimethylallyl pyrophosphate. We identified five out of the seven steps of the cytosolic MVA pathway; two passed the threshold for clustering and belonged to cluster I-2a. We identified all seven enzymes of the plastid-localized MEP pathway; two passed the threshold for clustering and belonged to clusters II-1a and II-2. In addition, we identified four enzymes operating immediately downstream of the MEP pathway, producing precursors for carotenoids and chlorophylls, as well as for quinones and tocopherol. The cytosolic MVA pathway was maximally expressed in the first 2 cm and then rapidly decreased below detection (Figure 11E). By contrast, the MEP and post-MEP enzymes peaked around 4 to 5 cm and then gradually decreased, but with considerable accumulation levels remaining (Figure 11E). Clearly, the MEP enzymes correlated much better with chloroplast biogenesis than did the MVA pathway, suggesting that plastid isoprenoids were synthesized within the organelle. By contrast, mitochondrial biogenesis followed the MVA pathway (cf. Figures 11D and 11E), suggesting that mitochondrial terpenoid precursors (e.g., ubiquinone) originate from the cytosolic pathway. Chlorophyll biosynthetic enzymes (13 out of 16 steps) were well covered and fell mostly in cluster II-1 or II-2. Our data indicate an extended period of chlorophyll synthesis between 2 and 9 cm (Figure 11F); this corresponded well with the long expansion of the thylakoid system and the induction kinetics of the MEP pathway. We also identified four chloroplast enzymes involved in carotenoid synthesis, each with low expression levels. The most abundant protein, phytoene dehydrogenase, with the most reliable quantification, strongly peaked around 4.5 cm from the base, similar to the MEP and post-MEP pathway. Just recently it was shown in Arabidopsis that two unusual members of the LHC family, LIL3-1 and LIL3-2 (also named stress-enhanced protein 3 [SEP3]; Engelken et al., 2010), which have two-transmembrane helices, rather than the three found in the major and minor LHCs, play a role in stabilizing geranylgeranyl reductase, an enzyme that catalyzes the conversion of GG-diphosphate to phytol-diphosphate (Phytol-PP) in the plastid (Tanaka et al., 2010). Phytol-PP is a substrate for both tocopherol synthesis and for chlorophyll synthesis. The maize LIL3 homologs (GRMZM2G477236_P01 and GRMZM2G027640_P01) have a distinctly different expression pattern (cluster II-1) compared with the major and minor LHCs of PSII and PSI (cluster II-3c or II-3d). Consistent with the newly discovered role in Arabidopsis, LIL3 proteins showed maximal accumulation around the transition zone (see Supplemental Figure 6 online), similar to enzymes in the tetrapyrrole and MEP (isoprenoid) pathway (Figure 11F). In addition to the two LIL3 proteins, Arabidopsis has additional two-helix proteins, one-helix LHC proteins named OHP, and three-helix early light-induced proteins (ELIPs) (Engelken et al., 2010). The functions of these are still unknown in plants, despite dozens of published studies in particular on the ELIPs (Hutin et al., 2003; Engelken et al., 2010). In our current data set, we identified an additional seven OHP/LIL/ELIP homologs (see Supplemental Data Sets 1A and 1B online; PPDB). Three of these, ELIP1/2, SEP4, and OHP2-like, passed the threshold for cluster analysis. Interestingly, LIL3 (mentioned above) and ELIP followed the same developmental expression pattern, peaking in the sink region before the 4-cm point and decreasing gradually toward the tip and with consistent higher levels in the BS strands (cluster II-1b). By contrast, SEP4 and OHP-like proteins had very similar expression patterns and were part of cluster II-3c, similar to PSII proteins and other MC chloroplast markers (see Supplemental Figure 6 online). Our results therefore suggest that ELIP1/2 likely functions in the plastid isoprenoid pathway, which is in line with the initially proposed function in barley (Hordeum vulgare) to play a regulatory role during early chloroplast development (Grimm and Kloppstech, 1987). By contrast, SEP4 and OHP are unlikely to be involved in biosynthesis of isoprenoids/tetrapyrroles but have a function directly relating to PSII or a stress defense associated with the linear electron transport activity in the MC chloroplasts. Redox Regulation and Reactive Oxygen Species Defense Superoxide (O2 −) is generated when oxygen functions as terminal acceptor and is converted into hydrogen peroxide (H2O2) by superoxide dismutases (SODs). H2O2 is also generated by peroxisomal glycolate oxidase and cell wall–bound peroxidases. H2O2 is detoxified in the ascorbate and gluthatione cycle and by peroxiredoxins and glutaredoxins in particular within the plastid. The thioredoxin system is central in redox regulation. We identified 82 proteins assigned to these redox and reactive oxygen species (ROS)-related functions; 39 proteins passed the minimum criteria for clustering, evenly divided over cluster I and II. Importantly, most proteins in cluster I were extraplastidic, but most in cluster II were plastidic. Cytosolic, plastidic Cu,Zn-SODs, and mitochondrial MnSOD belonged to clusters I-2a, II-1, and I-1, respectively. Plastid SOD was 5- to 10-fold more abundant than extraplastidic SODs, peaked around 4.5 cm from the base, and was clearly enriched in BS strands (Figure 13A), suggesting a particularly high O2 − production during the onset of BS thylakoid formation. The cytosolic ascorbate cycle enzymes were highly expressed in the first 2 to 3 cm (cluster I) (Figure 13B), showing positive correlation with mitochondrial respiration. By contrast, plastid ascorbate peroxidase, peroxiredoxins (B, II-E, and Q), glutathione peroxidase 2, and two glutaredoxins were all part of cluster II-3c,d,e and showed a strong positive correlation with photosynthetic light reactions (Figure 13B). These correlation patterns suggest that plastid ROS is detoxified mainly within the organelle. Perhaps surprisingly, γ-glutamylcysteine synthase, a key step in synthesis of glutathione, was part of cluster I-1, suggesting that glutathione is in particular important as a redox buffer when photosynthesis is not yet operational. Figure 13. Open in new tabDownload slide Expression of Proteins Involved in ROS Detoxification and Redox Regulation along the Developmental Gradient in Leaves and BS Strands. Leaves, closed symbols; BS strands, open symbols. (A) Expression pattern of plastid Cu,Zn-SOD. (B) Cumulative accumulation of the ascorbate detoxification system within and outside of the plastid. (C) Expression of plastid thioredoxins f2, m2, m4, and plastid thioredoxin reductase, as well as extraplastidic thioredoxin h. Figure 13. Open in new tabDownload slide Expression of Proteins Involved in ROS Detoxification and Redox Regulation along the Developmental Gradient in Leaves and BS Strands. Leaves, closed symbols; BS strands, open symbols. (A) Expression pattern of plastid Cu,Zn-SOD. (B) Cumulative accumulation of the ascorbate detoxification system within and outside of the plastid. (C) Expression of plastid thioredoxins f2, m2, m4, and plastid thioredoxin reductase, as well as extraplastidic thioredoxin h. The plant thioredoxins (TRXs) consist of six distinct types, namely, the plastid-localized type f, m, x, y, and z and the extraplastidic types h and o, as well as several types of TRX reductases (Schürmann and Buchanan, 2008). We identified most of the TRX system, which generally showed high expression of plastid-localized proteins toward the tip (II-3), with clear cell-specific expression patterns for TRXm2,x,y1 (low in BS strands) and TRXf (high in BS strands; II-3d,e). Plastid TRXm4 proteins decreased from base to tip (I-1), completely opposite from the other plastid thioredoxins but similar to extraplastidic TRXh proteins (Figure 13C). Thus, thioredoxins have specialized functions within and across different (sub)cellular compartments. N, S, and Amino Acid Metabolism Shows Distinct Developmental and Differentiation Patterns More than 100 enzymes, as well as several transporters involved in amino acid, N, and S metabolism were present in our data sets. A reconstruction of the main pathways, including plots with expression kinetics in the leaf and BS strands for key enzymes, is provided in Supplemental Figures 7 and 8 online. We already referred to a number of enzymes and transporters in previous sections on photorespiration, transport, and mitochondria. For a detailed description of N, S, and amino acid metabolism, see Supplemental Results online. As expected, initial steps in N assimilation and S assimilation occurred in MC and BSC chloroplasts near the tip, respectively, consistent with our observations for purified MC and BSC chloroplasts from leaf tips (Friso et al., 2010). Glu, Gln, and Asp are the initial amino acids that are made where inorganic nitrogen is assimilated, and enzymes involved in their biosynthesis as well as N assimilation belonged to cluster II-3. This is consistent with the relatively large need for reducing equivalents in N assimilation. In S assimilation, the first steps in plastid-localized sulfate assimilation were localized in the leaf tip, nearly exclusively in the BS strand, consistent with Friso et al. (2010). However, biosynthetic enzymes for other amino acids, including incorporation of sulfide into Cys and ammonium into Arg, were found to be most abundant at the base or around the transition zone. Formation of Primary and Secondary Cell Walls and Lignification Plant cell walls consist of primary and secondary walls. Primary walls originate at cytokinesis and are further modified during cell expansion, whereas secondary walls are laid down after cessation of cell expansion. Cell wall biogenesis enzymes peaked within the first 4.5 cm, coinciding with cell elongation and cell wall thickening (Figure 2D). Strong BS strand enrichment was observed for the cellulose synthases and cell wall degradation enzymes involved in remodeling; this enrichment is consistent with the formation of the thick-walled xylem and the thick cell walls of the BS sheath cells. The expression patterns of quantified enzymes in cellulose biosynthesis and nucleotide-sugar conversions are shown in Supplemental Figure 9 online and those for lignin biosynthesis are in Supplemental Figure 10 online. Lipid and Fatty Acid Metabolism: Structural Components, Membrane Remodeling, and Signaling Molecules Excluding isoprenyl lipids (e.g., chlorophylls, quinones, and carotenoids), we identified ~131 enzymes in lipid metabolism. Enzymes for most lipid categories, a dozen unassigned GDSL-motif lipases and lipid transfer proteins, as well as enzymes involved in β-oxidation of lipids were detected. Accumulation patterns of these enzymes likely reflect cell growth and differentiation and organelle and membrane biogenesis. The largest protein mass investment (from total NadjSPC) across the leaf gradient (and BS strands) was in plastid-localized fatty acid (FA) metabolism (28%) and a group of enzymes involved in oxylipin synthesis (39%). The expression patterns for enzymes in metabolism of phospholipids, fatty acids, and wax are shown in Figure 14A and for brassinosteroids and jasmonic acid (JA) in Figure 14B. Forty proteins passed the threshold for clustering, most of them involved in FA synthesis, with 27 in cluster I and the remainder mostly in cluster II-1. Plastid-localized FA metabolism was high in the sink region and decreased toward the tip (Figure 14A). By contrast, extraplastidic FA synthesis, which included enzymes involved in very long FA chains, as well as wax synthesis enzymes mostly for generation of the cuticle, showed a clear peak expression at 4.5 cm from the base (Cluster II-1) (Figure 14A). Interestingly, as the outlier in plastid FA metabolism, acetyl-CoA carboxylase, which generates malonyl-CoA, showed a sharp peak expression at 4.5 cm (cluster II-1), suggesting a specific bottleneck for this enzymatic step (Figure 14A). Most leaf cellular membranes other than plastid membranes are enriched for phospholipids. We identified 13 enzymes in phospholipid metabolism; their accumulation levels were high in the first 2.5 cm and then decreased rapidly within the sink zone, thus showing a different expression pattern than FA metabolism, reflecting investment of phospholipids into extraplastidic membranes near the leaf base (Figure 14A). Figure 14. Open in new tabDownload slide Expression Patterns in Leaf Sections for Enzymes in Metabolism of Phospholipids, FAs, Wax, Brassinosteroids, and JA. Phospholipids, fatty acids, and wax (A) and brassinosteroids and JA (B). ACCase, acetyl-CoA carboxylase; AOS, allene oxide synthase; AOC, allene oxide cyclase; LOX, lipoxygenase; 5x and 10x, the NadjSPC values are multiplied by 5 or 10 to make the data points better visible. Figure 14. Open in new tabDownload slide Expression Patterns in Leaf Sections for Enzymes in Metabolism of Phospholipids, FAs, Wax, Brassinosteroids, and JA. Phospholipids, fatty acids, and wax (A) and brassinosteroids and JA (B). ACCase, acetyl-CoA carboxylase; AOS, allene oxide synthase; AOC, allene oxide cyclase; LOX, lipoxygenase; 5x and 10x, the NadjSPC values are multiplied by 5 or 10 to make the data points better visible. The brassinosteroids, lipoxygenase-derived oxylipins including JA, and phosphatidylinositols are important lipid regulators of leaf development and cellular differentiation. The brassinosteroid pathway showed highest expression at the base followed by rapid decrease toward the tip (Figure 14B) (cluster I-2A). This localization in the rapidly expanding and differentiating leaf zone fits well with the known stimulatory effect of brassinosteroids on cell expansion, xylem differentiation and expansion, and the suppression of phloem differentiation (Kim and Wang, 2010). JA is a lipid plant hormone derived from linolenic acid in the plastid, and it is best known for its role in defense; however, its role in leaf development or cell differentiation is not clear (Browse, 2009). We identified three enzymes (allene oxide synthase in multiple isoforms, allene oxide cyclase, and 12-oxophytodienoate reductase 3) specifically involved in JA synthesis. Only plastidic allene oxide synthase passed the threshold for clustering, and its expression was highest in the sink (cluster I-2a). Both plastid-localized lipoxygenases (LOX2) and extraplastidic lipoxygenases (LOX1,5) were quantified, with the plastid enzymes expressed higher toward the tip (cluster II-3) and the extraplastidic enzymes showing highest expression around the sink-source transition zone (cluster I-1) (Figure 14B). Furthermore, we found clear evidence for β-oxidation in the first centimeter of the leaf blade, whereas the glyoxylate cycle (e.g., peroxisomal citrate synthase) was expressed mostly between 2 and 5 cm (see Supplemental Data Set 1B online). This lack of coexpression suggests that the β-oxidation here is not coupled to the glyoxylate cycle, but rather relates to the synthesis of JA (or other oxylipins), which requires β-oxidation (Browse, 2009). We did not observe a systematic under- or overexpression in the BS strand when enzymes within pathways were summed; however, isoforms within sets of homologs frequently showed differential under- or overexpression in BS strands, indicative of cell type–specific specialization. Synthesis of Nucleic Acids, Nucleotides, Vitamins, and Cofactors Leaves also synthesize nucleic acids, nucleotides, vitamins, and other cofactors [e.g., NAD(P), FAD, and FMN], and metabolic enzymes for many of these components were identified. Thiamine biosynthesis enzymes (THI1 and THIC) were the most abundant among the vitamin metabolic enzymes, which is perhaps not surprising as it is an important cofactor in major metabolic pathways (Goyer, 2010). In Arabidopsis, there is only one THI1 gene, but its protein is dually targeted to plastids and mitochondria (Chabregas et al., 2003). Interestingly, we identified two maize homologs for THI1, one expressed highly at the base (cluster I-1) and one high in the source region (II-2), possibly localized in mitochondria and plastids, respectively. Inner plastid envelope protein MPBQ/MSBQ methyl transferase (VTE3) involved in tocopherol synthesis (Cheng et al., 2003) is strikingly enriched in BS strands (cluster II1b). This is very interesting in the light of the observed connection between tocopherol synthesis and a plasmodesmata phenotype in the sucrose export defective (sxd1) maize mutant (Provencher et al., 2001). Specifically, sxd1 showed deposition of callose at the BS-VP interface, resulting in a loss of symplasmic cell-to-cell transport across the BS-VP interface of the leaf blade. A connection between tocopherol and phloem loading was also observed in Arabidopsis (Maeda et al., 2006) and potato (Solanum tuberosum; Hofius et al., 2004). We identified 56 proteins (groups) involved in nucleic acid and nucleotide metabolism, of which 17 passed the minimal criteria for clustering mostly falling in cluster I (high expression in the base). The few enzymes that were highly expressed in the source region were involved in the breakdown of pyrophosphate into inorganic phosphate or phoshorylation of adenylates (e.g., conversions of AMP to ADP). Auxin and Trehalose The phytohormone auxin regulates many plant biological processes, including cell division, elongation, differentiation, etc. Importantly, it has a clear role in formation of the vascular system in Arabidopsis and rice and is therefore relevant for formation of the Kranz anatomy (Zhao, 2010). We identified several proteins involved in auxin regulation or signaling, but we did not observe biosynthetic enzymes. These auxin-related proteins were extraplastidic and expressed mainly in the base and middle part of the leaf, consistent with a role in maturation of the vascular system. Trehalose is a nonreducing dissacharide and is synthesized from UDP-glucose and glucose-6P by trehalose-6P-synthase and trehalose-6P-phosphatase. The precursor of trehalose, trehalose-6P (T6P), is a signaling molecule in plants with a role in coordinating carbon supply with biosynthetic process involved in growth and development (Paul et al., 2010). T6P could be a significant player in sink-source transitions and possibly C4 differentiation. We identified (albeit only with few adjSPC) trehalose-6P-synthase and trehalose-6P-phosphatase and trehalase-like proteins only in the BS strand samples mostly in the source region and not in the total leaf samples, suggesting their relative enrichment in the BS strands. It remains to be determined if T6P is a sugar critical for C4 leaf development and differentiation. Other Cellular Functions and Pathways Extraplastidic proteins involved in DNA/RNA-related functions, cellular organization (e.g., actins, tubulins, and profilins), and signaling (mostly G-proteins, calcium binding proteins, and 14-3-3 proteins) were strongly enriched at the base and belonged to cluster I-1 or I-2a (Figures 2B and 2C), without strong candidates that were under- or overexpressed in the BS strand. These proteins, as well as hundreds of proteins with miscellaneous (190 proteins) or without assigned functions (585 proteins), can be further explored through the protein expression viewer (http://ppdb.tc.cornell.edu/dbsearch/plotgradient.aspx) or by interrogating the PPDB and Supplemental Data Sets 1A and 1B online. DISCUSSION Cellular differentiation in the maize leaf occurs in a basipetal direction such that the most mature cells are at the tip of the blade and the least mature cells are at the base (Freeling, 1992; Tsiantis and Langdale, 1998). Therefore, a single individual young maize leaf blade contains all developmental, structural, and metabolic transitions, including cellular differentiation related to C4 function, as was recognized previously (Martineau and Taylor, 1985; Nelson and Langdale, 1992; Evert et al., 1996), with the exception of cell fate determination, which occurs already in the leaf meristem (Freeling, 1992; Barton, 2010). In this study, we used the developmental continuum in maize leaves of defined developmental and physiological stage to compare induction and differentiation kinetics of key C4 structural features to metabolic characteristics, biogenesis components, and organelle accumulation in maize leaves. The presented systems analysis builds on previous studies concerning many aspects of C4 leaf development and differentiation, evolution of C4 species, as well as in vitro and in vivo biochemical analysis of individual enzymes and processes (Nelson and Langdale, 1992; Sage and Monson, 1999; Buchanan et al., 2000; Winter and Huber, 2000; Edwards et al., 2001a, 2001b; Bowsher et al., 2008; Bräutigam et al., 2008; Majeran and van Wijk, 2009). We attempted to integrate as much as possible of this existing information into our analysis and interpretations. We demonstrated that the large-scale proteome analysis by spectral counting yielded meaningful results as evidenced by the coherent coexpression patterns of known markers, excellent coexpression of proteins involved in the same process, as well as good correlations between biological replicates. Furthermore, in many cases, we followed the expression patterns of processes that included cumulative information of multiple proteins, thus reducing possible noisy signals from individual proteins, in particular those of low abundance with lower number of spectral counts. For instance, the expression pattern of cytosolic translation was based on the accumulative spectral counts of 168 protein accessions (ribosomal proteins and different types of translation factors), with the vast majority showing similar accumulation patterns and clustering. The complete maize genome sequence is now available (Schnable et al., 2009); therefore, it has become possible to identify, annotate, track, and study individual members of small and large gene families. The presence of gene families is very significant in the evolution of C4, and it has been proposed that gene duplication followed by neofunctionalization and inactivation, as well as modification of cis-elements, was the predominant mechanisms in C4 evolution (Marshall et al., 1996; Monson, 2003; Sage, 2004; Westhoff and Gowik, 2004; Christin et al., 2009; Wang et al., 2009). In this discussion, we will focus on the timeline for C4 differentiation and its integration with leaf development and formation of critical cellular structures, the sink-source transition, and other aspects of metabolism; these metabolic and structural transitions are summarized in Figure 15. Furthermore, we will point out several genes/proteins that are likely making important contributions to the C4 differentiation process. We will end this section with a brief discussion on the possible drivers of C4 leaf differentiation. Figure 15. Open in new tabDownload slide Summarizing Overview of the Observed Transients along the Leaf Developmental Gradient and Differentials between the BS Strand and the MCs. The top part of the figure summarizes anatomical and structural features. The bottom part of the figure summarizes observations based on proteomics data. Processes and features in blue are prevalent in the sink part of the leaf, while those in red are prevalent in the source part of the leaf. The vertical dashed line indicates the sink-source transition point. Lines in orange and green refer to processes or proteins, respectively, that are most prominent in the sink (cluster I) and source region (cluster II). *Mitoch./plastid ratio, refers to cross-section ratio of mitochondria/plastids; SAM, S-adenosylmethionine. Figure 15. Open in new tabDownload slide Summarizing Overview of the Observed Transients along the Leaf Developmental Gradient and Differentials between the BS Strand and the MCs. The top part of the figure summarizes anatomical and structural features. The bottom part of the figure summarizes observations based on proteomics data. Processes and features in blue are prevalent in the sink part of the leaf, while those in red are prevalent in the source part of the leaf. The vertical dashed line indicates the sink-source transition point. Lines in orange and green refer to processes or proteins, respectively, that are most prominent in the sink (cluster I) and source region (cluster II). *Mitoch./plastid ratio, refers to cross-section ratio of mitochondria/plastids; SAM, S-adenosylmethionine. Resolving the Transitions in Leaf Development and C4 Differentiation: Identification of Five Key Stages To establish a functional C4 maize leaf, several structural and metabolic features need to be in place (see Introduction), and the order in which they are established is critical (Nelson and Langdale, 1992; Edwards et al., 2001a, 2001b; Bräutigam et al., 2008; Leegood, 2008). Through integration of the kinetics of the formation of several of these structural features and developmental and metabolic processes deduced from the proteome analysis, summarized in Figure 15, we identified five major transition phases: Phase 1 In the first centimeter above the ligule, PDs in the interface between future BSCs and MCs were already established, prior to maturation of phloem and xylem and prior to any buildup of photosynthetic capacity. This is logical; since most metabolic exchange between cells occurs via symplastic transport (Sowiński et al., 2008), the presence of PDs is essential to deliver sugars, amino acids, and other metabolites to the developing cells and is a prerequisite for C4 photosynthesis. Moreover, in this first centimeter, a large portion of the cellular protein content was invested into the cytosolic translational machinery (30%), as well as signaling, nuclear gene expression, and other developmental functions (15%). Enzymes in phospholipid metabolism were at their highest in the first 2 cm, most reflecting the buildup of the extraplastidic cellular membranes. The first 2 cm were completely heterotrophic, as evidenced by the lack of accumulation of the photosynthetic apparatus and maximal levels of sucrose degradation enzymes, whereas mitochondrial proteome content, including abundance of respiratory enzymes and various mitochondrial transporters, was maximal (6%) in this region of the leaf blade. Plastids were present in the form of a homogeneous population of proplastids; they accumulated transitory starch at the end of the light period, consistent with low levels of starch metabolic enzymes. Phase 2 Maturation of protoxylem and protophloem occurred during the first 3 cm of the leaf. These vascular elements are required for efficient local delivery and redistribution of reduced carbohydrates and other nutrients and therefore are a prerequisite for rapid buildup of photosynthetic capacity. The orientation of cells in a Kranz anatomy could be recognized between 2 and 3 cm, and the differences between BSC and MC plastids became visible in the form of small granal stacks in the MC plastids. However, expression levels of proteins involved in dark and light reactions of photosynthesis, as well as the various envelope transporters that coexpressed with photosynthesis, were minimal. Moreover, differential accumulation of proteins between BS strand and total leaf was observed for only a small fraction of identified proteins; these were in particular sink transporters enriched in BS strands (e.g., GLT1-1 and PPT1-2). Phase 3 Between 3 and 5 cm, the plastids were rapidly building up their translational apparatus as well as other components needed for proteome homeostasis. Indeed, the contribution of the total cellular proteome invested in plastid proteome biogenesis reached maximal levels at the 4.5-cm point. By 5 cm, cell elongation was nearly completed, but cell wall thickening was still in progress; consistent with this, accumulation of cellulose synthases and enzymes involved in wax and extraplastidic FA synthesis peaked at 4.5 cm. Importantly, BSC and MC C4 differentiation began in earnest; this was much less obvious in the first 4 cm, mostly because investment in the classical C4 functions was really very low. Indeed, only after 4 cm from the base did we observed a rapid accumulation of the photosynthetic machinery, the C4 malate-pyruvate shuttle, and the envelope transporters involved in export of triosephosphates (TPT1) as well as BSC-enriched source transporters. In agreement with this, induction of photorespiratory enzymes started at 4 cm from the base. Thus, around 4 to 5 cm, when the tissue emerged into direct light, no longer wrapped inside the older leaves, the tissue strongly accelerated its transition from heterotropic to autotropic tissue, with rapid buildup of the photosynthetic apparatus (both dark and light reactions) and completion of cell expansion. Accumulation of plastid isoprenoid and tetrapyrrole biosynthetic enzymes and regulators sharply peaked between 4 and 5 cm. The results clearly suggest that the cells put the major phase of plastid expansion on hold until about the 4-cm point, after which plastid development and differentiation is unleashed. Phase 4 In the zone between 5 and 8 cm, cell elongation was completed by 6 cm, closely followed by completion of the gas-impermeable cell wall around the BSCs. The cell wall provided not only the necessary structural support but also reduced losses of CO2 in advance of completion of the buildup of photosynthetic capacity and the C4 cycle. Furthermore, specialization between BSCs and MCs became more and more pronounced, with the differentiation process continuing to the tip of the leaf. Both from the 14C labeling and the accumulation of increasing levels of photosynthetic enzymes and sucrose synthesis (reaching ~50% of the maximum relative concentration around 7 cm) and the strong reduction of sucrose degradation enzymes, it was clear that the tissue was now mostly autotrophic. Phase 5 Between 8 cm and the leaf tip, the cells continued building the photosynthetic apparatus and completing C4 differentiation. Notably, expansion of BSC chloroplasts and accumulation of thylakoid membranes slowed down at 8 cm, whereas MC chloroplasts continued to expand until the leaf tip. Total accumulation levels of chloroplast biogenesis functions gradually decreased toward the tip; however, specific proteins remained high in either BSC or MC chloroplasts, suggesting involvement in the C4 differentiation process. Examples are stromal protease PreP1 and ClpC chaperones in BS strands. Several ROS defense components, in particular those involved with H2O2 detoxification and plastid redox regulators (e.g., thioredoxin f2 and m2), increased toward the tip, reflecting their important role in plastid metabolism. Collectively, the image and proteome analysis show that vascular maturation is complete prior to significant induction of C4 photosynthetic development and differentiation. The proteome analysis showed dramatic shifts in protein investments among the developmental leaf gradient. The proteome profiles and image analysis showed that these investments served in particular to enable plastid division and biogenesis, as well as formation of cellular structures (e.g., cell walls and cellular membranes). Thus, our results point to highly coordinated metabolic (from heterotrophic to autotrophic) and structural transitions along the leaf gradient. Of course the underlying regulatory network driving and coordinating these transitions remains to be determined (see below), but this study provides a framework to associate potential regulators to developmental and metabolic transitions. Mitochondria clearly are not static organelles, and their relative importance in cellular metabolism changed along the sink-source transition. Moreover, mitochondria make different contributions to the BS strand and the rest of the leaf, in particular due to induction of photorespiration in the BSCs (Thompson et al., 1998; Bowsher and Tobin, 2001). A fascinating and unresolved question is what determines the dominance of mitochondria in the early sink tissue and in the vascular bundle along the leaf gradient? BSC and MC Chloroplast Development and C4 Differentiation The development of proplastids into chloroplasts is still poorly understood, not only in maize and other grasses, but also in dicotyledons such as Arabidopsis, even if many genes have been shown to be important for chloroplast biogenesis (Vothknecht and Westhoff, 2001; Stern et al., 2004; Lopez-Juez and Pyke, 2005; López-Juez, 2007; Kessler and Schnell, 2009; Waters and Langdale, 2009). The proplastid-to-chloroplast transition involves different signaling pathways, import of nuclear-encoded proteins, expression of plastid-encoded genes, and coordinated synthesis or import of many cofactors, including carotenoids, chlorophylls, quinones, and metals. Understanding how all these processes are coordinated is a huge challenge, and systems biology at different molecular levels (i.e., transcripts, proteins, metabolites, and inorganic compounds) combined with protein interaction network analysis may be able to resolve this complex developmental program in the long term. Systems-level analysis of maize chloroplast biogenesis has the additional complexity of dimophorphic chloroplasts due to C4 differentiation. Many prior studies on maize plastid development focused on the major photosynthetic complexes and enzymes using protein immunoblots, pulse labeling, native gel electrophoresis, and various activity assays for O2 evolution, CO2 fixation, and chlorophyll accumulation (Martineau and Taylor, 1985; Schuster et al., 1985; Bassi and Simpson, 1986; Oswald et al., 1990; Kubicki et al., 1994; Bassi et al., 1995). Many of these studies were specifically discussed in our previous proteomics studies on isolated, fully developed BSC and MC chloroplasts (Majeran et al., 2005, 2008; Friso et al., 2010). Collectively, these studies certainly provide a general overview of chloroplast development along the leaf, but the spatial resolution along the leaf gradient was relatively low, and none of these studies was able to measure the development of the various functions of chloroplast simultaneously. Due to the dramatic improvement of proteomics and mass spectrometry technologies, as well as the availability of the sequenced maize genome, we were able to obtain a systems view of plastid biogenesis along the developmental gradient of the leaf and of the BS strand. Importantly, we observed two types of behavior for plastid proteins: (1) plastid proteins did not significantly accumulate until after the 4 cm point; this applies to the majority of proteins of the photosynthetic apparatus, irrespective of sample type; and (2) plastid proteins accumulated in the first 4 cm to significant levels; these were in particular proteins involved in chloroplast biogenesis. This suggests that at the initial stage of chloroplast development in the leaf blade, most of the biogenesis was focused on establishing the basic functions of a chloroplast, rather than specific cell-type differentiation related to C4 functions. Investigation of early differentially expressed plastid-localized proteins showed mostly proteins of unknown function or proteins involved in RNA metabolism or translation (e.g., ribosomal proteins). The microscopy analysis showed that around 2 to 3 cm, the first sign of C4 plastid differentiation in BSCs and MCs occurred, indicating that a signaling pathway that drives C4 differentiation must be active. This region of the leaf is heterotrophic, not yet exposed to direct light (still wrapped in older leaves), whereas maturation of the xylem and phloem has reached completion. We did not identify specific transcription factors or other regulatory proteins that are obvious candidates for regulation of C4 differentiation. The fact that maturation of the vascular system coincides with the first signs of C4 differentiation would be compatible with a mobile signal, such as sugars. A more general discussion on signaling is provided below. One of the most striking examples of coordinated differential plastid protein accumulation was found within the thylakoid membrane system, namely, the differentiated expression of PSII and the NDH complex, involved in respectively linear and cyclic electron flow (Takabayashi et al., 2005). This is particularly interesting as it involved dozens of both plastid-encoded and nuclear-encoded proteins. Within the first 4 cm, the ratio between the NDH and PSII complex was identical in the BS strand and in the whole leaf. By contrast, during leaf development toward the tip, the PSII/NDH ratio in the BS strand decreased ~4-fold, whereas the NDH/CF ratio increase 2-fold. This shows that PSII accumulation is negatively regulated, whereas NDH is upregulated during development of chloroplasts in the BS strand. Since this involves both plastid- and nuclear-encoded proteins, it is likely that transcriptional (transcription and transcript stability), translational, and posttranslational (assembly and proteolysis) regulation is involved. Moreover, plastid-nuclear signaling is likely to contribute to coordination of both genomes. This regulatory network is a tangible target for future study. Transporter Discovery and Expression Patterns High fluxes of metabolic transport are required to support C4 photosynthesis, but C4 chloroplast transporters are poorly understood (Majeran and van Wijk, 2009; Weber and von Caemmerer, 2010). We determined the expression patterns of many chloroplast metabolite transporters, including several for which the substrates are unknown despite the fact that these transporters are highly expressed. Tight coexpression with BSC marker proteins suggested at least three BS strand-enriched transporters (MEP1, 2, and 4) likely involved in C4 photosynthesis, the malate-pyruvate shuttle, and photorespiration. Functional characterization in vivo and in vitro is urgently needed. Moreover, we identified several proplastid envelope transporters in the sink region of the leaf that also await further characterization. Generally, our results for the differential expression patterns at the leaf tip were consistent with our previous analysis of isolated BSC and MC chloroplasts (Friso et al., 2010). However, for a number of envelope proteins, such as MEP1,2,3 and PPT1-1, we observed strongly enriched BS strand accumulation (compared with the respective leaf section), whereas they were enriched in isolated MC chloroplasts compared with BSC chloroplasts (Friso et al., 2010). Since the data sets in these two independent studies both appeared robust, a possible explanation is that the nonphotosynthetic plastids within the vascular bundle (Williams, 1974; Walsh and Evert, 1975) do make a significant contribution to these transporters (but not to many other functions, such as photosynthesis). Clearly, our observations warrant a functional analysis of these mysterious nongreen plastids within the vascular bundle. Moreover, future analysis of the plastid envelope transporters must also consider their function within the vascular bundle. Drivers and Coordinators/Integrators of Leaf Development and the C4 Differentiation Plant growth, development, and differentiation must depend on signaling systems that provide information on internal and external conditions. Information on the status of cellular metabolites such as sugars and amino acids is of crucial importance, and plants possess sophisticated signaling systems that respond to metabolite concentrations (Smeekens et al., 2010). In the base of the leaf, there must be strong competition for carbon and nitrogen, and clearly most of these resources are used for nuclear gene expression and formation of cellular membranes and cell walls. It is known that the availability of sugars is a driver of growth and that sugars act both as substrates for metabolism and as signaling molecules. Several signaling pathways are now known in plants that promote growth and include the hexokinase (HXK) glucose sensing, T6P signaling, and the Target of Rapmycin (TOR) kinase system (Breuninger and Lenhard, 2010; Smeekens et al., 2010). The sucrose nonfermenting-1 related protein kinase family (SnRK1) also plays several important roles in metabolic regulation in plants, including regulation of carbon and amino acid metabolism (Halford and Hey, 2009). The C/S1 bZIP transcription factor network plays a role inhibiting growth. In this study, we identified HXK1 and discussed its enzymatic function, several enzymes in trehalose metabolism, an upstream activator of the TOR system, as well as three SnRK1-interacting proteins, but no bZIP transcription factors. RMZM2G164547_P01 was by far the most abundant of the three SnRK1-interacting proteins; its expression peaked 2 to 3 cm from the base but was relatively low in the BS strands. Whereas its precise function is unknown, this expression pattern does correlate well with the early sink-source transitions. Translationally controlled tumor protein (TCTP) is a ubiquitous, surprisingly abundant, upstream activator of the TOR protein kinase signaling pathway, the major regulator of cell growth in animals and fungi (Bommer and Thiele, 2004). Silencing of TCTP in Arabidopsis slows vegetative growth, causes reduced leaf expansion because of smaller cell size, and has an effect on auxin homeostasis (Berkowitz et al., 2008). We identified two abundant maize TCTP homologs (GRMZM2G075624_P01 and GRMZM2G108474_P01) that decreased from base to tip, without any specific under- or overexpression in the BS strand (cluster I-2a), which is consistent with its suggested role in regulating cellular growth. Genetic inactivation of TCTP in maize could provide insight into the relationship between cell growth and C4 differentiation. Thus, our maize proteome analysis identified several different components of growth regulation; these identified components could be the starting point for investigating how cellular growth and C4 differentiation are coupled. Identification of drivers of C4 differentiation is the key for truly understanding C4 biology and for engineering C4 pathways. Given that C4 photosynthesis evolved independently several dozen times (Sage, 2004), it has been suggested that upstream master switches (e.g., transcription factors) could be in place that initiate a cascade of events to establish the features of the C4 syndrome. An example for such a transcription factor appeared to be GOLDEN2 (g2) or GLK2, which is a Myb family member; in maize, loss of G2 function specifically disrupted photosynthetic development in BSCs, and, indeed, G2 was expressed in BSCs (Cribb et al., 2001; Rossini et al., 2001). Whereas this transcription factor appears to be a BS-specific transcription factor that activates multiple nuclear photosynthetic genes, it does not appear to be a master regulator responsible for C4 differentiation, since it is expressed in the source zone, late in chloroplast development. Indeed, so far, no master regulator has been identified and no mutants have (yet) been discovered that are consistent with such master regulators. Even if master transcription factors and common C4 cis-elements were discovered, they would explain only part of the C4 syndrome as multiple conditions are required for efficient C4 photosynthesis (an extensive discussion was presented in Sage, 2004). In addition to the anatomical changes (e.g., increased PD density and/or cross section between the BSC and MC interface and decreased vein spacing), modification of the properties of specific enzymes (e.g., change of K m, redox properties, and phosphorylation sites) are also necessary, for instance (1) to cope with the much higher cellular concentrations of malate and 3-PGA, which will either overstimulate enzyme activities or inhibit enzyme activities by allosteric control or (2) to establish sufficiently large concentration gradients between BSCs and MCs to facilitate sufficient metabolic flux by diffusion through the symplastic transport system and (3) to relocate proteins to different subcellular locations and introduce cell type–specific expression through modified promoter regions. For instance, C4-PEPC (generating OAA as part of the malate-pyruvate shuttle) is less sensitive to malate inhibition than is C3-PEPC (involved in generation of OAA for the TCA cycle), while sensitivity to G6P is enhanced (Svensson et al., 2003). In a number of cases, the new C4 function of proteins must be present in addition to the original C3-type function; in such cases, gene duplications are necessary. Examples are NADP-MDH and PPDK, each with a MC-specific C4 form and a BS-specific C3 form (see Friso et al., 2010 and references therein). Finally, additional coordination of expression or induction of C4 development/differentiation has been suggested to occur by mobile metabolite signals (e.g., various sugars) from the vascular system and/or the BSCs; this argument is strengthened by the observation that MCs appear to lose MC-type expression patterns as judged by limited transcript analysis (e.g., RBCS and PPDK) (not by activity measurements) if they are not directly positioned next to BSCs, such as in the husk leaves (Ewing et al., 1998; Taniguchi et al., 2000). Our proteomics data set provides expression data for some 4300 proteins; in many cases, proteins are represented by one or more homologs. In a number of cases, homologs show differential developmental expression patterns. Based on these differential expression patterns, one can speculate on the differential functions (e.g., sink versus source activity) within these groups of homologs; these can be investigated in silico for the presence of regulatory residues (e.g., predicted phosphorylation sites, predicted redox-active Cys residues, etc.) and tested in vitro for differential activities (e.g., substrate specificity of transporters) or enzyme kinetic properties (K m, k cat, and V max). Such detailed analyses will ultimately be needed to truly understand C4 metabolism and make appropriate choices for metabolic engineering of C4 features, in, for example, the C4 rice projects (Matsuoka et al., 2001; Hibberd et al., 2008; Taniguchi et al., 2008; Weber and von Caemmerer, 2010; Zhu et al., 2010). Resources for the Plant Community The large-scale LM and TEM image collection and maize proteome information, which provided the basis for this study, also constitute a valuable new resource to explore additional questions that are beyond the focus of this article. For instance, the LM image analysis will allow calculation of cellular and some organellar dimensions, as well as intracellular spaces along the leaf gradient that will be needed to calculate better diffusion rates in flux models of, for example, photosynthesis and cell–cell exchange of metabolites and inorganic molecules (e.g., CO2) (Jenkins et al., 1989; Farquhar et al., 2001; Zhu et al., 2007; Chen et al., 2008), as well as functional-structural or three-dimensional virtual plant modeling (Vos et al., 2010). We functionally annotated >4300 identified maize proteins and assigned many to a subcellular localization (in particular, plastids, mitochondria, peroxisomes, and vacuoles) based on experimental data for maize proteins, in combination with information for their best homologs in Arabidopsis, sorghum (Sorghum bicolor), rice, and others. Our maize protein annotations will be transferred to the maize community database MaizeGDB (http://www.maizegdb.org/) to contribute to the functional annotation of the maize genome and is also available via Supplemental Data Sets 1A and 1B online and the PPDB. Last but not least, the more than 7.5 million high-resolution MS/MS spectra obtained in this study constitute the largest collection of maize proteome data to date. Together with our previous set of 5.5 million MS/MS spectra of the proteome of isolated maize chloroplasts from the leaf tip (Friso et al., 2010), this will allow improved annotation of maize gene models and discovery of unannotated genes (summarized under the term proteogenomics) (Ansong et al., 2008; Armengaud, 2009; Castellana and Bafna, 2010) as well as detection of posttranslation modifications, such as N-terminal acetylation, trimethylation of Lys residues (in particular for histones and the translational machineries), and mRNA editing (Zybailov et al., 2009a). Conclusions The analysis presented in this study provides a well-defined molecular template for the structural and metabolic transitions that occur during C4 leaf differentiation. The molecular template presented in this study will allow targeting of specific developmental transitions in the search for regulators. In this study, we also identified components in several signaling networks, such as TCTP in the TOR kinase system and several SnRK1-interacting proteins. These could be used as starting point to address regulation of cell growth and how this is intertwines with C4 differentiation. Inspection of chloroplast proteases in our recent BSC and MC chloroplast analyses (Majeran et al., 2008; Friso et al., 2010) and this study show several proteases that are differentially expressed and could play a role in C4 differentiation. One of the most studied proteins in maize carbon metabolism is sucrose synthase (SUS), primarily involved in sucrose degradation in sink tissues (Huber and Hardin, 2004). Consistently, SUS showed a very pronounced expression profile in our data set, peaking at the base (0 to 3 cm) and undetectable near the tip. The pronounced loss of SUS protein must be the result of regulated degradation, possibly through the proteasome (Hardin and Huber, 2004), and is an example of how proteolysis is needed in leaf development. Finally, the LM and TEM images and the large proteomics data sets with their functional annotation provide a resource for the plant community. METHODS Plant Growth, 14C-Labeling, LM, TEM, and Image Analysis Maize (Zea mays) plants (MO16 and B73) were grown for 9 d in a growth chamber (12 h light/12 h dark, 400 μmol photons·m−2·s−1), 31°C (day)/22°C (night). The 3rd leaves were harvested and separated from their plant at their ligule, 2 to 3 h after the onset of the light period. One set of leaves were harvested 2 h before the end of the light period to study starch localization by TEM analysis. For radiolabel transport experiments, 14CO2 was injected into a cuvette enclosing leaf 2 or the tip of leaf 3. After 15 min of labeling, followed by a 45-min chase, leaf 3 was frozen in dry ice, lyophilized, and autoradiographed with x-ray film (Turgeon and Medville, 1998). For LM and TEM, tissue was fixed in glutaraldehyde/paraformaldehyde and embedded in Spurr resin (Reidel et al., 2009). ImageJ software (http://rsbweb.nih.gov/ij/index.html) was used to aid in image analysis and quantification. Samples for Proteome Analysis For the total leaf proteome, one individual leaf was selected per biological replicate to maximize the spatial resolution. Each leaf was sectioned into 1-cm segments, and the total cellular proteome was rapidly extracted in presence of SDS and protease inhibitors. For the total BS strand proteome, four sections were cut from the 3rd leaf: 0 to 1.5 cm, 2.5 to 4 cm, 4 to 5 cm, and 12 to 13 cm; sections from ~100 leaves were pooled. BS strands were then isolated, and contamination from MCs was assessed as described by Majeran et al. (2005). Equal amounts of protein extracts for all sections from each leaf or BS strand series were separated by SDS-PAGE gels (one gel per biological replicate) and stained by Coomassie Brilliant Blue R 250. For analysis of soluble BS strand proteome as a pilot study, maize inbred W22 was grown in a growth chamber for 12 d (16 h light/8 h dark, 400 μmol photons·m–2·s–1, and constant 28°C). For the soluble BS strand proteome, three sections were cut from each 3rd leaf as follows: the first 2 cm from the ligule, followed by a 1.5-cm gap, then the next 2 cm of the greening section, and the fully differentiated last 3 cm of the tip. Soluble proteins of the isolated BS strands were extracted by grinding in liquid nitrogen and the extraction buffer supplemented with a cocktail of protease inhibitors. Membrane and soluble proteins were separated using ultracentrifugation at 55,000 rpm for 20 min. The soluble proteins from the supernatant were then concentrated. For quantitative proteome analysis, equal amounts of proteins (200 μg) from each section were separated by one-dimensional Tricine gels. Gels were stained with Coomassie Brilliant Blue R 250. Proteome Analysis by NanoLC-LTQ-Orbitrap Each gel lane was then cut in eight slices (leaf), 10 slices (BS strand), or 11 to 14 slices (soluble BS strand). Proteins were digested with trypsin, and the extracted peptides were analyzed by nanoLC-LTQ-Orbitrap MS using data dependent acquisition and dynamic exclusion, as described by Friso et al. (2010). The complete analysis was performed in two (leaf and total BS strands) or three (soluble BS strand proteins) independent biological replicates. In total, 293 MS runs were performed, with extensive blanks between each sample analysis to avoid carryover of peptides that could bias quantification. Processing of the MS Data, Database Searches, Quantification of Identified Proteins, and Data Submission to PPDB and PRIDE Peak lists (mgf format) were generated using DTA supercharge (v1.19) software (http://msquant.sourceforge.net/) and searched with Mascot v2.2 (Matrix Science) against maize genome release 4a.53 (with 53,764 models) from http://www.maizesequence.org/ supplemented with the plastid-encoded proteins (111 protein models) and mitochondrial-encoded proteins (165 protein models). Details for calibration and control of false positive rate can be found in Friso et al. (2010). MS-based information of all identified proteins was extracted from the Mascot search pages and filtered for significance (e.g., minimum ion scores, etc), ambiguities, and shared spectra as described by Friso et al. (2010). All filtered results were uploaded into the PPDB (http://ppdb.tc.cornell.edu/; Sun et al., 2009). Identified proteins were further processed and quantified as described by Friso et al. (2010). A summary of the complete workflow is displayed in Figure 3. Expression patterns of all identified proteins can be generated explored through the protein expression viewer (http://ppdb.tc.cornell.edu/dbsearch/plotgradient.aspx). All MS data (the mgf files reformatted as PRIDE XML files) were uploaded in the Proteomics Identifications database (PRIDE) at http://www.ebi.ac.uk/pride/. Hierarchical Cluster Analysis To group proteins with similar accumulation pattern across the leaf and BS strand sections, hierarchical clustering was employed using the statistics toolbox of MATLAB version 7 (Mathworks) as described by Olinares et al. (2010). Only those 1044 proteins that had at least an average of four adjSPC/sections across both biological replicates in the leaf (total adjSPC/protein ≥24 adjSPC) and two adjSPC/section across both biological replicates in the BS strand (total adjSPC/protein ≥8 adjSPC) were included. Accession Numbers Accession numbers for sequence data from this article can be found in Supplemental Data Sets 1A and 1B online. All MS data (the mgf files reformatted as PRIDE XML files) are available via PRIDE at http://www.ebi.ac.uk/pride/ under accession numbers 10979-11094, 11259-11338, and 11578-11673. Supplemental Data The following materials are available in the online version of this article. Supplemental Figure 1. Representative Seedling and 14C Labeling of Maize Leaves to Show Sink-Source Transition in the 3rd Leaf Selected for Proteome and Image Analysis. Supplemental Figure 2. Structural Analysis of the Maize Leaf along the Developmental Gradient. Supplemental Figure 3. Markers of BS Strands. Supplemental Figure 4. Plastid Envelope Transporters Involved in Transport of Copper, S-Adenosylmethionine, or Unknown Substrates Passing the Minimum Threshold for Clustering. Supplemental Figure 5. Mitochondrial Transporters Passing the Minimum Threshold for Clustering. Supplemental Figure 6. Accumulation Patterns of Unusual LHC Proteins, LIL3, ELIP1/2, SEP4, and OHP2-Like, along the Leaf Developmental Gradient. Supplemental Figure 7. The Expression Patterns of Enzymes Involved in N-Assimilation. Supplemental Figure 8. The Expression Patterns of Enzymes Involved in S-Assimilation and the SAM Cycle. Supplemental Figure 9. The Expression Patterns of Quantified Enzymes in Cellulose Biosynthesis and Nucleotide-Sugar Conversions Supplemental Figure 10. The Expression Patterns of Enzymes Involved in Lignin Biosynthesis. Supplemental Results. N, S, and Amino Acid Metabolism Show Distinct Developmental and Differentiation Patterns. Supplemental Data Set 1A. Protein Identification, Their Annotated Function, and Location. Supplemental Data Set 1B. Protein Quantification. Supplemental Data Set 1C. List of Full Names and Abbreviations of Proteins Mentioned in the Text or Figures. Acknowledgments This work was supported by a grant to K.J.V.W., B.T., and Q.S. from the National Science Foundation (PGRP-0701736). We thank Richard Medville at Electron Microscopy Services for the excellent microscopy analysis. We thank Tim Nelson and Tom Brutnell for discussions. References 1. Ansong C. Purvine S.O. Adkins J.N. Lipton M.S. Smith R.D. ( 2008 ). Proteogenomics: Needs and roles to be filled by proteomics in genome annotation . Brief. Funct. Genomics Proteomics 7 : 50 – 62 . Google Scholar Crossref Search ADS WorldCat 2. Armengaud J. ( 2009 ). A perfect genome annotation is within reach with the proteomics and genomics alliance . Curr. Opin. Microbiol. 12 : 292 – 300 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Bäckman H.G. Pessoa J. Eneqvist T. Glaser E. ( 2009 ). Binding of divalent cations is essential for the activity of the organellar peptidasome in Arabidopsis thaliana, AtPreP . FEBS Lett. 583 : 2727 – 2733 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Barton M.K. ( 2010 ). Twenty years on: The inner workings of the shoot apical meristem, a developmental dynamo . Dev. Biol. 341 : 95 – 113 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Bassi R. Marquardt J. Lavergne J. ( 1995 ). Biochemical and functional properties of photosystem II in agranal membranes from maize mesophyll and bundle sheath chloroplasts . Eur. J. Biochem. 233 : 709 – 719 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Bassi R. Simpson D.J. ( 1986 ). Differential expression of LHCII genes in mesophyll and bundle sheath cells of maize . Carlsberg Res. Commun. 51 : 363 – 370 . Google Scholar Crossref Search ADS WorldCat 7. Belacel N. Wang Q. Cuperlovic-Culf M. ( 2006 ). Clustering methods for microarray gene expression data . OMICS 10 : 507 – 531 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Berkowitz O. Jost R. Pollmann S. Masle J. ( 2008 ). Characterization of TCTP, the translationally controlled tumor protein, from Arabidopsis thaliana . Plant Cell 20 : 3430 – 3447 . Google Scholar Crossref Search ADS PubMed WorldCat 9. Bommer U.A. Thiele B.J. ( 2004 ). The translationally controlled tumour protein (TCTP) . Int. J. Biochem. Cell Biol. 36 : 379 – 385 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Bowsher C. Steer M. Tobin A. ( 2008 ). Plant Biochemistry . ( New York : Garland Science ). Google Scholar 11. Bowsher C.G. Tobin A.K. ( 2001 ). Compartmentation of metabolism within mitochondria and plastids . J. Exp. Bot. 52 : 513 – 527 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Bräutigam A. Hoffmann-Benning S. Hofmann-Benning S. Weber A.P. ( 2008 ). Comparative proteomics of chloroplast envelopes from C3 and C4 plants reveals specific adaptations of the plastid envelope to C4 photosynthesis and candidate proteins required for maintaining C4 metabolite fluxes . Plant Physiol. 148 : 568 – 579 . Erratum. Plant Physiol. 148: 1734 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Bräutigam A. et al. . ( 2010 ). An mRNA blueprint for C4 photosynthesis derived from comparative transcriptomics of closely related C3 and C4 species . Plant Physiol. 155 : (in press) . Google Scholar OpenURL Placeholder Text WorldCat 14. Bréhélin C. Kessler F. van Wijk K.J. ( 2007 ). Plastoglobules: versatile lipoprotein particles in plastids . Trends Plant Sci. 12 : 260 – 266 . Google Scholar Crossref Search ADS PubMed WorldCat 15. Breuninger H. Lenhard M. ( 2010 ). Control of tissue and organ growth in plants . Curr. Top. Dev. Biol. 91 : 185 – 220 . Google Scholar Crossref Search ADS PubMed WorldCat 16. Browse J. ( 2009 ). Jasmonate passes muster: A receptor and targets for the defense hormone . Annu. Rev. Plant Biol. 60 : 183 – 205 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Buchanan B. Gruissem W. Jones R.L. ( 2000 ). Biochemistry and Molcular Biology of Plants . ( Rockville, MD : American Society of Plant Physiologists ). Google Scholar 18. Carpaneto A. Geiger D. Bamberg E. Sauer N. Fromm J. Hedrich R. ( 2005 ). Phloem-localized, proton-coupled sucrose carrier ZmSUT1 mediates sucrose efflux under the control of the sucrose gradient and the proton motive force . J. Biol. Chem. 280 : 21437 – 21443 . Google Scholar Crossref Search ADS PubMed WorldCat 19. Carpita N.C. McCann M.C. ( 2008 ). Maize and sorghum: Genetic resources for bioenergy grasses . Trends Plant Sci. 13 : 415 – 420 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Castellana N. Bafna V. ( 2010 ). Proteogenomics to discover the full coding content of genomes: A computational perspective . J. Proteomics 73 : 2124 – 2135 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Chabregas S.M. Luche D.D. Van Sluys M.A. Menck C.F. Silva-Filho M.C. ( 2003 ). Differential usage of two in-frame translational start codons regulates subcellular localization of Arabidopsis thaliana THI1 . J. Cell Sci. 116 : 285 – 291 . Google Scholar Crossref Search ADS PubMed WorldCat 22. Chen C.P. Zhu X.G. Long S.P. ( 2008 ). The effect of leaf-level spatial variability in photosynthetic capacity on biochemical parameter estimates using the Farquhar model: A theoretical analysis . Plant Physiol. 148 : 1139 – 1147 . Google Scholar Crossref Search ADS PubMed WorldCat 23. Chen Z.H. Walker R.P. Técsi L.I. Lea P.J. Leegood R.C. ( 2004 ). Phosphoenolpyruvate carboxykinase in cucumber plants is increased both by ammonium and by acidification, and is present in the phloem . Planta 219 : 48 – 58 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Cheng W.H. Im K.H. Chourey P.S. ( 1996 ). Sucrose phosphate synthase expression at the cell and tissue level is coordinated with sucrose sink-to-source transitions in maize leaf . Plant Physiol. 111 : 1021 – 1029 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Cheng Z. Sattler S. Maeda H. Sakuragi Y. Bryant D.A. DellaPenna D. ( 2003 ). Highly divergent methyltransferases catalyze a conserved reaction in tocopherol and plastoquinone synthesis in cyanobacteria and photosynthetic eukaryotes . Plant Cell 15 : 2343 – 2356 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Chow W.S. Kim E.H. Horton P. Anderson J.M. ( 2005 ). Granal stacking of thylakoid membranes in higher plant chloroplasts: The physicochemical forces at work and the functional consequences that ensue . Photochem. Photobiol. Sci. 4 : 1081 – 1090 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Christin P.A. Samaritani E. Petitpierre B. Salamin N. Besnard G. ( 2009 ). Evolutionary insights on C4 photosynthetic subtypes in grasses from genomics and phylogenetics . Genome Biol. Evol. 1 : 221 – 230 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Covshoff S. Majeran W. Liu P. Kolkman J.M. van Wijk K.J. Brutnell T.P. ( 2008 ). Deregulation of maize C4 photosynthetic development in a mesophyll cell-defective mutant . Plant Physiol. 146 : 1469 – 1481 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Cribb L. Hall L.N. Langdale J.A. ( 2001 ). Four mutant alleles elucidate the role of the G2 protein in the development of C4 and C3 photosynthesizing maize tissues . Genetics 159 : 787 – 797 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 30. Dong M. Yang L.L. Williams K. Fisher S.J. Hall S.C. Biggin M.D. Jin J. Witkowska H.E. ( 2008 ). A “tagless” strategy for identification of stable protein complexes genome-wide by multidimensional orthogonal chromatographic separation and iTRAQ reagent tracking . J. Proteome Res. 7 : 1836 – 1849 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Edwards G.E. Franceschi V.R. Ku M.S. Voznesenskaya E.V. Pyankov V.I. Andreo C.S. ( 2001b ). Compartmentation of photosynthesis in cells and tissues of C4 plants . J. Exp. Bot. 52 : 577 – 590 . Google Scholar OpenURL Placeholder Text WorldCat 32. Edwards G.E. Furbank R.T. Hatch M.D. Osmond C.B. ( 2001a ). What does it take to be C4? Lessons from the evolution of C4 photosynthesis . Plant Physiol. 125 : 46 – 49 . Google Scholar Crossref Search ADS WorldCat 33. Eisen M.B. Spellman P.T. Brown P.O. Botstein D. ( 1998 ). Cluster analysis and display of genome-wide expression patterns . Proc. Natl. Acad. Sci. USA 95 : 14863 – 14868 . Google Scholar Crossref Search ADS WorldCat 34. Engelken J. Brinkmann H. Adamska I. ( 2010 ). Taxonomic distribution and origins of the extended LHC (light-harvesting complex) antenna protein superfamily . BMC Evol. Biol. 10 : 233 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Evert R.F. Russin W.A. Bosabalidis A.M. ( 1996 ). Anatomical and ultrastructural changes associated with sink-to-source transition in developing maize leaves . Int. J. Plant Sci. 157 : 247 – 261 . Google Scholar Crossref Search ADS WorldCat 36. Ewing R.M. Jenkins G.I. Langdale J.A. ( 1998 ). Transcripts of maize RbcS genes accumulate differentially in C3 and C4 tissues . Plant Mol. Biol. 36 : 593 – 599 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Farquhar G.D. von Caemmerer S. Berry J.A. ( 2001 ). Models of photosynthesis . Plant Physiol. 125 : 42 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Freeling M. ( 1992 ). A conceptual framework for maize leaf development . Dev. Biol. 153 : 44 – 58 . Google Scholar Crossref Search ADS PubMed WorldCat 39. Friso G. Majeran W. Huang M. Sun Q. van Wijk K.J. ( 2010 ). Reconstruction of metabolic pathways, protein expression, and homeostasis machineries across maize bundle sheath and mesophyll chloroplasts: Large-scale quantitative proteomics using the first maize genome assembly . Plant Physiol. 152 : 1219 – 1250 . Google Scholar Crossref Search ADS PubMed WorldCat 40. Glaser E. Nilsson S. Bhushan S. ( 2006 ). Two novel mitochondrial and chloroplastic targeting-peptide-degrading peptidasomes in A. thaliana, AtPreP1 and AtPreP2 . Biol. Chem. 387 : 1441 – 1447 . Google Scholar Crossref Search ADS PubMed WorldCat 41. Goyer A. ( 2010 ). Thiamine in plants: Aspects of its metabolism and functions . Phytochemistry 71 : 1615 – 1624 . Google Scholar Crossref Search ADS PubMed WorldCat 42. Grimm B. Kloppstech K. ( 1987 ). The early light-inducible proteins of barley. Characterization of two families of 2-h-specific nuclear-coded chloroplast proteins . Eur. J. Biochem. 167 : 493 – 499 . Google Scholar Crossref Search ADS PubMed WorldCat 43. Halford N.G. Hey S.J. ( 2009 ). Snf1-related protein kinases (SnRKs) act within an intricate network that links metabolic and stress signalling in plants . Biochem. J. 419 : 247 – 259 . Google Scholar Crossref Search ADS PubMed WorldCat 44. Hardin S.C. Huber S.C. ( 2004 ). Proteasome activity and the post-translational control of sucrose synthase stability in maize leaves . Plant Physiol. Biochem. 42 : 197 – 208 . Google Scholar Crossref Search ADS PubMed WorldCat 45. Hibberd J.M. Sheehy J.E. Langdale J.A. ( 2008 ). Using C4 photosynthesis to increase the yield of rice-rationale and feasibility . Curr. Opin. Plant Biol. 11 : 228 – 231 . Google Scholar Crossref Search ADS PubMed WorldCat 46. Hofius D. Hajirezaei M.R. Geiger M. Tschiersch H. Melzer M. Sonnewald U. ( 2004 ). RNAi-mediated tocopherol deficiency impairs photoassimilate export in transgenic potato plants . Plant Physiol. 135 : 1256 – 1268 . Google Scholar Crossref Search ADS PubMed WorldCat 47. Huang S. Taylor N.L. Narsai R. Eubel H. Whelan J. Millar A.H. ( 2009 ). Experimental analysis of the rice mitochondrial proteome, its biogenesis, and heterogeneity . Plant Physiol. 149 : 719 – 734 . Google Scholar Crossref Search ADS PubMed WorldCat 48. Huber S.C. Hardin S.C. ( 2004 ). Numerous posttranslational modifications provide opportunities for the intricate regulation of metabolic enzymes at multiple levels . Curr. Opin. Plant Biol. 7 : 318 – 322 . Google Scholar Crossref Search ADS PubMed WorldCat 49. Hummel M. Rahmani F. Smeekens S. Hanson J. ( 2009 ). Sucrose-mediated translational control . Ann. Bot. (Lond.) 104 : 1 – 7 . Google Scholar Crossref Search ADS WorldCat 50. Hutin C. Nussaume L. Moise N. Moya I. Kloppstech K. Havaux M. ( 2003 ). Early light-induced proteins protect Arabidopsis from photooxidative stress . Proc. Natl. Acad. Sci. USA 100 : 4921 – 4926 . Google Scholar Crossref Search ADS WorldCat 51. Inoue H. Higuchi K. Takahashi M. Nakanishi H. Mori S. Nishizawa N.K. ( 2003 ). Three rice nicotianamine synthase genes, OsNAS1, OsNAS2, and OsNAS3 are expressed in cells involved in long-distance transport of iron and differentially regulated by iron . Plant J. 36 : 366 – 381 . Google Scholar Crossref Search ADS PubMed WorldCat 52. Jenkins C.L. Furbank R.T. Hatch M.D. ( 1989 ). Mechanism of c(4) photosynthesis: a model describing the inorganic carbon pool in bundle sheath cells . Plant Physiol. 91 : 1372 – 1381 . Google Scholar Crossref Search ADS PubMed WorldCat 53. Kanervo E. Singh M. Suorsa M. Paakkarinen V. Aro E. Battchikova N. Aro E.M. ( 2008 ). Expression of protein complexes and individual proteins upon transition of etioplasts to chloroplasts in pea (Pisum sativum) . Plant Cell Physiol. 49 : 396 – 410 . Google Scholar Crossref Search ADS PubMed WorldCat 54. Kapri-Pardes E. Naveh L. Adam Z. ( 2007 ). The thylakoid lumen protease Deg1 is involved in the repair of photosystem II from photoinhibition in Arabidopsis . Plant Cell 19 : 1039 – 1047 . Google Scholar Crossref Search ADS PubMed WorldCat 55. Kato Y. Miura E. Ido K. Ifuku K. Sakamoto W. ( 2009 ). The variegated mutants lacking chloroplastic FtsHs are defective in D1 degradation and accumulate reactive oxygen species . Plant Physiol. 151 : 1790 – 1801 . Google Scholar Crossref Search ADS PubMed WorldCat 56. Kato Y. Sakamoto W. ( 2009 ). Protein quality control in chloroplasts: A current model of D1 protein degradation in the photosystem II repair cycle . J. Biochem. 146 : 463 – 469 . Google Scholar Crossref Search ADS PubMed WorldCat 57. Kato Y. Sakamoto W. ( 2010 ). New insights into the types and function of proteases in plastids . Int. Rev. Cell Mol. Biol. 280 : 185 – 218 . Google Scholar Crossref Search ADS PubMed WorldCat 58. Kessler F. Schnell D. ( 2009 ). Chloroplast biogenesis: Diversity and regulation of the protein import apparatus . Curr. Opin. Cell Biol. 21 : 494 – 500 . Google Scholar Crossref Search ADS PubMed WorldCat 59. Kim J. Rudella A. Ramirez Rodriguez V. Zybailov B. Olinares P.D. van Wijk K.J. ( 2009 ). Subunits of the plastid ClpPR protease complex have differential contributions to embryogenesis, plastid biogenesis, and plant development in Arabidopsis . Plant Cell 21 : 1669 – 1692 . Google Scholar Crossref Search ADS PubMed WorldCat 60. Kim J.Y. Mahé A. Guy S. Brangeon J. Roche O. Chourey P.S. Prioul J.L. ( 2000 ). Characterization of two members of the maize gene family, Incw3 and Incw4, encoding cell-wall invertases . Gene 245 : 89 – 102 . Google Scholar Crossref Search ADS PubMed WorldCat 61. Kim T.W. Wang Z.Y. ( 2010 ). Brassinosteroid signal transduction from receptor kinases to transcription factors . Annu. Rev. Plant Biol. 61 : 681 – 704 . Google Scholar Crossref Search ADS PubMed WorldCat 62. Kubicki A. Steinmüller K. Westhoff P. ( 1994 ). Differential transcription of plastome-encoded genes in the mesophyll and bundle-sheath chloroplasts of the monocotyledonous NADP-malic enzyme-type C4 plants maize and Sorghum . Plant Mol. Biol. 25 : 669 – 679 . Google Scholar Crossref Search ADS PubMed WorldCat 63. Kusano T. Tateda C. Berberich T. Takahashi Y. ( 2009 ). Voltage-dependent anion channels: Their roles in plant defense and cell death . Plant Cell Rep. 28 : 1301 – 1308 . Google Scholar Crossref Search ADS PubMed WorldCat 64. Laby R.J. Kim D. Gibson S.I. ( 2001 ). The ram1 mutant of Arabidopsis exhibits severely decreased beta-amylase activity . Plant Physiol. 127 : 1798 – 1807 . Google Scholar Crossref Search ADS PubMed WorldCat 65. Lalonde S. Wipf D. Frommer W.B. ( 2004 ). Transport mechanisms for organic forms of carbon and nitrogen between source and sink . Annu. Rev. Plant Biol. 55 : 341 – 372 . Google Scholar Crossref Search ADS PubMed WorldCat 66. Leegood R.C. ( 2008 ). C4 photosynthesis: minor or major adjustment to a C3 theme? In Charting New Pathways to C4 Rice , Sheehy J.E. Mitchell P.L. Hardy B. , eds ( London : World Scientific ), pp. 81 – 94 . Google Scholar 67. Leegood R.C. Walker R.P. ( 2003 ). Regulation and roles of phosphoenolpyruvate carboxykinase in plants . Arch. Biochem. Biophys. 414 : 204 – 210 . Google Scholar Crossref Search ADS PubMed WorldCat 68. Linka N. Weber A.P. ( 2010 ). Intracellular metabolite transporters in plants . Mol. Plant 3 : 21 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat 69. Liu H. Sadygov R.G. Yates J.R. III ( 2004 ). A model for random sampling and estimation of relative protein abundance in shotgun proteomics . Anal. Chem. 76 : 4193 – 4201 . Google Scholar Crossref Search ADS PubMed WorldCat 70. Liu X. Yu F. Rodermel S. ( 2010 ). Arabidopsis chloroplast FtsH, var2 and suppressors of var2 leaf variegation: A review . J. Integr. Plant Biol. 52 : 750 – 761 . Google Scholar Crossref Search ADS PubMed WorldCat 71. Livingston A.K. Cruz J.A. Kohzuma K. Dhingra A. Kramer D.M. ( 2010 ). An Arabidopsis mutant with high cyclic electron flow around photosystem I (hcef) involving the NADPH dehydrogenase complex . Plant Cell 22 : 221 – 233 . Google Scholar Crossref Search ADS PubMed WorldCat 72. Long T.A. Brady S.M. Benfey P.N. ( 2008 ). Systems approaches to identifying gene regulatory networks in plants . Annu. Rev. Cell Dev. Biol. 24 : 81 – 103 . Google Scholar Crossref Search ADS PubMed WorldCat 73. López-Juez E. ( 2007 ). Plastid biogenesis, between light and shadows . J. Exp. Bot. 58 : 11 – 26 . Google Scholar Crossref Search ADS PubMed WorldCat 74. Lopez-Juez E. Pyke K.A. ( 2005 ). Plastids unleashed: Their development and their integration in plant development . Int. J. Dev. Biol. 49 : 557 – 577 . Google Scholar Crossref Search ADS PubMed WorldCat 75. Lu P. Vogel C. Wang R. Yao X. Marcotte E.M. ( 2007 ). Absolute protein expression profiling estimates the relative contributions of transcriptional and translational regulation . Nat. Biotechnol. 25 : 117 – 124 . Google Scholar Crossref Search ADS PubMed WorldCat 76. Lunn J.E. Furbank R.T. ( 1997 ). Localisation of sucrose-phosphate synthase and starch in leaves of C4 plants . Planta 202 : 106 – 111 . Google Scholar Crossref Search ADS PubMed WorldCat 77. Maeda H. Song W. Sage T.L. DellaPenna D. ( 2006 ). Tocopherols play a crucial role in low-temperature adaptation and phloem loading in Arabidopsis . Plant Cell 18 : 2710 – 2732 . Google Scholar Crossref Search ADS PubMed WorldCat 78. Majeran W. Cai Y. Sun Q. van Wijk K.J. ( 2005 ). Functional differentiation of bundle sheath and mesophyll maize chloroplasts determined by comparative proteomics . Plant Cell 17 : 3111 – 3140 . Google Scholar Crossref Search ADS PubMed WorldCat 79. Majeran W. van Wijk K.J. ( 2009 ). Cell-type-specific differentiation of chloroplasts in C4 plants . Trends Plant Sci. 14 : 100 – 109 . Google Scholar Crossref Search ADS PubMed WorldCat 80. Majeran W. Zybailov B. Ytterberg A.J. Dunsmore J. Sun Q. van Wijk K.J. ( 2008 ). Consequences of C4 differentiation for chloroplast membrane proteomes in maize mesophyll and bundle sheath cells . Mol. Cell. Proteomics 7 : 1609 – 1638 . Google Scholar Crossref Search ADS PubMed WorldCat 81. Marshall J.S. Stubbs J.D. Taylor W.C. ( 1996 ). Two genes encode highly similar chloroplastic NADP-malic enzymes in Flaveria. Implications for the evolution of C4 photosynthesis . Plant Physiol. 111 : 1251 – 1261 . Google Scholar Crossref Search ADS PubMed WorldCat 82. Martineau B. Taylor W.C. ( 1985 ). Photosynthetic gene expression and cellular differentiation in developing maize leaves . Plant Physiol. 78 : 399 – 404 . Google Scholar Crossref Search ADS PubMed WorldCat 83. Matsuoka M. Furbank R.T. Fukayama H. Miyao M. ( 2001 ). Molecular engineering of C4 photosynthesis . Annu. Rev. Plant Physiol. Plant Mol. Biol. 52 : 297 – 314 . Google Scholar Crossref Search ADS PubMed WorldCat 84. Monson R. ( 2003 ). Gene duplication, neofunctionalization, and the evolution of C4 photosynthesis . Int. J. Plant Sci. 164 : S43 – S54 . Google Scholar Crossref Search ADS WorldCat 85. Nelson T. Langdale J. ( 1992 ). Developmental genetics of C4 photosynthesis . Annu. Rev. Plant Physiol. Plant Mol. Biol. 43 : 25 – 47 . Google Scholar Crossref Search ADS WorldCat 86. Nilsson Cederholm S. Bäckman H.G. Pesaresi P. Leister D. Glaser E. ( 2009 ). Deletion of an organellar peptidasome PreP affects early development in Arabidopsis thaliana . Plant Mol. Biol. 71 : 497 – 508 . Google Scholar Crossref Search ADS PubMed WorldCat 87. Ohsugi R. Huber S.C. ( 1987 ). Light modulation and localization of sucrose phosphate synthase activity between mesophyll cells and bundle sheath cells in C(4) species . Plant Physiol. 84 : 1096 – 1101 . Google Scholar Crossref Search ADS PubMed WorldCat 88. Old W.M. Meyer-Arendt K. Aveline-Wolf L. Pierce K.G. Mendoza A. Sevinsky J.R. Resing K.A. Ahn N.G. ( 2005 ). Comparison of label-free methods for quantifying human proteins by shotgun proteomics . Mol. Cell. Proteomics 4 : 1487 – 1502 . Google Scholar Crossref Search ADS PubMed WorldCat 89. Olinares P.D. Ponnala L. van Wijk K.J. ( 2010 ). Megadalton complexes in the chloroplast stroma of Arabidopsis thaliana characterized by size exclusion chromatography, mass spectrometry and hierarchical clustering . Mol. Cell. Proteomics 9 : 1594 – 1615 . Google Scholar Crossref Search ADS PubMed WorldCat 90. Oswald A. Streubel M. Ljungberg U. Hermans J. Eskins K. Westhoff P. ( 1990 ). Differential biogenesis of photosystem-II in mesophyll and bundle-sheath cells of ‘malic’ enzyme NADP(+)-type C4 plants. A comparative protein and RNA analysis . Eur. J. Biochem. 190 : 185 – 194 . Google Scholar Crossref Search ADS PubMed WorldCat 91. Paul M.J. Jhurreea D. Zhang Y. Primavesi L.F. Delatte T. Schluepmann H. Wingler A. ( 2010 ). Upregulation of biosynthetic processes associated with growth by trehalose 6-phosphate . Plant Signal. Behav. 5 : 386 – 392 . Google Scholar Crossref Search ADS PubMed WorldCat 92. Peltier J.B. Ripoll D.R. Friso G. Rudella A. Cai Y. Ytterberg J. Giacomelli L. Pillardy J. van Wijk K.J. ( 2004 ). Clp protease complexes from photosynthetic and non-photosynthetic plastids and mitochondria of plants, their predicted three-dimensional structures, and functional implications . J. Biol. Chem. 279 : 4768 – 4781 . Google Scholar Crossref Search ADS PubMed WorldCat 93. Pontén F. et al. . ( 2009 ). A global view of protein expression in human cells, tissues, and organs . Mol. Syst. Biol. 5 : 337 . Google Scholar Crossref Search ADS PubMed WorldCat 94. Provencher L.M. Miao L. Sinha N. Lucas W.J. ( 2001 ). Sucrose export defective1 encodes a novel protein implicated in chloroplast-to-nucleus signaling . Plant Cell 13 : 1127 – 1141 . Google Scholar Crossref Search ADS PubMed WorldCat 95. Quintana L.F. Campistol J.M. Alcolea M.P. Bañon-Maneus E. Sol-González A. Cutillas P.R. ( 2009 ). Application of label-free quantitative peptidomics for the identification of urinary biomarkers of kidney chronic allograft dysfunction . Mol. Cell. Proteomics 8 : 1658 – 1673 . Google Scholar Crossref Search ADS PubMed WorldCat 96. Reidel E.J. Rennie E.A. Amiard V. Cheng L. Turgeon R. ( 2009 ). Phloem loading strategies in three plant species that transport sugar alcohols . Plant Physiol. 149 : 1601 – 1608 . Google Scholar Crossref Search ADS PubMed WorldCat 97. Rolland F. Baena-Gonzalez E. Sheen J. ( 2006 ). Sugar sensing and signaling in plants: Conserved and novel mechanisms . Annu. Rev. Plant Biol. 57 : 675 – 709 . Google Scholar Crossref Search ADS PubMed WorldCat 98. Rossini L. Cribb L. Martin D.J. Langdale J.A. ( 2001 ). The maize golden2 gene defines a novel class of transcriptional regulators in plants . Plant Cell 13 : 1231 – 1244 . Google Scholar Crossref Search ADS PubMed WorldCat 99. Rudella A. Friso G. Alonso J.M. Ecker J.R. van Wijk K.J. ( 2006 ). Downregulation of ClpR2 leads to reduced accumulation of the ClpPRS protease complex and defects in chloroplast biogenesis in Arabidopsis . Plant Cell 18 : 1704 – 1721 . Google Scholar Crossref Search ADS PubMed WorldCat 100. Rumeau D. Peltier G. Cournac L. ( 2007 ). Chlororespiration and cyclic electron flow around PSI during photosynthesis and plant stress response . Plant Cell Environ. 30 : 1041 – 1051 . Google Scholar Crossref Search ADS PubMed WorldCat 101. Rylott E.L. Metzlaff K. Rawsthorne S. ( 1998 ). Developmental and environmental effects on the expression of the C3-C4 intermediate phenotype in moricandia arvensis . Plant Physiol. 118 : 1277 – 1284 . Google Scholar Crossref Search ADS PubMed WorldCat 102. Sage R.F. ( 2004 ). The evolution of C4 photosynthesis . New Phytol. 161 : 341 – 370 . Google Scholar Crossref Search ADS WorldCat 103. Sage R.F. Monson R.K. ( 1999 ). C4 Plant Biology . ( New York : Academic Press ). Google Scholar 104. Sakamoto W. ( 2006 ). Protein degradation machineries in plastids . Annu. Rev. Plant Biol. 57 : 599 – 621 . Google Scholar Crossref Search ADS PubMed WorldCat 105. Schnable P.S. et al. . ( 2009 ). The B73 maize genome: Complexity, diversity, and dynamics . Science 326 : 1112 – 1115 . Google Scholar Crossref Search ADS PubMed WorldCat 106. Schürmann P. Buchanan B.B. ( 2008 ). The ferredoxin/thioredoxin system of oxygenic photosynthesis . Antioxid. Redox Signal. 10 : 1235 – 1274 . Google Scholar Crossref Search ADS PubMed WorldCat 107. Schuster G. Ohad I. Martineau B. Taylor W.C. ( 1985 ). Differentiation and development of bundle sheath and mesophyll thylakoids in maize. Thylakoid polypeptide composition, phosphorylation, and organization of photosystem II . J. Biol. Chem. 260 : 11866 – 11873 . Google Scholar Crossref Search ADS PubMed WorldCat 108. Sheen J. ( 1999 ). C4 gene expression . Annu. Rev. Plant Physiol. Plant Mol. Biol. 50 : 187 – 217 . Google Scholar Crossref Search ADS PubMed WorldCat 109. Sjögren L.L. Stanne T.M. Zheng B. Sutinen S. Clarke A.K. ( 2006 ). Structural and functional insights into the chloroplast ATP-dependent Clp protease in Arabidopsis . Plant Cell 18 : 2635 – 2649 . Google Scholar Crossref Search ADS PubMed WorldCat 110. Slewinski T.L. Meeley R. Braun D.M. ( 2009 ). Sucrose transporter1 functions in phloem loading in maize leaves . J. Exp. Bot. 60 : 881 – 892 . Google Scholar Crossref Search ADS PubMed WorldCat 111. Smeekens S. Ma J. Hanson J. Rolland F. ( 2010 ). Sugar signals and molecular networks controlling plant growth . Curr. Opin. Plant Biol. 13 : 274 – 279 . Google Scholar Crossref Search ADS PubMed WorldCat 112. Sowiński P. Szczepanik J. Minchin P.E. ( 2008 ). On the mechanism of C4 photosynthesis intermediate exchange between Kranz mesophyll and bundle sheath cells in grasses . J. Exp. Bot. 59 : 1137 – 1147 . Google Scholar Crossref Search ADS PubMed WorldCat 113. Stern D.B. Hanson M.R. Barkan A. ( 2004 ). Genetics and genomics of chloroplast biogenesis: Maize as a model system . Trends Plant Sci. 9 : 293 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat 114. Sun Q. Zybailov B. Majeran W. Friso G. Olinares P.D. van Wijk K.J. ( 2009 ). PPDB, the Plant Proteomics Database at Cornell . Nucleic Acids Res. 37 ( Database issue ): D969 – D974 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 115. Sun X. Ouyang M. Guo J. Ma J. Lu C. Adam Z. Zhang L. ( 2010 ). The thylakoid protease Deg1 is involved in photosystem-II assembly in Arabidopsis thaliana . Plant J. 62 : 240 – 249 . Google Scholar Crossref Search ADS PubMed WorldCat 116. Svensson P. Bläsing O.E. Westhoff P. ( 2003 ). Evolution of C4 phosphoenolpyruvate carboxylase . Arch. Biochem. Biophys. 414 : 180 – 188 . Google Scholar Crossref Search ADS PubMed WorldCat 117. Takabayashi A. Kishine M. Asada K. Endo T. Sato F. ( 2005 ). Differential use of two cyclic electron flows around photosystem I for driving CO2-concentration mechanism in C4 photosynthesis . Proc. Natl. Acad. Sci. USA 102 : 16898 – 16903 . Google Scholar Crossref Search ADS WorldCat 118. Tanaka R. Rothbart M. Oka S. Takabayashi A. Takahashi K. Shibata M. Myouga F. Motohashi R. Shinozaki K. Grimm B. Tanaka A. ( 2010 ). LIL3, a light-harvesting-like protein, plays an essential role in chlorophyll and tocopherol biosynthesis . Proc. Natl. Acad. Sci. USA 107 : 16721 – 16725 . Google Scholar Crossref Search ADS WorldCat 119. Taniguchi Y. Ohkawa H. Masumoto C. Fukuda T. Tamai T. Lee K. Sudoh S. Tsuchida H. Sasaki H. Fukayama H. Miyao M. ( 2008 ). Overproduction of C4 photosynthetic enzymes in transgenic rice plants: an approach to introduce the C4-like photosynthetic pathway into rice . J. Exp. Bot. 59 : 1799 – 1809 . Google Scholar Crossref Search ADS PubMed WorldCat 120. Taniguchi M. Izawa K. Ku M.S. Lin J.H. Saito H. Ishida Y. Ohta S. Komari T. Matsuoka M. Sugiyama T. ( 2000 ). The promoter for the maize C4 pyruvate, orthophosphate dikinase gene directs cell- and tissue-specific transcription in transgenic maize plants . Plant Cell Physiol. 41 : 42 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat 121. Thimm O. Bläsing O. Gibon Y. Nagel A. Meyer S. Krüger P. Selbig J. Müller L.A. Rhee S.Y. Stitt M. ( 2004 ). MAPMAN: A user-driven tool to display genomics data sets onto diagrams of metabolic pathways and other biological processes . Plant J. 37 : 914 – 939 . Google Scholar Crossref Search ADS PubMed WorldCat 122. Thompson P. Bowsher C.G. Tobin A.K. ( 1998 ). Heterogeneity of mitochondrial protein biogenesis during primary leaf development in barley . Plant Physiol. 118 : 1089 – 1099 . Google Scholar Crossref Search ADS PubMed WorldCat 123. Tsiantis M. Langdale J.A. ( 1998 ). The formation of leaves . Curr. Opin. Plant Biol. 1 : 43 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat 124. Turgeon R. ( 1989 ). The sink-source transition in leaves . Annu. Rev. Plant Physiol. Plant Mol. Biol. 40 : 119 – 138 . Google Scholar Crossref Search ADS WorldCat 125. Turgeon R. Medville R. ( 1998 ). The absence of phloem loading in willow leaves . Proc. Natl. Acad. Sci. USA 95 : 12055 – 12060 . Google Scholar Crossref Search ADS WorldCat 126. Usuda H. Edwards G.E. ( 1980 ). Localization of glycerate kinase and some enzymes for sucrose synthesis in c(3) and c(4) plants . Plant Physiol. 65 : 1017 – 1022 . Google Scholar Crossref Search ADS PubMed WorldCat 127. Vierstra R.D. ( 2009 ). The ubiquitin-26S proteasome system at the nexus of plant biology . Nat. Rev. Mol. Cell Biol. 10 : 385 – 397 . Google Scholar Crossref Search ADS PubMed WorldCat 128. Vos J. Evers J.B. Buck-Sorlin G.H. Andrieu B. Chelle M. de Visser P.H. ( 2010 ). Functional-structural plant modelling: A new versatile tool in crop science . J. Exp. Bot. 61 : 2101 – 2115 . Google Scholar Crossref Search ADS PubMed WorldCat 129. Vothknecht U.C. Westhoff P. ( 2001 ). Biogenesis and origin of thylakoid membranes . Biochim. Biophys. Acta 1541 : 91 – 101 . Google Scholar Crossref Search ADS PubMed WorldCat 130. Walsh A. Evert R.F. ( 1975 ). Ultrastructure of metaphloem sieve elements in Zea mays . Protoplasma 83 : 365 – 388 . Google Scholar Crossref Search ADS WorldCat 131. Wang X. Gowik U. Tang H. Bowers J.E. Westhoff P. Paterson A.H. ( 2009 ). Comparative genomic analysis of C4 photosynthetic pathway evolution in grasses . Genome Biol. 10 : R68 . Google Scholar Crossref Search ADS PubMed WorldCat 132. Waters M.T. Langdale J.A. ( 2009 ). The making of a chloroplast . EMBO J. 28 : 2861 – 2873 . Google Scholar Crossref Search ADS PubMed WorldCat 133. Weber A.P. von Caemmerer S. ( 2010 ). Plastid transport and metabolism of C3 and C4 plants—Comparative analysis and possible biotechnological exploitation . Curr. Opin. Plant Biol. 13 : 257 – 265 . Google Scholar Crossref Search ADS PubMed WorldCat 134. Westhoff P. Gowik U. ( 2004 ). Evolution of C4 phosphoenolpyruvate carboxylase. Genes and proteins: a case study with the genus Flaveria . Ann. Bot. (Lond.) 93 : 13 – 23 . Google Scholar Crossref Search ADS WorldCat 135. Williams E. ( 1974 ). Fine structure of vascular and epidermal plastids of the mature maize leaf . Protoplasma 79 : 395 – 400 . Google Scholar Crossref Search ADS WorldCat 136. Wingler A. Walker R.P. Chen Z.H. Leegood R.C. ( 1999 ). Phosphoenolpyruvate carboxykinase is involved in the decarboxylation of aspartate in the bundle sheath of maize . Plant Physiol. 120 : 539 – 546 . Google Scholar Crossref Search ADS PubMed WorldCat 137. Winter H. Huber S.C. ( 2000 ). Regulation of sucrose metabolism in higher plants: Localization and regulation of activity of key enzymes . Crit. Rev. Biochem. Mol. Biol. 35 : 253 – 289 . Google Scholar Crossref Search ADS PubMed WorldCat 138. Wise R.R. ( 2006 ). The diversity of plastid form and function. In The Structure and Function of Plastids , Wise R.R. Hoober J.K. , eds ( Dordrecht, The Netherlands : Springer ), pp. 3 – 26 . Google Scholar 139. Wormit A. Trentmann O. Feifer I. Lohr C. Tjaden J. Meyer S. Schmidt U. Martinoia E. Neuhaus H.E. ( 2006 ). Molecular identification and physiological characterization of a novel monosaccharide transporter from Arabidopsis involved in vacuolar sugar transport . Plant Cell 18 : 3476 – 3490 . Google Scholar Crossref Search ADS PubMed WorldCat 140. Zelitch I. Schultes N.P. Peterson R.B. Brown P. Brutnell T.P. ( 2009 ). High glycolate oxidase activity is required for survival of maize in normal air . Plant Physiol. 149 : 195 – 204 . Google Scholar Crossref Search ADS PubMed WorldCat 141. Zhao Y. ( 2010 ). Auxin biosynthesis and its role in plant development . Annu. Rev. Plant Biol. 61 : 49 – 64 . Google Scholar Crossref Search ADS PubMed WorldCat 142. Zhu X.G. de Sturler E. Long S.P. ( 2007 ). Optimizing the distribution of resources between enzymes of carbon metabolism can dramatically increase photosynthetic rate: A numerical simulation using an evolutionary algorithm . Plant Physiol. 145 : 513 – 526 . Google Scholar Crossref Search ADS PubMed WorldCat 143. Zhu X.G. Shan L. Wang Y. Quick W.P. ( 2010 ). C4 rice - An ideal arena for systems biology research . J. Integr. Plant Biol. 52 : 762 – 770 . Google Scholar Crossref Search ADS PubMed WorldCat 144. Zybailov B. Coleman M.K. Florens L. Washburn M.P. ( 2005 ). Correlation of relative abundance ratios derived from peptide ion chromatograms and spectrum counting for quantitative proteomic analysis using stable isotope labeling . Anal. Chem. 77 : 6218 – 6224 . Google Scholar Crossref Search ADS PubMed WorldCat 145. Zybailov B. Friso G. Kim J. Rudella A. Rodríguez V.R. Asakura Y. Sun Q. van Wijk K.J. ( 2009b ). Large scale comparative proteomics of a chloroplast Clp protease mutant reveals folding stress, altered protein homeostasis, and feedback regulation of metabolism . Mol. Cell. Proteomics 8 : 1789 – 1810 . Google Scholar Crossref Search ADS WorldCat 146. Zybailov B. Sun Q. van Wijk K.J. ( 2009a ). Workflow for large scale detection and validation of peptide modifications by RPLC-LTQ-Orbitrap: Application to the Arabidopsis thaliana leaf proteome and an online modified peptide library . Anal. Chem. 81 : 8015 – 8024 . Google Scholar Crossref Search ADS WorldCat Author notes 1 These authors contributed equally to this work. 2 Current address: Institut des Sciences du Vegetal, UPR 2355, Centre National de la Recherche Scientifique, Avenue de la terrasse 23, F-91198 Gif-sur-Yvette Cedex, France. 3 Address correspondence to kv35@cornell.edu. The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantcell.org) is: Klaas J. van Wijk (kv35@cornell.edu). Online version contains Web-only data. www.plantcell.org/cgi/doi/10.1105/tpc.110.079764 © 2010 American Society of Plant Biologists This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Structural and Metabolic Transitions of C4 Leaf Development and Differentiation Defined by Microscopy and Quantitative Proteomics in Maize
JF - The Plant Cell
DO - 10.1105/tpc.110.079764
DA - 2010-12-23
UR - https://www.deepdyve.com/lp/oxford-university-press/structural-and-metabolic-transitions-of-c4-leaf-development-and-21VyjMmzBO
SP - 3509
EP - 3542
VL - 22
IS - 11
DP - DeepDyve
ER -