Insights Into the Regulation of the Expression Pattern of Calvin-Benson-Bassham Cycle Enzymes in C₃ and C₄ Grasses

Abstract

C₄ photosynthesis is characterized by the compartmentalization of the processes of atmospheric uptake of CO₂ and its conversion into carbohydrate between mesophyll and bundle-sheath cells. As a result, most of the enzymes participating in the Calvin-Benson-Bassham (CBB) cycle, including RubisCO, are highly expressed in bundle-sheath cells. There is evidence that changes in the regulatory sequences of RubisCO contribute to its bundle-sheath-specific expression, however, little is known about how the spatial-expression pattern of other CBB cycle enzymes is regulated. In this study, we use a computational approach to scan for transcription factor binding sites in the regulatory regions of the genes encoding CBB cycle enzymes, SBPase, FBPase, PRK, and GAPDH-B, of C₃ and C₄ grasses. We identified potential cis-regulatory elements present in each of the genes studied here, regardless of the photosynthetic path used by the plant. The trans-acting factors that bind these elements have been validated in A. thaliana and might regulate the expression of the genes encoding CBB cycle enzymes. In addition, we also found C₄-specific transcription factor binding sites in the genes encoding CBB cycle enzymes that could potentially contribute to the pathway-specific regulation of gene expression. These results provide a foundation for the functional analysis of the differences in regulation of genes encoding CBB cycle enzymes between C₃ and C₄ grasses.

Introduction

C₄ plants achieve higher photosynthetic efficiency by concentrating CO₂ around RubisCO. In contrast with enzymes participating in C₃ photosynthesis, C₄-enzymes are compartmentalized to specific cell types, namely mesophyll (M) and bundle-sheath (BS) cells. Enzymes enriched in M cells include phosphoenolpyruvate carboxylase (PEPC) and pyruvate orthophosphate dikinase (Ppdk), whereas decarboxylating malic enzymes (NAD or NADP-Me) and RubisCO are enriched in the BS cells (Sheen and Bogorad, 1987; Hibberd and Covshoff, 2010; Berry et al., 2011).

During C₄ evolution a change in localization of the enzymes involved in CO₂ assimilation resulted in the compartmentalization of these reactions in either the M or BS cell types. A number of regulatory elements conferring a M or BS specific expression pattern have been identified in the regulatory sequences of the genes encoding PEPC, Ppdk; or NADP-ME, NAD-ME, and RubisCO (Nomura et al., 2000; Berry et al., 2011; Williams et al., 2016; Reyna-Llorens et al., 2018). To further interrogate those regulatory elements, a combination of comparative transcriptomics to identify differential expression of genes (Bräutigam et al., 2011; Aubry et al., 2014; Xu et al., 2016) and DNAse-seq to map differences in open chromatin regions between M and BS cells (Burgess et al., 2019) have been used. These studies have led to the identification of putative cis-regulatory elements and the trans-acting transcription factors binding to those elements, and have shown that the motifs conferring differences in expression in the C₄ species have been recruited from pre-existing sequences in C₃ species, rather than being generated de novo during the evolution of the C₄ condition (Niklaus and Kelly, 2019).

Calvin Benson-Bassham (CBB) cycle enzymes, including RubisCO, are expressed in both C₃ and C₄ species. Similar to RubisCO, most of the CBB cycle enzymes are enriched in BS cells in C₄ species (Sheen and Bogorad, 1987; John et al., 2014; Rao et al., 2016). Unlike RubisCO, little is known about the changes in the regulatory sequences of the other 10 genes encoding CCB cycle enzymes that enable such compartmentalization, limiting our ability to develop strategies to manipulate this pathway to improve photosynthetic efficiency. Here, we present a bioinformatics analysis of the regulatory sequences of genes encoding CBB cycle enzymes with the aim of identifying regulatory elements that are common to C₃ and C₄ species, or C₄-specific regulatory elements that control photosynthesis and contribute to C₄ compartmentalization. We selected four of the CBB cycle enzymes known to be redox-regulated by the ferredoxin/thioredoxin (Fd/TRX) system (Michelet et al., 2013) and that function exclusively in the CBB cycle: SBPase, FBPase (chloroplastic variant), PRK and GAPDH-B. Given the numerous independent origins of C₄ photosynthesis that might have led to parallel evolution of cis-regulatory elements (Sage et al., 2012), in this paper we focus on a small subset of eight grasses from the Poaceae family whose genomes have been sequenced and annotated.

In this study we have identified putative regulatory elements that are common in both C₃ and C₄ species as well as C₄-specific elements. We have also used existing data to explore the expression patterns of the trans-acting factors that have been shown or proposed to bind to these elements, suggesting a possible role in the compartmentalization of CBB cycle enzymes in C₄ plants. The results presented here provide the basis for future functional studies.

Materials and Methods

DNA Sequences

Genomic sequences encoding CBB cycle enzymes of Oryza sativa (Ouyang et al., 2007), Hordeum vulgare (Beier et al., 2017; Mascher et al., 2017), Brachypodium distachyon (International Brachypodium Initiative, 2010), Zea mays (Schnable et al., 2009; Hirsch et al., 2016), Sorghum bicolor (McCormick et al., 2018), Setaria viridis (v2.1, DOE-JGI)¹ and Panicum virgatum (v1.0, DOE-JGI, see footnote) were obtained from Phytozome12 (Goodstein et al., 2011). Arabidopsis thaliana genes (AT3G55800—SBPase, AT3G54050—chlFBPase, AT1G32060—PRK, and AT1G42970—GAPDHB) were used to identify orthologs in every species. For the genomic sequences encoding CBB cycle enzymes of Dichanthelium oligosanthes (Studer et al., 2016), the A. thaliana coding sequences were aligned against the D. oligosanthes genome using BLAST (Altschul et al., 1990) to find orthologous genes. Sequences used are included in Supplementary Material.

Motif Prediction in Conserved Non-coding Sequences (CNS)

Genomic sequences were aligned using mVISTA (Frazer et al., 2004) and aligned CNSs were used as input for motif prediction using MEME (v5.1.1; Bailey et al., 2009). Motif site distribution was set to zoops and maximum motif width to the size of the shorter CNS. Predicted motifs were used as input in FIMO (Grant et al., 2011) to scan the regulatory sequences of orthologous genes in other species.

Motif Scanning of Genomic Sequences

A collection of 529 plant transcription factor motifs validated in A. thaliana (O’Malley et al., 2016) were used to scan for motifs using FIMO (Grant et al., 2011) with default parameters.

Data Processing and Visualization

Data processing and visualization were performed using R 3.6.0 (R Core Team, 2019). The dplyr package (Wickham et al., 2019) was used to filter the identified motifs by q < 0.05, genomic feature, and by species. The UpSetR package (Gehlenborg, 2019) was used to generate Figure 2A showing all possible interactions; and the ggplot2 (Wickham, 2016) and the gggenes packages (Wilkins, 2019) were used to generate Figure 2B.

Transcriptomics Analysis

Transcriptomic data from RNAseq experiments in which mesophyll and bundle sheath cells were separated in P. virgatum (Rao et al., 2016), S. viridis (John et al., 2014), Panicum hallii (Washburn et al., 2017), and Setaria italica (Washburn et al., 2017) were obtained from NCBI (BioProject accession numbers: PRJNA293441, PRJEB5074, PRJNA475365). A classification-based quantification was performed using kallisto (Bray et al., 2016) with the transcriptomes and genome annotation obtained from Phytozome 12 (Goodstein et al., 2011; Bennetzen et al., 2012). In short, a kallisto index was built with the reference transcriptome of each species, and kallisto quant was used to quantify abundance of pair-end reads with default parameters. Differential expression analysis was performed with R packages DESeq (Anders and Huber, 2010) using estimateSizeFactors, estimateispersions and nbinomTest functions; DESeq2 (Love et al., 2014) using DESeq function, and edgeR (Robinson et al., 2009; McCarthy et al., 2012) using estimateCommonDisp, estimateTagwiseDisp and exactTest functions. P-values were adjusted with the Hochberg method in the three analyses, and only genes with adjusted p < 0.05 in at least one of the analyses were included in Table 1. Parallel (Tange, 2018) was used at every step to run jobs in parallel.

To construct Supplementary Table 1, we used the 57 A. thaliana transcription factors that have been shown to bind the identified transcription factor binding sites (TFBS, 50 shared by different orthologous genes, Figure 2; plus 7 absent from C₃ or C₄ species, Supplementary Figure S5). We identify the orthologous genes in grass species and evaluate their enrichment in M or BS cells using publicly available transcriptomic data for S. viridis (John et al., 2014), S. italica (Washburn et al., 2017), P. virgatum (Rao et al., 2016), P. halli (Washburn et al., 2017), Z. mays (Chang et al., 2012), and S. bicolor (Döring et al., 2016). All these databases separate M and BS cells from whole leaves. We identified 10 orthologous genes significantly enriched (adj. p < 0.05) in P. virgatum, which corresponded to 8 genes in A. thaliana. For P. halli we identified 21 orthologs corresponding to 11 A. thaliana genes. For S. viridis we identified 53 orthologs corresponding to 26 A. thaliana genes. For S. italica we identified 46 orthologs corresponding to 22 A. thaliana genes. For Z. mays we identified 10 orthologs corresponding to 4 A. thaliana genes. For S. bicolor we identified 2 orthologs corresponding to 2 A. thaliana genes. In Table 1, we only included the A. thaliana genes for which the log2 fold was at least 1, and with consistent data from at least two species. We also removed Z. mays and S. bicolor orthologous genes as their transcriptomic data did not add any information on the A. thaliana genes included on Table 1.

Results

To account for the numerous independent origins of C₄ photosynthesis, we focus on a small subset of eight grasses: Oryza sativa, Hordeum vulgare, Brachypodium distachyon, Dichanthelium oligosanthes, Zea mays, Sorghum bicolor, Panicum virgatum, and Setaria viridis. All of these plant species belong to the Poaceae family and shared a common ancestor around 50 million years ago. O. sativa, H. vulgare, B. distachyon, and D. oligosanthes perform C₃ photosynthesis, whereas Z. mays, S. bicolor, P. virgatum, and S. viridis perform C₄ photosynthesis. Notably, D. oligosanthes belongs to the PACMAD clade (Figure 1), to which all selected C₄ species belong, and shares a common ancestor with them around 15 million years ago (Studer et al., 2016). To identify conserved regulatory regions in genes encoding CBB cycle enzymes of C₃ and C₄ grasses, we aligned each gene against its orthologous gene in a representative C₃ species (B. distachyon; Figure 1A and Supplementary Figures S1A, S2A, S3A) and against its orthologous gene in a representative C₄ species (S. bicolor; Figure 1C and Supplementary Figures S1C, S2C, S3B). The genomic sequence including potential promoters [2000 base pair (bp)] upstream from the annotated transcription start site (or start codon otherwise) and potential terminators (1,000 bp downstream from the end of 3′UTR or stop codon) was used to allow for the identification of putative regulatory regions outside coding sequences. Regions showing between 50 and 100% identity were plotted and conserved regions with over 70% identity were colored depending on the genomic feature (Figures 1A,C; coding sequences in purple, untranslated regions [UTRs] in cyan, and intergenic regions and introns in pink) As expected, most of the coding sequences were conserved among all orthologous genes, whereas only parts of the introns and intergenic sequences showed over 70% identity. We defined those regions as conserved non-coding sequences (CNS). For SBPase, we identified one CNS located at the last intron of most orthologs (Figures 1A,C), and two CNSs found only in SBPase orthologous genes from PACMAD species (C₄ species + D. oligosanthes; Figure 1C). In addition, we found one CNS located at the 5′ intergenic region of all PRK genes (Supplementary Figure S1), and two CNSs located at the 5′ intergenic region of FBPase genes from PACMAD species (Supplementary Figure S2). To further characterize these CNSs, they were subjected to motif prediction using MEME (Bailey et al., 2009), which generated a position weight matrix for the predicted motifs (Figure 1B and Supplementary Figures S1B, S2B). We used these motifs to scan the orthologous genes of other species, and identified the PRK CNS in the intergenic regions of PRK orthologs in non-grasses species (Supplementary Dataset 1). These results indicate that there are conserved potential cis-regulatory sequences shared between C₃ and C₄ species. However, this alignment approach is based on sequence identity over at least 50 bp; so it was possible that smaller motifs, such as transcription factor binding sites (TFBS) could have been disregarded.

FIGURE 1

FIGURE 2

To evaluate the presence of TFBS in the regulatory regions of genes encoding CBB cycle enzymes, a dataset containing validated TFBS in Arabidopsis thaliana (O’Malley et al., 2016) was used to scan the putative regulatory sequences (intergenic regions, untranslated regions, and introns) of orthologous genes, i.e., SBPase orthologs across the subset of eight grass species were scanned at the same time. We first determined the A. thaliana TFBS shared between orthologous genes, and used those to compare between the genes encoding the selected four CBB cycle enzymes (Figure 2A). This way, we identified one TFBS present in all of the potential regulatory sequences (common_CBB, in Figure 2A) that was bound by VRN1 in A. thaliana. This TFBS was also identified it in the putative regulatory regions of genes encoding photorespiratory (GDCH) and housekeeping proteins (CBP20) (Supplementary Dataset 2), suggesting that it might play a regulatory role not limited to photosynthetic genes. We also identified 13 putative TFBS shared between SBPase, FBPase, and GAPDHB orthologous genes (common_SFG), 9 TFBS shared between GAPDHB and SBPase orthologous genes (common_GS), one shared between GAPDHB and PRK orthologous genes (common_GP), and one TFBS shared between GAPDHB and FBPase orthologous genes (common_GF). In addition, 17, 2, and 6 putative TFBS were shared between GAPDHB orthologs (common_GAPDHB), FBPase orthologs (common_FBP), and PRK orthologs (common_PRK); but not between any other group of orthologous genes. Notably, these common sequences can be found in potential regulatory sequences of other orthologous genes from some but not in all of the species in the study. The fact that all the A. thaliana TFBS found in SBPase orthologs are shared with other genes (common_SFG, common_GS) whereas most of the TFBS found in PRK orthologs are not shared (common_PRK) suggests different mechanisms for the regulation of the expression of the genes encoding CBB cycle enzymes. Most of the trans-acting factors binding to the identified TFBS belong to the Apetala2/Ethylene-Response-Factor (AP2/ERF) family, and often recognize similar binding sites. A comparison between the location of the A. thaliana TFBS at the putative regulatory regions of the orthologs in the selected species (Figure 2B and Supplementary Figure S4) revealed that the TFBS tend to cluster together in discrete regions of the putative regulatory sequences, although the genomic coordinates of these clusters change between species.

We used a similar approach to identify C₄-specific TFBS contributing to the difference in expression pattern between C₃ and C₄ species. After scanning together orthologous genes (i.e., all SBPase orthologous genes with the A. thaliana validated TFBS), we selected TFBS absent from the putative regulatory regions of genes encoding C₃-enzymes (Supplementary Figure S5). Using this approach (choosing absent motifs from genes encoding C₃ enzymes rather than present motifs in all genes encoding C₄ enzymes), it was possible to account for the multiple independent origins of C₄ photosynthesis. Three TFBS were found absent from SBPase C₃-genes, bound by At5g66940, BZR1, and CEJ1 in A. thaliana; one absent from FBPase C₃-genes (bound by TCX2) and one absent from PRK C₃-genes (bound by At3g12130). In addition, using the same approach to identify TFBS absent from genes encoding C₄ enzymes revealed two C₃-specific motifs in the 5′ intergenic region of PRK (bound by AREB3 and bZIP16). These results suggest that there are C₃- and C₄-specific TFBS that might contribute to the compartmentalization of C₄ CBB cycle enzymes.

To further understand how the identified TFBS might regulate the expression pattern of CBB cycle enzymes, we obtained the transcriptomic data from a collection of RNA-seq experiments on C₄ species where samples were taken separately from mesophyll and bundle sheath cells (John et al., 2014; Rao et al., 2016; Washburn et al., 2017), and assessed the expression pattern of the trans-acting factors. Despite the complexities of using data from different experiments, and the limited validation of the interaction between trans-acting factors and the identified TFBS (i.e., only validated in the C₃ species A. thaliana), we identified ten trans-acting factors differentially enriched in M and BS cell types (Table 1). These trans-acting factors were the orthologs of the validated trans-acting factors in A. thaliana, and could be classified into three categories in regard to BS-specific enrichment (Table 1): (1) putative activators, if all the orthologs were consistently enriched in BS over M cells, such as the orthologs of At5g66940, whose TFBS are only found in SBPase orthologs of C₄ species; (2) putative repressors, if all the orthologs were consistently enriched in M over BS cells, for example the orthologs of At3g12130, whose TFBS are only found in PRK orthologs of C₄ species; (3) broad regulators, if enrichment of orthologous genes was inconsistent within species (some were enriched in BS over M cells while others were enriched in M over BS cells), such as the orthologs of TCX2, which are found enriched in both M and BS cells, and whose TFBS are only found in FBPase orthologs of C₄ species.

Discussion

In this study, we have used publicly available data to analyze putative regulatory regions of genes encoding a selected subset of CBB cycle enzymes (SBPase, FBPase, PRK, and GAPDHB) in C₃ and C₄ species. We used two different approaches to identify potential regulatory elements that might contribute to the compartmentalization of CBB enzymes in C₄ species. The alignment of the genomic regions of the orthologs encoding the selected CBB cycle enzymes allowed us to identify conserved non-coding sequences (CNSs) shared by C₃ and C₄ orthologous genes, whereas the scanning of putative regulatory regions with TFBS validated in A. thaliana, allowed us to identify putative C₄-specific regulatory elements. The results presented here provide new information on putative regulatory elements of the genes encoding SBPase, FBPase, PRK, and GAPDHB in both C₃ and C₄ species and although we do not provide experimental evidence in this paper the results form the basis for future functional studies.

The alignment of the genomic regions of orthologous CBB genes revealed a number of CNSs shared between C₃ and C₄ species. We identified a highly conserved sequence in the 5′ intergenic region of every PRK gene (Supplementary Figure S1). This CNS stands out because of its length (113 bp) and the level of conservation, as it can be found in C₃ species even outside of Poaceae (Supplementary Dataset S1). These attributes suggest that this region could have contributed to the regulation of PRK expression throughout evolutionary history. In contrast, the CNS identified in the last intron of SBPase orthologous genes (Figure 1) was only found in species belonging to Poaceae, suggesting that the possible contribution to the regulation of SBPase genes is limited to Poaceae species (Supplementary Dataset S1). Nevertheless, the location of this conserved region highlights the relevance of searching for regulatory elements outside of the up- and down-stream non-coding sequences of genes (Rose, 2019). Additionally, we also identified CNSs conserved only within the more closely related species of the PACMAD clade (D. oligosanthes, Z. mays, S. bicolor, P. virgatum, and S. viridis) but not within the more distant related species of the BOP clade (O. sativa, H. vulgare, and B. distachyon; Figure 1 and Supplementary Figure S2). The fact that these CNSs are only shared between the species of the PACMAD clade, including D. oligosanthes which performs C₃ photosynthesis, suggests that these CNSs do not play a role in C₄ compartmentalization and instead they are a result of shared evolutionary history. However, the significance of the contribution of these conserved regions to the levels or patterns of expression of these genes remains to be elucidated experimentally.

Using a different approach based on validated TFBS and their trans-acting factors in A. thaliana, we identified putative (i.e., non-validated in grasses) TFBS shared by the genes encoding CBB cycle enzymes in both C₃ and C₄ species (Figure 2A), as well as C₄-specific (C₃-absent) putative TFBS (Supplementary Figure S5). We found three putative TFBS absent from SBPase C₃-genes, one absent from FBPase C₃-genes, and one absent from PRK C₃-genes. The identification of A. thaliana TFBS in genes encoding C₄ CBB cycle enzymes supports the hypothesis that C₄ genes co-opted regulatory elements of C₃ genes to establish their restricted expression pattern (Brown et al., 2011; Xu et al., 2016; Borba et al., 2018; Reyna-Llorens et al., 2018). Notably, we did not identify any C₃- or C₄-specific putative TFBS in the regulatory regions of GAPDHB orthologs (Supplementary Figure S5) but found more shared TFBS between C₃ and C₄GAPDHB orthologs (Figure 2A). Despite being expressed in BS cells, which should allow for CBB cycle function in those cells, GAPDHB is enriched in M cells (Majeran et al., 2005; Rao et al., 2016). The lack of C₄-specific putative TFBS in GAPDHB regulatory regions suggests that its expression might be regulated similarly in both C₃ and C₄ plants. Most of the identified TFBS are recognized by members of the AP2/ERF family in A. thaliana, which supports the results of a recent study in which this family of TFBS was enriched in the regulatory regions of C₄ photosynthetic genes (Burgess et al., 2019). We realized that these putative TFBS were often quite similar and cluster together at specific locations in the genome and this warrants further investigation to explore the functional significance. Despite the similarities, we only identified one TFBS, bound by VRN1 in A. thaliana, in the putative regulatory regions of every gene selected for this study, but its presence in other non-photosynthetic genes indicates that VRN1 is unlikely to be exclusive to the regulation of the expression of genes encoding CBB cycle enzymes. Furthermore, the variety in the putative TFBS identified in different sets of orthologous genes indicates differences in the regulatory networks controlling their expression. These results suggest that there is no “master” transcriptional regulator coordinating the expression of the genes encoding CBB cycle enzymes, in contrast to what has been reported in other metabolic pathways (Okada et al., 2009; Nützmann et al., 2018). In addition, the lack of a unique, “master” regulator would emphasize the importance of the simultaneous manipulation of multiple targets to increase CBB cycle efficiency (Simkin et al., 2015, 2017; López-Calcagno et al., 2020).

Based on data validated in the model plant A. thaliana, we used a computational approach to identify cis-regulatory elements whose putative trans-acting factors might play a role in C₄ compartmentalization. These data have been used to investigate the putative role of orthologous genes in other crops (Capote et al., 2018; Moon et al., 2018; Burgess et al., 2019; Zeng et al., 2019; DeMers et al., 2020; Elzanati et al., 2020; Gray et al., 2020; Zhou et al., 2020), and allow us to generate a compelling hypothesis, as it is expected that similar DNA-binding domains of trans-acting factors would have similar DNA sequence preferences (Lambert et al., 2019). However, several complementary experimental approaches will be needed to provide evidence of functional significance in C₄ plants. To confirm the TFBS in different species, transcription factor binding assays such as DAP-seq (O’Malley et al., 2016) could be developed in some of the grass species examined in this study. To assess the chromatin accessibility of potential regulatory regions, experiments such as DNAse-seq (Zhang and Jiang, 2015) or ATAC-seq (Buenrostro et al., 2015; Bajic et al., 2018; Maher et al., 2018), could be implemented. To enhance our ability to detect regulatory elements within coding sequences (Reyna-Llorens et al., 2018), functional assays that discriminate between conserved sites with a regulatory role and conserved sites with a coding sequence role could be developed. Finally, the generation of transcriptomic data from different species using a comparable sampling process, should allow us to unveil consistent pattern of expression among different species.

Taking all of our results together, we propose that the compartmentalization of the CBB cycle enzymes investigated in this study has occurred through the recruitment of TFBS whose trans-acting factors are enriched in either one of the C₄ cell types. The expression pattern of any gene is determined by a combination of the TFBS present and the corresponding trans-acting factors binding to these regulatory regions at any given time. It then follows that the expression pattern of any gene can be changed either by recruiting new TFBS or by altering the expression pattern of the trans-acting factors. Thus, to enrich the expression of C₄ enzymes in BS cells, new TFBS could be recruited into gene regulatory regions of C₄ species to confer BS-specific expression. Alternatively, trans-acting factors could become enriched in BS cells to promote the expression of C₄ enzymes in BS cells (as the predicted putative activators), or these factors could become enriched in M cells to repress the expression of C₄ enzymes in M cells (predicted putative repressors). This transcriptional regulation would likely be complemented by regulation at post-transcriptional and/or post-translational level to achieve a precise regulation of the expression pattern of CBB cycle enzymes.

To our knowledge, and excluding the extensive work on RubisCO (discussed in Hibberd and Covshoff, 2010; Berry et al., 2011; Schlüter and Weber, 2020), this is the first study to focus specifically on the differences in the regulatory sequences of CBB cycle genes between C₃ and C₄ species. These results provide a hypothetical foundation for future functional analysis. Future experiments should include the in vivo validation of the trans-acting factors binding to cis-regulatory elements, and the resultant regulation of CBB cycle genes; the transfer of C₄-specific transcription factors into C₃ species to establish a C₄-like expression pattern; or the precise genome editing of the cis-elements to evaluate their contribution to compartmentalization in C₄ plants.

References

• 1

  AltschulS. F.GishW.MillerW.MyersE. W.LipmanD. J. (1990). Basic local alignment search tool.J. Mol. Biol.215403–410.

  • Google Scholar

• 2

  AndersS.HuberW. (2010). Differential expression analysis for sequence count data.Genome Biol.11:R106.

  • Google Scholar

• 3

  AubryS.KellyS.KümpersB. M. C.Smith-UnnaR. D.HibberdJ. M. (2014). Deep evolutionary comparison of gene expression identifies parallel recruitment of trans-factors in two independent origins of C4 photosynthesis.PLoS Genet.10:e1004365. 10.1371/journal.pgen.1004365

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  BaileyT. L.BodenM.BuskeF. A.FrithM.GrantC. E.ClementiL.et al (2009). MEME Suite: tools for motif discovery and searching.Nucleic Acids Res.37W202–W208.

  • Google Scholar

• 5

  BajicM.MaherK. A.DealR. B. (2018). Identification of open chromatin regions in plant genomes using ATAC-Seq.Methods Mol. Biol.1675183–201. 10.1007/978-1-4939-7318-7_12

  • CrossRef
  • Google Scholar

• 6

  BeierS.HimmelbachA.ColmseeC.ZhangX. Q.BarreroR. A.ZhangQ.et al (2017). Construction of a map-based reference genome sequence for barley, Hordeum vulgare L.Sci. Data4:170044.

  • Google Scholar

• 7

  BennetzenJ. L.SchmutzJ.WangH.PercifieldR.HawkinsJ.PontaroliA. C.et al (2012). Reference genome sequence of the model plant Setaria.Nat. Biotechnol.30555–561.

  • Google Scholar

• 8

  BerryJ. O.PatelM.ZielinskiA. (2011). “Chapter 12 C4 gene expression in mesophyll and bundle sheath cells,” in C4 Photosynthesis and Related CO₂ Concentrating Mechanisms, edsRaghavendraA. S.SageR. F. (Dordrecht: Springer), 221–256. 10.1007/978-90-481-9407-0_12

  • CrossRef
  • Google Scholar

• 9

  BorbaA. R.SerraT. S.GórskaA.GouveiaP.CordeiroA. M.Reyna-LlorensI.et al (2018). Synergistic binding of bHLH transcription factors to the promoter of the maize NADP-ME gene used in C4 photosynthesis is based on an ancient code found in the ancestral C3 state.Mol. Biol. Evol.351690–1705. 10.1093/molbev/msy060

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  BräutigamA.KajalaK.WullenweberJ.SommerM.GagneulD.WeberK. L.et al (2011). An mRNA blueprint for C4 photosynthesis derived from comparative transcriptomics of closely related C3 and C4 species.Plant Physiol.155:142. 10.1104/pp.110.159442

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11

  BrayN. L.PimentelH.MelstedP.PachterL. (2016). Near-optimal probabilistic RNA-seq quantification.Nat. Biotechnol.34525–527. 10.1038/nbt.3519

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12

  BrownN. J.NewellC. A.StanleyS.ChenJ. E.PerrinA. J.KajalaK.et al (2011). Independent and parallel recruitment of preexisting mechanisms underlying C(4) photosynthesis.Science3311436–1439. 10.1126/science.1201248

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13

  BuenrostroJ. D.WuB.ChangH. Y.GreenleafW. J. (2015). ATAC-seq: a method for assaying chromatin accessibility genome-wide.Curr. Protoc. Mol. Biol.10921.29.21–21.29.29.

  • Google Scholar

• 14

  BurgessS. J.Reyna-LlorensI.StevensonS. R.SinghP.JaegerK.HibberdJ. M. (2019). Genome-wide transcription factor binding in leaves from C3 and C4 grasses.Plant Cell312297–2314. 10.1105/tpc.19.00078

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15

  CapoteT.BarbosaP.UsieA.RamosA. M.InacioV.OrdasR.et al (2018). ChIP-Seq reveals that QsMYB1 directly targets genes involved in lignin and suberin biosynthesis pathways in cork oak (Quercus suber).BMC Plant Biol.18:198. 10.1186/s12870-018-1403-5

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16

  ChangY.-M.LiuW.-Y.ShihA. C.-C.ShenM.-N.LuC.-H.LuM.-Y. J.et al (2012). Characterizing regulatory and functional differentiation between maize mesophyll and bundle sheath cells by transcriptomic analysis.Plant Physiol.160:165. 10.1104/pp.112.203810

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17

  DeMersL. C.RedekarN. R.KachrooA.TolinS. A.LiS.Saghai MaroofM. A. (2020). A transcriptional regulatory network of Rsv3-mediated extreme resistance against Soybean mosaic virus.PLoS One15:e0231658. 10.1371/journal.pgen.0231658

  • CrossRef
  • Google Scholar

• 18

  DöringF.StreubelM.BräutigamA.GowikU. (2016). Most photorespiratory genes are preferentially expressed in the bundle sheath cells of the C4 grass Sorghum bicolor.J. Exper. Bot.673053–3064. 10.1093/jxb/erw041

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 19

  ElzanatiO.MouzeyarS.RocheJ. (2020). Dynamics of the transcriptome response to heat in the moss, Physcomitrella patens.Int. J. Mol. Sci.21:1512. 10.3390/ijms21041512

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20

  FrazerK. A.PachterL.PoliakovA.RubinE. M.DubchakI. (2004). VISTA: computational tools for comparative genomics.Nucleic Acids Res.32W273–W279.

  • Google Scholar

• 21

  GehlenborgN. (2019). UpSetR: a More Scalable Alternative to Venn and Euler Diagrams for Visualizing Intersecting Sets. Available online at: https://rdrr.io/cran/UpSetR/(accessed May 21, 2020).

  • Google Scholar

• 22

  GoodsteinD. M.ShuS.HowsonR.NeupaneR.HayesR. D.FazoJ.et al (2011). Phytozome: a comparative platform for green plant genomics.Nucleic Acids Res.40D1178–D1186.

  • Google Scholar

• 23

  GrantC. E.BaileyT. L.NobleW. S. (2011). FIMO: scanning for occurrences of a given motif.Bioinformatics271017–1018. 10.1093/bioinformatics/btr064

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24

  GrayS. B.Rodriguez-MedinaJ.RusoffS.ToalT. W.KajalaK.RuncieD. E.et al (2020). Translational regulation contributes to the elevated CO₂ response in two Solanum species.Plant J.102383–397. 10.1111/tpj.14632

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25

  HibberdJ. M.CovshoffS. (2010). The regulation of gene expression required for C4 photosynthesis.Annu. Rev. Plant Biol.61181–207.

  • Google Scholar

• 26

  HirschC. N.HirschC. D.BrohammerA. B.BowmanM. J.SoiferI.BaradO.et al (2016). Draft assembly of elite inbred line PH207 provides insights into genomic and transcriptome diversity in maize.Plant Cell282700–2714. 10.1105/tpc.16.00353

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  International Brachypodium Initiative (2010). Genome sequencing and analysis of the model grass Brachypodium distachyon.Nature463763–768. 10.1038/nature08747

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28

  JohnC. R.Smith-UnnaR. D.WoodfieldH.CovshoffS.HibberdJ. M. (2014). Evolutionary convergence of cell-specific gene expression in independent lineages of C4 grasses.Plant Physiol.16562–75. 10.1104/pp.114.238667

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  LambertS. A.YangA. W. H.SasseA.CowleyG.AlbuM.CaddickM. X.et al (2019). Similarity regression predicts evolution of transcription factor sequence specificity.Nat. Genet.51981–989. 10.1038/s41588-019-0411-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 30

  López-CalcagnoP. E.BrownK. L.SimkinA. J.FiskS. J.Vialet-ChabrandS.LawsonT.et al (2020). Stimulating photosynthetic processes increases productivity and water-use efficiency in the field.Nat. Plants61054–1063. 10.1038/s41477-020-0740-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  LoveM. I.HuberW.AndersS. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.Genome Biol.15:550.

  • Google Scholar

• 32

  MaherK. A.BajicM.KajalaK.ReynosoM.PauluzziG.WestD. A.et al (2018). Profiling of accessible chromatin regions across multiple plant species and cell types reveals common gene regulatory principles and new control modules.Plant Cell3015–36. 10.1105/tpc.17.00581

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  MajeranW.CaiY.SunQ.Van WijkK. J. (2005). Functional differentiation of bundle sheath and mesophyll maize chloroplasts determined by comparative proteomics.Plant Cell17:3111. 10.1105/tpc.105.035519

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  MascherM.GundlachH.HimmelbachA.BeierS.TwardziokS. O.WickerT.et al (2017). A chromosome conformation capture ordered sequence of the barley genome.Nature544427–433.

  • Google Scholar

• 35

  McCarthyD. J.ChenY.SmythG. K. (2012). Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation.Nucleic Acids Res.404288–4297. 10.1093/nar/gks042

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36

  McCormickR. F.TruongS. K.SreedasyamA.JenkinsJ.ShuS.SimsD.et al (2018). The Sorghum bicolor reference genome: improved assembly, gene annotations, a transcriptome atlas, and signatures of genome organization.Plant J.93338–354. 10.1111/tpj.13781

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37

  MicheletL.ZaffagniniM.MorisseS.SparlaF.Perez-PerezM. E.FranciaF.et al (2013). Redox regulation of the Calvin-benson cycle: something old, something new.Front. Plant Sci.4:470. 10.3389/fpls.2013.00470

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 38

  MoonS.ChandranA. K. N.AnG.LeeC.JungK. H. (2018). Genome-wide analysis of root hair-preferential genes in rice.Rice11:48.

  • Google Scholar

• 39

  NiklausM.KellyS. (2019). The molecular evolution of C4 photosynthesis: opportunities for understanding and improving the world’s most productive plants.J. Exp. Bot.70795–804. 10.1093/jxb/ery416

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  NomuraM.KatayamaK.NishimuraA.IshidaY.OhtaS.KomariT.et al (2000). The promoter of rbcS in a C3 plant (rice) directs organ-specific, light-dependent expression in a C4 plant (maize), but does not confer bundle sheath cell-specific expression.Plant Mol. Biol.4499–106.

  • Google Scholar

• 41

  NützmannH.-W.ScazzocchioC.OsbournA. (2018). Metabolic gene clusters in eukaryotes.Annu. Rev. Genet.52159–183. 10.1146/annurev-genet-120417-031237

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42

  OkadaA.OkadaK.MiyamotoK.KogaJ.ShibuyaN.NojiriH.et al (2009). OsTGAP1, a bZIP transcription factor, coordinately regulates the inductive production of Diterpenoid Phytoalexins in rice.J. Biol. Chem.28426510–26518. 10.1074/jbc.m109.036871

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 43

  O’MalleyR. C.HuangS.-C.SongL.LewseyM. G.BartlettA.NeryJ. R.et al (2016). Cistrome and Epicistrome features shape the regulatory DNA landscape.Cell1651280–1292. 10.1016/j.cell.2016.04.038

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 44

  OuyangS.ZhuW.HamiltonJ.LinH.CampbellM.ChildsK.et al (2007). The TIGR rice genome annotation resource: improvements and new features.Nucleic Acids Res.35D883–D887.

  • Google Scholar

• 45

  R Core Team (2019). R: A Language and Environment for Statistical Computing.Vienna: R Core Team.

  • Google Scholar

• 46

  RaoX.LuN.LiG.NakashimaJ.TangY.DixonR. A. (2016). Comparative cell-specific transcriptomics reveals differentiation of C4 photosynthesis pathways in switchgrass and other C4 lineages.J. Exper. Bot.671649–1662. 10.1093/jxb/erv553

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 47

  Reyna-LlorensI.BurgessS. J.ReevesG.SinghP.StevensonS. R.WilliamsB. P.et al (2018). Ancient duons may underpin spatial patterning of gene expression in C4 leaves.Proc. Natl. Acad. Sci. U.S.A.115:1931. 10.1073/pnas.1720576115

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48

  RobinsonM. D.MccarthyD. J.SmythG. K. (2009). edgeR: a bioconductor package for differential expression analysis of digital gene expression data.Bioinformatics26139–140. 10.1093/bioinformatics/btp616

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49

  RoseA. B. (2019). Introns as gene regulators: a brick on the accelerator.Front. Genet.9:672. 10.3389/fgene.2018.00672

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50

  SageR. F.SageT. L.KocacinarF. (2012). Photorespiration and the evolution of C4 photosynthesis.Annu. Rev. Plant Biol.6319–47.

  • Google Scholar

• 51

  SchlüterU.WeberA. P. M. (2020). Regulation and evolution of C4 photosynthesis.Annu. Rev. Plant Biol.71183–215.

  • Google Scholar

• 52

  SchnableP. S.WareD.FultonR. S.SteinJ. C.WeiF.PasternakS.et al (2009). The B73 maize genome: complexity, diversity, and dynamics.Science3261112–1115.

  • Google Scholar

• 53

  SheenJ. Y.BogoradL. (1987). Regulation of levels of nuclear transcripts for C4 photosynthesis in bundle sheath and mesophyll cells of maize leaves.Plant Mol. Biol.8227–238. 10.1007/bf00015031

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54

  SimkinA. J.Lopez-CalcagnoP. E.DaveyP. A.HeadlandL. R.LawsonT.TimmS.et al (2017). Simultaneous stimulation of sedoheptulose 1,7-bisphosphatase, fructose 1,6-bisphophate aldolase and the photorespiratory glycine decarboxylase-H protein increases CO₂ assimilation, vegetative biomass and seed yield in Arabidopsis.Plant Biotechnol. J.15805–816. 10.1111/pbi.12676

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 55

  SimkinA. J.McauslandL.HeadlandL. R.LawsonT.RainesC. A. (2015). Multigene manipulation of photosynthetic carbon assimilation increases CO₂ fixation and biomass yield in tobacco.J. Exp. Bot.664075–4090. 10.1093/jxb/erv204

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56

  StuderA. J.SchnableJ. C.WeissmannS.KolbeA. R.MckainM. R.ShaoY.et al (2016). The draft genome of the C3 panicoid grass species Dichanthelium oligosanthes.Genome Biol.17:223.

  • Google Scholar

• 57

  TangeO. (2018). GNU Parallel 2018.Morrisville: Lulu.

  • Google Scholar

• 58

  WashburnJ. D.KothapalliS. S.BroseJ. M.CovshoffS.HibberdJ. M.ConantG. C.et al (2017). Ancestral reconstruction and C3 bundle sheath transcript abundance in the paniceae grasses indicate the foundations for all three biochemical c4 sub-types were likely present in the most recent ancestor.bioRxiv [Preprint]. 10.1101/162644

  • CrossRef
  • Google Scholar

• 59

  WickhamH. (2016). ggplot2: Elegant Graphics for Data Analysis.New York, NY: Springer-Verlag.

  • Google Scholar

• 60

  WickhamH.FrançoisR.HenryL.MüllerK. (2019). dplyr: a Grammar of Data Manipulation. Available online at: https://CRAN.R-project.org/package=dplyr(accessed May 21, 2020).

  • Google Scholar

• 61

  WilkinsD. (2019). gggenes: Draw Gene Arrow Maps in ‘ggplot2’. Available online at: https://rdrr.io/cran/gggenes/(accessed May 17, 2020).

  • Google Scholar

• 62

  WilliamsB. P.BurgessS. J.Reyna-LlorensI.KnerovaJ.AubryS.StanleyS.et al (2016). An untranslated cis-element regulates the accumulation of multiple C4 enzymes in Gynandropsis gynandra Mesophyll cells.Plant Cell28:454. 10.1105/tpc.15.00570

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 63

  XuJ.BräutigamA.WeberA. P. M.ZhuX.-G. (2016). Systems analysis of cis-regulatory motifs in C4 photosynthesis genes using maize and rice leaf transcriptomic data during a process of de-etiolation.J. Exper. Bot.675105–5117. 10.1093/jxb/erw275

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  ZengZ.ZhangW.MarandA. P.ZhuB.BuellC. R.JiangJ. (2019). Cold stress induces enhanced chromatin accessibility and bivalent histone modifications H3K4me3 and H3K27me3 of active genes in potato.Genome Biol.20:123.

  • Google Scholar

• 65

  ZhangW.JiangJ. (2015). Genome-wide mapping of DNase I hypersensitive sites in plants.Methods Mol. Biol.128471–89. 10.1007/978-1-4939-2444-8_4

  • CrossRef
  • Google Scholar

• 66

  ZhouP.LiZ.MagnussonE.Gomez CanoF.CrispP. A.NoshayJ. M.et al (2020). Meta gene regulatory networks in maize highlight functionally relevant regulatory interactions.Plant Cell321377–1396. 10.1105/tpc.20.00080

  • Pubmed Abstract
  • CrossRef
  • Google Scholar