Constructing the bundle sheath towards enhanced photosynthesis

This article comments on: van Rooijen R, Schulze S, Petzsch P, Westhoff P. 2020. Targeted misexpression of NAC052, acting in H3K4 demethylation, alters leaf morphological and anatomical traits in Arabidopsis thaliana. Journal of Experimental Botany 71, 1434–1448.

The enzyme Rubisco plays a central role in photosynthesis by capturing atmospheric carbon dioxide (CO 2 ) in an organic form. This enzymatic step is the basis of C 3 photosynthesis. However, Rubisco also reacts with oxygen (O 2 ) to generate a toxic by-product, which must be salvaged by an energyintensive process known as photorespiration (Portis and Parry, 2007). Many lineages of plants have evolved means to prevent photorespiration by increasing the relative concentration of CO 2 around Rubisco. In plants, the C 4 cycle is the most prevalent of these mechanisms which involves spatial separation of the reactions of photosynthesis. CO 2 is initially fixed in mesophyll (M) cells as a four carbon (C 4 ) intermediate, diffused deeper inside the leaf into bundle sheath (BS) cells where the C 4 intermediate is decarboxylated for refixation by Rubisco (Hatch and Slack, 1966). Therefore, C 4 photosynthesis requires BS cell-specific expression of Rubisco, but also several other enzymes that must be M or BS specific in order to operate the cycle. In conjunction, a unique cellular arrangement in most C 4 leaves (termed Kranz anatomy) has evolved to facilitate this molecular CO 2 pump (El-Sharkawy and Hesketh, 1965;Hatch, 1987). Compared with C 3 plants, C 4 Kranz anatomy generally comprises denser venation, increased BS cell size, number, and chloroplast content, a greater reliance on the BS for photosynthesis, fewer M cells, and more plasmodesmata connections between the M and BS (Box 1). As Kranz anatomy is a multifaceted trait, identifying its genetic determinants has been a bottleneck in C 4 photosynthetic research.
The transcription factors SCARECROW (SCR) and SHORTROOT (SHR) have been implicated with BS specification in C 4 maize and C 3 Arabidopsis (Slewinski et al., 2012;Cui et al., 2014). However, recent studies have shown that SCR/SHR regulate cell patterning in more broad contexts, such as root, epidermal, stomatal, and M cell patterning, suggesting that these two factors cannot fully account for the developmental changes in the BS to enable C 4 photosynthesis (Hughes et al., 2019). Paralogues of the maize GOLDEN2 (G2) transcription factor family regulate dimorphic chloroplast differentiation in BS and M cells (Wang et al., 2013). Overexpression of GOLDEN2-LIKE1 in C 3 rice led to increased chloroplast development in the vascular bundles of rice seedlings (Nakamura et al., 2009). Thus, GOLDEN and GOLDEN-LIKE transcription factors play a role in plastid morphogenesis that probably aided in increasing the photosynthetic capacity of the BS. Beyond a few characterized regulatory steps that occur at multiple levels of gene expression, our current understanding of the steps required to engineer Kranz anatomy and cell-specific expression of C 4 cycle enzymes is limited (Reeves et al., 2017;Sedelnikova et al., 2018).

NAC052, a H3K4 demethylase: identified as a novel genetic regulator of bundle sheath anatomy in Arabidopsis
In this issue of the Journal of Experimental Botany, van Rooijen et al. (2019) used activation tagging to identify a regulator influencing the number and chloroplast content of BS cells in A. thaliana (Box 2). In activation tagging, a promoter is randomly inserted into a reference genome, which results in transcriptional changes of genes in close proximity to the insertion site (Tani et al., 2004). In this study, the authors used an A. thaliana reference line from Döring et al. (2019) which was transformed with the promoter of the C 4 Flaveria trinervia GLYCINE DECARBOXYLASE P-SUBUNIT gene (pGLDPA Ft ) to drive BS-preferential expression of a chloroplast-targeted green fluorescent protein (pGLDPA Ft ::RbcS.TP-sGFP). In order to identify regulators that influence the morphology of the BS, they used a second BS-preferential promoter from the F. trinervia GLYCINE DECARBOXYLASE T-SUBUNIT gene (pGLDT Ft ) as an activation tag. Altered GFP fluorescence relative to the reference line allowed screening to find individual lines with altered BS-related phenotypes. Genomic analysis of one such line revealed that the reference promoter had inserted in the coding sequence of the gene encoding NAC052, a transcriptional repressor involved in H3K4 demethylation (Ning et al., 2015).
The insertion event resulted in a 5′-truncated transcript variant of NAC052 leading to a partial deletion of its DNAbinding domain. The mutation led to changes in GFP fluorescence as well as changes in BS anatomy and leaf and whole-plant morphology, such as a greater number of BS cells and chloroplasts as compared with the reference line. The JMJ14-NAC052 module is involved in post-transcriptional gene silencing by acting as a H3K4 demethylase which promotes transgene transcription by preventing DNA methylation (Butel et al., 2017). Furthermore, the activation tagging mutation event was reconstructed by expressing pGLDT Ft ::5′-truncatedNAC052 in the pGLDPA Ft ::RbcS.TP-sGFP reference line. This recapitulated the previous chlorotic and wrinkled leaf edge phenotypes and caused a greater accumulation of BS cells as compared with the reference line. Further validation of the involvement of NAC052 in leaf development was confirmed by expressing the full reading frame of the NAC052 transcript under the control of the GLDT Ft promoter in the pGLDPA Ft ::RbcS.TP-sGFP reference background (pGLDT Ft ::NAC052). The lines showed enhanced GFP signal intensity and more BS cells. As an additional line of evidence of NAC052 function, the endogenous NAC052 was mutated with clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9). The CRISPR/Cas9 mutant line was small and had chlorotic leaf edges, but it did not show the wrinkled leaf edges. However, in contrast to the activation-tagged mutant, the GFP signal intensity was decreased in the CRISPR/ Cas9 mutant as compared with the reference line.
To assess the genetic impact of ectopic NAC052 expression, mRNA sequencing of the reference and all the transgenic lines was undertaken to identify its downstream gene regulatory targets. Comparative transcriptomics of the lines showed differential transcript abundance of genes involved in leaf cell wall organization and arabinogalactan metabolism, which are mediators between the cell wall, the plasma membrane, and Box 1. Comparison of C 3 and C 4 leaf anatomy A schematic of transverse cross-sections of mature C 4 Flaveria trinervia and C 3 Arabidopsis thaliana leaves. Cell outlines: upper and lower epidermis (black), vasculature (grey), bundle sheath (green), and mesophyll (pink). The middle layer of mesophyll cells (pink) highlights the difference in cell number between veins in C 3 and C 4 species. The dark green -color in C 4 plants represents higher photosynthetic capacity of BS cells. the cytoplasm. In summary, van Rooijen et al. (2019) associate NAC052 with leaf developmental patterns that alter anatomy specifically related to Kranz-like features. This opens up exploration into the role of other trans-factors that may have arisen from existing regulatory networks to transition from an ancestral C 3 state to a derived C 4 photosynthetic state.

Future perspectives
The study from van Rooijen et al. (2019) is an advancement in our current understanding of BS anatomical regulation and furthers investigation into the role of post-transcriptional gene silencing in leaf development. Incorporation of bisulfite sequencing and methylome data sets might uncover underlying epigenetic patterns affecting BS anatomy and function across C 3 and C 4 species.
Here, van Rooijen et al. found that NAC052 had a transcriptionally repressive role in C 3 A. thaliana, which caused a boost in BS number and chloroplast content when ectopically expressed. Extension of their methodology to a C 4 species would allow association of NAC052 with traits of Kranz anatomy. Coupled with putative regulators of cell and plastid division genes, misexpression of NAC052 might shed insights into a mechanistic understanding of gene-regulatory networks that enhance the photosynthetic capacity of the BS. It would be particularly interesting to see if this could lead to trait stacking for efforts to engineer C 4 photosynthesis in C 3 crops, such as increased vein density or metabolic flux between M and BS cells from more plasmodesmata connections.
To sum up, van Rooijen et al. (2019) highlight how highthroughput phenotyping of transgenic activation-tagged lines can uncover novel gene-regulatory networks. This expands Box 2. Activation tagging of NAC052 alters leaf anatomy and the expression of its target genes (A) Whole-plant morphology of transgenic lines transformed with various versions of NAC052, a H3K4 demethylase, into a reference background containing a bundle sheath-localized GFP signal. (B) Screening lines for differences in GFP intensity allowed detection of the enhanced number of chloroplasts in the bundle sheath and overall number of bundle sheath cells. (C) Comparative transcriptomics identified putative regulatory targets of NAC052.
knowledge on how to manipulate the role and structure of the BS in C 3 species. Use of forward genetics like this seems to be a promising approach to unravel the complexity of Kranz anatomy. Hopefully this will lead to further reports that identify genetic determinants underpinning the regulation of Kranz traits in C 4 photosynthesis.