Periodic Reporting for period 1 - Trans C4 (Deciphering Regulatory DNA and Transcription Factor Binding Sites in C3 and C4 Species with Varying Water Use Efficiencies)
Berichtszeitraum: 2018-06-26 bis 2020-06-25
Water-related stress is the number one limiting factor for plant productivity and also human well-being. Unfortunately, one-third of the current world population faces water shortages and by 2025, two-thirds are expected to experience water stress conditions, and 1.8 billion people will be subjected to absolute water scarcity. Unless ameliorating measures are taken, almost half the world's population will be highly water-stressed by 2030. Moreover, the UN World Water Development Report 2015 estimates that global water demand will increase by 55%, with agriculture the greatest user. To make matters worse by 2050 agricultural production has to increase by 60% to provide feed and fuel. At least one in nine worldwide do not have enough food to eat, hence improving crop plants to be both productive and water-efficient users is a matter of life.
The C4 photosynthetic pathway not only boosts plant productivity by ~50% but also increases water use efficiency. C4 photosynthesis is a remarkable trait that is thought to have evolved in response to environmental factors including increased aridity and seasonality. In contrast to C3 crops where CO2 fixation is catalysed by RuBisCO in mesophyll cells of leaves, most C4 plants use two distinct and specialized cell types for photosynthesis. All C4 plants concentrate CO2 in leaves, increasing productivity by ~50%, but also maintaining lower stomatal conductance than C3 species. For example, under heat stress induced by a temperature rise from 20°C to 30°C, C3 plants double water loss via transpiration whilst C4 plants are able to decrease the diffusive efflux of water vapour by 50%, and are therefore considered as water-efficient users. In summary, C4 photosynthesis allows increased water use efficiency but also underpins significant improvements in crop yield. A full understanding of this remarkable phenomenon would facilitate water efficient and productive crops to be engineered in the future. The experimental programme is designed to provide significant new insight into the mechanisms underpinning differences in gene expression associated with contrasting photosynthetic and water use efficiencies.
The C4 photosynthetic pathway not only boosts plant productivity by ~50% but also increases water use efficiency. C4 photosynthesis is a remarkable trait that is thought to have evolved in response to environmental factors including increased aridity and seasonality. In contrast to C3 crops where CO2 fixation is catalysed by RuBisCO in mesophyll cells of leaves, most C4 plants use two distinct and specialized cell types for photosynthesis. All C4 plants concentrate CO2 in leaves, increasing productivity by ~50%, but also maintaining lower stomatal conductance than C3 species. For example, under heat stress induced by a temperature rise from 20°C to 30°C, C3 plants double water loss via transpiration whilst C4 plants are able to decrease the diffusive efflux of water vapour by 50%, and are therefore considered as water-efficient users. In summary, C4 photosynthesis allows increased water use efficiency but also underpins significant improvements in crop yield. A full understanding of this remarkable phenomenon would facilitate water efficient and productive crops to be engineered in the future. The experimental programme is designed to provide significant new insight into the mechanisms underpinning differences in gene expression associated with contrasting photosynthetic and water use efficiencies.
C4 photosynthesis allows increased water use efficiency but also underpins significant improvements in crop yield. A full understanding of this remarkable phenomenon would facilitate water efficient and productive crops to be engineered in the future. In this programme, we used sorghum (Sorghum bicolor), which uses the more efficient C4 pathway, and is highly drought-tolerant. To better understand molecular events associated with the induction of C4 photosynthesis, changes in the transcriptome of sorghum leaves were assessed during de-etiolation. Samples were taken at 0, 0.5 2, 4, 6, and 12 h after leaves were transferred from dark to light and samples assessed for chlorophyll content, mRNA abundance and accessibility of DNA for binding by transcription factors.
Consistent with expectations from other systems, chlorophyll content increased exponentially over this period. To determine the dynamics of mRNAs, three biological replicates per each time point were assessed, quality controlled, and subjected to deep sequencing. Reads obtained were checked using FASTQ, trimmed using Trimmomatic v0.32 and then Salmon was used to quantify transcript abundance. Genes that responded to light were identified using DESeq2.
To provide a broad overview of the types of genes that were found in each of these clusters, Gene Ontology (GO) term analyses was performed. To investigate the response of sorghum photosynthesis genes, Clusters in which the GO terms relating to photosynthetic processes were over-represented were analysed in more detail. This included comparison of these new data with existing publicly available datasets for specific cell types of the sorghum leaf (Covshoff et al., 2013, Emms et al., 2016). Re-analysis of these data showed that triplicate samples generated 21 to 25 million reads per replicate, and the same pipeline as was used above implemented.
The complementary datasets were then interrogated for transcription factors that showed behaviours that could explain the induction of photosynthesis gene expression in sorghum. Thus, to provide greated insight into which transcription factors may be important, we implemented ATAC-seq across the same time course to define the DNA regions that were accessible to transcription factor binding, and the actual cis-elements bound by transcription factors in vivo.
Consistent with expectations from other systems, chlorophyll content increased exponentially over this period. To determine the dynamics of mRNAs, three biological replicates per each time point were assessed, quality controlled, and subjected to deep sequencing. Reads obtained were checked using FASTQ, trimmed using Trimmomatic v0.32 and then Salmon was used to quantify transcript abundance. Genes that responded to light were identified using DESeq2.
To provide a broad overview of the types of genes that were found in each of these clusters, Gene Ontology (GO) term analyses was performed. To investigate the response of sorghum photosynthesis genes, Clusters in which the GO terms relating to photosynthetic processes were over-represented were analysed in more detail. This included comparison of these new data with existing publicly available datasets for specific cell types of the sorghum leaf (Covshoff et al., 2013, Emms et al., 2016). Re-analysis of these data showed that triplicate samples generated 21 to 25 million reads per replicate, and the same pipeline as was used above implemented.
The complementary datasets were then interrogated for transcription factors that showed behaviours that could explain the induction of photosynthesis gene expression in sorghum. Thus, to provide greated insight into which transcription factors may be important, we implemented ATAC-seq across the same time course to define the DNA regions that were accessible to transcription factor binding, and the actual cis-elements bound by transcription factors in vivo.
Chlorophyll content increased exponentially over the 24 hours transition from dark to light period (Figure 1A). Principal Component Analysis (PCA) of samples over this time-course showed good clustering of biological replicates, and a clear separation along two main components (Figure 1B). To identify transcripts that were differentially expressed over this time-course, the data were subjected to the DESeq2 pipeline. Through pairwise comparisons between consecutive time-points, 9091 transcripts were identified as being differentially expressed. Of these, 1032 transcripts showed a response within 30 minutes of exposure to light, and the greatest response was seen between 0.5 and 2 hours when 6323 transcripts were differentially expressed (Figure 1C).
Three distinct algorithms were used to partition these differentially expressed genes into specific behaviours (Figure 2A). Clust, K-means and the WGCNA pipeline all detected similar groups of transcripts although the latter two identified more genes in each cluster. Because Clust appeared to produce the most conservative groupings we used its outputs for subsequent analysis. Clust classified 4064 genes into 20 clusters, which could then be manually placed into five groups (Figure 2A).
To provide a broad overview of the types of genes that were found in each of these clusters, Gene Ontology (GO) analyses were performed. Consistent with previous knowledge, GO terms for biological processes of photosynthesis (GO:0015979) were over-represented in the light-upregulated Clusters 4 and 9 (Figure 3). In addition, Cluster 9 was enriched in GO terms for such as plastid organization (GO:0009657) and chloroplast organization (GO:0009658), light reactions (GO:0019684), plastid membrane organization (GO:0009668) (Figure 3). To investigate the response of sorghum photosynthesis genes, Clusters in which the GO terms relating to photosynthetic processes were over-represented were analysed in more detail.
We next used the data to test the extent to which genes encoding components of the core C4 cycle also populated these photosynthesis clusters (Figure 4). Genes associated with the C4 cycle were classified as such based on previous work with sorghum. Notably, all genes associated with the core C4 cycle in mesophyll cells showed similar induction dynamics as those defined by the photosynthesis clusters (Figure 4).
To follow the downstream signals, we searched for genes of transcription factor (TF) families by comparing our significant gene list to the sorghum bicolor TFs in the Plant Transcription Factor Database (http://planttfdb.cbi.pku.edu.cn/). Although, the light-responsive pattern of some of TFs could not be clustered, others fitted well in clusters C2-5, and C9-10. The ATAC-seq data are currently being analysed, and so it is not possible to provide a summary of the findings that relate to transcription factor binding over the same time-course.
Three distinct algorithms were used to partition these differentially expressed genes into specific behaviours (Figure 2A). Clust, K-means and the WGCNA pipeline all detected similar groups of transcripts although the latter two identified more genes in each cluster. Because Clust appeared to produce the most conservative groupings we used its outputs for subsequent analysis. Clust classified 4064 genes into 20 clusters, which could then be manually placed into five groups (Figure 2A).
To provide a broad overview of the types of genes that were found in each of these clusters, Gene Ontology (GO) analyses were performed. Consistent with previous knowledge, GO terms for biological processes of photosynthesis (GO:0015979) were over-represented in the light-upregulated Clusters 4 and 9 (Figure 3). In addition, Cluster 9 was enriched in GO terms for such as plastid organization (GO:0009657) and chloroplast organization (GO:0009658), light reactions (GO:0019684), plastid membrane organization (GO:0009668) (Figure 3). To investigate the response of sorghum photosynthesis genes, Clusters in which the GO terms relating to photosynthetic processes were over-represented were analysed in more detail.
We next used the data to test the extent to which genes encoding components of the core C4 cycle also populated these photosynthesis clusters (Figure 4). Genes associated with the C4 cycle were classified as such based on previous work with sorghum. Notably, all genes associated with the core C4 cycle in mesophyll cells showed similar induction dynamics as those defined by the photosynthesis clusters (Figure 4).
To follow the downstream signals, we searched for genes of transcription factor (TF) families by comparing our significant gene list to the sorghum bicolor TFs in the Plant Transcription Factor Database (http://planttfdb.cbi.pku.edu.cn/). Although, the light-responsive pattern of some of TFs could not be clustered, others fitted well in clusters C2-5, and C9-10. The ATAC-seq data are currently being analysed, and so it is not possible to provide a summary of the findings that relate to transcription factor binding over the same time-course.