by Alt-Jan Van Dijk
Plant Research International, Wageningen University
Progress in sequencing technology has had a clear impact on flowering research. For example, ChIP-seq has been applied to study molecular aspects of transcriptional regulation of flowering. Also, RNA-seq data for different stages of flower development have been generated in various species (e.g. Singh et al., 2013, Wang et al., 2014, Mantegazza et al., 2014, Han et al., 2017). However, what is still lacking to a large extent is the high-resolution characterization of spatial variation in gene expression. This holds true for plant transcriptomes in general but also specifically for flower development.
A paper by Giacomello et al., 2017 presents an approach to profile plant transcriptomes with high spatial resolution. To do so, it employs semi-randomized primers with barcodes, which indicate the position of a primer on an ~1,000 spot array. Each spot is 100 μm in diameter and provides spatial resolution. After fixation and permeabilization of tissue sections on the array, polyadenylated transcripts are captured by the primers. The transcripts are then converted into cDNA and analyzed by sequencing.
Amongst others, Giacomello and colleagues applied their approach to analyze Arabidopsis inflorescence tissue. Comparison with the AtGenExpress Development dataset (Schmid et al., 2005) for five broadly defined tissue domains (stem, meristem, flowers of stage 9, 11 and 12) revealed a reasonable overlap. Note that, of course, the key difference between AtGenExpress and the data by Giacomello et al. is that this new data provides a much higher spatial resolution. Some of the patterns provided for individual genes are rather convincing. For example, they observed ubiquitous expression of the housekeeping gene TUB2, whereas the floral organ identity genes showed expression specifically in flowers. In spots under stamens, markedly higher expression was observed for AP3, PI, and AG than for AP1 and AP2, in agreement with the known expression patterns of these genes. When going through results from individual replicates, for example for AP3 and PI, there also seems to be quite a bit of variation between the replicates. Whether this is evidence of true biological variation or still might indicate technical variation, is not clear to me.
Be that as it may, this study not only presents high resolution data on spatial transcriptomes in floral tissues. In addition, it demonstrates how such data can be analysed. To do so, two key steps are taken. First, not just gene expression levels are used as the variable of interest, but ‘pathway scores’ which reflect the expression level of groups of genes that constitute pathways. Second, the influence of location on expression of the pathways is analysed by not just comparing each pair of locations to each other. Instead, a model is built to analyze the influence that various factors have on expression levels. These factors involve in particular the spatial location, both at the tissue level and at the level of the different spots on the array. One of the reported findings is the enrichment of the stamen filament development pathway in floral stage 11, and that the pollen exine formation pathway was altered in floral stages 10 and 11. Note that these stages indeed produce exine, one of the major constituents of the pollen wall that is deposited on the pre-pollen cells.
This paper represents the next step in the application of sequencing technology to study flowering. It is interesting to see how more and more data relevant for the study of flowers and their development is being generated using sequencing-related techniques. Time course data has been available for a while. Given the increased resolution with which spatial aspects of transcriptome expression can be measured, an important next step will be to measure the expression with both high temporal and high spatial resolution. In addition, it will be exciting to see if applications of sequencing such as ChIP-seq also can be given high spatial resolution.
Giacomello S, Salmén F, Terebieniec BK, et al. 2017. Spatially resolved transcriptome profiling in model plant species. Nature Plants 3:17061. doi: 10.1038/nplants.2017.61
Han Y, WanH, Cheng T, Wang J, Yang W, Pan H, Zhang Q. 2017. Comparative RNA-seq analysis of transcriptome dynamics during petal development in Rosa chinensis. Scientific Reports 7: 43382. doi:10.1038/srep43382
Mantegazza O, Gregis V, Chiara M, Selva C, Leo G, Horner DS, and Kater MM. 2014. Gene coexpression patterns during early development of the native Arabidopsis reproductive meristem: novel candidate developmental regulators and patterns of functional redundancy. Plant Journal 79:861-877. doi:10.1111/tpj.12585
Schmid M, Davison TS, Henz SR, Pape UJ, Demar M, Vingron M, Schölkopf B, Weigel D, Lohmann JU. 2005. A gene expression map of Arabidopsis thaliana development. Nature Genetics 37, 501. doi:10.1038/ng1543
Singh VK, Garg R, Jain M. 2013. A global view of transcriptome dynamics during flower development in chickpea by deep sequencing. Plant Journal 11, 691. doi
Wang H, You C, Chang F, Wang Y, Wang L,Qi J, Ma H. 2014. Alternative splicing during Arabidopsis flower development results in constitutive and stage-regulated isoforms. Frontiers in Genetics 5: doi:25. 10.3389/fgene.2014.00025