Molly Campbell (MC): Please can you tell us more about Drop-seq, how it works and why you decided to apply it in plant biology for the first time?
Christine Shulse (CS): Drop-seq works by capturing individual cells in tiny droplets along with a barcoded bead. The cell breaks open inside the droplet, releasing its RNA, which is then captured by the barcoded bead. A researcher can quickly capture and barcode the RNA from thousands of cells, and ultimately pool them for a single sequencing run. The barcode allows the researcher to determine which RNA molecules came from the same cell. Plants are a major scientific focus area within the Biosciences Area at Berkeley Lab due to their use as feedstocks for biofuels and their ability to respond to environmental change. Therefore we wanted to see if Drop-seq, this really powerful and open-source tool for single-cell RNA sequencing, would work on plants.
MC: Why is it important for us to gain insight into the roles of certain genes in plants, what applications may this knowledge have?
CS: There are many reasons why we would want to know the roles of specific genes in plants – for instance, if we wanted to engineer a drought-tolerant biofuel feedstock or crop species, we would need to know which genes play a role in drought tolerance. The same could apply for engineering plant species that are resistant to pests, etc. The great thing about Drop-seq is that we can use it to identify these genes AND identify the cell type in which they are expressed.
MC: As the technology was first published in a paper in 2015, why has it taken four years for Drop-seq to be applied in plant biology? What deterred plant scientists from using it?
CS: Plant cells, unlike animal cells, have a cell wall that must be removed before single-cell sequencing. Additionally, plant cells are generally large and non-uniform in size. I think this deterred some plant scientists from attempting to adapt Drop-seq for use with plants.
MC: What were the main successes and challenges you faced in this research project?
CS: One of our main successes was of course the technical success of generating >12,000 single cell profiles from Arabidopsis root. Once we had those, we were also successful in reconstructing endodermis development by ordering the single cell transcriptional profiles based in part on their expression of genes known vary in expression through development. We then looked at genes with increasing or decreasing expression over this “pseudotime” we were able to identify additional genes that may be important in endodermal development. My biggest initial challenge was generating large amounts of healthy protoplasts, which was solved by partnering with our co-authors at UC Davis, who have a well-developed protocol for protoplasting Arabidopsis root.
MC: How can you be sure that the process of protoplasting has not altered the cell to the extent that you cannot gain insight into normal functioning?
CS: In order to make sure that the protoplasting did not result in major changes in gene expression, we combined the transcriptomes of all our single cells into a “pseudobulk” transcriptome, and compared it to a traditional bulk transcriptome generated from a whole-root sample where we did not do any protoplasting prior to RNA extraction. We found a high correlation between the single cell “pseudobulk” transcriptome and the traditional transcriptome, so we felt confident that the protoplasting did not significantly alter gene expression.
MC: Please can you tell us about your research findings and their significance?
CS: First of all we found that Drop-seq can be used to quickly generate expression profiles from individual plant cells. Secondly, we identified genes as markers for specific cell types in the root – in other words, they were highly expressed in a specific cell type (such as cortex, hair, endodermis, etc. cells) but not highly expressed in other cell types. These genes may play important roles in the development or function of these cell types and are candidates for further investigation. Thirdly, by profiling roots grown with and without sucrose, we found that roots grown without sucrose were enriched in meristematic (immature) cells. This implies that sucrose drives root maturation in some fashion.
MC: What advice would you give to other plant biologists considering using Drop-seq in their research?
CS: Go for it! I am excited to see this technology used in other plant species and to answer other questions in plant biology. As for advice, I would say identify a protoplasting protocol specific for your plant and tissue of interest, and do the proper controls to ensure that the protoplasting is not affecting your expression profiles.
MC: What are your next research steps?
CS: The next steps for Drop-seq in plant biology will be to extend its use to non-model plants, and to use it to assess the effects of various environmental conditions, such as drought, salt, pathogen infection, or pH levels.
Christine Shulse was speaking with Molly Campbell, Science Writer for Technology Networks.