Objective 1: Define and collect the different embryogenic cell types present in microspore culture using GFP-based reporters, time-lapse imaging, and cell sorting.
The three GFP reporters mark an overlapping set of embryogenic cells, each with different fates i) differentiated embryos (DR5, LEC1 and GRP) and ii) unorganized callus (LEC1 and GRP) (Figure 1).
Time lapse imaging:
Time-lapse imaging of the embryo-expressed GFP reporter lines was used to determine the fate of the different embryogenic cell types in culture. We designed an efficient immobilization system to fix the position ofembryogenic structures and to follow them in time without interfering with their development. We optimized the imaging parameters by testing different microscopes/imaging platforms and different software packages for data handling and image processing. Time-lapse imaging of GFP-positive embryogenic structures at day 5 indicated that only the compact structures develop into differentiated embryos (Figure 2). Embryogenic callus-like structures never form differentiated embryos; the majority of these both stop growing and lose their embryo identity, visualized by a loss GFP embryo marker expression.
Cell sorting:
We tested two systems for their utility in sorting GFP-positive and GFP-negative cells from microspore culture: 1) the BioSorter (Union Metrica); and 2) Fluorescence Activated Cell Sorting or FACS. In addition, we tested filtering as a means to enrich for differentiated embryos in older cultures. The BioSorter was not useful for sorting GFP-positive and negative populations in microspore embryo cultures. Using FACS, we could separate 100% GFP-positive and 100% GFP-negative cell populations in the GRP:GFP and LEC1:GFP lines until day 6 of culture (Figure 3). However, DR5: GFP expression was too weak to allow cell sorting. Filtering was the most efficient method for isolating older embryogenic structures, around 94% of the filtrate corresponded to GFP-positive embryogenic structures (compact and callus-like), while the remaining 6% was pollen(Figure 4).
Objective 2: Transcriptome analysis
We used freshly isolated microspores, which can be easily isolated in large numbers, to optimise different steps in the mRNA-seq protocol, including mRNA isolation, mRNA amplification and mRNA-seq library construction before proceeding with the FACS samples.
After many trials we decided to use the Picopure kit (Thermo Fisher)) for RNA isolation in combination with the MessageAmp kit (Thermo fisher) for mRNA amplification. 40,000 cells/sample was set as the lower limit for each sample. We collected GFP-positive and GFP-negative cells by FACS or filtering during 6 months. Samples were collected from microspore culture with and without TSA, at day 0, 2, 4 and six days after stress treatment, using three replicates from two different Brassica napus lines. Most samples are ready for sequencing. Objective 3: Determine the function of candidate genes
Due to the large amount of time needed to set up the FACS system and to collect the samples we did not have time to study the function of candidate genes.
New objective: Study the cellular characteristics of embryogenic cells.
We prepared samples for light and transmission electron microscopy (Figure 5). We observed differences between the different types of embryogenic structures. Embryogenic calli, which do not form embryos, are characterized by loss of cell adhesion (Figure 6), most likely resulting from a decrease in the amount of pectins, arabinogalactan proteins and callose in the cell wall. Moreover in embryogenic calli we observed a massive increase in endoplasmic reticulum, ER stress and autopgahy-related markers (Figure 7), including lipohagy (Figure 8), as well as cell death. These abnormalities may explain why these structures do not continue with the embryogenic pathway.
