The hot, the cold and the tulip : the regulation of flowering time and dormancy release

The ornamental geophyte Tulipa gesneriana is the most cultivated bulbous species in the Netherlands. It is widely grown in the field for vegetative propagation purposes and in greenhouses for the production of high quality cut flowers. Over the last decade, the tulip bulb industry is affected by the rapid climate change the world is facing. Temperature is rising and influences the vegetative to reproductive phase change (floral induction) inside the tulip bulbs in spring and processes that are occurring during winter, such as dormancy release. In this thesis the two temperature-dependent processes related to tulip flowering, being floral induction and dormancy release, were investigated in detail with a special focus at the molecular level. Flowering time has been studied in a broad range of species, including the model species Arabidopsis thaliana and Oryza sativa. The current understanding of this process can be translated to non-model species, such as tulip, through a ‘bottom-up’ and ‘top-down’ approach (Chapter 2). For the ‘bottom-up’ approach conservation of molecular pathways is assumed and researchers make use of sequence homology searches to identify candidate genes. The ‘top-down’ approach starts from large scale data mining, such as RNA-sequencing (RNA-seq) data or microarrays, followed by the association between phenotypes, genes and gene expression patterns. Here, a comparison with data from model plant species is made at the end of the process and this also leads to the identification of candidate genes for a particular process. Large scale genomics data mining in tulip is only possible via transcriptome analysis with RNA-seq derived data, because no full genome-sequence is present at this moment. Genome sequencing remains a challenge for species with a large and complex genome, containing probably a large number of repetitive sequences, which is the case for tulip and lily. In chapter 3 a high quality transcriptome of tulip and lily is presented, which is derived from a collection of different tissues. In order to obtain good transcriptome coverage and to facilitate effective data mining, different filtering parameters were used. This analysis revealed the limitations of commonly applied methods used in de novo transcriptome assembly. The generated transcriptome for tulip and lily is made publicly available via a user friendly database, named the ‘Transcriptome Browser’. The molecular regulation of the temperature-dependent floral induction was studied through the use of RNA-seq (Chapter 4). A better understanding of this process is needed to prevent floral bud blasting (dehydration of the flower) in the future. The development at the shoot apical meristem (SAM) was morphologically investigated in two contrasting temperature environments, high and low. Meristem-enriched tissues were collected before and during the start of flower development. The start of flower development is morphologically visible by rounding of the SAM and correlates with the up-regulation of TGSQA, an AP1-like gene. A ‘top-down’ approach was used to identify possible regulators of the floral induction in tulip. However, Gene ontology (GO)-enrichment analysis of the differentially expressed genes showed that the floral induction, maturation of the bulb and dormancy establishment are occurring around the same period in time. Therefore a ‘bottom-up’ approach was followed to identify specific flowering time regulators based on knowledge obtained from other species. Expression analysis in tulip, heterologous analysis in Arabidopsis and yeast two hybrid-based protein-protein interaction studies revealed that Tulipa gesneriana TERMINAL FLOWER 1 (TgTFL1) is likely a repressor of flowering, whereas Tulipa gesneriana SUPPRESSOR OF OVEREXPRESSION OF CONSTANS-LIKE2 (TgSOC1L2) acts probably as a floral activator. Another well-known flowering time regulator is FLOWERING LOCUS T (FT), which is a member of the PEBP gene family found in Arabidopsis and many more plant species. In tulip and lily, a total of four highly similar sequences to FT and HEADING DATA 3A (Hd3a) were identified (Chapter 5). Overexpression of Lilium longiflorum FT (LlFT) and TgFT2 in Arabidopsis resulted in an early flowering phenotype, but upon overexpression of TgFT1 and TgFT3 a late flowering phenotype was observed. The tulip PEBP genes TgFT2 and TgFT3 have a similar expression pattern during development, but show a different behaviour in Arabidopsis. Therefore the difference within the amino acid sequence was investigated, which resulted in the identification of two important amino acids for the FT function, which appeared to be mutated in TgFT3. Interchanging of these amino acids between TgFT2 and TgFT3 resulted in conversion of the phenotype, showing the potential importance of these positions in the protein and these specific amino acids for the molecular mode of action of these two proteins. Based on all the data, LlFT is considered to play a role in creating meristem competency to flowering related cues and TgFT2 to act as a florigen involved in the floral induction. The function of TgFT3 is not clear, but phylogenetic analysis suggests a bulb specific function. After the floral induction and completion of flower development inside the tulip bulb, a period of prolonged cold is required for proper flowering in spring. Low temperature stimulates the re-mobilization of carbohydrates from the scale tissues to the sink organs, such as the floral stem, floral bud and leaves. Not many details are known about the molecular and metabolic changes during this cold period. In chapter 6, first insights are shown on the development of the different tissues inside the bulbs. The floral bud appears to be the least active tissue in comparison with the floral stem and leaves, suggesting a type of floral bud dormancy in tulip. However, metabolic changes are suggesting that the floral bud is still showing active cell division and/or preparation for elongation by turgor-driven cell wall extension. Dormancy of all tissues seems to be released ten weeks after planting and is correlated with the increase of glucose levels. In the leaves, from this same moment, photosynthesis related genes are up-regulated suggesting that the leaves are preparing for photosynthesis while still beneath the soil surface. At the end of the thesis a glance is given at different perspectives of the tulips life cycle, categorizing tulip as a perennial, biennial or annual plant species, respectively. The perennial way of life is applicable when growing bulbs from seeds, while biennial and annual are more in relation to vegetative propagation. Also the importance of bulb size is highlighted, because it will determine if the bulbs are able to flower or not the following spring. Two scenarios are discussed related to availability of energy in the presence of carbohydrates and meristem incompetency to floral inducing signals. Throughout all research done for this thesis, it became clear that tulip bulbs and seeds have a lot in common. By combining the knowledge of processes in different plant species or developmental systems it is possible to understand how flowering and dormancy release are regulated and this provides us with novel insights how these processes are regulated in bulbous plant species, such as tulip.