Morphogenesis of multicellular organisms is based on differentiation of individual cells. This requires that individual cells take different developmental fates. This has to be organised on a level that transcends individual cells. How can this be achieved? There are generally two mechanisms: either the cells are already different from the very beginning and this difference is first not visible, but nevertheless present and will be just uncovered during the course of the development. Or the cells are principally able to take any fate and they assign individual fates by communicating with their neighbours. Plants with their pronounced flexibility are expected to use the second strategy. However, there exist cases, where the pattern of cell divisions appears to be very strict, almost stereotypical (so called cell lineages), and this led to the model that here cell fate depends on the sequence of cell divisions (whereby during specific, formative divisions the daughter cells get asymmetric sets of developmental factors that will drive them to different routes of differentiation). Especially the pattern of cell differentiation in the root of the model plant Arabidopsis was thought to be a case of such cell-lineage driven differentiation. By laser ablation and the use of patterning mutants it was later possible to show that the observed cell lineages just covered the fact that also here cells determine their fate by exchanging signals and some of the signals could be even identified.
To find some background on the concept of cell lineage, how it was disproved and how such signals were identified refer to Ilias: Fakultät für Chemie und Biowissenschaften - current semester - BIO_MA_FOR_1201_Plant_Cell_Biology
Your task is to design a strategy that follow cell lineage. The question is, whether meristemoids in cotyledons are laid down by cell lineage very early in embryogenesis in the model plant Arabidopsis. Meristemoids are small islands of dividing cells that produce a defined outcome, for instance a guard cell apparatus or a trichome. In adult leaves, the pattern of guard cells is regulated by inhibitory signals from already existing guard cells (a peptide called epidermal patterning factor 1), which are perceived by a receptor (called too many mouths) that, upon binding of the ligand will activate a MAPK-cascade that will block the cell cycle, such that no additional guard cell can be formed. But what about the guard cells on the cotyledon - they must be determined much earlier, already during embryogenesis. When? Where?
To approach this, it is necessary to define a cell-type specific promotor and place it upstream of a reporter (for instance GFP). This can be done, using the GENEVESTIGATOR search engine. You have to get an account, make sure that you use your kit.student account, because then you will get free access as member of an academic institution. You have to install JAVA on your computer, then the engine (everything is quite nicely explained). The Genevestigator software compiles numerous expression data from many model organisms, including Arabidopsis and Rice. You can, for instance, find out the expression pattern for a gene of interest (which tissues, which conditions). You can also search for candidate genes that are expressed in a particular tissue or under specific conditions. The software is intuitive and after playing around a bit, you can look rapidly for what you need.
There are experiments, where people have got protoplasts from guard cells (you can get those by searching for "conditions"), the programme provides you the GenBank accession numbers and when you mouse over the name, you will also find out what gene it is.
Next step is to get the sequence from GenBank for your candidate. You have then to find out the start codon, and then you have got already your promotor of choice.
Now, you have to design primers for genomic GATEWAY PCR to amplify the promotor, find out, which promotor would be suitable on the GATEWAY site in Gent (search for promoter-reporter or transcriptional reporter) - there are several possibilities.
Then, you are set - the description, how you actually do the transformation, is not the objective of this question.