With a rapidly growing population and decreasing farmland, the world will need to dramatically increase crop yields. The cells in the edible parts of corn, beans, rice, wheat, and peas undergo a specialized process called endoreduplication where the cells replicate their DNA many times. Although individual cells undergo different amounts of endoreduplication, the proportion of cells remains constant from plant to plant. When geneticists try to increase endoreduplication, they find that the response of seemingly identical cells is somewhat random: some cells respond strongly, some moderately, and some not at all. This observation raises one of the biggest questions in developmental biology: how do identical cells make different decisions? The proposed research will investigate the counterintuitive hypothesis that individual plant cells make these decisions about how much to endoreduplicate utilizing small, random molecular differences. This hypothesis is surprising: biologists would normally expect randomness to be suppressed, since development is highly reproducible, always creating the same proportion of endoreduplicated cells. Instead, randomness may in fact be required to produce this regular portion of endoreduplicated cells. Testing this hypothesis will require advanced use of computers to examine individual cells over time. University students and middle school girls will simultaneously be involved in this analysis as well as being trained in computational skills.
Endoreduplicating cells of crop plants are buried deep within the seed coat where they are experimentally inaccessible. Endoreduplication will thus be studied in Arabidopsis thaliana sepals, whose epidermal cells are accessible for live imaging as they endoreduplicate. The transcription factor Arabidopsis thaliana MERISTEM LAYER1 (ATML1) drives sepal endoreduplication, and may cause it to occur randomly. Since ATML1 is expressed in all epidermal cells, it is unclear how ATML1 causes only a few cells to endoreduplicate. Preliminary data show that ATML1 triggers endoreduplication in a dosage-dependent manner: high ATML1 makes all epidermal cells endoreduplicate, low ATML1 prevents any endoreduplication, and individual sepal cells have varying levels of ATML1. The hypothesis tested here is that randomly fluctuating ATML1 causes different levels of endoreduplication in different cells; a few cells exceed ATML1's threshold, triggering endoreduplication, while other cells remain below the threshold and do not endoreduplicate. In Aim 1, we will determine whether ATML1 fluctuations initiate endoreduplication. The associated education goal will be to work with undergraduate students to analyze images generated in this aim. For Aim 2, we will determine whether ATML1 feedback loops drive endoreduplication by amplifying differences in ATML1 expression and, in an associated educational aim, students will write and publish a computational modeling glossary to better understand the newly generated models. Finally, in Aim 3 we will determine the molecular mechanism through which ATML1 induces endoreduplication in a dosage-dependent manner. We will also develop an outreach to middle school students for the analysis of images from Aim 3 in a hands-on, web-based image analysis workshop.
