Large-scale gene expression profiling of the transition from cell division to endoreduplication in developing soybean seeds

Soybean seed development can be divided into two major phases. The first phase is characterized by the normal mitotic cell-cycle and rapid cell division. Cell division declines as seeds enter the phase of cell expansion characterized by the endocycle—replication of the genome and increase in cell size in the absence of mitotic cell divisions. At this stage cells accumulate storage proteins and oils and increase their size. The first goal of this project was to select the genes differentially expressed between the cell division and cell expansion phases. The second goal was to perform transcriptional profiling of a large subset of such genes at several time-points throughout the transition from cell division to endoreduplication to answer such questions as what are the main transcriptional patterns in the soybean seeds during the transition from cell-division to endoreduplication; what genes follow these patterns; what transcriptional patterns could be related to the transition being studied; are there novel genes that follow these patterns; what are the relationships between the expression patterns of certain known and novel genes? The first goal was accomplished through the use of Suppression Subtractive Hybridization [Diatchenko et al., 1996]. Two subtracted cDNA libraries, each containing approximately 9,000 clones harbored in bacterial freezer cultures, have been prepared and analyzed. These libraries were used for the selection of a set of probes for preparation of a cDNA microarray. The microarray was used for transcriptional profiling studies of soybean seeds collected at 7, 14, 18, and 22 DAF (Days After Flowering). Microarray probes that corresponded to the transcripts whose levels significantly changed between at least one pair of time-points were identified via sequencing and sequence alignments against the major public sequence databases. Clustering analysis of the transcription profiles generated by these probes was carried out to answer the questions posed above. In addition, almost 80 novel transcripts were identified and present a pool of candidates for further biochemical and molecular biological analyses