Expression Profiles and Signaling Networks During Abiotic Stress: The Function of Noise
Hans J. Bohnert,Mark Fredricksen,Qingqiu Gong,Shisong Ma,Saori Miyazaki,Cecile Sano
2003-01-01
Abstract:Investigations of plant responses to abiotic stresses has recently been advanced by new technologies that promise to move from the analysis of single factors to an integrative understanding of entire systems. We will discuss results from relevant new technologies. First, collections of mutants specifically generated to report altered response to, for example, ionic and osmotic imbalances or pathogen attack. Literally hundreds of such mutants have been characterized to some degree. For many genes that have been recognized in these mutant lines a clear relationship to a specific function in the stress response is difficult to assess. A picture emerges suggesting that any deviation from cellular homeostasis that is perceived by different signaling events leads to responses that are not necessarily specific to the particular stress. Instead, there seems to be an inbuilt element of noise that may serve in eliminating unspecific stress responses and focusing on a specific reaction sequence. A second change includes availability and analysis of stress -, developmental stage-, and tissue-specific transcript populations. Their value depends on, and is greatly enhanced by, genome sequence information, large-scale EST collections, and comparative approaches that extend information known from model species by investigating relatives that are characterized by different adaptations. The size of such collections now permits comparisons that address plant life style based on transcript abundance and specificity, for example, in the comparison between Arabidopsis thaliana and its more stress-tolerant relative Thellungiella halophila (salt cress). A third revolutionary tool, microarrays and oligochips, provide recordings of dynamic changes in transcript abundance and regulation relative to the steady-state condition or another baseline. The design of experiments, in addition to necessities that establish statistical confidence, for transcript profiling is of crucial importance for the veracity of the message that can be extracted. For example, shock treatments, which plants rarely encounter in nature, generate a very different response profile compared to that obtained after gradual imposition of stress. Also, measurements that record the progression of abiotic stress responses over time are more indicative of the capacity of a species, mutant, or breeding line to mount a productive defense. A comparison of abiotic stress response time course experiments with corn, Arabidopsis, Thellungiella, yeast, and rice - including glycophytes and halophytes - seems to indicate some differences in the early stress response that could be at the basis of halophytism. Differences have been seen in signaling components, and in how the "downstream" components, for example ion transporters, are either synthesized or regulated. The expression profiles recorded in different models seem to indicate that noise in transcript expression, and the ability to decode the initial response distinguish salinity stress tolerant and sensitive species. (Supported by NSF, DBI-0223905, the U Arizona and UIUC Colleges of Agriculture, and UIUC internal funds).