Hybrid Modelling to Investigate Plant Immunity

Plant defence responses, which are often associated to the development of a form of programmed cell death called the hypersensitive response (HR), need to be tightly regulated to be not only efficient but also beneficial. Activation of plant immunity is achieved by multiple transcriptional regulators that reprogram cell transcription to favour defence over routine cellular requirements. MYB30 and MYB96 are two Arabidopsis MYB transcription factors that positively regulate plant defence and HR responses. Both MYB30 and MYB96 modulate plant metabolism by activating the production of very long chain fatty acids (VLCFAs) following bacterial inoculation (Raffaele et al., 2008) (Seo et al., 2011). Moreover, as it is usually the case for key cellular regulators, the activity of MYB30 is tightly regulated through protein-protein interactions and post-translational modifications (Froidure et al., 2010; Marino et al., 2013; Raffaele and Rivas, 2013). Furthermore, MYB96 directly interacts with MYB30 in the plant cell nucleus and MYB30 and MYB96 are able to regulate each other’s expression, suggesting a complex molecular interaction between these two transcription factors. Taken together, these results show that MYB30 and MYB96 act in cooperation as the main components of a MYB transcriptional rheostat controlling a regulatory network that promotes and controls the initiation of hypersensitive cell death trough the production of VLCFAs.
To gain a better understanding of the MYB30-/MYB96-triggered defence response, we have initiated mathematical modelling of this system. We begun model reconstruction by listing all interactions involved in the system and obtained two types of biological network with different characteristics. First, a regulatory network composed of various types of biological entities (genes, proteins, biochemical reactions …) that are connected through directed interactions. Second, a metabolic network composed of balanced biochemical reactions. Since different methods exist dedicated to the analysis of these two types of networks, we chose to use a hybrid model where the two networks are modelled independently and integrated during simulation. To model the regulatory network, we use logical modelling in which biological entities can display a finite number of states and are updated using logical rules (Morris et al., 2010). The metabolic network is modelled using a constraint-based method called Flux Balance Analysis (FBA) (Orth et al., 2010). The integration of qualitative states of regulatory network entities and FBA is made through their common biological entities: biochemical reactions. Indeed, qualitative states of reactions can be translated into quantitative constraints that are added to the FBA model as linear equations. The usage of hybrid modelling allows using appropriate simulation methods for different model modules. Moreover, it facilitates curation of individual model parts (Le Novère, 2015).
The reconstruction and the simulations of this model are expected to help us investigate the specific roles of MYB30 and MYB96 during plant defence, as well as their effect on metabolism through lipid production.