Firstly, I confirmed that BPS1/BPS2/BPS3 are indeed BIK1 substrates: they associate with BIK1, when co-expressed in N. benthamiana, and are phosphorylated by BIK1 in vitro. Due to time / technical constraints, most of my subsequent work focused on BPS1. Thus, I found that in planta BPS1 is a highly phosphorylated protein. To find BIK1-dependent phosphosites on BPS1, I performed in vitro kinase assays and identified phosphosites by mass spectrometry. I also demonstrated that BIK1 and BPS1 associate within the same complex in Arabidopsis; the interaction being independent of PAMP treatment.
Secondly, I verified whether BPS1 is genetically involved in PTI. I analysed several PTI responses in bsp1 mutants. I found that reactive oxygen species (ROS) production was reduced in bsp1 compared to wild-type. bsp1 mutants demonstrated reduced seedling growth inhibition and impaired stomata closure in response to PAMP treatment. Together, these data suggest that BPS1 acts genetically as a positive regulator of PTI.
Then, I asked which proteins BPS1 associates with in planta? I purified FLAG-tagged BPS1 and identified its associated proteins with or without PAMP treatment. I found BPS1 interacted with proteins of the FLS2/BAK1 PRR complex after PAMP treatment, thus confirming its involvement in very early PTI signalling. I found that BPS1 is phosphorylated in vitro by the co-receptor protein BAK1, and identified the corresponding phosphosites using mass spectrometry. Thus, the above data place BPS1 at the level of the PRR complex and associated cytoplasmic kinases.
The fact that PTI responses triggered by different PAMPs are affected in bps1 mutant suggests that BPS1 acts on a common regulator of different pathways. Thus, BIK1 itself becomes a good candidate for being a BPS1 target. BIK1 is known to be tightly regulated at the protein level; with a balance between its “inactive” and “activated” form being important for the amplitude of immune signalling. Therefore, I checked BIK1 protein levels in bps1 and BPS1 overexpression lines. Before PAMP treatment, BIK1 protein level is reduced in BPS1 overexpression line, while it is increased in bps1 mutant line. This result is consistent with the observed reduced ROS production in BPS1 overexpression line, as ROS production is BIK1-dependent. However, after PAMP treatment, in bps1 mutant there is a clear reduction in the accumulation of activated BIK1 (phosphorylated form). Moreover, inhibition of the proteasome leads to the restoration of activated BIK1 levels in bps1. Thus, in bps1, there is a lack of activated, signalling-competent form of BIK1. This fits with the observed reduced PTI outputs in bps1 line. Together, these results demonstrate that BPS1 acts on a convergent point of PTI signalling, BIK1, thus affecting defence responses triggered by different PAMPs. Moreover, BPS1 protein level is important for correct accumulation of both “inactive” and “activated” forms of BIK1.
I presented my results at the SEB Cell Symposium 2017 “From proteome to phenotype: role of post-translational modifications”, Edinburgh and at the International Conference on Arabidopsis Research (ICAR) 2018 in Turku, Finland. I regularly presented my project at Prof. Zipfel’s lab meetings as well as at TSL seminars.
