|Title||Visualizing brassinosteroid receptor hetero-oligomers in Arabidopsis roots|
|Source||University. Promotor(en): Sacco de Vries, co-promotor(en): Jan Willem Borst. - S.l. : s.n. - ISBN 9789461736543 - 195|
|Publication type||Dissertation, internally prepared|
|Keyword(s)||brassinosteroïden - biochemische receptoren - arabidopsis - wortels - beeldanalyse - signaalpeptide - signaaltransductie - fluorescentiemicroscopie - brassinosteroids - biochemical receptors - roots - image analysis - signal peptide - signal transduction - fluorescence microscopy|
Living matter is continuously challenged by the dynamics of its environment and intrinsic fluctuations. In the course of evolution, cells have developed mechanisms to detect and adapt to environmental and endogenous cues by the use of a wide array of receptors (Afzal et al., 2008). These receptors perceive specific signals, which, in turn, initiate a sequence of molecular events within the cells that convert signal perception into an adequate physiological response. Collectively, these processes of signal perception, signal transmission and cell adaptation represent so-called signal transduction pathways.
For the perception of signals such as hormones or pathogens cells are equipped with receptors that are often located at the cell surface. In plants, many of these receptors belong to the class of leucine-rich repeat receptor-like kinases (LRR-RLKs) (Shiu and Bleecker, 2001). They comprise an extracellular LRR domain for ligand binding, a transmembrane domain, which anchors them within the plasma membrane (PM) of their host cells, and an intracellular kinase domain for transducing the event of ligand binding into the cell interior. One of the best-described plant LRR-RLKs is the Brassinosteroid insensitive 1 (BRI1) receptor. Since the discovery in 1997 (Li and Chory, 1997) its mode of action has been studied extensively and has resulted in the elucidation of a complete set of molecular components constituting the brassinoteroid (BR) signal transduction pathway (Clouse, 2011).
BRs, the ligands of BRI1, are a group of polyhydroxy lactones that are structurally similar to animal steroid hormones (Grove et al., 1979). This class of phytohormones regulates several aspects of plant growth and development (Kutschera and Wang, 2012). During the last decade it has been shown that BRI1 indeed perceives BRs at the PM (Kinoshita et al., 2005), however, initiation of BR signal transduction requires interaction of BRI1 with other, non-ligand binding receptors (Nam and Li, 2002; Wang et al., 2008; Gou et al., 2012). These coreceptors belong to the family of Somatic embryogenesis receptor-like kinases (SERKs) and have a related structural architecture to BRI1, but with a smaller extracellular domain. Three members of this protein family are involved in BR signaling: SERK1, SERK3 (also known as BAK1 for BRI1-associated kinase 1), and SERK4 (also known as BKK1 for BAK1-like kinase 1). Besides their role as coreceptors of BRI1, the SERKs have also been implicated in various other signaling processes like somatic embryogenesis, male fertility, cell death regulation and plant immunity (Chinchilla et al., 2009).
In the first Chapter of this thesis, the BR signaling pathway was introduced in further detail and it was highlighted how genetic and biochemical approaches attributed to the identification of cellular components that link signal perception of BRs at the PM to BR dependent transcriptional regulation in the nucleus. Based on these findings a model for BRI1-mediated signal transduction was established, which often serves as a paradigm for plant PM receptor signaling. Even though the molecular determinants of BR signaling have been revealed, full mechanistic detail is still missing. The aim of this thesis was to describe BRI1-mediated signal transduction and the respective role of SERK3, the main coreceptor of BR signaling (Albrecht et al., 2008), at (sub)cellular level in Arabidopsis roots. For this purpose different fluorescence imaging techniques were applied, which allowed investigating the spatiotemporal localization and interaction dynamics of BRI1 and SERK3 in their natural environment.
One of the main microscopic methods applied throughout this thesis was fluorescence lifetime imaging microscopy (FLIM). Most imaging approaches, like confocal microscopy, only rely on fluorescence intensities as read-outs. However, the fluorescence lifetime τ is an additional parameter of fluorescence microscopy. This parameter is sensitive to the local environment of fluorescent probes and therefore can be exploited to illuminate cellular processes in live cells and tissues. In Chapter 2, the theoretical background of FLIM was introduced and it was illustrated how this technique can be used to reveal protein-protein interactions in Arabidopsis mesophyll protoplasts based on Förster resonance energy transfer (FRET). Next to a protocol for protoplast isolation and transient transfection, we provided a tutorial for analyzing time-resolved fluorescence intensity images using the software package SPCImage (Becker & Hickl). By determining the fluorescence lifetimes of a FRET donor fluorophore in the absence and the presence of a FRET acceptor chromophore physical interaction between the fluorescently tagged proteins of interest can be deduced. If the two proteins of interest, and thus the conjugated fluorophores, reside in close proximity FRET can occur and will result in a decrease of donor fluorescence lifetime. Besides the applicability to live cells and organisms, another important advantage of FRET-FLIM is the possibility to spatially resolve protein interactions within the two-dimensional imaging plane.
In Chapter 3, this technique was applied to live Arabidopsis roots. In our attempt to visualize the molecular events upon initiation of BR signaling, we performed FRET-FLIM on a double transgenic plant line expressing BRI1-GFP (Friedrichsen et al., 2000) and SERK3-mCherry. In accord with the current model of BR signal transduction (Jaillais et al., 2011a), a time-dependent and ligand-induced hetero-oligomerization between BRI1 and SERK3 was observed, similar to previous reports using coimmunoprecipitation (Wang et al., 2005; 2008; Albrecht et al., 2012). In addition, the spatially resolved FLIM images enabled us to localize these BRI1-SERK3 receptor complexes to restricted areas within the PM of live epidermal root cells, a cell file known to exhibit active BR signaling (Hacham et al., 2011). Application of brefeldin A (BFA), a fungal toxin reported to inhibit recycling (Nebenführ et al., 2002), allowed the visualization of intracellular receptor oligomers, which were most likely endocytosed from the PM. In contrast to the established BRI1 signaling model, FRET-FLIM revealed that a substantial amount of the BRI1-SERK3 hetero-oligomers was preformed. Constitutive receptor oligomerization is a well-established concept in animal signaling research (Gadella and Jovin, 1995; Martin-Fernandez et al., 2002; Issafras et al., 2002; Van Craenenbroeck et al., 2011), however in the plant field only a single study reported similar findings (Shimizu et al., 2010).
Besides the physical interaction between BRI1 and SERK3, also their localization and colocalization patterns were investigated (Chapter 3). As expected, most of the fluorescently tagged receptors localized to the PM. The intracellular fraction of BRI1-GFP mainly resided in punctate endosomal structures as documented previously (Geldner et al., 2007; Viotti et al., 2010; Irani et al., 2012). Similar endomembrane compartments were also observed for SERK3-mCherry, though to a lesser extent. In contrast to BRI1, for SERK3 an additional intracellular compartment was elucidated, the tonoplast. A further difference in the localization patterns of BRI1 and SERK3 was revealed when BFA was applied. Whereas BRI1-GFP strongly accumulated in BFA bodies, SERK3-GFP was only marginally affected, which hints at a differential endocytic pathway for both receptors. Although BRI1 and SERK3 showed distinct localization patterns, the two fluorescently tagged proteins also overlapped to some degree. Comparative colocalization analysis revealed that both the PM and the intracellular overlap between both LRR-RLKs is responsive to the BR signaling status. Application of brassinolide (BL), an endogenous BRI1 ligand, as well as BFA, which was demonstrated to elevate BR signaling activity (Geldner et al., 2007; Irani et al., 2012), resulted in an increased number of colocalizing BRI1 and SERK3 proteins. Thus FRET-FLIM and confocal imaging based colocalization analysis indicated that activation of the BR signaling system is accompanied by spatially distinct association of the two signal transduction inducing receptors BRI1 and SERK3.
As just illustrated, fluorescence microscopy is a valuable tool for investigating signal transduction processes in the natural environment of the executing molecular components. Unfortunately, a major drawback of the various techniques is that often only qualitative read-outs are obtained. Therefore we examined (Chapter 4) two different quantitative colocalization approaches in their ability to discriminate varying colocalizing proteinpopulations. First, the cytosolic colocalization of BRI1-GFP with the endosomal markerproteins ARA6 and ARA7 was investigated. Both tested and freely available ImageJ plugins Coloc2 and PSC Colocalization (French et al., 2008) revealed that BRI1-GFP preferentially localized to ARA7-mRFP labeled endosomal compartments. This finding was confirmed by manual counting of the respective endosomal structures and verified the reliability of the two quantitative colocalization methods. A biological explanation of the obtained result is given by the identity of the labeled endomembrane compartments. ARA7 localizes to both early endosomes (EEs), which enable recycling to the PM, and late endosomes (LEs; also known as multivesicular bodies [MVBs]), which are determined for vacuolar fusion. In contrast, ARA6 labels mainly LEs/MVBs. Thus both markers overlap to a certain degree during the maturation of LE but still have distinct localization patterns (Ueda et al., 2004; Ebine et al., 2011). Since BRI1 undergoes constitutive recycling (Geldner et al., 2007), our finding of preferential colocalization between BRI1 and ARA7 is plausible. In addition, similar observations were recently also reported for Flagellin sensing 2 (FLS2), an LRR-RLK involved in plant immunity, which also exhibits constitutive recycling (Beck et al., 2012).
After establishing the applicability of both colocalization approaches, we also intended to confirm our previous observation of increased BRI1 and SERK3 colocalization in response to BFA obtained with the Coloc2 plugin (Chapter 3). The application of PSC Colocalization indeed confirmed our initial colocalization results. The elevated colocalization of BRI1 and SERK3 upon drug treatment mostly like reflects the PM-stabilizing effect of BFA (Irani et al., 2012), which may also account for SERK3, since both Manders’ colocalization coefficients were increased. Nonetheless, a difficulty of quantitative colocalization analysis is the interpretation of colocalization coefficients obtained for individual images. However, they enable to assess image data sets, recorded under the same imaging conditions, in a comparative manner and that way allows drawing quantitative conclusions (Dunn et al., 2011). Colocalization analysis is not the only approach that suffers from qualitative read-outs and interpretations. The same accounts for FRET-FLIM studies. In particular the observation of preformed BRI1-SERK3 hetero-oligomers triggered our interest in developing a quantitative FLIM analysis procedure, which would be able to resolve ligand-independent and ligand-induced receptor complex populations. The details of our approach, which is based on time-correlated single photon count (TCSPC) measurements, were described in Chapter 4. Using this novel FLIM analysis procedure enabled us to estimate the different populations of BRI1 and SERK3 complexes. Upon BL stimulation around 10% of PM-located BRI1-GFP receptors were in complex with SERK3-mCherry. This finding is in line with recently reported data based on an in silico modeling approach (van Esse et al., 2012) and semi-quantitative coimmunoprecipitation (Albrecht et al., 2012), which suggested that active BR signal transduction involves between 1-10% of BRI1 receptors. Unfortunately, there are no quantitative data available for constitutive BRI1-SERK3 hetero-oligomers, even though their existence was proposed (Wang et al., 2005). Based on our imaging approach and analysis procedure we estimate that approximately 70% of PM BRI1-SERK3 heterooligomers are preformed. Finding such a considerable amount of preformed BRI1-SERK3 receptor complexes in the PM of root epidermal cells was intriguing since it contradicts the current view on BR signaling, which assumes a strictly ligand-dependent association of the two LRR-RLKs (Jaillais et al., 2011a). This posed the question when or where these preformed complexes are established. To address this point we investigated in Chapter 5 which cellular compartments harbor individual BRI1 and SERK3 receptors, and in which organelles these two receptors colocalize. Comparative colocalization analysis in live Arabidopsis roots revealed that both LRR-RLKs follow the traditional secretory and retrograde transport routes. These observations confirmed and extended previous findings for BRI1 using live cell (Friedrichsen et al., 2000; Geldner et al., 2007; Viotti et al., 2010; Irani et al., 2012) and electron microscopy (Viotti et al., 2010). For SERK3, to date only localization to EEs was suggested (Russinova et al., 2004).
Using the transient expression system of Arabidopsis protoplasts we could moreover show that both receptors also colocalize in the various endomembrane compartments of anterograde and retrograde trafficking. However, using electron microscopy a striking difference between their localization in retrograde endosomal compartments was elucidated. Whereas BRI1 was previously shown to reside at the membranes of the enclosed vesicles (Viotti et al., 2010), SERK3 was visualized at the limiting membrane of prevacuolar compartments (PVCs). This finding also explains, why SERK3, but not BRI1, was observed at the tonoplast (Chapter 3). Fusion of MVBs with the vacuole results in the release of BRI1 along with the inner MVB vesicles into the vacuolar lumen. PVC-localized SERK3 instead is incorporated into the tonoplast after membrane fusion. Collectively, the colocalization analysis of BRI1 and SERK3 with respect to endomembrane compartments revealed that subpopulations of both LRR-RLKs probably follow the same route to the PM, but that after endocytosis from the PM, possibly during the maturation of TGN/EEs to LEs/MVBs, a separation occurs. Still, these findings do not answer where or when BRI1-SERK3 hetero-oligomers are established. For that reason we applied FRET-FLIM on the subcellular compartment, in which BRI1 and SERK3 colocalized for the first time, the endoplasmic reticulum (ER). Similar to our observations at the PM of root epidermal cells (Chapter 3), most of the ER membrane did not show BRI1-SERK3 receptor complexes. Still, in restricted ER membrane regions strongly reduced donor fluorescence lifetimes were observed, indicating that BRI1-SERK3 hetero-oligomers are established already in the ER before entering the anterograde trafficking pathway. Finally, using a heat-shock inducible plant system we could confirm the establishment of BRI1-SERK3 hetero-oligomers shortly after biogenesis on their way to the PM. Thus, the observed preformed receptor complexes in the PM of root epidermal cells (Chapter 3) mostly likely originated from the ER and were inserted via targeted transport into the PM, the site where they fulfill their function as BR signaling units.