ABSTRACT
A large part of a plants’ developmental plasticity relies on the activities of the phytohormone auxin and the regulation of its own distribution. This process involves a cohort of transcriptional and non-transcriptional effects of auxin on polar auxin transport, regulating the abundancy, biochemical activity and polar localization of the molecular components, predominantly PIN auxin exporters. While the transcriptional auxin signaling cascade has been well characterized, the mechanism and role of non-transcriptional auxin signaling remains largely elusive. Here, we addressed the potential involvement of auxin-induced Ca2+ signaling in auxin’s inhibitory effect on PIN endocytic trafficking. On the one hand, exogenous manipulations of Ca2+ availability and signaling effectively antagonized auxin effects suggesting that auxin-induced Ca2+ signaling is required for inhibition of internalization. On the other hand, we addressed the auxin-mediated inhibition of PIN internalization in the auxin signaling (tir1afb2,3) or Ca2+ channel (cngc14) mutants. These mutants were strongly defective in auxin-triggered Ca2+ signaling, but not in auxin-inhibited internalization. These data imply that, while Ca2+ signaling may be required for normal PIN trafficking, auxin-mediated increase in Ca2+ signaling is not a direct part of a downstream mechanism that mediates auxin effects on Brefeldin A-visualized PIN intercellular aggregation. These contrasting results obtained by comparing the mutant analysis versus the exogenous manipulations of Ca2+ availability and signaling illustrate the critical importance of genetics to unravel the role of Ca2+ in a process of interest.
INTRODUCTION
Calcium cross-links pectinate polymers and is therefore an important structural determinant of the cell wall (Feng et al., 2018; Cosgrove and Anderson, 2020), and inside the cell it directly modulates the biochemical activities of proteins, Ca2+ sensors (relays and responders) and phospholipids, impacting on numerous cellular processes (Himschoot et al., 2017; Kudla et al., 2018). The pleiotropic activities of Ca2+ necessitate submicromolar range Ca2+ levels in the cytoplasm, while in the apoplast and in the lumen of organelles, Ca2+ levels are several orders of magnitude higher (Stael et al., 2012; Costa et al., 2018). Such steep concentration gradients allow to trigger a local significant increase in Ca2+ levels by the simple opening of a few channels in response to a specific stimulus (Demidchik et al., 2018). In a typical signaling cascade, the cytoplasmic increase of Ca2+ is decoded by specialized proteins that translate the Ca2+ signal into defined cellular responses, such as the modulation of channels or kinases (Kudla et al., 2018).
The plant Ca2+ signaling toolkit is strongly diversified in comparison to the one in animals (Edel et al., 2017), most prominently reflected in the existence of plant and animal specific Ca2+ signaling components, such as inositol (1,4,5)-triphosphate receptors, and ryanodine receptors. Despite this important diversification, the current commonly used Ca2+ pharmacology consists of very general inhibitors or chelators, or inhibitors that were designed to target mammalian Ca2+ channels and signaling components (De Vriese et al., 2018). In most cases, the molecular targets of the inhibitors are not well enough conserved or even absent in plants, making it difficult to make strong claims based on inhibitor studies.
The signaling function of Ca2+ is currently best understood in the context of abiotic and biotic stress responses (Kudla et al., 2018), guard cell movement (Konrad et al., 2018) and in pollen tubes (Guo and Yang, 2020). In contrast, the function of the since long described auxin-induced Ca2+ response remains largely elusive (Vanneste and Friml, 2013; Shih et al., 2015; Dindas et al., 2018). Only recently, this for long overlooked aspect of auxin signaling regained attention with the identification of the non-selective cation channel CNGC14 as a critical component of auxin-induced Ca2+ entry (Shih et al., 2015; Dindas et al., 2018). CNGC14 activity was proposed to participate in root gravitropism (Shih et al., 2015) and root hair development (Dindas et al., 2018; Brost et al., 2019). Additionally, manipulations of Ca2+ availability and channels revealed connections to polar auxin transport (Dela Fuente and Leopold, 1973) and polarization of auxin transporters (Zhang et al., 2011; Li et al., 2019), indicating an important interplay between auxin and Ca2+.
In contrast to our poor understanding of auxin-induced Ca2+ signaling, the mechanism of auxin-induced transcriptional changes has been characterized in great detail (Lavy and Estelle, 2016; Roosjen et al., 2018; Powers and Strader, 2020). The canonical pathway for auxin-induced transcription involves the auxin-stabilized interaction between TIR1/AFB F-box proteins and Aux/IAA transcriptional corepressors (Dharmasiri et al., 2005; Kepinski and Leyser, 2005). This results in the ubiquitination and proteolysis of the latter (Nemhauser, 2018). Consequently, the transcriptional repressive effect imposed by Aux/IAA on ARF transcription factors is released, and auxin-induced transcription can proceed (Pierre-Jerome et al., 2016; Roosjen et al., 2018). Recently, a non-transcriptional branch of TIR1-based auxin signaling was demonstrated to effect acute inhibition of elongation (Fendrych et al., 2018; Gallei et al., 2020). Moreover, the non-transcriptional repertoire of TIR1/AFBs was recently further expanded by the observation that tir1/afb2,3 mutants are defective in auxin-induced Ca2+ signaling (Dindas et al., 2018). Additionally, auxin signals converging on pavement cell morphogenesis (Xu et al., 2010), lipid composition and distribution (Pan et al., 2009; Li et al., 2015; Platre et al., 2019), cell division during lateral root formation (Huang et al., 2019), suppression of auxin biosynthesis (Wang et al., 2020), stability and polarity of PIN proteins (Abas et al., 2006; Sauer et al., 2006; Baster et al., 2013; Prat et al., 2018; Mazur et al., 2020), and their internalization (Paciorek et al., 2005; Robert et al., 2010; Platre et al., 2019), possibly act via alternative auxin perception mechanisms, such as the receptor like kinase family TRANSMEMBRANE KINASE1-4 (Cao et al., 2019; Huang et al., 2019; Platre et al., 2019).
Based on the non-transcriptional character of the inhibition of internalization by NAA (Robert et al., 2010; Zhang et al., 2020), we postulated that auxin-induced Ca2+ signaling could be a signaling component in this response to NAA. Therefore, followed two strategies. On the one hand, we manipulated NAA-induced Ca2+ signaling using inhibitors or washing seedlings in Ca2+ free medium. We validated their effects on NAA-induced Ca2+ signaling, and assessed their impact on NAA-inhibited internalization. This approach provided a convincing and tight correlation between NAA-induced Ca2+ signaling and NAA’s ability to inhibit internalization. However, this correlation between Ca2+ signaling and inhibition of internalization could not be confirmed in tir1/afb and cngc14, two mutants that are specifically defective auxin-induced Ca2+ signaling. This discrepancy in outcome between both approaches calls for extreme caution when analyzing the role of Ca2+ in a process of interest as current pharmacology or manipulations of Ca2+ availability are prone to pleiotropic, misleading effects.
RESULTS
Characterization of NAA-induced Ca2+ signaling
The synthetic auxin, 1-naphthaleneacetic acid (NAA) is widely used in auxin biology as a proxy for the endogenous auxin indole-3-acetic acid (IAA) and is a more potent inhibitor of internalization than IAA (Paciorek et al., 2005). Because we wanted to evaluate the role of Ca2+ signaling in inhibition of internalization, we set out to characterize the NAA-induced Ca2+ response in Arabidopsis root meristems in more detail. The cytoplasmic, intensiometric Ca2+ indicator R-GECO1 (Keinath et al., 2015) reported an instant cytosolic Ca2+ elevation in response to 10µM NAA application (Figure 1A,B; Supplemental Movie S1). Also at lower concentrations (1µM and 0.1µM), NAA triggered rapid Ca2+ signaling, albeit with smaller amplitude (Figure 1B), illustrating a dose-dependence of the maximal response, similar to the one reported for the natural auxin indole-3-acetic acid (IAA) (Dindas et al., 2018). The onset of the Ca2+ increase started within seconds after NAA application, and reached a maximum after ±70 sec, followed by a gradual attenuation response. Similarly, subcellular targeting of ratiometric Ca2+ sensors revealed a rise in Ca2+ concentrations in the cytoplasm (visualized with NES-YC3.6 (Krebs et al., 2012)), in the cytosol near the plasma membrane (visualized with PM-YC3.6-Lti6b (Krebs et al., 2012)), in the endoplasmic reticulum lumen (visualized with CRT-D4ER (Bonza et al., 2013)) and in mitochondria (visualized with 4mt-YC3.6 (Loro et al., 2012)) (Supplemental Figure S1A-D). The signals in the ER and mitochondria were slightly delayed compared to the other reporters (Supplemental Figure S1C,D), suggesting that these organelles may act as Ca2+ sinks for attenuation of the cytoplasmic Ca2+ signal. The lack of Ca2+ response after benzoic acid (BA) treatment (Supplemental Figure S1E,G), shows that the Ca2+ response to NAA was not a response to acidification associated with NAA treatment. The NAA-induced Ca2+ response could be inhibited pharmacologically by the Ca2+ channel inhibitors Bepridil and Nifedipine (De Vriese et al., 2018; De Vriese et al., 2019) (Supplemental Figure S1F,G; Supplementary Movies S2,S3).
Additionally, we modulated the available extracellular Ca2+ by washing the seedlings with 0.5xMS medium that lacked CaCl2 (hereafter referred to as CaMIN). This simple treatment was sufficient to reduce the NAA-induced Ca2+ response in comparison to normal 0.5xMS medium (≈1.5mM CaCl2, hereafter CaPLUS) (Figure 1C). Jointly, these data suggest that NAA triggers a complex Ca2+ response that largely depends on extracellular Ca2+.
NAA-mediated Inhibition of internalization correlates with Ca2+ signaling
The synthetic auxin NAA interferes rapidly via a non-transcriptional pathway with the internal accumulation of plasma membrane proteins (hereafter referred to as internalization) in Brefeldin A (BFA)-induced intracellular endosomal aggregates (so-called BFA bodies) (Paciorek et al., 2005; Robert et al., 2010). This effect of NAA was concentration dependent, showing the maximum inhibitory effect at 10µM (Supplemental Figure S2A,B), correlating with the dose-dependence of the maximal NAA-induced Ca2+ response (Figure 1C,D). Given the immediacy of both processes, we postulated that auxin-induced Ca2+ responses reflects an auxin signaling cascade involved in NAA-regulated internalization.
Indeed, when we used the organic Ca2+ channel blockers Nifedipine and Bepridil at concentrations that interfered with NAA-induced Ca2+ entry (Supplemental Figure 1F,G), PIN1 internalization was restored in BFA/NAA co-treated roots (Supplemental Figure S2C). A similar nullification on NAA-inhibited PIN1 internalization was achieved using the membrane-permeable calmodulin inhibitor W-7 (Supplemental Figure S2C). This suggests that Ca2+ increase and its downstream signaling is required for NAA’s inhibitory effect on PIN1 internalization. Given the potential off-target effects of the Ca2+ channel inhibitors (De Vriese et al., 2018), we also evaluated the effect of CaMIN on NAA-inhibited PIN1 internalization. Similarly to Bepridil, Nifedipine and W-7, a 30min CaMIN pretreatment was sufficient to restore PIN internalization in BFA/NAA co-treated roots (Figure 2A,B). To evaluate the specificity of the treatment to Ca2+ availability, we analyzed PIN1 internalization in CaMIN supplemented with either 1.5mM CaCl2 (comparable to 0.5xMS) or 1.5mM MgCl2 (Figure 2A,B). The addition of CaCl2 fully restored the NAA sensitivity of PIN1 internalization. In contrast, the internalization in roots treated with MgCl2 supplemented CaMIN could not restore the NAA sensitivity, indicating the specificity of Ca2+ in this process. This effect of CaMIN on NAA-inhibited internalization could also be observed for other plasma membrane cargoes such as PIN2, AUX1 in the protophloem and NPSN12 (WAVE131-YFP) (Figure 2C-F). Jointly, these findings strongly support a notion that Ca2+ is required for NAA’s inhibitory effect on internalization of plasma membrane proteins.
The TIR1/AFB-CNGC14 module is not required for NAA’s inhibitory effect on internalization
TIR1/AFB-based auxin perception was demonstrated to be required for IAA-induced Ca2+ signaling via the Ca2+ permeable cation channel CNGC14 (Shih et al., 2015; Dindas et al., 2018). Assuming an auxintrigerred Ca2+ signal being part of the mechanism for regulation of PIN internalization, we predicted that the defective auxin-induced Ca2+ signaling in tir1/afb or cngc14 mutants would result in NAA-insensitive PIN internalisation. Contrary to expectations based on CaMIN and Ca2+ signaling inhibitors, none of the tested mutants showed NAA-insensitive internalisation in CaPLUS, suggesting that the TIR1/AFB-CNGC14 Ca2+ signaling module is not essential for this effect. Surprisingly, PIN1 internalization was inhibited by NAA in tir1afb1afb3, tir1afb2afb3 and three different cngc14 alleles also in CaMIN conditions, unlike in WT and other tir1/afb mutants (Figure 3B-E; Supplemental Figure S3A-E). In contrast to tir1/afb and cngc14 mutants, the NAA-sensitive internalization in CaMIN was not observed in the auxin biosynthesis defective mutant yuc3,5,7,8,9 (yucQ) (Chen et al., 2014) (Supplemental Figure S3F). Thus, the restored NAA sensitivity in CaMIN-treated tir1/afb and cncg14 mutants does not seem to be related to changes in auxin levels but is rather specific to TIR1/AFB-CNGC14-mediated auxin signaling. Jointly, these data demonstrate that in CaPLUS, NAA inhibits internalisation independently of TIR1/AFB-CNGC14-mediated auxin signaling.
The discrepancy between the exogenous Ca2+ manipulations and the mutant analyses, raised the hypothesis that NAA may induce Ca2+ signaling via a TIR1/AFB-CNGC14-independent mechanism. Therefore, we transformed the ratiometric Ca2+ indicator NES-YC3.6 in tir1/afb2,3 and cngc14-1 and analysed the Ca2+ responses to a pulse of NAA in these mutants. We found that the NAA-induced Ca2+ response was strongly reduced in tir1afb2,3 triple mutants and was completely absent from cngc14-1 (Figure 3A), suggesting that NAA elicits Ca2+ signaling through the same mechanism as IAA. Also in the absence of exogenously supplied Ca2+, the NAA-induced Ca2+ responses in tir1afb2,3 and cngc14-1 mutants were strongly defective (Figure 3B), excluding compensatory Ca2+ signaling mechanisms under low Ca2+ availability, as reported for pathogen-triggered immunity-associated Ca2+ signaling in cngc2 and cngc4 mutants (Tian et al., 2019). These data show that NAA-induced Ca2+ signaling is fully dependent on the TIR1/AFB-CNGC14 module. Notably, the YC3.6 Ca2+ indicator did not allow detecting the obvious differences in the NAA-induced Ca2+ response between CaMIN and CaPLUS that we observed when using R-GECO1 (Figure 1D). This probably reflects fundamental differences in Ca2+ binding properties and dynamic ranges that exist between both Ca2+ indicators’ properties (Nagai et al., 2004; Zhao et al., 2011; Keinath et al., 2015; Waadt et al., 2017). This also suggests that the reduction in Ca2+ response caused by CaMIN treatment is weaker than the one in tir1afb2afb3 and cngc14-1. The normal NAA sensitivity of PIN internalisation in these mutants, that are strongly defective in auxin-induced Ca2+ signaling, thus provides further evidence that auxin-induced Ca2+ signaling is not an essential part of the mechanism by which auxin affects trafficking. This is in direct contrast to the conclusions drawn using exogenous Ca2+ signaling manipulations, highlighting an important lack of specificity of these treatments.
DISCUSSION
Auxin-mediated regulation of trafficking is considered an important aspect of auxin’s self-regulating properties in plant growth and development. However, the underlying molecular mechanisms remain largely unclear. The synthetic auxins such as NAA displays a strong effect on trafficking, as illustrated by strong interference with accumulation of plasma membrane cargoes in BFA bodies (Paciorek et al., 2005). This effect was reported to be very fast, not requiring transcriptional changes and independent of canonical auxin signaling (Robert et al., 2010). Instead, an extracellular auxin perception mechanism was proposed based on the activities of AUXIN BINDING PROTEIN1 (Robert et al., 2010). Using updated genetic tools, however, the latter was put into question (Gao et al., 2015). Given the immediacy of the response, we hypothesized that auxin-induced Ca2+ signaling could be involved in auxin’s inhibitory effect on internalization. We established that NAA activates Ca2+ responses at the plasma membrane via a TIR1/AFB-CNGC14-dependent mechanism, similarly as was recently described for the endogenous auxin IAA (Shih et al., 2015; Dindas et al., 2018). Using mutants and pharmacology we interfered with auxin-induced Ca2+ signaling and evaluated of the impact on internalization. The pharmacological interference revealed a good correlation between intensity of auxin-induced Ca2+ signaling and NAA’s ability to inhibit internalization, supporting our original notion. In striking contrast, genetic disruption of auxin-induced Ca2+ signaling did not affect NAA’s ability to inhibit internalization. These data unequivocally show that auxin-induced Ca2+ signaling does not inhibit internalization. It does however, not exclude roles for TIR1/AFB-CNGC14-mediated Ca2+ signaling in the auxin-regulated vacuolar trafficking of PIN proteins, that depends on TIR1/AFB function (Baster et al., 2013), or PIN polarization (Sauer et al., 2006; Prat et al., 2018; Mazur et al., 2020; Mazur et al., 2020). The contrasting results obtained using mutants versus the currently available Ca2+ pharmacology or manipulating Ca2+ availability illustrate that Ca2+ signaling in plants is a highly interconnected system, not allowing for easy, specific manipulations. Instead, the current lack of highly specific pharmacology imply that the most reliable conclusions can be drawn through genetic analysis.
CONCLUSION
Auxin is well-known to trigger a Ca2+ response. However, its physiological role remains poorly understood. Here, we evaluated auxin-induced Ca2+signaling in the context of regulating internalization of plasma membrane cargoes. Pharmacological evidence indicated its requirement. However, genetics demonstrated unequivocally that NAA’s effect on internalization is independent of auxin-induced Ca2+ signaling.
MATERIALS AND METHODS
Plant Growth Conditions
Arabidopsis thaliana seeds were sterilized by using bleach gas (8mL concentrated HCl to 150mL bleach) overnight, afterwards the seeds were sown on Petri dishes (12 cm X 12 cm) containing sterile half-strength Murashige and Skoog (½ MS) medium (0.5 x MS salts, 0.8% sucrose, 0.5 g/L 2-(N-morpholino) ethanesulfonic acid, pH 5.7, and 1% w/v agar), and grown under continuous light (21°C, continuous light), after 2 days vernalization at 4°C in the dark. To prepare 2L liquid CaMIN medium following components were dissolved in MilliQ: 100mL MS basal salt micronutrient solution (catalog Nr), 20 g sucrose, 0.20 g myoinositol, 1.00 g MES, 1.652g NH4NO3, 0.180g MgSO4, 1.920g KNO3, 0.152g KH2PO4 and pH 5.7. The liquid CaPLUS medium contains an additional 0.332g CaCl2 per 2L.
Chemicals
The following hormones/chemicals were used: 10μM NAA (catalog Nr N0903; Duchefa Biochemie). 25μM Brefeldin A (BFA, catalog Nr B6542-25MG). All hormones/drugs were dissolved in 100% dimethylsulfoxide (DMSO; catalog Nr D4540-500ML) and obtained from Sigma Aldrich.
Plant Lines Used
The Arabidopsis used as control in this study was Columbia (Col-0) ecotype. We used the following mutants and transgenic lines were described previously: tir1-1 (Ruegger et al., 1998), afb2-3 (Parry et al., 2009), afb3-4 (Parry et al., 2009), tir1-1afb2-3 (Parry et al., 2009), tir1-1afb3-4 (Parry et al., 2009), tir1-1afb1-3afb3-4 (Parry et al., 2009), tir1-1afb2-3afb3-4 (Parry et al., 2009), yucQ (Chen et al., 2014), cngc14-1 and cngc14-1 (Shih et al., 2015), cngc6,9, cngc6,14, cngc9,14 and cngc6,9,14 (Brost et al., 2019), WAVE131-YFP (Geldner et al., 2009), R-GECO1 (Keinath et al., 2015), NES-YC3.6 (Krebs et al., 2012), PM-YC3.6 (Krebs et al., 2012), CRT-D4ER (Bonza et al., 2013), 4mt-YC3.6 (Loro et al., 2012). The NES-YC3.60 (Krebs et al., 2012) reporter was transformed directly into tir1-1afb2-3afb3-4 (Parry et al., 2009) and cngc14-1 (Shih et al., 2015). Transformants were selected based on uniform expression, and expression levels comparable to the control NES-YC3.60. Analyses of the Ca2+ response was done on T2 generation seedling roots showing strong and uniform expression.
Immunodetection and Confocal Microscopy
The seedlings used for Immunodetection are 3-day-old and pre-treated in liquid CaPLUS, CaMIN in presence or absence of chemicals as indicated for 30 minutes. The samples were fixed by paraformaldehyde (4%) in PBS for 1 hour in vacuum. The following steps of the immunostaining were performed by the immuno-robot InsituProII, as described by Sauer et al. (2006). The dilutions of the primary antibodies were: goat anti-PIN1 (1:600) (sc-27163, SantaCruz), rabbit anti-PIN2 (1:600) (Abas et al., 2006) and anti-AUX1 (1:400) (AS 12 1868, Agrisera). The secondary antibodies used were AlexaFluor488 donkey anti-goat (1:600) (A-11055, ThermoFisher) and AlexaFluor555 donkey anti-rabbit (1:600) (A-31572, ThermoFisher).
Microscopy and image analysis
R-GECO1-based Ca2+ imaging experiments and analysis were performed as described (Himschoot et al., 2018). Yellow Cameleon-based experiments were performed and analyzed as described (Behera et al., 2013). Confocal laser scanning microscopes Leica SP2 (Leica) or Zeiss 710 CLSM microscopes were used to analyze the immunolocalisations, and imaging of R-GECO1. Fluorescence emission of Alexa488 (ex 488 nm/em 500-545nm), Alexa555 (ex 561nm/em 555-610nm), YFP (ex 514nm/em 520-565nm), was detected using a 63x water objective. Images were analyzed using Fiji (Schindelin et al., 2015). Fiji was used to rotate, crop images and label the region of all the roots for quantification. The proportion of cells with BFA bodies was scored manually and calculated by using Excel. The BoxPlotR was used to generate the box plots figures (Spitzer et al., 2014).
Statistical analysis
For statistical analysis of the Ca2+ imaging data, unpaired two-tailed t-tests with Welch correction for unequal standard deviations between populations where performed using GraphPad Prism (GraphPad Prism 8 for Windows 64-bit, version 8.4.1).
For statistical analysis of the immunolocalization experiments, a logistic regression was performed to compare the presence of BFA bodies in root cells of treated versus untreated roots or wild type versus mutant. A random effect was added to the model for the experiments with multiple repeats to take into account the correlation between measurements done at the same time. The analysis was performed with the glimmix procedure from SAS (Version 9.4 of the SAS System for windows 7 64bit. Copyright 2002-2012 SAS Institute Inc. Cary, NC, USA (www.sas.com)). Maximum likelihood estimation was done with the default estimation method. A Wald-type test was performed to estimate the treatment/genotype effect on the presence of BFA bodies in the root cells.
Acknowledgements
We thank Yunde Zhao (UC San Diego, USA), Marc Estelle (UC San Diego, USA), Petra Dietrich (F.A. University of Erlangen-Nurnberg, Germany) and Niko Geldner (University of Lausanne, Switserland) for sharing published materials, and NASC for providing seeds. We thank Veronique Storme for help with the statistical analyses. This work was supported by grants of the China Scholarschip Counsil (CSC) (to R.W. and J.C.), the special research fund Ghent University (to E.H.), the European Research Counsil (ERC) (to D.V.D and J.F.), the Deutsche Forschungsgemeinschaft (DFG) through Grants within FOR964 (M.K. and K.S.), Piano di Sviluppo di Ateneo 2019 (University of Milan) (to A.C.), and by a PhD fellowship from the University of Milan (to MG).
Footnotes
↵* Joint first authors